PHOTOSYNTHESIS AND RELATED PROCESSES, I ^ PHOTOSYNTHESIS and Related Processes By EUGENE I. RABINOWITCH Research Associate, Solar Energy Conversion Research Project, Massachusetts Institute of Technology, Cambridge, Massachusetts. Y l^M^U JxIm^ J. iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii liiiiiiiiiiiiiiiiiiiiminii Chemistry of Photosynthesis, , Chemosynthesis and Related j Processes in Vitro and in Vivo iiiiiniiiiiiiiiiniiiiiiiiiiiiiiiii null iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiMii 0) are called endothermal (because they consume heat, if carried out at a constant temperature). For reactions wliich occur with an increase m free energy {AF > 0), the term endergonic has been suggested. The total energy (or "heat effect") is a characteristic constant of a chemical reaction (at a given temperature) ; while the free energy depends on the concentrations of the reaction components and reaction products. Photosynthesis is a strongly endothermal, and (with the usual concen- trations of carl)on dioxide and oxygen), an even more strongly endergonic process. Certain bacteria can live autotrophicully, without carrying out true photosynthesis. Some of them sjmthesize organic matter, in the dark, with the help of the free energy of unstable organic or inorganic chemical 4 PHOTOSYNTHESIS IN NATURE CHAP. 1 systems; these are called "chemo-autotrophic" bacteria. Others, so- called purple bacteria, use light for the synthesis of organic matter from carbon dioxide and organic or inorganic hydrogen donors, for example, hydrogen sulfide or fatty acids. In this variation of photosynthesis (called "photoreduction" by Gaffron, cf. page 129), light energy is utilized mainly for temporary activation and not for permanent conversion into chemical energ3^ The energy of the organic matter produced by purple bacteria is only to a small part converted light energy; most, if not all of it is chemical energy transferred from one unstable chemical sys- tem to another. The existence of these bacteria is possible only because the crust of the earth has not yet settled into a complete chemical equilibrium, and spots of high chemical potential can still be found here and there (particularly in volcanic regions). Conceivably, these peculiar modes of autotrophic life (we will speak more of them in Chapter 5) may have played a greater role in earlier geological ages, when the chemical activity on the surface of the earth was more widespread and violent. They are consequently of considerable interest in speculations as to the origin and development of life on this planet. In the contemporary cycle of the living matter on earth, these processes are of no consequence. Photosynthesis by green plants alone prevents the rapid disappearance of all life from the face of the earth. We cannot definitely assert that photosynthesis takes exactly the same course and leads to the same primary product in all organisms from the lowly diatoms to the highly organized flowering plants. Differences in structure and composition of the photosynthetic organs of different species (described in Chapters 14 and 15) make minor variations in the mecha- nism of photosynthesis probable. However, the universal occurrence of chlorophyll in all photosynthesizing plants and the similarities between the kinetic relationships governing photosynthesis in unicellular algae (e. g., Chlorella), and in higher land plants (e. g., wheat) {cf. Vol. II, Chapters 27 and 28) indicate that the general characteristics of the process must be the same throughout the plant world. After the completion of photosynthesis, the plants and animals begin to aminate, halogenate, polymerize, oxidize, reduce or dismute the first products of photosynthesis, thus producing fats, proteins, nucleo- proteids, pigments, enzymes, vitamins, cellulose and other structural ma- terials. In due time, before or after the death of the organism, all these compounds will be oxidized and decomposed. This decomposition goes by many different paths. Only one of them, the oxidation of sugars by the respiratory system, appears as a direct reversal of photosynthesis; and even in this case, it is doubtful whether the analog}' extends beyond the over-all result {cf. Chapter 9, section 4). The reservoir of life is fed by a single channel, through which matter is pumped up from the low-lying TOTAL YIELD OF PHOTOSYNTHESIS ON EARTH 5 sea of the stable inorganic world, to the high plateau of organic life; it finds its way back in hundreds of streams or meandering rivulets, which set into rotation, as tlie.y hurry down towards the sea, thousands of little wheels of life. B. The Total Yield of Photosynthesis on Earth * To acquire an adequate notion of the importance of photosynthesis in the chemical household of the earth, it is interesting to estimate the total turnover of matter and energy involved in this process. This can be done, of course, only very approximately. The total yield of photosynthesis on earth was first evaluated by Liebig in his famous book. Chemistry in its Application to Agriculture and Physiology, whose first edition appeared in 1840. He estimated that, if all the land were a single meadow with a yearly crop of 5 metric tons (1 ton = 10'' g.) per hectare (10^ sq. meters), all carbonic acid in the air would be used up in from 21 to 22 years. Considering the carbon dioxide content of the air (c/. Table l.IV), this statement is equivalent to the assertion that the plants utilize 10" tons of carbon dioxide and produce 3 X 10'" tons of organic carbon annually. Arrhenius (1908) and Cia- mician (1913) made similar calculations, but assumed an average crop of only 2.5 tons per hectare of land. (They gave Liebig's authority for this figure; but according to Schroeder 1919, this was a misquotation). They thus obtained a yearly yield of only 1.8 X 10'" tons of organic carbon. Ebermayer (1885) substituted a more elaborate picture for Liebig's simplified assumption of "all land a single meadow." He distinguished between wooded areas, cultivated fields, steppes and barren lands, and added to the crop the roots and stubbles remaining in the fields. In this way, he arrived at a figure of 2.4 X 10'" tons of organic carbon for the annual production of organic matter by the plants. The next attempt, on the basis of improved statistical data, was made by Schroeder (1919); the results are condensed in table l.I. Schroeder's total of 1.63 X 10'" tons of carbon for the whole surface of the land, corresponds to an average of 1.1 tons per hectare. Assuming that 15% of the organic matter synthesized by the plants is used up by their own respiration, this total can be revised upward to 1.9 X 10'" tons of carbon per annum. In this estimate, the production of organic matter in the oceans was altogether neglected. Schroeder made an estimate for the benthos, i. e., the ground-attached algal vegetation of the continental ledge, and found its contribution negligible compared to that of the land plants. As to the free-swimming plankton, he saw no way of estimating its yield but * Bibliography, page 1 1 . G PHOTOSYNTHESIS IN NATURE CHAP. I Table l.I Carbon Fixation by Land Plants (after Schroeder 1919) Area, millions of sq. km. Rate of carbon fixation, millions of tons per annum riant habitat Estimated from to Probable mean value Woods Farmland Steppes Deserts 44 27 31 47 9000 3500 500 100 13000 4500 2200 500 11000 4000 1100 200 Total: 149 13100 20200 16300 suspected that it too, is relatively small. Later (1919-), he conceded that the carbon fixation by the plankton may as much as double the yield calculated for the land plants alone, thus bringing the annual rate of carbon fixation by all plants to 3.8 X 10^*^ tons. It seems that Schroeder's plankton correction still was much too small. Recent experiments make it probable that the oceans account for much more than one-half the total organic synthesis on earth. Table l.II, Table l.II Carbon Dioxide Reduction by the Plankton (from Riley) Location Carbon per annum per sq. m., g. Method Reference Long Island Sound 600-1000] Gross production W. Atlantic 23°-38° N. W. Atlantic 38°-4r N. 530 [ 320 and oxygen liber- ation in experi- Riley (1938, 1939, 1941) Dry Tortugas 60-430 mental bottles W. Atlantic 3°-13° N. 278 [Oo] deficit Seiwell (1935) English Channel 60-98 Changes in [CO.], [O2], [P] and [N]. Cooper (1933) Barents Sea 170-330 Consumption of Kreps and Verbinskaya phosphorus (1932) Average : 375 compiled b}^ Riley (1941) shows that the production of organic matter by the plankton does not change much from the Equator to the Polar Circle; and that the average of all measurements is as high as 375 g. of organic carbon annually per sc}. meter', corresponding to 3.75 tons per hectare. If this yield of organic carbon is taken as representative of all oceans, TOTAL YIELD OF PHOTOSYNTHESIS ON E.\RTH 7 multiplied by their area (361 X lO*' sq. km.) and corrected for 15% respiration losses, the result is 15.5 X 10^° tons, or eight times the yield of carlion fixation on land as calculated by Schroeder! Even a deduction of 10 or 20% for the Polar Sea, and generally less fertile waters, would not change this result significantly. Thus, the most probable value of the rate of carbon fixation on earth is 15-20 X 10^° tons annually, with at least four-fifths (and perhaps nine-tenths) of this amount contributed by the oceans. Table l.III Carbon Fixation by L.\ND AND Sea Plants Plant habitat Area, millions of sq. km. .Average carbon fixation per hectare per year, tons Total carbon fixation per year, tons Oceans Land 361 149 3.75 1.3 15.5 X 10»» 1.9 X IQi" It is interesting to compare this rate of carbon transformation by the plants with the total quantity of carbon available on earth. Table LlV T.\BLE l.IV Carbon Reserves of the Earth Region of earth or atmosphere Total amount of carbon, tons Lithosphei-e (earth's crust 16 km. deep) Hj'drosphere (oceans and seas) Troposphere (air up to 11 km. height) 2-8 X 10'6 5 X 10" 6 X lO'i lists some geochemical data (cf., for example, Vernadsky 1930). The large amount of carbon in the carbonate rocks is almost unavailal)le to the plants. The large reservoir of dissolved carbonates and bicarbonates, on the other hand, is fully available to aquatic plants, and stands in a continuous exchange with the gaseous carlion dioxide in the atmosphere, thus helping to maintain the concentration of the latter on a constant level, and contributing indirectly to the food supply of land plants. Ac- cording to the figures given in Table l.III, the land plants assimilate a quantity of carbon equivalent to the total amount of carbon dioxide in the air above the continents, in less than ten years. An approximate confirmation of this estimate Avas provided by (Jut (1938). who measured the carbon dioxide concentration of the air at different heights above a pine forest and calculated that each day the trees consume all carbon dioxide from an air column 50 meters high. Since the atmosphere corre- 8 PHOTOSYNTHESIS IN NATURE CHAP. 1 sponds to an air layer approximately 8 X 10^ meters thick (under stand- ard conditions), Gut's calculations indicate a yearly consumption of about 22% of all carbon dioxide in the air column above the forest. However, the rapid exchange of carbon dioxide between atmosphere and hydrosphere makes the separate comparison of the carbon utilization by land plants with the amount of atmospheric carbon dioxide, and of the carbon utilization by sea plants with the quantity of dissolved carbonates irrelevant. Instead, we have to compare the total carbon fixation by all plants, on land and in the sea, with the total carbon reserve available in both the troposphere and the hydrosphere. The plants assimilate a quantity of carbon equal to this reserve in 300 or 400 years. To maintain the cycle, an equivalent quantity of carbon dioxide must be liberated during the same period bj- respiration, and by the decay of vegetable and animal matter. Veniadsky (1930) estimated the weight of the "biosphere"' on earth as 10'^ tons, and postulated its renewal "several times a year." In other words, he assumed that an average carbon atom remains in organic combination for only a few months. These as- sumptions call for a rate of photosynthesis of the order of at least 10" and more probably 10^* tons of organic carbon per year (i. e.. from 60 to 600 times more than was allowed above) and for the consumption, each year, of the whole amount of carbon available in the air and in the water of the oceans. In another place, Vernadsky speaks of the living organisms transforming in a single year "more than the total quantity of carbon on the earth." Even if he means only the carbon content of the air and water, this estimate appears much too high, if compared with the better supported figures of Liebig, Schroeder, and Riley. To make his figures plausible, Vernadsky criticizes the values of Liebig for average crops because of the neglect of roots and stubbles, and points out that some plants can produce crops far in excess of these averages. He quotes, for example, 50 tons of dry organic matter per hectare (not including wood and leaves) harvested from a banana plantation; 250 tons fresh tubers which Manihoi/utilissima Pohl can produce per hectare, as well as other types of plants which can yield as much as 50 or 60 tons of organic carbon per hectare annually. Table l.III shows, however, that unless Vernadsky is willing to postulate that the average annual plankton production in the sea is not 3.75 tons, as estimated by Riley, but 30 or .50 tons of carbon per hectare, his assumption of a yearly fixation of 10" tons carbon is impossible. Our estimates of the rate of carbon fixation on earth can be checked by calculations of an entirely different and not less interesting type — based on the amount of available sun energy and its average utilization in photosynthesis. The energy of the sun radiation reaching the upper boundaries of the atmosphere is 1.25 X 10-'^ cal. per annum, and only about 40% of this energy penetrates to the surface of the earth, the rest being scattered and absorbed by the clouds and by the atmosphere. Of the 5 X 10'-^ cal. which reach the earth's surface, 50% are in the form of infrared and extreme red radiations, which are not used in photosynthesis, and at least 20% are absorbed by rocks, sand and ploughed fields, or TOTAL YIELD OF PHOTOSYNTHESIS OX EARTH 9 reflected by ice and snow, so that not much more than 2 X 10'^ cal. can be allocated to plant-covered land and plankton-filled sea. Of this amount, at least lO^c are lost by reflection on water surface, and further losses must occur through the absorption of visible light by water (c/. Vol. II, Chapter 22) and dissolved ions. On land, too, 10 or 20% of all radiation which falls on plant -covered areas is lost by diffuse reflection. (Woods and forests are more efficient light traps than meadows and fields, and therefore appear as dark spots on aerial maps.) Altogether, not more than 1.5 X 10-^ cal. are absorbed annually by the plant pigments, and can thus be utilized in photosynthesis. It will be shown in volume II, chapter 28 that, under natural conditions, the energy conversion yield of green plants is of the order of 2%. This brings us to the figure 3 X lO'-^ cal. for the probable annual energy accumulation by photosynthesis, corresponding to the formation of 3 X 10" tons of organic carbon. (The heat of combustion of organic matter is approximately 10^" cal. per ton of carbon contained in it.) This agrees with the value derived above from crop estimates, and confirms the utter impossibility of the much larger figures of Vernadsky. The reduction of carbon dioxide by green plants is the largest single chemical process on earth. To make clearer what a yield of 10" tons per year means, we may compare it with the total output of the chemical, metallurgical, and mining industries on earth, which is of the order of 10^ tons annually. Ninety per cent of this output is coal and oil, i. e., products due to photosynthesis in earlier ages. Similarly impressive is the comparison of the energy stored annually by the plants, with the energy available from other sources. The energy converted by photo- synthesis is about one hundred times larger than the heat of combustion of all the coal mined on earth in the same period, and ten thousand times larger than the energy of falling water utilized in the whole world. On the other hand, about three hundred times more solar energy is spent on the evaporation of water from the oceans and continents than is utilized in photosynthesis. Plants alone spend, according to the estimates of Schroeder (1919-), l(i X 10-^ cal. annually on transpiration, or more than ten times more than on photosynthesis. Since aciuatic plants have no need for transpiration, all this energy is used up by land plants. The ratio of transpiration energy to the energy of photosynthesis is, for land plants, of the order of 50 or 100 to 1. The carbon dioxide cycle in, nature is of great importance for the climate of the earth; even small changes in its photostationary state, which probably have occurred in the history of the earth, must have had far-reaching consequences. It is probable, for example, that in the pre- glacial age, the carbon dioxide concentration in the air was higher than now, and the climate warmer (because carbon dioxide prevents the escape 10 PHOTOSYNTHESIS IN NATURE CHAP. 1 of infrared radiation from the earth). Photosynthesis was therefore more intense than now, and the cover of vegetation denser. The removal of carbon dioxide by the minerahzation of coal and oil, the formation of carbonate rocks and tlie increase in the concentration of alkaline earth ions in the oceans, have caused a decrease in photosynthesis and drop in temperature, and thus contributed to the advent of the glacial period. We cannot dwell here much longer on the role of photosynthesis in the evolution of the earth and its influence on the geochemical distribution of the elements. It must be mentioned, however, that in addition to the carbon dioxide cycle, photosynthesis gives rise also to a natural cycle of oxygen. The atmosphere contains approximately 2.8 X 10^^ tons of this element. If the living organisms consume annually the equivalent of 15 X 10^" tons of carbon dioxide, i. e., 12 X 10^° tons of oxygen, they must renew all the oxygen in the air in a little over two thousand years, and decompose all the water in the oceans in about two million years. This is still a short period compared with the age of life on earth ; we can thus conclude that all oxygen now present on the surface of the earth, as H2O or O2, has repeatedly passed, in previous geological ages, from the atmos- phere through the biosphere into the hydrosphere and back. Since this cycle is composed of a photochemical forward reaction and a nonphoto- chemical back reaction, it does not lead to a thermodynamic equilibrium, but rather to a steady " photostationary state." This fact may be of importance for the distribution of oxygen isotopes between air and water. It has been revealed by the measurements of Dole (1936), Greene and Voskuyl (1936) and others, that the oxygen in the air is about 7.5 X 10~^ atomic weight units heavier than oxygen in the water of the oceans. If air and water were in a thermodynamic isotopic equilibrium at the average temperature prevailing at sea level, the difference should be three times smaller. Among several attempts to explain this discrepancy, Greene and Voskuyl suggested that photosynthesis may play a part in it. They pointed out that, if plants convert oxygen from carbon dioxide (two O atoms) and water (one O atom) into free oxygen without discrimination between O^^ and O^^, the oxygen produced in this way must be 1 X 10~^ atomic weight units heavier than oxygen in water (because the heavy isotope is more abundant in carbon dioxide than in water). However, this explanation presumes that oxygen in the air is the product of a single photosynthesis, whereas we have seen above that it must have undergone repeated back and forth transfers. Furthermore, it has been long suspected, and recently confirmed by exi)eriments with radioactive tracers, (c/. page ')'■>), that all oxygen produced in photosynthesis comes from water. Thus, the hypothesis of Greene and Voskuyl is untenable. However, photosynthesis may have influenced the oxygen isotope distribution between air and water in a different way. It was suggested above that this distribution corresponds to a photostationary state, and not to a thermodynamic equilibrium. If oxygen is produced indiscriminately from HoO'8 and H2O"' in photosynthesis (as seems to be indicated by the results of Vinogradov and Teis 1941) but the heavy isotope reacts slower in tlic thermal liack reaction, tiiis must bring about an accunmlation of the lieavy isotope in the atmosphoro, and may thus account for tlie liighcr density of atmospheric oxygen. BIBLIOGRAPHY TO CHAPTER 1 11 Bibliography to Chapter 1 Total Yield of Photosynthesis on Earth 1840 Liebii;:, J., Chemie in ihrer Anwendung auf Agriculhir xind Physiologie. 1st ed., Vieweg, Braunschweig, 1840. 1885 Ebermayer, E., Die Bcschaffcnheit der WaUUuft und die Bedeidiing der atmospharischen Kohlensdure fiir die W aldveg elation. Enke, Stutt- gart, 1885. 1908 Arrhenius, S., Werden der Welten. Akaderaische Verlagsgesellschaft, Leipzig, 1908, ]). 51. 1913 Ciamician, G., Die Photochemie d^r Zukunft, Enke, Stuttgart, 1913, p. 17. 1919 Schroeder, H., X at urwissenschaften, 7, 8. Schroeder, H., ibid., 7, 976. 1930 Vernadsky, V. I. (Wernadsky), Geochemie. Akademische Verlags- gesellsc'liaft, Leipzig, 1930; La biosphere. Pari.s, 1930. 1932 Krei)s, E., and Verbinskaj^a, X., /. conseil intern, exploration mer, 7, 25. 1933 Cooper, L. H. X.. ./. Marine Biol. Assoc. United Kingdom, 20, 197. 1935 Seiwell, H. R., ./. conseil intern, exploration mer, 10, 20. 1936 Dole, M., /. Chem. Phys., 4, 268. Greene, C. H., and Voskuyl, R. J., /. Am. Chem. Soc, 58, 693. 1938 Gut, Ch., J.forestier suisse, 89, 195, 236. Riley, G. A., /. Marine Research Sears Foundation, 1, 335. 1939 Riley, G. A., ibid., 2, 145. 1941 Riley, G. A., Bull. Bingham Oceanogr. Coll., 1,1. Vinogradov, A. P., and Teis, R. V., Compt. rend. acad. sci. L'RSS, 33, 490. Chapter 2 THE DISCOVERY OF PHOTOSYNTHESIS * 1. Precursor: Stephen Hales Stephen Hales (1677-1761), the minister-naturalist of Teddington, England, an illustrious contemporary of Newton, made numerous ex- periments on the evolution and absorption of gases by different sub- stances of animal or vegetable origin, and decided thereupon that air must be an important constituent of all organic matter. (He meant by this, that air, one of the four original elements of Aristotle, must be added to the list of alchemistic "first principles" — mercury, oil, salt, sulfur — which were supposed at that time to contribute to the composition of all material bodies.) Discussing further, in his Vegetable Staticks (1727), the importance of leaves for plants. Hales wrote: "Plants very probably draw through their leaves some part of their nourishment from the air," and he added, "may not light also, by freely entering surfaces of leaves and flowers, contribute much to ennobling the principles of Vegetables?" 2. The Background : The Birth of Pneumochemistry No further elaboration of Hales' vision was possible until the existence and nature of different kinds of "air" became known through the work of the great "pneumochemists" of the second half of the eighteenth century. Black discovered "fixed air" (that is, carbon dioxide) in 1754; Scheele prepared chlorine in 1774, and found in 1773 that atmospheric air is composed of two gases, one inert and the other capable of main- taining combustion. (This observation was not made public until 1777, thus depriving Scheele of the priority in the discovery of oxygen.) In 1775, Priestley obtained "dephlogisticated air" (that is, oxygen) from mercurous oxide; and later the same author described nitrous oxide, sulfur dioxide, hydrochloric acid gas, and carbon monoxide. "Inflammable air" (that is, hydrogen, although, for a while, methane was often con- fused with it) was discovered by Cavendish in 1766; in 1784, the same author proved that water is a compound of hydrogen and oxygen. Be- tween 1772 and 1782, Lavoisier discovered the composition of the air and propounded the new doctrine of oxidation and respiration, interpret- ing the.se processes as combinations of different substrates with oxygen, * Bibliography, page 27. 12 PRIESTLEY 13 and this concept rapidly displaced the old phlogiston theory of Stahl. In 1781 Lavoisier showed that "fixed air" is a compound of carbon and oxygen. Such was the background of rapid progress in the chemistry of gases between 1750 and 1775, which made the discovery of photosynthesis possible, yes, almost inevitable. Of three men whose names are associated with this discovery, Priestley, Ingen-Housz and Senebier, two were clerics, like Hales, and one a physi- cian; all were typical amateur naturalists of the Age of Enlightenment. There their similarity ended; for it would be difficult to find men more unlike each other than the militant nonconformist minister, Joseph Priestley, the pompous, brilliant, court physician, Jan Ingen-Housz, who was equally at home in Amsterdam, London, Paris and Vienna, and Jean Senebier, a plodding, provincial pastor from the pious and savant town of Geneva. 3. The Purification of Air by Plants: Priestley Joseph Priestley (1733-1804) was undoubtedly the greatest scientist of the three, although he considered science the least important of his many occupations. His foremost interest was in theology and philoso- FiG. 1. — Joseph Prie.stley. phy; his ardent nonconformism brought him into perpetual conflict with authorities and into disrepute as a sympathizer with the French Revolu- tion. In the stormy days of 1791, his house in Birmingham was sacked bj' 14 DISCOVERY OF PHOTOSYNTHESIS CHAP. 2 the mob, whereupon the French RepubHc made him an honorary citizen, and offered him refuge in France; however, he preferred to exile himself to America, where he spent the last ten yesivs of his life in the little town of Northumberland, Pennsylvania. Throughout his life, Priestley enjoj^ed playing about with gases; his writings reveal him as one of the most skillful and successful experi- mentalists of all times; but his powers of logic and analysis were reserved for philosoph}^ and theology, and in the presentation of his experiments he stuck to the old phlogiston theory — long after the new concepts of Lavoisier had found general acceptance. Thus, the wide implications of Priestley's greatest discovery — oxygen — escaped him; and he was even less aware of the general import of his experiments with green plants, and quite willing to leave their exploitation to others. The following quota- tions (1772) show the extent of his contribution to the discovery of photosynthesis : " I have been so happy as by accident to hit upon a method of restoring air which has been injured by the burning of candles and to have discovered at least one of the restor- atives which Nature employs for this purpose. It is vegetation. One might have imagined that sinee common air is necessary to vegetable as well as to animal life, both plants and animals had affected it in the same manner; and I own that I had that expectation when I first put a sprig of mint into a glass jar standing inverted in a vessel of water; but when it had continued growing there for some months, I found that the air would neither extinguish a candle, nor was it at all inconvenient to a mouse which I put into it. "Finding that candles would burn very well in air in whicli plants had grown a long time ... I thought it was possible that plants might also restore the air which had been injured by the burning of candles. Accordingly, on the 17th of August 1771 I put a sprig of mint into a quantity of air in which a wax candle had burned out and found that on the 27th of the same month another candle burnt perfectly well in it." This momentous observation was described in the Philosophical Trans- actions of the Royal Society in 1772; and reprinted in the first volume of Priestley's famous work, Experiments and Observations on Different Kinds of Air, which appeared in 1776. Other experiments, described in the same ])aper, showed that plants thrive particularly well in air made "obnoxious" by the exhalations of animals. The discovery of the complementary character of the chemical functions of plants and animals made a great impression on Priestley's contemporaries, and brought him in 1773 the award of the Copley medal of the Royal Society. However, Priestley did not return to the study of air improvement by plants until 1777; and the first new results on this subject appeared in his second physicochemical work, Experiments and Observations Relating to Various Branches of Natural Philosophy, whose first volume was published in 1779. These new observations were both interesting and confusing. Several scientists abroad, notably Scheele, PRIESTLEY 15 had failed to confirm the beneficient effect of plants on air; and now Priestley too found himself unable to obtain positive results regularly. Instead of his earlier categorical statements, he now wrote: "Upon the whole, I think it probable that the vegetation of healthy plants, growing in situations natural to them, has a salutarv effect on the air." Before ^ DoC/YJ/i PiFLOa/STOX. //'//'//,v// //;•/..;// Fig. 2. — A contemporary cartoon of Joseph Prie.stley. he succeeded in finding an explanation for the irregular results (which, we think now, might have been caused by poor illumination), Priestlej^'s attention was diverted by an observation which he called "the most extraordinar}^ of all ixvy unexpected discoveries." He found, namely, that a "green matter" deposited on the walls of many of his water con- tainers, formed bubbles of pure "dephlogisticated air" (that is, oxygen), whenever it was illuminated by the sun. At first, Priestle}^ thought this 16 DISCOVERY OF PHOTOSYNTHESIS CHAP. 2 matter to be of vegetable nature. How close he was, at this point, to the final discovery of photosj^nthesis ! However, he let himself be deceived by microscopic observations which revealed no organic forms in the green matter, and bj^ the formation of this matter in closed vessels, and decided that it was a thing "sui generis," of mineral rather than organic character. Subsequent observations diverted him even further away from the right track — he noticed that water decanted from the green matter also evolved "purified air" upon shaking, and that even pure pump water, not visibly contaminated with green matter, produced "purified air" upon prolonged standing in sunlight. The picture thus became more and more confused, as Priestley acknowledged in the following sentences: "It will probably be imagined that the result of the experiments recited in this section throws some uncertainty on the result of those from which I have con- cluded that air is ameliorated by the vegetation of plants, and especially as the water by which they were confined was exposed to the open air and the sun in the garden. To this I can only say that I have represented the naked facts, as I have observed them ; and having not great attach- ment to any particular hypothesis, I am very willing that my reader should draw his own conclusions for himself." (However, he added some arguments which made him believe that the previously observed plant effects were genuine and not due to the illumination of water.) Obviously, Priestley's careful attention to factual details and his re- fusal to attach too much importance to hypotheses — in short, his purely experimental approach — preve ted, at that time, his complete realization of the true nature of photosynthesis. His doubts were cleared away two years later, when he published the second volume of Observations and Experiments in Natural Philosophy. By then, the green matter was definitely identified as vegetable in nature. (It is interesting to reflect that the same unicellular green algae, which have recently become the fa- vorite subjects of photosynthetic study, served in the discovery of this phe- nomenon over a century and a half ago.) The action of water decanted from the green algal deposits, which baffled Priestley in 1779, was ex- plained in 1781 as an effect of supersaturation with ox\^gen; the formation of green deposit in closed vessels was attributed to imperfect closure and contamination of water with "seeds" before corking. Thus, the picture of oxygen being formed by the cooperation of green vegetable matter and sunlight, emerged clearly from the temporary confusion. These results were obtained and published by Priestley in 1781; and in the meantime, a development had taken place, which Priestley himself had described, almost prophetically, four years earlier, when he wrote (in the preface to the first volume of Different Kinds of Ai7-) : "I do not think it at all degrading to the business of experimental philosophy to compare it, as I often do, to the diversion of hunting, where it sometimes happens that INGEN-HOUSZ 17 those who have beat the ground the most, and are consequently the best acquainted with it, weary themselves without starting any game; when it may fall in the way of a mere passenger, so that there is but little room for boasting in the most successful termination of the chase." The "mere passenger" 4. The Importance of Sunlight: Ingen-Housz in this case turned out to be the Dutchman, Ingen-Housz. Three years older than Priestley (he was born in 1730 in the Dutch town of Breda), he was a man of the world and until then had found httle time for experimental research. He practiced medicine, first in his home town in Holland, then in England and Austria, where he be- ^^t t Fig. 3. — Jan Ingen-Housz (from a bust bj- Seifert). came court physician to the Empress Maria Theresa, and was named Aulic counselor of the Empire, in recognition of his services in saving, by inoculation, the children of the Empress during an epidemic of small- pox. According to Ingen-Housz' own story, his interest in the chemical function of plants was aroused by the speech which the then President 18 DISCOVERY OF PHOTOSYNTHESIS CHAP. 2 of llie Royal Society, John Piingle, made on the occasion of the presenta- tion of tiic Cople.y medal to Priestley in 177;i In this speech, Pringle extolled the importance of Priestley's disco\'ery, which showed that the apparently useless or even poisonous plants "do not grow in vain," and that "salutary gales convey to the woods that flourish in the most remote and unpeopled regions, our vitiated air, for our relief, and for their nourishment." However, Ingen-Housz' plans for studying this mutual well-being of animals and plants, even if the.y were actually conceived in 1773, did not mature for seven years, and thus not until after the publication of Priestley's observations on the "green matter" which evolved purified air in sunlight. Ingen-Housz — the passenger — was spending the year, 1779, in England on a leave of absence from Vienna. In June of this year he retired to a "small villa" in the Enghsh countryside, and there, working "from morning till night," he performed, in less than three months, more than five hundred experiments. With amazing speed, the results of these experiments were ready for presentation to the public in October of the same year, in the form of a book entitled Experiments upon Vegetables, discovering Their Great Power of Purifying the Common Air in Sunshine and Injuring it in the Shade and at Night. The book was dedicated to John Pringle, whose presidential address seven years earlier had first aroused Ingen-Housz' interest in plant chemistry. Ingen-Housz explained the hurried publication by the necessity of returning to Vienna; but he also realized the importance of his discoverj'^ and was decided not to let anybody deprive him of it or even as much as share in it. The progress achieved by Ingen-Housz in these three months was indeed remarkable. Despite numerous errors which can be found in his book and which were caused partly by haste and partly by the ready belief of the author in his conjectures (so different from the conscientious- ness of a Priestley), Ingen-Housz has clearly established in this book the fundamental facts of photosynthesis. He wrote: "I observed that plants not only have a faculty to correct bad air in six or ten days, by growing in it, as the experiments of Dr. Priestley indicate, but that they perform this important office in a complete manner in a few hours; that this wonderful operation is by no means owing to the vegetation of the plant, but to the influence of the light of the sun upon the plant. I found that plants have, moreover, the most surprizing faculty of elaborating the air which they contain, and undoubtedly absorb continually from the common atmosphere, into real and fine dephlogisticated air; that they pour down con- tinually a shower of this depurated air, which . . . contributes to render the atmosphere more fit for animal life; that this operation . . . begins only after the sun has for some time made his appearance above the horizon . . .; that this operation of the plants is more or less brisk in proportion to the clearness of the day and the exposition of the plants; that plants shaded by high buildings, or growing under a dark shade of other plants, do not perform this office, but, on the contrary, throw out an air hurtful to animals; . . . that this operation of plants diminishes towards the close of the day, and SENEBIER 19 ceases entirely at sunset; that this office is not performed by the whole plant, but only by the leaves and the green stalks; that even the most poisonous plants perform this office in common with the mildest and most salutary; that the most part of leaves pour out the greatest quantity of this dephlogisticated air from their under surface . . .; that all plants contaminate the surrounding air by night; . . . that all flowers render the surrounding air highly noxious, equally by night and by day; that roots and fruits have the same deleterious quality at all times; . . . that the sun by itself has no power to mend the air without the concurrence of plants." As revealed by this summary, Ingen-Hoiisz' main achievements were the discovery of the importance of light for the "dephlogistication" of air by plants, and the proof that the plants improve the air not merely by absorbing its "mephitic" constituents (as suggested by Priestley), but that they also actively produce "vital air" (oxj^gen). He proved this fact by demonstrating the formation of oxygen bubbles by submerged leaves, a technique which was to be widely used later in the study of photo- sjmthesis. (Priestley had observed the oxygen bubble formation with his "green matter" before; but, at that time, he did not yet realize that the absorption of "bad air" by land plants and the liberation of "good air" by "green matter" were merely two aspects of one and the same phenomenon.) Ingen-Housz laid great emphasis, as shown by the title of his book, on his discovery of the "poisoning" of the air by plants in the dark, and he dwelt in detail on the gruesome dangers of keeping large trees in in- habited rooms, or of spending nights in a closed space containing large quantities of flowers, fruits or vegetables. This point became a subject of violent controversy between Ingen-Housz and Senebier, the latter denying any gas liberation by plants in darkness, and insisting that plants produce no dangerous "effluvia" when they grow in the open air. The fact of plant respiration was correctly observed by Ingen-Housz; but its dangerous aspects were obviously over-stressed, perhaps for the sake of philosophical satisfaction which he derived from the apposition of the wholesome influence of plants during the day and their poisonous activity during the night. The difference between inert gases (nitrogen and car- bon dioxide) and active poisons did not become clear to the chemists until much later. 5. The Part of Carbon Dioxide : Senebier Ingen-Housz' haste in asserting his priority proved well founded. Three years later, in 1782, Jean Senebier published, in his home town of Geneva, three volumes of Memoires physico-chimiques sur Vinfluence de la lumiere solaire pour modifier les etres des trois regnes de la nature et surtout ceux du regne vegetal. After acknowledging in the preface the priorit}^ and the importance of the work of Ingen-Housz, Senebier explained that he started his experiments before the appearance of the latter 's book, 20 DISCOVERY OF PHOTOSYNTHESIS CHAP. 2 being qualified for this work by his previous occupation with the effects of hght. He proceeded with the detailed description of his experiments and conclusions, without specifically acknowledging the similarity (or disagreement) between his results and those of Ingen-Housz. He sug- gested, not unreasonably, that the importance of the subject makes its study by two independent observers worth while. Ever after, he found himself exposed to the merciless irony and clever insinuations of Ingen- Housz, whose wrath would not be assuaged by the long-winded explana- tions of the Swiss pastor. The subsequent publications of both adver- FiG. 4. — Jean Senebier. saries, the second \olume of Ingen-Housz' French edition of Experiments on Vegetables (1789) and Senebier's Recherches siir Vinfluence de la lumiere solaire pour metamorphoser Vair fixe en air pur par vegetation (1783) and Experiences sur U action de la lumiere solaire dans la vegetation (1788), are filled with acid polemics, and make sad reading. Senebier served his cause badly by the extreme profuseness of his writing. In addition to the above-mentioned Memoires physico-chimiques (1782), Recherches (1783), and Experiences (1788), he published a Physio- logic vegetale in five volumes in 1800. The absence of adequate sum- maries, and the trite rhetorics embellishing the extensive descriptions SENEBIER 21 of his experiments, make these thirteen vohimes very tedious reading. No wonder the historians were not too kind to Senebier. Harvey-Gibson, in his Outlines of the History of Botany, denied him the recognition of the one important step which he made beyond the confirmation of the dis- coveries of Ingen-Hoiisz, and which was credited to him by Sachs and Pfeffer — the realization of the part plaj'ed by "fixed air" (carbon dioxide) in photosynthesis. The words "fixed air" do not occur at all in Ingen-Housz' first book (1779), and in the second one (1789) they appear almost exclusively in connection with the description of the products of plant respiration. He found the effect of large ciuantities of "fixed air" to be deleterious to the air-purifying activity of plants, and that of sniall quantities indefinite. If chemical equations had been known at that time, Ingen-Housz would have written the equation of photosynthesis in the following form: plants (2.1) Air + light > something "phlogisticated" in the plants + dephlogisticated air He had no definite conception as to what kind of "air" is used for this transformation. He even considered the possibilit.v of substituting, for "common air" in (2.1), "inflammable air" (hydrogen), or water vapor. Senebier, on the other hand, was aware even in his first work, pub- lished in 1782, of the accelerating effect of "fixed air" on the production of "pure air" by plants and wrote, "II paroit clairement, que Fair, fourni par les feuilles exposees sous I'eau au soleil, est I'effet d'une combination particuliere de I'air fixe, operee dans la feuille par le moyen du soleil," and in another place, " Je n'admets pas que Fair commun de I'atmosphere se tamise dans les feuilles des \'egetaux pour y deposer sa portion phlogis- ticiue, et en sortir air dephlogistique apres cette depuration." This last remark was directed against the hypothesis of Priestley, but apjilied ('([ually well to the ideas of Ingen-Housz. As alternative to the picture of a transformation of ordinary air into pure "dephlogisticated air," Senebier suggested "que Fair fixe, dissous dans I'eau, est la nourriture que les plantes tirent de I'air qui les baigne, et la source de I'air pur qu'elles fournissent par I'elaboration qu'elles lui font subir." The transformation of "fixed air" into "pure air" is the main subject of his second book (1783), as shown by its title, and in his Experiences (1788) a special chapter deals with the proof that "I'air rendu par les plantes exposees au soleil, est le produit de I'elaboration de I'air fixe par le moyen de la lumiere." Senebier shows convincingly, in polemics against Ingen-Housz, who even in 1784 denied that fixed air is necessary for the "purification" of air bj' plants, that no "dephlogisticated air" is formed from distilled water, even if it is saturated with common air, provided the latter is free from "fixed air." 22 DISCOVERY OF PHOTOSYNTHESIS CHAP. 2 The equation (2.1), by which we have summarized the discoveries of Ingen-Housz, was thus improved by Senebier in one point, and could now be written as follows: plant (2.2) Fixed air + light > something phlogisticated in the plant + dephlogisticated air 6. The Assimilation of Carbon Priestlej^ and Ingen-Housz assumed that the plants, in the process of transforming "impure air" into "purified air," acquire nourishment for themselves. (If they would have adhered strictly to the phlogiston theory, they should have concluded that this nourishment is pure phlo- giston.) Senebier, who discovered that the substrate of transformation is fixed air, could have gone a step further, and used Lavoisier's theory of the composition of fixed air to deduce that carbon is the acquisition which the plants make by absorbing carbon dioxide and exhaling oxygen. How- ever, ■ Senebier again showed himself no match for the quick-witted Ingen-Housz. In 1779. the Dutchman had reaped the seeds sown b}' Priestley's experiments, and left to the Swiss pastor the work of gleaning; in 1796, he was the first to harvest the fruits of Lavoisier's theory. He, who as late as 1784, denied that fixed air has anything to do with photo- sj-nthesis, wholeheartedly espoused this doctrine in 1796, in a new book Food of Plants and Renovation of the Soil. He hinted that he had believed in this doctrine since 1779, and left Senebier and his polemics with him about this point unmentioned. Nevertheless, this last book of Ingen- Housz is another remarkable achievement, and has been a milestone in the development of the science of plant nutrition. The mechanism of photosynthesis is presented in this book, for the first time, in terms of the new chemical theory, calling "fixed air" — carbonic acid, and "dephlogis- / ticated air" — oxygen, and proclaiming that plants acquire their carbon ' (whose important role in the composition of organic matter was well recog- nized by then) by the decomposition of carbonic acid from the air. Hales' doctrine of "aerial nourishment" of plants has thus received its first concrete interpretation. While Senebier thought that carbon dioxide from the air is first dissolved in soil water and reaches the leaves through the roots, Ingen-Housz suggested that plants receive only their "juices" from the soil, and obtain both their carbon, and their oxygen, from the air. He had the erroneous idea that while carl)()n is obtained from carbonic acid during the day, oxygen is derived from the same source during the night. (This is one example of the many "wild guesses" which can be found in the writings of Ingen-Housz.) One question which worried Ingen-Housz was the supply of carbon dioxide. While de Saussure thought that carbon dioxide is a permanent, although minor and variable component of the atmosphere, Lavoisier found no carbon dioxide at all DE SAUSSURE 23 in the common air. In his introduction to the German translation of Ingen-Housz' pamphlet, von Humboldt communicated analyses which purported to show the constant occurrence of carbon dioxide in common air, in concentrations from 0.7 to 1.4%. The true carbon dioxide content of the air — which is fairly constant at 0.03% in the open country, but may rise considerably above this value in the neighborhood of populated places, industrial establishments and active volcanoes — was established early in the next century by Dalton, de Saussure, and Boussingault (c/. the review by Letts and Black, 1900). After the appearance of Ingen-Housz' Food of Plants, the problem of oxygen liberation by illuminated plants became merged with the problem of carbon assimilation from the air and the synthesis of organic matter. The two terms under which this process is generally known, ''photosyn- thesis" and "assimilation" have their origin in these two aspects of the problem. Neither is entirely satisfactory, since "photosynthesis" could etymologically mean any synthesis under the influence of light, and "as- similation" (or even "assimilation of carbon") covers an even wider variety of phenomena. We shall use the term "photosynthesis" because it has practically acquired a very definite meaning and is only seldom ap- plied to any photochemical reaction other than the synthesis of organic matter by plants in light. In the light of Ingen-Housz' realization of the nutritive role of "air refining" by plants, we can again rewrite equation (2.2), in the form plant (2.3) Carbonic acid + light > organic matter + oxygen 7. The Participation of Water : de Saussure Quantitative treatment soon proved equation (2.3) to be incomplete. The recognition of this incompleteness came when weighing was added to volume measuring in the study of photosynthesis. This progress was due to another scholar from Geneva, Nicolas Theodore de Saussure (1767- 1845) (son of the physicist w^ho invented the hygrometer), a quiet and retiring man and a skillful and conscientious experimentalist. His results were published in 1804, under the title, Recherches chimiques sur la vegeta- tion. This is the first modern book on plant nutrition, full of careful analyses of gases, humus, and ash, and almost devoid of any speculations, or even hypotheses. The measurements of de Saussure proved definitely the correctness of Ingen-Housz' doctrine of aerial nutrition, and showed what elements the plants acquire from the soil. They confirmed the surmise of Senebier, that i)lants find enough nourishment in the small amount of carbon dioxide regularly present in the air, and showed that this is the only source of their carbon supply. De Saussure made the first comparison of the amounts of carbon dioxide absorbed and of oxygen 24 DISCOVERY OF PHOTOSYNTHESIS CHAP. 2 liberated by the plants; and most important of all, he proved that the increase in dry weight caused by the assimilation of a certain quantity of carbon dioxide, is considerably larger than the weight of carbon contained in it. Since the equivalent of all the oxygen contained in the decomposed carbon dioxide is evolved into the air, the large weight increase cannot be attributed to a coassimilation of oxygen from this source. The plants grew in pure water and air containing air carbon dioxide; therefore the only other possible source of weight increase was water (taken up in a form not removable by drying). De Saussure thought at that time that the assimilation of water is an independent process, merely coupled with the decomposition of carbon dioxide. However, the experimental results do not warrant such a separation of the two processes; what they prove is the participation of water in photosynthesis. They require an amplifica- tion of the over-all equation (2.3), which we may now write in the form: plant (2.4) Carbon dioxide + water + light > organic matter + oxygen Perhaps because water came into the picture several years later than carbon dioxide, the concept of photosynthesis as decomposition of carbon dioxide light (2.5) CO. > C + O2 followed by hydration of carbon (2.6) n C + m H2O > Cn(H20)„ has persisted in the literature for a long time. We shall see in chapter 3, that photosynthesis is better interpreted as a reaction between carbon dioxide and water — more exactly, as reduction of carbon dioxide by water (or oxidation of water by carbon dioxide). The study of photosynthesis, after the above described rapid start in the quarter century between 1779 and 1804, lapsed into an almost com- plete quiescence for the next fifty years. Liebig (1803-1873) in his fa- mous Chemistry in its Application to Agriculture and Physiology, severely criticized the methods used by plant physiologists of that time in dealing with the problems of material exchange between plants and the surround- ing media. Boussingault (1802-1887) was the first to improve these methods; to him is due the first redetermination of the gas exchange in l)hotosynthesis, with a precision better than that achieved bv de Saussure in 1804 (cf. C^hapter 3). 8. The Conversion of Light Energy: Robert Mayer If the time between 1804 and 1864 (the year when Boussingault pub- lished his first paper on photosynthesis) was sterile in the development of tlie physiology and chemistry of ])hotosynthesis, it saw the foundation of ROBERT MAYER 25 its physics laid b}' another surgeon, Julius Kobert Mayer of Heilbronn. In 1845, three years after he first enounced the law of conservation of energy, Mayer, in a pamphlet entitled The Organic Motion in its Relation Fig. 5. — Robert Mayer. to Metabolism dwelt upon the consequences of this law for life processes on earth, and w^rote: "Die Xatur hat sich die Aufgabe gestellt, das der Erde zustromende Licht im Fluge zu erliaschen und die beweglichste aller Kriifte, in starre Form umgewandelt, auf- zuspeichern. Zur Erreichung dieses Zweckes hat sie die Erdkruste mit Organisinen iiberzogen, welche lebend das SoiinenHcht in sich aufnehmen and unter Verwendung dieser Kraft eine fortlaufende Summe chemischer DiiTerenz erzeugen. Diese Organismen sind die Pflanzen: die Pflanzenwelt bildet ein Reservoir, in welchem die fliichtigen Sonnenstiahlen fixiert und zur Xutzniessung geschickt niederge- legt werden; eine okonomische Fiirsorge, an welche die physische Existenz des Men- schengeschlechtes unzertrennlich gekniipft ist. '•Die Pflanzen nehmen eine Kraft, das Licht, auf. und bringen eine Kraft hervor: die cheniisrhe Differenz." The last quotation means that our equation (2.4) can be amplified, to read plant (2.7) Carbon dioxide + water + light > organic matters- oxygen + chemical energy The work of Robert Mayer can be considered as the concluding chap- ter in the history of the discovery of photosynthesis. Qualitatively, 2G DISCOVERY OF PHOTOSYNTHESIS CHAP. 2 equation (2.7) is complete; its further elaboration could result only from detailed quantitative investigations, of the kind inaugurated by Bous- singault in 1864. The discovery of photosynthesis has engendered bitter rivalries, and historians have kept the flame of hate alive long after the principal actors in the drama have been silenced by death. The biographers of Priestlej' have felt that the discovery has been unfairly snatched away from this great experimentalist; the biographer of Ingen-Housz, Wiesner (1905), had no doubts that all honors belong to his hero, who "always kept his fights noble and magnanimous, even after Priestley had attempted base- lesslj^ to defame his reputation and Senebier had undertaken to put him- self into the possession of his discoveries by dishonest means. The most he has ever allowed himself has been gentle, nicely expressed irony." The friends and compatriots of Senebier, including de Saussure, and later N. Pringsheim (the latter mainly for reasons of his disagreement with everything Sachs stood for), were inclined to attribute to him, if not the whole, at least a large part of the discovery. We are now sufficiently remote from these controversies to be only mildly interested in the questions of priority and personal ambitions, and to recognize the discover}^ of photosynthesis as the inevitable conse- quence of the two great achievements of science in the period between 1770 and 1840 — the discovery of chemical elements, and the creation of the concept of energy. The work of the five men whose names are associ- ated with the foundations of photosynthesis, Priestley, Ingen-Housz, Senebier, de Saussure and Robert Mayer, has evolved from this back- ground, which two of them, Priestley and Mayer, themselves helped to create. After 1860, the development of plant physiology in general, and of photosynthesis in particular, took a rapid spurt, under the leadership of such men as Sachs, Pfeffer and Timiriazev. From this time on, the literature on photos,ynthesis has grown precipitously, initil now a complete review of all papers on this subject, scattered through botanical, agri- cultural, chemical and physical journals, appears almost impossible.* This broad development of the subject of photosynthesis after 1860, which has branched into manj'- different fields, including in addition to plant physiology also organic chemistrj-, physical chemistry and physics (not to speak of ecology and other branches of botany), makes it advisable to close here the historical introduction, and to deal henceforth with the different aspects of photosynthesis in a logical rather than chronological order. * A list of about 900 investigations published before 1925 can be found in W. Stiles' monograph, Photosj/nlhesis. The subject was treated in an even more comprehensive manner one year later, in the well-known monograph, of the same title, by H. A. Spoehr (1926j, which, however, does not contain a list of bibliographical references. BIBLIOGRAPHY TO CHAPTER 2 27 Bibliography to Chapter 2 Discovery of Photosynthesis (a) Original Works 1727 Hales, St., Statical Essays, Containing Vegetable Staticks, or, an Account of Some Statical Experiments on the Sap in Vegetation. W. Innys, London, 1727. 1772 Priestley, J., Phil. Trans. Roy. Soc. London, 62, 147. 1776 Priestley, J., Experiments and Observations on Different Kinds of Air. Vol. 1, J. Johnson, London, 1776. 1779 Priestley, J., Experiments and Observations Relating to Various Branches of Natural Philosophy. Vol. 1, J. Johnson, London, 1779. Ingen-Housz, J., Experiments upon Vegetables, Discovering their Great Power of Purifying the Common Air in Sunshine and Injuring It in the Shade and at Xighf. Elmsly & Payne, London, 1779. 1780 Ingen-Housz, J., Experiences sur vegetables etc. (French translation ))y the author, with additions). F. Didot le jeune, Paris, 1780. 1781 Priestley, J., Experiments atid Observations Relating to Various Branches of Xatural Philosophy. Vol. 2, J. Johnson, London, 1781. 1782 Senebier, J., Memoires physico-chimiques sur Vinfluence de la lumiere solaire pour modifier les etres de trois regnes, surtout ceux du regne vegetal. 3 vols., Chirol, Geneve, 1782. 1783 Senebier, J., Recherches sur Vinfluence de la lumiere solaire pour metamorphoser Voir fixe en air pur par vegetation. Chirol, Geneve, 1783. 1784 Ingen-Housz, J., Vermischte Schriften. 2 vols., Wapple, Wien, 1784. 1788 Senebier, J., Experiences sur Vaction de la huniere solaire dans la vegetation. Barde, Manget et Co., Geneve, 1788. 1789 Ingen-Housz, J., Experiences sur vegetables etc. Vol. 2, Barrois, Paris, 1789. 1796 Ingen-Housz, J., An Essay on the Food of Plants and the Renovation of Soils. (Appendix to 15th Chapter of the General Report from the Board of Agriculture) London, 1796; appeared as a separate book in German: Erndhrung der Pfianzen und Fruchtbarkeil des Bodens, with an introduction by F. A. von Humboldt, Leipzig, 1798. 1800 Senebier, J., Physiologic vegetate. 5 vols., Paschoud, Geneve, 1800. 1804 de Saussure, N. Th., Recherches chimiques sur la vegetation. Nyon, Paris, 1804. 1845 Mayer, J. R., Die Organische Bewegung in ihrem Zusammenhang mil dem Stoffwechsel. Heilbronn, 1845. Reprinted in Die Mechanik der Wdrme; gesammelte Schriften, Stuttgart, 1893, and in Ostwald's Klassiket der exacten X aturivissenschaften. No. 180, Akad. Verlags- gesellschaft; Leipzig, 1911. 28 DISCOVERY OF PHOTOSYNTHESIS CHAP. 2 (b) Historical Works and Reviews Sachs, J. v., Geschichte der Botanik, First German ed., Oldenbourg, Mlinchen, 1875, 1st English ed., Clarendon Press, London, 1890. Hansen, A., "Geschichte der Assimilation und Chlorophyllfunktion," Arbeit. Botan. Inst. Univ. Wiirzburg, 3, 123 (1884). Letts, E. A., and Black, R. F., "Carbonic Anhydride of the Atmos- phere," Sci. Proc. Roy. Dublin Soc, 9, II, 105 (1900). Green, J. R., A History of Botany in the Vnited Kingdom. Dent & Co., London, and Button, New York, 1914. Harvey-Gibson, R. J., Outlines of the History of Botany. A. & C. Black, London, 1919. about Priestley: Thorpe, T. E., Joseph Priestley. English Men of Science Series, Dent & Co., London, and Dutton, New York, 1906. Smith, E. F., Priestley in America, 1794-1804- Blakiston, Pliila- delphia, 1920. Holt, A., A Life of John Priestley. Oxford University Press, London, 1931. Aykroydt, W. R., Three Philosophers (Lavoisier, Priestley, Cavendish). Heinemann, London, 1935. about Ingen-Housz: Wiesner, J., Jan Ingen-Housz, sein Leben and sein Wirken als Natur- forscher und Arzt. Wien, 1905. PART ONE THE CHEMISTRY OF PHOTOSYNTHESIS AND RELATED PROCESSES Chapter 3 THE OVER-ALL REACTION AND THE PRODUCTS OF PHOTOSYNTHESIS In the first chapter, photosynthesis was characterized as the reversal of combustion, a process by which carbon dioxide and water are com- bined to form organic matter, Avhile oxygen escapes into the medium. This definition describes what may be called "normal photosynthesis" of the higher plants, mosses and algae. In recent years, some significant variations of this process have been discovered, which occur regularly in pigmented bacteria, and can be induced artificially in certain algae. The present chapter deals with the over-all reaction of normal photosynthesis, while chapters 5 and 6 will be devoted to its modifications in bacteria and algae. A. The Quantitative Balance of Photosynthesis * 1. The Photosynthetic Quotient The relative amounts of oxygen and carbon dioxide exchanged in photosynthesis were first determined by de Saussure in 1804. He found the volume of oxygen evolved, AO2, to be smaller (by as much as 30-40%) than the volume of carbon dioxide consumed by the plants, — ACO2. According to his analyses, the "missing" oxygen was converted into nitrogen. We cannot blame de Saussure for this error, for it was riot until sixty years later that the methods of quantitative plant physiology were sufficiently improved to allow a better determination of the "photo- synthetic quotient," Qpi (3.1) Qp = AO2/- ACO2 (The term "photosynthetic qutotient" has been used by many authors, for example, by Willstatter and Stoll, to designate the inverse ratio, — ACO2/AO2; this difference calls for care in the quotation of numer- ical results.) When Boussingault redetermined the photosynthetic quotient in 1864, he found Qp values varying between 0.8 and 1.2, with an average close to unity. In his measurements, the net production of oxygen was compared with the net consumption of carbon dioxide. However, even * Bibliography, page 56. 31 32 OVER-ALL REACTION OF PHOTOSYNTHESIS CHAP. 3 Ingen-Housz knew (or suspected) that plants continue to respire in light, so that their net gas exchange during the day is the l^alance of photo- synthesis and respiration. Tiie calculg-tion of true photosynthesis thus requires the application of a "respiration correction," which cannot be determined without certain arbitrary assumptions. Bonnier and Mangin (188G), who were the first to face this problem, used four methods for its solution. One method (which has since come into common use) was to determine the respiration in darkness, and to assume that its rate remains unchanged in light (c/. Chapter 20). The second method was based on the inhibition of photosynthesis by narcotics (which leaves the respiration almost unchanged, cf. Chapter 12); the third on the pre- vention of photosynthesis by deprivation of carbon dioxide, and the fourth on the comparison of gas exchange in leaves with high and low content of chlorophyll. By all four methods, Bonnier and Mangin ob- tained Qp values considerably above unity (1.1 to 1.3). These results were not confirmed by subseciuent investigators, notably Maquenne and Demoussy (1913) and Willstiitter and StoU (1918), who found that Q? is equal to unity within the limits of experimental error. Maquenne and Demoussy determined the respiration correction by experiments in the dark, while Willstatter and Stoll reduced it to insignificance by working in very strong light and with ample supply of carbon dioxide, so that photosynthesis was twenty or thirty times stronger than respiration. Table 3.1 gives a selection of their results together with those of some recent investigations, in which a different type of plant (lower algae) has been used instead of the higher land plants. Table 3.1 shows the remarkable constancy of the photosynthetic quotient — it is independent of light intensity, duration of illumination, temperature, and the concentrations of oxygen and carbon dioxide. Values slightly above 1 seem to predominate, although deviations from unity are hardly beyond the limits of experimental error. Table 3.1 shows also that the respiratory quotient: (3.2) Qr = ACO2/- aO.> is close to unity for most plants (although its deviations from the nor- mal value are more common than are those of the quotient Qp). Very few significant cases of abnormal Qp values have been found. Before discussing them, we may first inquire what the normal value, Qp = 1, means for the chemical mechanism of photosynthesis. It finds a natural explanation in the assumption that the product of photosynthesis is a carbohydrate, i. e.; a compound with the atomic ratio H : O = 2 : 1. We can thus elaborate equation (2.4), by writing: light (3.3) X CO., + y H2O > Cx(H..O)y + a; O2 plant THE PHOTOSYNTHETIC QUOTIENT 33 Table 3.1 The Photosynthetic Quotient (Qp = AO2/— ACO2) and the Respiratory Quotient (Qr = ACO2/- AO2) Species Ailanthus Aspidistra Begonia Cherry laurel Kidney bean (young) Pea Ricinus Sorrel Wheat Average (27 spp.) : Sambucus nigra Pelargonium zonale Sambucus nigra Aesculus hippocastanum Ilex aquifolium Leucobryum glaucum Hormidium flaccidum Chlorella pyrenoidosa Nitzschia closterium Nitzschia palea Observer" M.D. M.D. M.D. M.D. M.D. M.D. M.D. M.D. M.D. M.D. W.St. W.St. W.St. W.St. W.St. W.St. W.St. W.St. v.d.P. M.S.D.D, B. Remarks B. B. B. 25° C, 5% CO2, illumi- nated for 5 hrs., 45,000 lux Same after 12 hrs. in dark 25° C; 2 hrs., 45,000 lux, low O2 (2%) Same after 3.5 hrs. in dark 85° C, 45,000 lux, 6.5% CO2, 4.5 hrs. 10° C, 45,000 lux, 6.5% CO2 Leathery leaves, Qp > 1 according to Bonnier and Mangin Moss, 25° C, 5% CO2, 22,000 lux Alga (green) Green alga, low light Diatom, high light, 12-31° C. Same, low light, 12-28° C. Diatom, high light Same, low light Qp 1.02 1.00 1.03 0.97 1.12 1.00 1.03 1.04 1.02 1.04 1.00-1.05 0.98-1.02 1.01-1.02 1.01 1.00-1.01 0.99-1.02 1.00 1.01 0.92-1.07 0.98* 1.04 ±0.03 1.07 ±0.02 1.03 ±0.03 1.05 ±0.02 Qr 1.08 0.94 1.11 1.03 1.12 1.07 1.03 1.04 1.03 1.05 0.93 ±0.04 0.79 ±0.03 " M.D. = Maquenne and Demoussy (1913); W.St. = Willstatter and Stoll (191S); v.d.P. = van der Paauw (1932); B. = Barker (1935); M.S.D.D. = Manning, Stauffer, Duggar and Daniels (1938). 6 Average of 18 single determinations of quantum yields of photosynthesis by simultaneous measure- ments of ACO2 and AO2; single values varying from 0.3 to 1.5. In the same year, 1864, in which the correct value of the photosyn- thetic quotient was first determined by Boussingault, a direct proof of the formation of carbohydrates by photosynthesis was given by Sachs, who observed the growth of starch grains in the chloroplasts of photosyn- thesizing leaves. However, not all plants form starch after intense 34 OVER-ALL REACTION OF PHOTOSYNTHESIS CHAP. 3 photosynthesis — some do not form any visible deposits of reserve ma- terials at all, while others store oils or proteins. Since the latter products are more strongly reduced than the carbohydrates, their formation by photosynthesis would require more hydrogen and lead to the liberation of more oxygen, than does the synthesis of carbohydrates. This would mean an increased value of the photosynthetic quotient. It is therefore important that Barker (1935) found a normal value of the photosynthetic quotient also in oil-storing diatoms (c/. Table 3.1), even in those which have an abnormally low value of the respiratory quotient {Nitzschia palea). This difference illustrates the fact, already stated in the first chapter, that photosynthesis is a universal process, taking the same course in all plants, while the reverse process of oxidation can proceed by many different paths. Whenever compounds other than polymerized carbohydrates (starch, inulin, cellulose) are stored in an organism, the symmetry of photosynthesis and respiration must needs be disturbed. Thus, plants (or animals feeding on carbohydrates) which accumulate fats, may have Qr values above unity while the fats are formed, and below unity while they are consumed. Conversely, plants which store low-molecular organic acids or salts, (e. g., oxalates, tartrates or citrates), often have Qr values below unity during the deposition, and above unity during the consumption of these reserve materials. This is true, for example, of ripening fruits, and of succulents during the nightly accu- mulation of acids (c/. page 264). Succulents (e. g., Cacti) often have also abnormally large photosynthetic quotients (Aubert 1892); or, at least, they appear to be large if determined by the usual method of subtracting the gas exchange in the dark from the gas exchange in light. The quotient decreases, however, with prolonged illumination, as shown by Willstatter and Stoll (1918) in experiments with Opuntia. The change is associated with the gradual disappearance of organic acids, which these plants accumulate in darkness. This " deacidification " in light can be interpreted either as a photoxidation, producing free carbon dioxide, or as a photor eduction, converting the acids into carbohydrates. If the first hypothesis is correct, the high photosynthetic quotient is deceptive: the complete oxidation of acids of the type of malic acid produces more carbon dioxide than it consumes oxygen, and thus reduces the net carbon dioxide consumption from the atmosphere more than the net liberation of oxygen. If, however, the second hypothesis is correct, the high photosynthetic quotients are real, and the succulents carry out a "photosynthetic assimilation of organic acids" instead of, or in addition to, the usual photosynthetic assimilation of carbon dioxide. We will return to this problem in chapter 10. Abnormal photosynthetic quotients are sometimes shown also by nonsucculent plants at the start of illumination after a period of darkness. THE YIELD OF ORGANIC MATTER 35 This was first noticed by Kostychev in 1921. During this so-called in- duction period, which, depending on specific conditions, can last for seconds, minutes or even hours, the photosynthetic quotient may be larger or smaller than unity; it may even become negative, that is, plants may consume (or liberate) both oxygen and carbon dioxide at the same time. These phenomena must be ascribed to the restoration of enzymatic systems and the regeneration of intermediate products, which have been destroyed during the dark interval (c/. Vol. II, Chapter 33). Considering all known deviations of the photosynthetic quotient from unity, we see no reason to admit the steady photochemical production of compounds other than carbohydrates. However, the value Qp = 1 does not preclude the formation of organic acids or other "underreduced" compounds as intermediates in photosynthesis. Willstatter and StoU quoted the constancy of Qp as an argument against Liebig's theory of photosynthesis, in which plant acids were supposed to accumulate by photosynthesis in summer, and to be slowly transformed into carbo- hydrates in fall. However, this objection does not avail against those modifications of Liebig's theory which assume that the plant acids are only passing intermediates in the transformation of carbon dioxide into carbohydrates. The observed value of Qp proves that no underreduced (or overreduced) intermediate products accumulate during steady photo- synthesis; but as long as these intermediates are consumed at the same rate as they are produced, their presence cannot affect the photosynthetic quotient (except during the induction period). 2. The Yield of Organic Matter The photosynthetic quotient is the most easily measurable quanti- tative characteristic of photosynthesis. However, it is not suflacient to give a complete picture of the chemical reaction. It does not reveal the absolute value of x in eq. (3.3), or that of the ratio x : y. Thus, it does not allow one to identify the product of photosynthesis as a simple sugar {x = y) or as a polymer {y < x, e. g., y = 5/6 x for high polymers of hexoses). Furthermore, the photosynthetic quotient is not a very sensitive criterion of the exclusive production of carbohydrates. A deviation of Qp by 3% from unity — which is well within the limits of error of most experiments — may mean the formation of as much as 12% of protein (Smith, 1943), or 5% of fats. This makes it important to use other and more direct methods for the determination of the chemical nature of the "photosynthate." Interesting information could be pro- vided by the determination of the amount of water assimilated together with a known quantity of carbon dioxide, but this experiment encounters considerable difficulties, because of the abundant presence of water in all cells. The determination of the total increase in organic matter, caused by 36 OVER-ALL REACTION OF PHOTOSYNTHESIS CHAP. 3 the assimilation of a certain quantity of carbon dioxide, is easier; we remember that de Saussure used this method in 1804 to prove the par- ticipation of water in photosynthesis. Quahtatively, the proof was suc- cessful; but quantitatively, the two experiments performed by de Saussure disagreed. In the first of them, seven Vinca plants assimilated 314 mg. water together with 217 mg. carbon, corresponding to a molecular ratio X : y = 1.03; in the second, two Mentha plants assimilated 159 mg. water together with 159 mg. carbon, corresponding to x : y = 1.50. Similar experiments were carried out almost one hundred years later by Krasheninnikov (1901), who determined the total increase in the dry matter of illuminated detached leaves of five species, and the amount of absorbed carbon dioxide, and obtained x : y ratios between 0.87 and 1.23. Bose (1924) compared the increase in dry weight and the oxygen pro- duction of several plants of Hydrilla, and obtained x : y ratios of 0.92 or less. Smith (1943) made careful determinations of the carbon dioxide consumption and dry matter production by sunflower leaves, after illu- mination periods of the order of 1-3 hours. Table 3. II shows some typical results. Table 3.II Carbon Assimilation and Increase in Dry Weight of Sunflower Leaves (after Smith) Expt. No. AC, mg. carbon assimilated AW, mg. (increase in dry weight) AC/ATF Calcd. from ACO2 Detd. by analysis From AGO 2 By analysb 1 2 3 4 5 6 5.3 7.6 7.3 6.8 6.0 6.2 5.2 8.9 8.0 6.5 6.2 5.6 12.9 20.6 17.2 17.6 14.1 14.0 0.41 0.37 0.43 0.39 0.43 0.44 0.41 0.43 0.47 0.37 0.44 0.40 Average : 0.41 0.42 The average ratio AC/AW (0.415), corresponds to x : y = 1.06 and thus agrees well with the theoretical ratio for a simple sugar (0.40, X : y = 1.00) and even better with that for a disaccharide (0.42, X : y ^ 1.09). In the leaf as a whole, the proportion of carbon is con- siderably larger than in the newly formed " photosynthate " (about 0.51 if referred to dry weight without ash). The most satisfactory method of determining the nature of the products of photosynthesis is direct analysis. However, when Sapozh- nikov (1890) first determined the difference between the carbohydrate THE YIELD OF ORGANIC MATTER 37 content of sunflower leaves kept in the dark, and that of similar leaves which had been exposed to light for 3-5 hours, and compared the amounts of synthesized carbohydrates with the quantities of carbon dioxide con- sumed by the illuminated leaves, he found "carbohydrate deficiencies" of 5-35%. Similarly, Krasheninnikov (1901) was able to identify as carbohydrates only 50-75% of the dry matter synthesized by the leaves of bamboo, cherry laurel, sugar cane, linden and tobacco. These results could be taken as indications of a rapid transformation of the primary product of photosynthesis into compounds other than carbo- hydrates; this conclusion was supported by the observation of Ruben, Hassid and Kamen (1939) who found that, after one hour of photo- synthesis by barley leaves in radioactive carbon dioxide, only 25% of the assimilated radioactive carbon could be recovered in water-soluble carbohydrates, and not more than 10% in insoluble material (cellulose). Smith (1943), on the other hand, was able to recover, in the form of carbohydrates, practically all carbon assimilated by sunflower leaves in illumination periods of 1-3 hours (c/. Table 3. III). According to this Table 3.III Determination of Carbohydrates in the Photosynthate OF Sunflower Leaves (after Smith) Number of experi- ments Temp., °C. Illumi- nation dura- tion, min. Percentage of assimilated carbon found in Mono- saccha- rides Sucrose Non- identi- fied sugars Starch All soluble carbo- hydrates In- soluble carbo- hydrates" All carbo- hydrates 4 7 10 20 156 58 7 10 71 52 5 3 16 26 99 91 8 7 107 (±5) 98 (±3) " Assumed to have the elementary composition of cellulose. table, more sucrose and less monosaccharides (and less starch) are ob- tained at 10° C. than at 20°. The proportion of monosaccharides may rise to as much as 35% if several hours of respiration in the dark are allowed to pass between illumination and analysis. (This amylolysis in the dark may be caused by water deficiency, c/. page 333). The difference between the observations of Smith and those of the earlier investigators may be attributed to improved methods of analysis. However, the results may also depend on the plant species used and on the conditions of the experiment (duration and intensity of illumination, temperature, etc.). An explanation remains to be found for the failure of Ruben and coworkers to recover more than one quarter of assimilated radioactive carbon in the carbohydrates photosynthesized by barley leaves. 38 OVER-ALL REACTION OF PHOTOSYNTHESIS CHAP. 3 B. The Products of Photosynthesis* 1. The Carbohydrates Experiments described in the preceding section indicate that the direct products of photosynthesis belong to the class of carbohydrates. How- ever, by the time when the quantity of the photosynthate becomes suf- ficient for chemical analysis, the carbohydrate fraction is found to contain a variety of compounds of different degree of polymerization, and it is unlikely that they all are the primary products of photosynthesis. Which of them, if any, is the primary product, is a moot question. Before presenting the arguments advanced on behalf of different con- tenders for this distinction, it seems advisable to give a short review of the structure and properties of the most common plant carbohydrates — pentoses, hexoses and their various polymers. For a detailed presenta- tion of sugar chemistry, the reader is referred to the monographs by Pringsheim (1932), Bernhauer (1933), Armstrong and Armstrong (1934) and Micheel (1939). (a) Pentoses These Cs sugars occur abundantly in many plants, usually not in the free soluble form, but as the so-called pentosans, i. e., anhydrous com- pounds of the composition (C6H804)i. The pentosans are mostly found in the supporting structures — cell walls, wood fibers, etc., and are thus not directly associated with photosynthesis. However, they are more easily hydrolyzed than cellulose, and sometimes serve as reserve ma- terials, thus coming into closer relation with nutritional processes. Some authors, Nef (1910), for example, thought that the synthesis of pentoses must be independent of that of hexoses; others, as Lob and Pulvermacher (1910), suggested that pentoses are intermediates in the formation of hexoses. The production of starch from externally supplied pentoses (c/. page 260) indicates that plants contain enzymes capable of bringing about the conversion of pentoses into hexoses. On the other hand, it is known that pentoses can be produced by degradation of hexoses (by the intermediary of hexuronic acids) . According to Spoehr, Smith, Strain and Milner (1940), albino maize plants provided with su- crose as the only source of carbon, produce uronic acids and pentosans from this food. This supports the opinion of those authors, including Ravenna (1911) and Tollens (1914), who believed that pentoses in plants are secondary products of transformation of the hexoses. The pentosans deposited in the cell walls and wood fibers must be produced there from products transported by the sap, which usually contains only glucose, fructose and sucrose (and no free pentoses). * Bibliography, page 57. THE CARBOHYDRATES 39 (6) Hezoses (Monosaccharides) Among the compounds of this group, glucose and fructose are most widely distributed in plants. They are found in the sap of practically all leaves (as first proved by Brown and Morris in 1893) in quantities which depend on species as well as on the previous treatment of the plant. Starvation may reduce their concentration to zero, while intense photo- synthesis may raise it to 10 or 15% of the dry weight of the leaf. Free fructose is sometimes more abundant than free glucose; Brown and Morris (1893), for example, found in Tropaeolum majus leaves four times more fructose than glucose, and Gast (1917) found up to eight times more in leaves of five different species. Equal quantities of glucose and fructose are contained in cane sugar (sucrose), which is present in all green leaves, while the most common highly polymeric carbohydrates, starch and cellulose, are built entirely of glucose units, thus making the latter the most abundant single organic compound on earth. It is often forgotten that the photosynthesis by higher plants is far inferior, in its yield, to the photosynthesis by the microscopic organisms of the plankton. The tendency to extend to the whole plant world con- clusions reached in the study of the highly developed land plants, is not without its danger. It is therefore important to note that hexoses have been found also in many algae, although no systematic information about their distribution in those organisms is as yet available. Hexoses other than glucose and fructose are rare in green plants. Clements (1932) was unable to find mannose in leaves of 42 species. Galactose is only encountered in the esterified form, as galactosides, or in condensation products with other sugars; while sorbose was definitely identified only in fruit juices. Glucose (and other aldohexoses) act chemically as mixtures of three or four different tautomeric forms. A B C D 1 CHO CHOH CHOH-n CHOH — 2 3 4 5 6 CHOH I CHOH I CHOH I CHOH CH2OH Open-chain aldehyde (Baeyer 1870) COH CHOH i CHOH I CHOH I CH2OH Enol form (Fischer 1895) CHOH CHOH I CHOH I CH I CH2OH 1,5-Glucopyranose (a- and j3-glucose) (Haworth 1926) CHOH I CHOH O CH CHOH CH2OH 1 ,4-Glucof uranose (7-glucose) (ToUens 1883) Formulae 3.1. Tautomers of Glucose The most stable structure is the six-membered ring C. These formulae describe all aldohexoses; the particular spacial arrangements characterizing the a- and ^-glucose 40 OVER-ALL REACTION OF PHOTOSYNTHESIS CHAP. 3 are best brought out by Haworth's prospective formulae: Alpha OH OH Befa OH H l/H H0\ OH H OH H CHjOH H CHgOH Formulae S.II. a- and fi-Glucose (a- and /3-glucopyranoses) In ^-glucose, all substituents are in trans-positions, which probably gives the lowest energy and highest stability. Fructose is a 2-ketohexose, with the five possible structures: A CH2OH I CHOH I CHOH 5 6 CHOH CH2OH Open-chain ketone CHOH COH I CHOH CHOH I CHOH I CH2OH 1,2-Enol form CH2OH COH II COH I CHOH C CH2OH I COH — D CH2OH I CH CHOH CH2OH 2,3-Enol form CHOH I CHOH O I CHOH I CHOH- 2,6-Fructo- pyranose CHOH I CHOH COH — O CH2OH 2,5-Fructo- furanose Formulae 3. III. Tautomers of Fructose Of the two enols, Bi is identical with the enol of glucose; this relation is responsible for the slow interconversion of glucose and fructose in alkaline solutions. The di- phosphates of these two sugars also are identical, and this must be important for their interconversion in living plants. In polymerization, glucose usually acts in the pyranoid form, whereas fructose more often enters into polymers in the form of a five-membered furanoid ring. All hexoses in plants are optically active and belong to the c?-series. This shows that asymmetric synthesis takes place in the course of the reduction of carbon dioxide, probably through the intervention of an asymmetric enzyme. Mention must be made also of inositols, carbocyclic compounds which are isomeric with hexoses, (cf. Formula 3. IV), taste sweet and are included in the general classification of sugars under the name of "cycloses." Inositols are widely distributed in plants (particularly in seeds), and have been identified in leaves, e. g., by H. Miiller (1907), Tanret (1907) and Curtius and Franzen (1916) in quantities of about 0.05% of dry weight. Interesting is the phytin, a calcium-magnesium salt of inositol phosphate, C6H6-[OPO(OH)2]6, which was found in green leaves by Curtius and Franzen. THE CARBOHYDRATES 41 H O Hi HOCH2 HOCH2 H2COH H2COH H2 O H Formula S.IV. Inositol (c) Disaccharides Glucose and fructose are the constituents of the most common dimeric and polymeric sugars. The relationships between these compounds are represented by the following scheme: Monosaccharides : Disaccharides: Maltose Sucrose Polysaccharides: Glucose- i Maltose i Starch, Cellulose 'Fructose Inulin Sucrose is the most common disaccharide in green leaves; its concen- tration often exceeds that of the free hexoses (c/. Table 3. III). Sucrose is a product of condensation of one molecule of a-glucose in the pyranoid form and one molecule of fructose in the furanoid form: (3.4) a-Glucopyranose + fructofuranose -> sucrose + water CHjOH — f\ y A H u h OH H H OH OH H Formula 3.V. Siicrose Since starch occurs regularly in a large proportion of leaves, one would also expect to find maltose which is an intermediate between starch and glucose. The molecule of maltose contains two glucopyranose units bound by an oxygen "bridge." Brown and Morris (1893) found 0.7-5.3% and Gast (1917) up to 1% maltose in Tropaeolum leaves; but Davis, Daish and Sawyer (1916) identified it as a product of hydrolysis of starch during the slow drying of the leaves. If leaves are killed rapidly, no maltose is found in them. Davis (1916) and Daish (1916) showed that leaves contain maltase, which hydrolyzes maltose to glucose, and attributed to this fact the absence of maltose in hving leaves. However, the presence of diastase, which hydrolyzes starch, (c/. Brown 42 OVER-ALL REACTION OF PHOTOSYNTHESIS CHAP. 3 and Morris 1893, Sjoberg 1922) and of invertase, which splits sucrose into fructose and glucose (Robertson, Irvin and Dobson 1909), does not prevent leaves from containing large quantities of these polysaccharides. The relative quantities of different monosaccharides and polysaccharides in living plants must be determined by the rates of their formation and decomposition, which depend on the available quantities of different enzymes and the distribution of the latter in the tissue. (d) Polysaccharides Among the polysaccharides and their derivatives which occur in very large quantities in plants, some (for example the cellulose of the higher plants, and the alginic acid of algae) are too inert or too far removed from the site of photosynthetic activity, to be suspected of a direct relationship to photosynthesis. Starch is the only polymeric carbohydrate whose association with photosynthesis is evident. The occurrence of starch grains in chloroplasts, i. e., plant organs primarily concerned with photo- synthesis, has been known since 1837, when von Mohl first observed the chloroplasts under the microscope. Boehm (1856) confirmed that these grains consist of starch, by the well-known iodine color test. Gris (1857) noticed their growth during the day and dissolution during the night, and Sachs (1862, 1863, 1864) first postulated their direct associ- ation with photosynthesis. In a famous experiment, Sachs exposed one- half of a leaf to the sun and kept the other covered, and showed that after some time only the exposed half gave the color reaction with iodine. Pfeffer (1873) and Godlewski (1877) completed the proof by showing that no starch is formed by leaves illuminated in absence of carbon dioxide. As mentioned above, starch is a high polymeric form of glucose; it contains, in the native state, a small proportion of phosphoric acid (about 0.1% P2O5) which is probably important for its enzymatic transforma- tions. As in maltose, the a-glucose molecules in starch are bound to- gether by oxygen bridges : -o H OH Formula S.VI. Starch H O— OH A similar chain of jS-glucose molecules forms the basis of the structure of cellulose. Most of our knowledge of starch has been derived from the study of storage materials in seeds, tubers and roots; recently the preparation and FORMATION OF OILS AND PROTEINS 43 properties of leaf starch have been described by Spoehr and Milner (1935, 1936). Grafe and Vouk (1912, 1913) and Melchior (1924) found that in some plants inulin replaces starch. Inulin is a polymer oi fructose, constructed from fructofuranose units, in the same way that starch is built up from glucopyranose units. Varied reserve materials are encountered in algae. Starch is found in green and red algae {Chlorophyceae and Rhodophyceae) ; and glycogen (a form of starch common in animals) in blue-green algae (Cyanophyceae). Brown algae (Phaeophyceae) store pentosans and fucosans. 2. The Photosynthetic Formation of Oils and Proteins All storage materials mentioned so far were carbohydrates. How- ever, many algae, particularly the diatoms {Bacillariophyceae), but also some green algae (Vaucheria), store oils or fats instead of carbohydrates (Beijerinck 1904). In addition, the chromoplasts of most algae contain so-called pyrenoids, peculiarly shaped bodies (c/. page 357), usually con- sidered as masses of reserve proteins surrounded by starch sheaths. It has sometimes been suggested that these reserve materials may represent the direct products of photosynthesis in algae. Bond (1932) thought for example that diatoms may produce fats directly by photo- synthesis, according to the equation: light (3.5) 55 CO2 + 52 H2O > C66Hio40« + 78 O2 However, we have seen (page 34) that the photosynthetic quotient of an oil-storing diatom was found to be not larger than 1.05, while equatoin (3.5) would require a value of 1.42. Thus, the oil deposits of the diatoms — and probably the fat and protein stores of other algae as well — must be considered as products of comparatively slow secondary transforma- tions not directly associated with photosynthesis. Oily drops have been observed not only in algae, but also in the leaves of some higher plants. Briosi (1873) suggested that these drops are produced directly by photosynthesis; however, his conclusions were criticized by Holle (1877) and Godlewski (1877). Meyer (1917) observed, in illuminated leaves of Tropaeolum majus the temporary formation of what he described as ''droplets of an assimi- latory secretion." Later (1918) he suggested that oil drops formed in some green algae (e. g., Vaucheria terrestra) are of a similar nature, and proposed that this " assimilatory secretion" be considered as an immedi- ate product of photosynthesis. From observations of its chemical behavior (1917^), he concluded that it is not a fat, but may possibly consist of hexenaldehyde, a compound whose presence in green leaves 44 OVER-ALL REACTION OF PHOTOSYNTHESIS CHAP, 3 was at that time investigated by Curtius and Franzen (c/. page 252). Hexenaldehyde has the formula CeHioO, and its formation by photo- synthesis should lead to a photosynthetic quotient of 1.33. In addition to this high photosynthetic quotient, not confirmed by experiments, two other arguments speak against Meyer's interpretation. In the first place, the quantity of hexenic aldehyde, found by Curtius and Franzen, is much too small to account for the large volume of Meyer's "photo- synthetic secretion." In the second place, even this small quantity has recently been proved to be of a secondary origin, being apparently formed during the steam distillation of the leaf material (page 254). It seems probable that Meyer's ''oil drops" were the so-called "grana," whose occurrence in chloroplasts was first postulated by Meyer himself in 1883 and recently confirmed by many other observers (c/. Chapter 14). 3. The First Products of Photosynthesis and Their Transformations Having satisfied ourselves that no direct photochemical formation of fats or proteins needs to be postulated on the basis of available experi- mental material, we may now return to the problem of the "first carbo- hydrate," mentioned on page 38. This role has variously been claimed for glucose, sucrose, inositol and starch, usually on the basis of experi- ments on the absolute and relative concentrations of these carbohydrates in plants at different times of the day and season of the year. Both the concentration of soluble sugars in the cell sap and the quantity of solid starch in the chloroplasts, undergo wide variations with the intensity of photosynthesis. They may be reduced to zero by starvation, and can rise to 20 or 30% of the total dry weight after a period of intense photo- synthesis, particularly if translocation is interrupted, as in detached leaves. Brown and Morris (1893) found, in the leaves of Tropaeolum majus attached to the plant, 9.7% sugars and 1.2% starch at 5 a.m. and 9.6% sugars and 4.6% starch at 5 p.m.; but if the leaves were detached at 5 A.M. and left in sunlight until 5 p.m., the concentration of sugars increased to 17.2%, while that of starch was 3.9%. Numerous authors have determined the relative quantities of glucose, fructose, and sucrose in leaves, and the changes in these ratios caused by starvation and illumination; and several of them, e. g. Perrey (1882), Brown and Morris (1893), Parkin (1911), Mason (1916), Davis, Daish and Sawyer (1916), Davis and Sawyer (1916), Gast (1917), and Venezia (1938) have arrived at the conclusion that the disaccharide sucrose pre- cedes the monosaccharides in the order of synthesis. They based this conclusion either on the more widespread occurrence and larger absolute quantity of sucrose in leaves, or on the observation that the concentration of sucrose follows more closely the diurnal cycle of photosynthesis. FIRST PRODUCTS OF PHOTOSYNTHESIS 45 However, the primary formation of a disaccharide seems implausible a priori, and authors who argued in its favor, have neglected the rapidity with which the primary product of photosynthesis may undergo enzymatic isomerizations and polymerizations in leaves which are equipped with invertase, diastase, maltase and other carbohydrate-transforming en- zymes. Priestley (1924), Stiles (1925), Spoehr (1926), and Barton-Right and Pratt (1930) stressed the fact that the way in which the leaves are killed (by freezing, drying, boiling, or immersion into alcohol) affects the analytical results, thus proving that extensive enzymatic transformations can take place even during the preparation of the material. Dixon and Mason (1916), Priestley (1924) and Spoehr (1926) pointed out that a mechanism for rapid enzymatic conversion of primary products (e. g., hexoses) into storage materials (and sucrose may be a soluble storage material) can keep the concentration of the primary products approxi- mately constant, while that of the storage materials rises and falls with the intensity of photosynthesis. Contrary to the experimental results of the above-mentioned investigators, others — notably Weevers (1924), Tottingham, Lepkovsky, Schulz and Link (1926), Clements (1930), Barton-Right and Pratt (1930) and Kretovich (1935)— have obtained analytical evidence favoring the conclusion that the monosaccharides precede the more complex sugars in organic synthesis. Weevers (1924), for example, found both glucose and sucrose in the green {i. e., photo- synthetically active) spots of variegated leaves, and only sucrose in the yellow spots. The same author observed that when a leaf of Pelargonium was deprived of all its sugars by starvation for 48 hours, the first sugar to appear upon illumination was glucose, which was only later followed by sucrose and starch. Clements (1930) and Barton-Right and Pratt (1930) found by hourly analyses extending from sunrise to sunset, that glucose predominates in leaves early in the morning, while sucrose begins to accumulate (and often surpasses glucose in concentration) later in the day. These experiments support the plausible assumption that disac- charides are secondary products formed by condensation of simple hexoses. On the other hand. Smith (1944) has again found, in extending to several hours the duration of his experiments on the fate of carbon assimilated in sunflower leaves (c/. page 37), that sucrose (and starch) are formed immediately upon the beginning of illumination, while the relative quantity of monosaccharides is at first very small, and increases with time (e. g., from 4% of total carbohydrates after 27 minutes of illumination to 22% after 146 minutes). He concluded that the primary product of photosynthesis is a common precursor of sucrose and starch (perhaps a hexose monophosphate), and suggested that the free mono- saccharides found in the cell sap are secondary products, formed by the hydrolysis of sucrose. 46 OVER-ALL REACTION OF PHOTOSYNTHESIS CHAP. 3 There does not seem to be any basis for arguing whether fructose and glucose are independent products, or whether one is a "primary" and the other a "secondary" sugar. Some leaves contain more free glucose, others more free fructose; while glucose usually predominates in the polymeric carbohydrates. Endo (1936) found only glucose in some green algae {C odium latum) , and only fructose in others {Cladophora Wrightiana) . In vitro, glucose, fructose and mannose are interconvertible in alkahne solutions (the so- called Lobry de Bruin-van Eckenstein reaction). This conversion, which is supposed to proceed through the intermediary of the enols Bi and B2 (Formula 3.III), inevitably produces a certain proportion of mannose. The leaf cells are not alkahne, but neutral or acid; and the leaves apparently contain no mannose (page 39). These facts have been used as arguments against the glucose-fructose interconversion in the leaves. However, Spoehr and Strain (1929) showed that in presence of sodium phosphate the interconversion can be achieved, in vitro, also in neutral or even shghtly acid solution. Fructose, kept at 37° C. in shghtly acid Na2HP04 solution (pH 6.7) was found to contain, after 165 hours, 8.5% mannose and 28% glucose. An enzyme (isomerase) is known which converts glucose monophosphate into fructose monophosphate; while the diposphates of these two sugars are identical. Thus, the interconversion in Uving plants probably occurs by a combination of phosphatiza- tion with the action of specific enzymes. In addition to glucose and fructose, the distinction of being the first Ce products of photosynthesis was claimed also for the inositols. Crato (1892) and Kogel (1919) suggested that these cyclic compounds, con- sisting entirely of HCOH groups, are the parent substances of all other sugars in plants. Gardner (1943) suggested that the first carbohydrate formed by photosynthesis may be the triose, glycer aldehyde; but the only basis for this hypothesis was that trioses are the last carbohydrates which occur in respiration (c/. Chapter 9, page 223). If it is implausible that disaccharides could precede monosaccharides in the synthesis of carbohydrates, it is, a fortiori, even less probable that starch could be formed directly by photosynthesis (as has occasionally been suggested, e. g., by Baly). True, one could conceive of a mecha- nism of photosynthesis in which new — CHOH links would be added to a chain, growing from some "carrier" molecule, and hexoses and other simple sugars would be produced by an enzymatic breakdown of these chains only after they have grown to a considerable length. It is, how- ever, unlikely that this hypothetical intermediate long-chain product should be identical with leaf starch. The prevailing opinion is that the latter is only a temporary storage product, deposited in the chloroplasts during intense photosynthesis, when more sugar is formed than can be re- moved by translocation. Sapozhnikov (1889, 1890, 1891, 1893) found that detached leaves of Vitis vinifera and V. labrusca can accumulate up to 25-50% dry weight in starch before becoming "choked" with this product. According to Winkler (1898) the formation of starch grains sets in when the concentration of glucose exceeds 0.2%, and reaches a maximum at 10% glucose in the leaf sap. THE ENERGY OF PHOTOSYNTHESIS 47 In addition to "common-sense" arguments against the direct forma- tion of starch by photosynthesis, the fact that many plants do not contain any leaf starch at all, also favors this conclusion. It has been known since Meyer (1885) that starch is less common in the leaves of the monocotyledons, than in those of the dicotyledons. Another argument in favor of starch formation as a secondary process which is not a part of photosynthesis proper, is the capacity of plants to convert artificially supplied sugars (or similar compounds) into starch, without the help of light. The starch synthesis and starch dissolution (amylolysis) in leaves must be consid- ered a part of the "second stage" of plant nutrition, which follows photosynthesis proper. Important advances in this field have become possible by the successful enzymatic synthesis of glycogen from glucose phosphate by Cori and coworkers. However, it would lead us too far to enter here into this complex matter. Spoehr and Milner (1939) have studied the effects of oxygen, carbon dioxide, water and temperature on the rate of dissolution of starch amylolysis in vivo. (These effects are important for photosynthesis because they influence the mechanism by which the stomata of the higher plants are opened and closed, thus regulating the supply of carbon dioxide to the chloroplasts, cf. page 334.) Spoehr and coworkers (1940) also initiated a study of the organic nutrition of albino plants to estabhsh the food requirements of plants which have been denied the possibihty of preparing their own food from the air. Obviously, studies of this kind can indirectly help in the identification of the first product of photo- synthesis. Glucose and other sugars can supply the plants with all their food require- ments (except for nitrogen and mineral elements assimilated through the roots), so that the formation of hexoses constitutes an entirely sufficient interpretation of the over-all reaction of photosynthesis; but whether it is also the minimum possible explanation, is another question. At the present stage of our knowledge, all processes from the moment of the entrance of carbon dioxide into the plant to the completion of sugar synthesis must be included into the "over-all reaction of photo- synthesis," which thus becomes (3.6) 6 CO2 + 6 H2O > CeHisOe + 6 Oa We will often use the abbreviated equation (3.7) CO2 + H2O > ICH2O} + O2 where {CH2O} stands for a generalized link in a carbohydrate chain, C. The Energy of Photosynthesis * From the time of Robert Mayer, it was known that photosynthesis converts light into chemical energy. The energy stored in this way is equal to the heat of combustion of the primary products of photosynthesis. In the first approximation, the heat of combustion of organic compounds containing carbon, hydrogen and oxygen, depends only on their level of * BibUography, page 59. 48 OVER-ALL REACTION OF PHOTOSYNTHESIS CHAP. 3 reduction (cf. Chapter 9). Table 3. IV shows how rapidly the heats of combustion rise with the progress of reduction. Photosynthesis lifts the stable "food" of the plants, CO2 + H2O, to the carbohydrate level, as indicated by the arrow on the left side of table 3. IV. Starting from this Table 3.IV The Four Reduction Levels of Carbon Dioxide Number of bonds Heat of com- bustion (in the Reduction stage c— 0—0 C— H 0— H gaseous state) to H2O (liq.) and CO2 (gas) AHc kcal/mole /. Carbon dioxide CO2 + 2 H2O S. Formic acid HCOOH + 13^ H2O + 1^ O2 ..5. Formaldehyde CH2O + H2O + O2 4. Methanol CH3OH + 3^ H2O + 13/^ O2 5. Methane CH4 + 2 O2 4 3 2 1 1 2 3 4 1 2 3 4 4 3 2 1 74 134 183 211 level, the plants produce compounds whose energy content is higher than that of the carbohydrates (e. g., alcohols and fats) without fur- ther supply of external energy, by dismutations, that is, reactions in which one part of the carbohydrates is oxidized enabling another part to be reduced. If formaldehyde were the first product of photosynthesis (as suggested by Baeyer in 1870), the heat effect of this process would be close to 135 kcal per gram atom of assimilated carbon. However, the "formaldehyde hypothesis" has never been proved and is now considered improbable {cf. Chapter 10). Whether photosynthesis involves the formation of another reduction intermediate with an energy content as high as that of formaldehyde is unknown. The same can be said of the often pos- tulated formation of a peroxide as precursor of free oxygen {cf. Chapter 11), which would add approximately another 45 kcal to the chemical energy accumulated in the first stage of photosynthesis. If both an unstable aldehyde and an unstable peroxide were among the immediate products of photosynthesis, the true heat effect of this process would be as high as 180 kcal per mole of reduced carbon dioxide. However, by the time the synthesis reaches an analytically recognizable stage — that of sugar and oxygen — the energy accumulation has been stabilized at about 112 kcal per mole. As shown by table 3.V, this value does not depend greatly on the exact nature ofithel^firstlsugar." Formaldehyde is less stable by about 23 kcal than an HCOH link in a long-chain carbohydrate; when we pass from this "monose" to a "biose" (glycolaldehyde), and further to "trioses" (for example, glyceraldehyde), tetroses, pentoses and THE ENERGY OF PHOTOSYNTHESIS 49 Table 3.V Energies (A//c) and Free Energies (AFc) of Combustion of Carbohydrates TO Liquid H2O and CO2 Gas, at 25° C. — AFc per gram -AHc. atom C L'^iqI i*ior Compound Formula State Kl^tll pel gram atom carbon Observer" to H2O (1.), CO2 (1 atm.) 6 to HjO (1.), CO2 (3 X 10-« atm.) Formaldehyde HCHO gas 134.1 W.L. 124.6 129.4 Formaldehyde, 1 M/l (HCHO)aq. solution 119.7 124.5 Para-formaldehyde (HCHO)„ solid 122.1 D.;W.M.R. Glyceraldehyde C3H6O3 solid 112.7 N.H.J. Arabinose CsHioOs soUd 112.0 K.F. Xylose CsHioOs sohd 112.1 K.F. a-Glucose C6H12O6 soUd 112.3 S.K.L.;K.F. 115.1 119.9 a-Glucose, 1 M/l (C6Hi206)aq. solution 112.6 S.L. 114.8 119.6 Fructose CeHnOe solid < 111.8 E.B. Galactose CeHnOe solid 111.7 K.F. Sucrose Ci2H220n solid 112.5 V.K. V.F. 115.0 119.8 Maltose C12H22O11 solid 112.3 K.N.N.W. Starch (C6Hio06)n solid 112.8 S.L. Cellulose (C6Hio06)n sohd 112.9 S.L. InuUn (C6Hio06)n sohd 113.1 K.F. "D. = DeMpine (1897), E.B. = Emery and Benedict (1911), K.F. = Karrer and Fiorom (1923), K.N.H.W. = Karrer, Nageli, Hurwitz, Walti (1921), N.H.J. = Neuberg, Hoffmann, Jacoby (1931), S.K.L. = Stohmann, Kleber, Langbein (1S90), S.L. = Stohmann, Langbein (1892), V.K. = Verkade, Koops (1923), W.L. = v. Wartenberg, Lerner-Steinberg (1925), and W.M.R. = v. Wartenberg, Much- lewski, Riedler (1924). „ ^ . ^ fc Figures from G. S. Parks and H. M. Huffman, The Free Energies of Some Organic Compounds. hexoses, the comparatively high energy of the keto group is rapidly "diluted" by the lower energies of the alcoholic groups, until a Kmiting value of about 112 kcal per link is reached. (The heats of combustion of the inositols, the only compounds consisting entirely of HCOH groups, have not yet been determined.) The standard bond energy table (c/. Table 9. II) shows that the endothermal character of photosynthesis has- a double origin. In the first place, the C— H bond is less stable (by 12 kcal) than the O— H bond, so that the hydrogen atoms have^^to^be transferred, in photosyn- thesis, from a stronger to a weaker bond. In the second place, the C=0 double bond in carbon dioxide (which has to be "opened" in photosynthesis) is more stable (by as much as 72 kcal) than the 0^0 double bond formed in this process. The weakness of the 0=0 double bond is the most important cause of the tendency of most elements for 50 OVER-ALL REACTION OF PHOTOSYNTHESIS CHAP. 3 oxidation, and of the difficulty of reversing this oxidation and expelling oxygen from oxides or organic oxygen compounds. The values of AFc in the last two columns of table 3.V, serve to illus- trate the statement, made in chapter 1 (page 3) that the increase in free energy in photosynthesis is even larger than the increase in total energy, particularly if AF is calculated not for the "standard" pressure of one atmosphere, but for the actual partial pressure of carbon dioxide in the air, 3 X 10~* atm. (Only for formaldehyde vapor AFc is smaller than AHc, because in this case two gases, H2CO and O2, are converted into one gas, CO2, and a liquid, H2O, thus decreasing the molecular disorder.) Of course, an increase in free energy is possible only because photo- synthesis is not a spontaneous process in a closed system, but a photo- chemical reaction, maintained by a continuous supply of light energy (to make the system complete, the sun should be included in it). To sum up, table 3.V makes it probable that photosynthesis proceeds with an accumulation of at least 112 kcal per mole of reduced carbon dioxide: light (3.8) CO2 + H2O > {CH2OI + O2 - 112 kcal plant and, in the free atmosphere, with an increase in free energy by about 120 kcal per mole. In thermochemical equations, we will conform to the usage and designate the absorbed energy by a minus sign, and the released energy, by a plus sign. On the other hand, the heat effect AH of a reaction shall be considered as positive for endothermal and negative for exothermal reactions, in accordance with the notation of Lewis and Randall. In other words, the figure — 112 kcal in equation (3.8) represents minus AHc. The efficiency of photosynthesis as an energy-converting process depends on the amount of light required for the reduction of one mole of the substrate. Much study has been devoted to this problem (which will be discussed in Vol. II, Chapters 28 and 29). Anticipating the results we can state that the average conversion yield in direct sunlight is of the order of 3% of the absorbed light energy, or 2% of the incident visible light; but in weak light and in presence of ample carbon dioxide it may rise to as much as 30%. This indicates that the low energy conversion observed under natural conditions is caused by a limited capacity of the photosynthetic apparatus and consequent dissipation of energy absorbed in excess of this capacity, rather than by an obligatory utilization of a large part of light energy for the activation of the chemical process of photosynthesis. Fundamentally, the photosynthetic mechanism is cap- able of converting light into chemical energy with an efficiency of not less than 30%, and perhaps more. This is a much higher efficiency than has ever been achieved in photochemical processes in the laboratory (except for reactions which take place only in ultraviolet light). PHOTOSYNTHESIS AS A SENSITIZED OXIDATION-REDUCTION 51 An even more striking characteristic of photosynthesis has been claimed by Spessard (1940), who asserted that photosynthesis results in "conversion of hght into matter." His experiments purported to show that a sealed vessel containing photosynthesizing plants increases in weight with the progress of photosynthesis, and certainly will receive a less spectacular explanation. D. Photosynthesis as a Sensitized Oxidation-Reduction * After having described the over-all chemical reaction of normal photosynthesis by equations (3.6) and (3.7), we will now assign to this reaction its proper place in the general classification of chemical reac- tions, by identifying it as a sensitized photochemical oxidation-reduction. When the first light was thrown on the chemistry of photosynthesis by the investigations of Ingen-Housz and Senebier, it appeared as "de- composition of fixed air" {i. e., carbon dioxide) with the oxygen escaping into the air, and carbon retained by the plants. Even when de Saussure in 1804 added water to the reaction components, he did not doubt that all oxygen liberated in photosynthesis was the product of decomposition of carbon dioxide, while the role of water was vaguely described as "con- tributing its elements" to the formation of organic matter. Later, the "decomposition" of carbon dioxide was generally recognized as a re- duction of this compound, and different paths of reduction were devised, e. g., by Liebig (1843) and Baeyer (1870). According to Liebig, the plant acids— oxalic, malic, succinic, tartaric — are the main intermediates in the reduction of carbon dioxide to carbohydrates; while according to Baeyer, formic acid and formaldehyde are the two main stepping stones in this reduction. The question of the way in which water participates in the reduction was left aside by both authors. However, some chemists have looked on photosynthesis from a dif- ferent angle. As early as 1864, Berthelot suggested that water is decom- posed by photosynthesis into hydrogen and oxygen, while carbon dioxide is dissociated into carbon monoxide and oxygen, after which the two products unite to form a carbohydrate, (3.8a) CO + H2 > { CH2O ) Although this theory was vague, it clearly made both carbon dioxide and water subjects of primary transformations in photosynthesis. Fifty years later, Bredig (1914) and Hofmann and Schumpelt (1916) turned the spotlights entirely on the transformation of water. They suggested that the primary effect of light in photosynthesis is the decomposition of water into oxygen and hydrogen. The former escapes into the atmosphere, while the latter reduces carbon dioxide to the carbohydrate level (by a secondary process, not specifically defined). * Bibliography, page 60. 52 OVER-ALL REACTION OF PHOTOSYNTHESIS CHAP. 3 Willstatter and Stoll (1918) were opposed to concepts of this kind. They thought that the exact equivalence between the consumption of carbon dioxide and the evolution of oxygen can be understood only if one assumes that oxygen originates in the decomposition of carbon dioxide, or, since water has to be given a place in the scheme, in the decomposition of carbonic acid: light (3.9) H2CO3 > O2 + { CH2O j plant These three schemes of photosynthesis: (a) decomposition of carbon dioxide (with a subsequent reaction of one of the products with water) ; (6) decomposition of water (with a secondary reaction between one of the products and carbon dioxide); and (c) decomposition of carbonic acid (after a preliminary combination of CO2 and H2O to H2CO3), have been widely used in the literature; but it was some time before it became clear that all three of them implied, without telling it in so many words, that photosynthesis is an oxidation-reduction reaction between carbon dioxide and water. That photosynthesis is a reduction of carbon dioxide, was generally acknowledged; but that reduction presupposes a reductant and that in photosynthesis the only possible reductant is water (which is oxidized to oxygen) was ignored. It seemed strange to call "oxidation " a process in which free oxygen is produced; but the removal of hydrogen from the water molecule is oxidation by any general definition of this term. In the above-mentioned scheme (a), carbon dioxide is reduced to carbon, and the latter "hydrated" by water, a process which seems to imply no oxidation at all. In scheme (c), of Willstatter and Stoll, neither the hydration of carbon dioxide to carbonic acid, nor the decomposition of carbonic acid into formaldehyde and oxygen, seems to bear the charac- ter of oxidation-reduction. However, both the "hydration" of C to H2CO and the "decomposition" of H2CO3 into H2CO and O2, involve transfers of hydrogen atoms from oxygen to carbon, and this is the mark of an oxidation-reduction, even if the transfer occurs intramolecularly, i. e., between two atoms belonging to the same molecule, and not inter- molecularly, as in typical oxidation-reduction reactions. To say that photosynthesis is an oxidation-reduction reaction between water and carbon dioxide, is not to suggest an hypothesis, but to make a statement of fact. In recent years, the mechanisms of many biological oxidation-reduc- tions have been elucidated, and the transfer of hydrogen atoms (or elec- trons, cf. page 219) from molecule to molecule, has emerged as the most important elementary act in these processes. Thus Wieland (1913, 1914) explained respiration as the transfer of hydrogen atoms from a substrate (glucose, for instance) to oxygen; and Kluyver and Donker (1926) and PHOTOSYNTHESIS AS A SENSITIZED OXIDATION-REDUCTION 53 Kluyver (1930) interpreted different anaerobic fermentations as similar transfers of hydrogen to acceptors other than oxygen. The hypothesis that photosynthesis can be placed alongside with other biological oxidation-reductions and interpreted as an intermolecular exchange of hydrogen atoms between water and carbon dioxide, was first discussed by Thunberg (1923), but the credit for its clear formulation, based on the analaysis of the metabolism of sulfur bacteria (which will be discussed in chapter V), belongs to van Niel (1931). Starting from Kluyver and Donker's generalization of Wieland's ideas, van Niel pro- claimed photosynthesis to be a hydrogen transfer from water to carbon dioxide in the higher plants, and from other hydrogen "donors" to carbon dioxide in bacteria. In spontaneous metabolic proce.sses, the transfer of hydrogen atoms (or electrons) always occurs "downhill," that is, in the direction of decreasing oxidation-reduction potentials. The substance with higher (more positive) potential yields its hydrogen to the substance with the lower (more negative) potential. In photosynthesis, which is the reverse of respiration, the hydrogen atoms must be moved "uphill," from a system with a lower potential — O2/H2O — to the system with a higher potential — C02/{CH20} — light being relied upon to give the necessary "push." The reduction of carbon dioxide to the carbohydrate level requires the hydrogenation of two C=0 double bonds, and thus the transfer of four hydrogen atoms: H (3. 10) 0=C=0 + 4 H > HO— C— OH H The primary product of photosynthesis, according to (3.10) is formal- dehyde hydrate. However, van Kiel's theory does not require the forma- tion of this compound as an intermediate in photosynthesis, since it can equally well be applied to the reduction of a larger molecule (e. g., of a carboxylic acid, R-COOH), into which CO2 has been incorporated in a preliminary reaction step. The four hydrogen atoms required in (3.10) can be provided by either two or four water molecules : (3.11) 2H2O ^02 + 4H or (3.12) 4 H2O > 2 H2O2 + 4 H > O2 + 2 H2O + 4 H The over-all reaction of normal photosynthesis becomes, with (3.11): (3.13) CO2 + 2 H2O > H2C(OH)2 + O2 > {CH2OI + H2O + O2 and with (3.12): (3.14) CO2 + 4H2O >H2C(OH)2 + 2H20 + 02 > {CH2O! +3H2O + O2 54 OVER-ALL REACTION OF PHOTOSYNTHESIS CHAP. 3 We prefer the second alternative, because the probabiHty of one water molecule losing both its hydrogen atoms in succession seems to be smaller than that of two different water molecules contributing one hydrogen atom each (and the remaining hydroxyl radicals reacting to water and oxygen). One of the two water molecules which enter reaction (3.13), and three of the four which enter reaction (3.14), are recovered at the end, and could be cancelled out in the equations, if it were not desirable to underline that hydrogen atoms from several water molecules partici- pate in the reduction of one molecule of carbon dioxide. If we add the " intermolecular hydrogen transfer" theory as scheme (d) to the three schemes listed on page 52, and consider the origin of oxygen according to all four of them, we find that only the oldest scheme, (a), suggests that all oxygen comes from carbon dioxide; schemes (6) and (d) predict that all of it should come from water; while according to scheme (c), one part of oxygen must come from carbon dioxide and another part from water. The last conclusion is based on the consideration that, since the two hydroxyl groups in H2CO3 are equivalent, they must contribute equally to the production of oxygen. After the hydration (3.15) CO2 + H2O . 0=C(0H)2 one-half the oxygen atoms in the hydroxyl groups have their origin in water and one-half in carbon dioxide. If the hydration is followed im- mediately by decomposition into H2CO and O2, the proportion of oxygen which originated in water can be one-half (if all oxygen comes from the hydroxyl groups), one-third (if all three O atoms in H2CO3 contribute equally to the formation of oxygen), or one-fourth (if one oxygen atom in O2 must come from the C=0 group). However, reaction (3.15) is, usually, not a single transformation, but a series of repeated hydrations and dehydrations, and as a result the ratio of oxygen atoms in H2CO3 which originally belonged to water or carbon dioxide, gradually ap- proaches the ratio of these atoms in all available molecules of these two compounds. Since water is present in large excess, practically all oxygen atoms in H2CO3 will ultimately be contributed by water. However, the hydration and dehydration of carbon dioxide are slow reactions (c/. Chapter 8, page 175) and since plants apparently do not contain the enzyme (carbonic anhydrase) which accelerates them (c/. Chapter 15, page 380), the equilibration of CO2 and H2CO3 takes a measurable time, and the contribution of carbon dioxide to the oxygen production accord- ing to scheme (c) must onlyjgradually^drop^from an initial maximum (H, %, or 34) to zero. The existence of a heavy isotope of oxygen, 0^^, made possible a direct check of these predictions — a striking example of the possibilities inherent in the method of "isotopic tracers." Ruben, Randall, Kamen and Hyde PHOTOSYNTHESIS AS A SENSITIZED OXIDATION-REDUCTION 55 (1941) introduced heavy oxygen (0^^) into the carbon dioxide and water used for photosynthesis of Chlorella, and determined the concentration of the heavy isotope in the liberated oxygen. The results are given in table 3. VI. It shows that the proportion of 0^^ in oxygen is under all Table 3.VI IsoTOPic Ratio in Oxygen Evolved in Photosynthesis by Chlorella "» (after Ruben, Randall, Kamen and Hyde) Substrate Time between dissolving KHCO3 + K2CO3 and start of O2 collection, Time at end of O2 collection, min. Per cent 0" in Expt. No. HzO HCO3- + CO3- O2 min. 1 KHCO3, 0.09 M 0.85 0.20 K2CO3, 0.09.1/ 45 110 0.85 0.4P 0.84 110 225 0.85 0.55* 0.85 225 350 0.85 0.61 0.86 2 KHCO3, 0.14 M 0.20 K2CO3, 0.06 il/ 40 110 0.20 0.50 0.20 110 185 0.20 0.40 0.20 3 KHCO3, 0.06 M 0.20 0.68 K2CO3, 0.14 71/ 10 50 0.20 0.21 50 165 0.20 0.57 0.20 ■■ The volume of evolved oxygen was large compared with the amount of atmospheric oxygen present at the beginning of the experiment. ^ Calculated values. circumstances equal to its proportion in water, and independent of its concentration in the carbonate — thus disproving the hypotheses (a) and (c). As each experiment progresses, the isotopic exchange between carbon dioxide and water, brought about by reaction (3.15), tends to equalize the isotope distributions in the two reaction components; but since the concentration of HCOa" is high and that of CO2 low {cf. page 178), the equalization proceeds slowly and several collections of oxygen can be made before its completion. In chapter 26 (Volume II), we will encounter an additional quantitative argument against Willstatter and StoU's scheme (c)— the inability of the hydration reaction (3.15) to keep pace with photosynthesis in very intense light. The fact that most if not all oxygen molecules liberated by photo- synthesis, originate from water, was confirmed by Vinogradov and Teis (1941) who determined the density of water synthesized from this oxy- gen, and proved that the isotopic composition of the latter is similar to that of oxygen in natural water (rather than to that of oxygen in carbon dioxide). As to the two schemes, (h) and (d), which predict that all oxygen should come from water, the difference between them is that the older 56 OVER-ALL REACTION OF PHOTOSYNTHESIS CHAP. 3 one assumes the intermediate formation of molecular hydrogen, while the newer one postulates a transfer of hydrogen atoms from water to carbon dioxide (either directly, or through the intermediary of catalysts). We prefer the second theory because the intermediate formation of molecular hydrogen in photosynthesis could hardly have remained unnoticed. After having characterized in general terms, the chemical nature of photosynthesis, it seems appropriate to add a similar general description of its physical nature, by classifying photosynthesis as a sensitized photo- chemical reaction. It must be sensitized by a pigment, because the reac- tion substrate (CO2 + H2O) does not absorb visible light. The concept of sensitization is familiar from the photographic plate, from so-called " photodynamic effects" in biology, and from many photochemical reac- tions in vitro. In the exact sense of the term, sensitization means a photochemical reaction induced by a light-absorbing substance which is not itself permanently affected by the reaction. True sensitizers are thus substances which act as catalysts in light. However, substances are often called "sensitizers" even if they take an active part in the photochemical re- action (as this is probably the case in most photodynamic effects). We cannot entirely avoid using "sensitization" and "sensitizer" in the usual loose manner, and shall therefore speak of " photocatalysts " when desir- ing to emphasize that we are dealing with a case of "true" sensitization. Chlorophyll is a photocatalyst, since no decrease in the concentration of chlorophyll in leaves has been observed after intense photosynthesis (c/. Chapter 19, page 549). Therefore, only truly photocatalytic reac- tions can be adduced as imitations of photosynthesis in vitro. This has often been neglected by investigators who have attempted to reproduce photosynthesis outside the living cell, as will be demonstrated by many examples in chapter 4. Bibliography to Chapter 3 The Over-All Reaction and the Products of Photosynthesis A. The Quantitative Balance of Photosynthesis 1804 de Saussure, N. Th., Recherches chimiques sur la vegetation. Nyon, Paris, 1804. 1864 Boussingault, T. B., Ann. sci. nat. Botan., 1, 314. 1874 Boussingault, T. B., Agronomie, chimie agricole et physiologie. Vol. 5, Mallet-Bachelier, Paris, 1874. 1886 Bonnier, G., and Mangin, L., Ann. sci. nat. Botan. VII, 3, 1. 1890 Sapozhnikov, V., Ber. deut. botan. Ges., 8, 234. 1892 Aubert, E., Rev. gen. botan., 4, 203, 273, 321, 337, 373, 421, 457, 558. 1901 Krasheninnikov, T., Accumulation of Solar Energy in Plants. (Russ.) Moscow, 1901. 1913 Maquenne, L., and Demoussy, E., Compt. rend., 156, 506. BIBLIOGRAPHY TO CHAPTER 3 57 1918 Willstatter, R. and Stoll, A., Untersuchungen uber die Assimilation der Kohlensdure. Springer, Berlin, 1918; pp. 315-341. 1921 Kostychev, S., Ber. deut. hotan. Ges., 39, 319. 1924 Bose, J. C, The Physiology of Photosynthesis. Longmans, London, 1924, p. 198. 1932 van der Paauw, P., Rec. trav. hotan. neerland., 29, 497. 1935 Barker, H. A., Arch. MikrobioL, 6, 141. 1938 Manning, W. M., Stauffer, J. F., Duggar, B. M., and Daniels, F., J. Am. Chem. Soc, 60, 266. 1939 Ruben, S., Hassid, W., and Kamen, M. D., J. Am. Chem. Soc, 61, 661. 1943 Smith, J. H. C, Plant Physiol, 18, 207. B. The Products of Photosynthesis 1837 Mohl, H. von, Untersuchungen uber anatomische Verhdltnisse des Chlorophylls. (Dissertation of W. Michler, Univ. Tubingen, 1837.) 1856 Boehm, T. A., Sitzber. Akad. Wiss. Wien, Math. Naturw. Klasse, 22, I, 479. 1857 Gris, A., Ann. sci. nat. Botan., (IV), 8, 170. 1862 Sachs, J., Botan Z., 20, 365. 1863 Sachs, J., Jahrb. wiss. Botan., 3, 184. 1864 Sachs, J., Botan. Z., 22, 289. 1873 Pfefifer, W., Monatsh. Akad. Wiss., Berlin, 784. Briosi, G., Botan. Z., 31, 529, 545. 1874 Boehm, T. A., Sitzber. Akad. Wiss. Wien, Math. Naturw. Klasse, 69, I, 76. 1876 Boehm, T. A., ibid., 73, I, 39. 1877 HoUe, H. G., Flora, 35, 113, 154, 160, 184. Godlewski, E., ibid., 35, 215. 1882 Perrey, A., Compt. rend., 94, 1124. 1883 Boehm, T. A., Botan. Z., 41, 33, 49. 1885 Meyer, A., ibid., 43, 417, 433, 449, 465, 480, 497. 1889 Sapozhnikov, V., Ber. deut. botan. Ges., 7, 25. 1890 Sapozhnikov, V., ibid., 8, 233. 1891 Sapozhnikov, V., ibid., 9, 293. 1892 Crato, E., ibid., 10, 250. 1893 Sapozhnikov, V., ibid., 11, 391. Brown, H. T., and Morris, J. H., /. Chem. Soc, 63, 604. 1898 Winkler, F., Jahrb. wiss. Botan., 32, 525. 1904 Beijerinck, M. W., Rec trav. botan. neerland., 1, 28. 1907 MuUer, H., /. Chem. Soc, 91, 1767. Tanret, G., Compt. rend., 145, 1196. 1909 Robertson, R. A., Irvin, T. C, and Dobson, M. E., Biochem. J., 4, 258. 1910 Nef, T. U., Ann. Chemie (Liebig), 376, 1. Lob, W., and Pulvermacher, G., Biochem. Z., 23, 10. 58 OVER-ALL REACTION OF PHOTOSYNTHESIS CHAP. 3 1911 Ravenna, C, Gazz. chim. ital., 41 (2), 115. Parkin, T., Biochem. J., 6, 1. 1912 Grafe, V., and Vouk, V., Biochem. Z., 43, 424. Grafe, V., and Vouk, V., ibid., 56, 249. 1914 Tollens, B., Kurzes Handbuch der Kohlehydrate. 3rd ed., Barth, Leipzig, 1914. 1916 Curtius, Th., and Franzen, H., Sitzber. Heidelberger Akad. Wiss. Math, naturw. Klasse, Nr. 7. Davis, W. A., Daish, A. T., and Sawyer, J. C., /. Agr. Sci., 7, 255. Davis, W. A., and Sawyer, G. C., ibid., 7, 352. Davis, W. A., Biochem. J., 10, 31. Daish, A. J., ibid., 10, 49. Mason, T. G., Sci. Proc. Roy. Soc. Dublin, 15, 13. Dixon, H. M., and Mason, T. G., Nature, 97, 160. 1917 Gast, W., Z. physiol. Chem., 99, 1. Meyer, A., Ber. deut. botan. Ges., 35, 586. Meyer, A., ibid., 35, 676. 1918 Meyer, A., ibid., 36, 235. 1919 Kogel, P. R., Biochem. Z., 95, 313. Kogel, P. R., ibid., 97, 21. 1922 Sjoberg, K., ibid., 133, 218. 1924 Weevers, T., Proc. Acad. Sci. Amsterdam, 27, 1. Priestley, J. H., New Phytologist, 23, 255. Melchior, R., Ber. deut. botan. Ges., 42, 198. 1925 Stiles, W., Photosynthesis. Longmans, Green, London, 1925, pp. 151-160. 1926 Spoehr, H. A., Photosynthesis. Chemical Catalog Co., New York, 1926, pp. 215-220. Tottingham, W. E., Lepkovsky, S., Schulz, E. R., and Link, P., J. Agr. Research, 33, 59. 1929 Spoehr, H. A., and Strain, H. H., J. Biol. Chem., 85, 365. 1930 Barton- Wright, E. C, and Pratt, M. C, Biochem. J., 24, 1217. Clements, H. F., Botan. Gaz., 89, 241. • 1932 Clements, H. F., Plant Physiol., 7, 547. 1935 Spoehr, H. A., and Milner, H. W., /. Biol. Chem., Ill, 679. Kretovich, V. C, Z. physiol Chem., 231, 265. 1936 Endo, S., Science Repts. Tokyo Bunriku Daigaku, 2, 223, 231 ; cf. Chem. Abstracts, 30, 7150. Spoehr, H. A., and Milner, H. W., /. Biol. Chem., 116, 493. 1938 Venezia, M., Atti 1st. Veneto 2, 97, 357. 1939 Spoehr, H. A., and Milner, H. W., Proc. Am. Phil. Soc, 81, 37. 1940 Spoehr, H. A., Smith, J. H. C, Strain, H. H., and Milner, H. W., Carnegie Yearbook, 39, 147. 1943 Smith, J. H. C, Plant Physiol., 18, 207. Gardner, Th. S., J. Org. Chem., 8, 111. 1944 Smith, J. H. C, Plant Physiol., 19, 394. t BIBLIOGRAPHY TO CHAPTER 3 59 General References Structure of Carbohydrates and their Occurrence in Plants Czapek, F., Biochemie der Pflanzen. Vol. 2, 2nd ed., Fischer, Jena, 1920. Pringsheim, H., The Chemistry of the Monosaccharides and of the Polysaccharides. McGraw-Hill, New York, 1932. Bernhauer, K., Grundzuge der Chemie und Biochemie der Zuckerarten. Springer, Berlin, 1933. Armstrong, E. F., and Armstrong, K. F., The Carbohydrates. 4tli ed., Longmans, Green, London, 1934. Micheel, F., Chemie der Zucker und Polysaccharide. Akadem. Ver- lagsgesellschaft, Leipzig, 1939. Products of Photosynthesis in Algae Tilden, J. E., The Algae and their Life Relations. Univ. Minn. Press, Minneapolis, 1935. Fritsch, F. E., The Structure and Reproduction of Algae. Macmillan, Cambridge, 1935. C. Energy of Photosynthesis 1890 Stohmann, F., Kleber and Langbein, Z. physik. Chem., 6, 334, 341. 1892 Stohmann, F., and Langbein, /. prakt. Chem., (2), 45, 305. 1897 Del^pine, N., Compt. rend., 124, 1525. 1911 Emery, A, G., and Benedict, F. G., Am. J. Physiol, 28, 301. 1921 Karrer, P., Nageli, C., Hurwitz, 0., and Walti, A., Helv. Chim. Acta, 4, 678. 1923 Karrer, P., and Fioroni, W., ibid., 6, 396. Verkade, P. E., and Koops, J., Rec. trav. chim., 42, 223. 1924 V. Wartenberg, H., Muchlewski, and Riedler, Z. angew. Chem., 37, 457. 1925 V. Wartenberg, H., and Lerner-Steinberg, ibid., 38, 592. 1931 Neuberg, C., Hoffmann, E., and Jacoby, M., Biochem. Z., 234, 344. 1940 Spessard, E. A., Plant Physiol, 15, 109. General References Compilations of Heats of Reactions Kharash, M. S., Bur. Standards J. Research, 2, 359 (1929). Bichowski, F. R., and Rossini, F. D., The Thermochemistry of Chemi- cal Substances. Reinhold, New York, 1936. Roth, W. A. in Vol. IP and IIP of Landolt-Bornstein's Physikalisch.- Chem. Tabellen. Springer, Berlin, 1931, 1936. Compilations of Free Energies Parks, G. S., and Huffman, H. M., The Free Energy of Some Organic Compounds. Reinhold, New York, 1932. V. Stackelberg and Teichmann, in Vol. IP and Ulich, in Vol. IIP of Landolt-Bornstein's Physikalisch.-Chem. Tabellen. Springer, Ber- lin, 1931, 1936. 60 OVER-ALL REACTIOX OF PHOTOSYNTHESIS CHAP. 3 D. Photosynthesis as an Oxidation- Reduction 1804 de Saussure, N. Th., Recherches chimiques sur la vegetation. Nyon, Paris, 1804. 1843 Liebig, J., Ann. Chemie, 46, 58. 1864 Berthelot, M., Legons sur les methodes generales de synthase en chimie organique. Gauthier-Villars, Paris, 1864. 1870 Baeyer, A. von, Ber. deut. chem. Ges., 3, 63. 1913 Wieiand, H., ibid., 46, 3327. 1914 Wieiand, H., ibid., 47, 2085. Bredig, G., Umschau, 18, 362. 1916 Hofmann, K. A., and Schumpelt, K., Ber. deut. chem. Ges., 49, 303. 1918 Willstatter, R., and StoU, A., Untersuchungen uber die Assimilation der Kohlensdure. Springer, Berlin, 1918, pp. 236-246 and 319. 1923 Thunberg, T., Z. physik. Chem., 106, 305. 1926 Kluyver, A. J., and Donker, H. J. L., Chem. Zelle Gewebe, 13, 134. 1930 Kluyver, A. J., Arch. Mikrobiol, 1, 181. 1931 van Niel, C. B., ibid., 3, 1. 1941 Ruben, S., Randall, M., Kamen, M. and Hyde, J. L., J. Am. Chem. Soc, 63, 877. Vinogradov, A. P., and Teis, R. V., Compt. rend. acad. sci. URSS, 33, 490. Chapter 4 PHOTOSYNTHESIS AND RELATED PROCESSES OUTSIDE THE LIVING CELL Complete photosynthesis — that is, reduction of carbon dioxide to carbohydrates, and oxidation of water to oxygen, at low temperature and with no energy supply except visible light — has never been achieved outside the living cell. What has been accomplished were at best "par- tial" successes, reactions which in some future time may (or may not) be integrated into a complete reconstruction of photosynthesis in vitro. In attempting to divest photosynthesis of its association with the living state, one may begin with the living cell, tear it down, and observe the effects of this procedure on the different aspects of photosynthesis; or one may build up from simple light-sensitive oxidation-reduction systems to more complicated ones, with an eye on the maximum con- version of light into chemical energy. Some investigators have been too impatient to use either of these gradual approaches, and have spent unprofitable years in attempts to reproduce photosynthesis in toto by experiments which deserve to be called alchemistic rather than chemical. A. Photosynthesis by Dried Leaves, Isolated Chloro- PLASTS AND CHLOROPHYLL PREPARATIONS * 1. Leaf Powders and Isolated Chloroplasts In 1881, Engelmann stated that "as soon as the structure of the chlorophyll-bearing bodies is destroyed, the capacity for oxygen pro- duction ceases at once and forever." Since then, the association of photosynthesis with the living state of the cells has been one of the fundamental facts of plant physiology. It was observed again and again, that freezing, drying, boiling or poisoning puts an immediate end to photosynthesis. Many tissues can live and function outside the body; a hear-t will beat for days in a physiological salt solution, but the chloro- plasts cease functioning the moment the cell is destroyed or damaged. In 1901, Friedel reported that powdered leaves, dried at 100° C. and mixed with glycerol extracts from fresh leaves, produce oxygen and consume carbon dioxide in * Bibliography, page 94. 61 62 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 light. Similar observations were described by Macchiati (1903). However, Harroy (1901), Herzog (1902), Molisch (1904) and Bernard (1904, 1905) failed to confirm them, and Friedel found himself unable to reproduce his earlier positive results later in the same year (190P), a failure which he attributed to the inefficiency of autumnal leaves. While dead or broken cells certainly are unable to carry out complete photosynthesis, they may maintain a Umited capacity for evolving oxygen in light (without a concurrent reduction of carbon dioxide). This was first noticed in 1888 by Haberlandt and in 1896 by Ewart, who observed that isolated chloroplasts (obtained by grinding leaves under water) liberate a small quantity of oxygen when exposed to light. Similar observations with dried leaf powders were made by Molisch in 1904. The matter rested there for twenty years, until Molisch came back to it in 1925. He confirmed the evolution of oxygen by leaves which were dried for three or four days at 30-35° C, and often kept in a desic- cator for several weeks. Some leaves produced oxygen even after having been heated to 84° for five hours. To obtain oxygen, dried leaves had to be powdered under water and the mixture illuminated without straining. Leaves killed by freezing also produced oxygen in light. Inman (1935) repeated the experiments of Ewart and Molisch. Fresh leaves of Trifolium repens, Zea mais and Melilotus alba were ground with sand and the remaining whole cells removed by filtration. The suspension of broken and unbroken chloroplasts, obtained in this way, produced oxygen upon illumination. Positive results were obtained also with leaves dried for several days at 30-35° before powdering. The leaves lost their capacity for evolving oxygen more quickly if they were first macerated and then dried, instead of drying first and powdering after- wards. The alga, Nostoc, was able to evolve oxygen even after having been kept in the dry state for eighteen months. Addition of protein- digesting enzymes (e. g., trypsin) prevented the evolution of oxygen by leaf powders. (Photosynthesis by unbroken cells was not affected by trypsin.) The production of oxygen was limited to the pH range 5-7, with a sharp maximum at a pB. of 5.5. Inman (1938) stressed the similarity between the effects of temperature, acidity, and trypsin on the oxygen evolution by leaf triturates and on the denaturation of proteins, and suggested that the oxygen liberation is catalyzed by an enzyme. According to Inman (1938^), the evolution of oxygen can also be demonstrated with the isolated cell contents of the giant cells of Nitella and Valonia macrophysa, and with press juices from clover leaves and Euglena viridis. In all these experiments, oxygen formation could be proved only by the luminescence of Beijerinck's bacteria; the rate of its evolution was unknown, but certainly very small; and the whole process lasted LEAF POWDERS AND ISOLATED CHLOROPLASTS 63 for not more than an hour. All this seems to point to a decomposition of a limited quantity of a peroxide, either left in the leaves as a residue from normal metabolism, or accumulated by post mortem processes. Yamafuji and coworkers have observed that dried vegetable and animal tissues form small quantities of peroxide {cf. page 78). This peroxide, slowly accumulated in darkness or in diffuse light, could decompose suddenly upon exposure to strong light (particularly if chlorophyll is present as a sensitizer), thus producing the "burst" of oxygen observed by Molisch and Inman. If this explanation is correct, the oxygen evolution by dried leaves is not directly related to photosynthesis (except for the fact that it, too, may be sensitized by chlorophyll). However, the experiments of Hill (1937, 1939, 1940) favor another hypothesis — that the oxygen evolution by leaf powders bears a significant relation to the production of oxygen in photosynthesis. In Hill's first experiments (1937, 1939), a chloroplast suspension was obtained from leaves of Stellaria media, Lamium album and other plants, by grinding in a phosphate-buffered 10% sucrose solution (pH 7.9), and filtering through glass wool. The suspension was mixed with a solution of hemoglobin, under exclusion of air. The evolution of oxygen (in light of 40,000 lux) was measured spectrophotometrically by observing the conversion of hemoglobin into oxyhemoglobin. However, oxygen was found only when an aqueous leaf extract was added to the suspension. (This extract was prepared by grinding leaves under acetone, filtering, drying the filtrate and extracting the residue with water.) In further developing these experiments. Hill observed that the leaf extract can be replaced by a yeast extract, and that the efficiency of the latter was dependent on its content of organic iron compounds. Finally, he found that the oxygen evolution can also be brought about by the addition of ferric potassium oxalate, thus allowing him to dispense with cell extracts of unknown composition. The illumination of an air-free mixture of chloroplasts with ferric potassium oxalate and hemoglobin, causes a rapid appearance of oxyhemoglobin, and a reduction of Fe+++ to Fe++ (detectable, for example, by means of dipyridyl). In the dark, the original state is slowly restored again. The total quantity of oxyhemo- globin produced in this experiment, depends on the quantity of ferric salt taken, while the initial velocity of oxidation is determined by the quantity of the chloroplasts. The most active wave lengths are those around 600 mju (thus pointing to chlorophyll as the sensitizing agent). The maximum rate of oxygen evolution by a chloroplast suspension in the presence of ferric oxalate, was about equal to the rate of oxygen liberation by a suspension of whole cells (from the same leaves) in the presence of carbon dioxide, but was only one-tenth of the rate of photo- 64 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 synthesis of the intact leaf (if all three rates were related to the same quantity of chlorophyll). One may attempt to explain Hill's results by a chlorophyll-sensitized oxidation of a peroxide by ferric oxalate (cf. Eq. 4.1). As pointed out by Kautsky (1938) ferric oxalate itself oxidizes hydrogen peroxide in violet and ultraviolet light; this reaction could easily be sensitized by chlorophyll. However, the total quantity of oxygen obtained in Hill's experiments would require the presence in the chloroplasts of 0.1 mole per liter of the peroxide, which is not plausible. Furthermore, only one-half a gram atom of oxygen was produced for one gram atom of re- duced ferric iron. For the oxidation of a peroxide, this ratio would be 1 : 1. light (4.1) Fe+++ + i H2O2 > Fe++ + H+ + ^ O2 Thus, the substrate of oxidation must be an oxide {e. g. water) rather than a 'peroxide: light (4.2) Fe+++ + \ H2O > Fe++ + H+ + i O2 Equation (4.2) suggests that Hill's reaction is a chlorophyll-sensitized reversal of the familiar oxidation of ferrous to ferric iron by oxygen, just as photosynthesis is a reversal of the familiar process of combustion of carbohydrates. In photosynthesis, oxygen can be liberated regardless of its partial pressure in the atmosphere. In Hill's first experiments with isolated chloroplasts, oxygen was evolved (in absence of hemoglobin) only if the partial pressure of this gas was less than 1 mm. in experiments with leaf extracts, and 4 mm. in experiments with ferric oxalate. Hill and Scarisbrick (1940^) found, however, that the limitation was caused by the reoxidation of ferrous oxalate by oxygen. If potassium ferricyanide (which does not itself cause an evolution of oxygen by the chloroplasts in light) was added to the mixture, it reoxidized ferrous oxalate more rapidly than did oxygen, and thus allowed the latter to accumulate, independently of its partial pressure, until its total quantity was equiva- lent to the maximum quantity of oxyhemoglobin obtainable from the same preparation. If no exhaustion of the oxidant (ferric oxalate) was allowed to occur, the evolution of oxygen could be maintained for several hours; however, it gradually became weaker, and sank to zero after five or six hours of illumination. Hill and Scarisbrick (1940-) investigated the effects of different external factors on the initial rate of liberation of oxygen and reduction of ferric oxalate by chloroplasts. For the determination of ferrous oxalate, they used the reduction of methemoglobin to hemoglobin (a method which they considered more reliable than the complex formation LEAF POWDERS AND ISOLATED CHLOROPLASTS 65 with l,l'-dipyridyl). The amount of hemoglobin was measured spectro- photometrically, by oxidizing it to oxyhemoglobin. Thus, in practice, the evolution of oxygen was determined by measuring the rate of for- mation of oxyhemoglobin in anaerobic mixtures of chloroplasts with ferric oxalate and hemoglobin, while the reduction oj ferric oxalate was determined by measuring the same rate in aerobic mixtures of chloroplasts with ferric oxalate and methemoglobin. The rates obtained by the second method were lower, but better re- producible than those calculated from anaerobic experiments. The cause of this difference was the slowness of the methemoglobin-ferrous oxalate reaction, which "limited" the over-all process, making it less sensitive to poisons (but more sensitive to temperature). Thus, values obtained in anaerobic experiments, though less consistent, are more significant, being free from such artificial limitation. 10000 20000 30000 40000 Light intensity, foot -candle Fig. 6. — Effect of varying light intensity on the rate of evolution of oxygen by chloroplasts (after Hill and Scarisbrick 1940^). Fe concentration, 4 X 10"'' mole per 1. Chloroplast suspension, 0.4 ml. Circles, extreme values; dots, mean values. The effect of light intensity on the rate of oxyhemoglobin production by chloroplasts from Stellaria media is illustrated by figure 6. It shows the phenomenon of "light saturation," typical of true photosynthesis (Vol. II, Chapter 28) and occurring in the same region (~40,000 lux) of intensities. The occurrence of light saturation shows that Hill's reac- tion includes a nonphotochemical process of limited velocity, in addition to the photochemical process proper. We know now that different enzy- matic reactions may become "limiting" in photosynthesis under appro- priate conditions, as, for instance, if slowed down by a specific poison. 66 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 The reaction responsible for the inhibition of photosynthesis by cyanide probably is associated with the entry of carbon dioxide into the photo- synthetic apparatus (cf. Chapter 12). Since carbon dioxide does not participate in Hill's reaction, it appears natural that this reaction is not affected by cyanide. (Less easily understandable is the indifference of this reaction to hydroxylamine, which, according to page 313, is a specific poison for the oxygen-liberating enzymatic system in photo- synthesis.) Urethans, on the other hand, which inhibit photosynthesis in the " light-Hmited " as well as in the "enzyme-limited" state, also inhibit Hill's reaction. Ethylurethane, for example, causes a 50% in- hibition in a 0.6% solution, while phenylurethan produces the same effect in a concentration of only 4 X 10"^%. (The ratio of the efficiencies is 1500, as compared with 450 in true photosynthesis, cf. table 12. VIII). The temperature coefficient of hemoglobin oxidation by ferric oxalate in the presence of chloroplasts is 1.3 to 1.4 (while that of methemoglobin reduction is 1.8 to 1.9). This value is smaller than the temperature coefficient of photosynthesis in the light-saturated state ( > 2) (cf. Vol. II, Chapter 31), but figure 6 shows that the conditions of Hill's measurements were not those of complete light saturation. These results cannot be fully appreciated in the present chapter, because the corresponding relationships in photosynthesis will first be discussed in chapters 12, 28 and 31. The assumption that Hill's reaction represents "one half of complete photosynthesis" (photochemical oxi- dation of water, with ferric oxalate as a substitute oxidant taking the place of carbon dioxide), seems to be in agreement with most observations (the most notable exception being the insensitivity of Hill's reaction to hydroxylamine). If this interpretation is correct, it means that broken or dried chloroplasts retain an important part of their normal photo- catalytic capacity — they can still produce oxygen from water in light. However, they are not able any more to transfer hydrogen to carbon dioxide as acceptor, and thus cannot synthesize organic matter. In chapter 7, we shall discuss several alternative theories of the primary photochemical reaction in photosynthesis — some envisaging a direct participation of carbon dioxide in this reaction, others interpreting it as a photoxidation of water, with the hydrogen being first transferred to an unknown intermediary acceptor. Hill's experiments fit best into this second picture. If, in living cells, the hydrogen atoms find their way from the primary acceptor to carbon dioxide, with the help of a nonphotochemical enzymatic apparatus, it appears plausible that this apparatus may be destroyed by drying or crushing the cells, while the photochemical mechanism remains more or less intact. A recent observation of Frenkel (cf. page 204) makes it probable (but by no means certain) that the first transformation of carbon dioxide EXPERIMENTS WITH CHLOROPHYLL PREPARATIONS 67 in photosynthesis takes place outside the chloroplasts. This may explain why isolated chloroplasts are unable to use carbon dioxide as oxidant, even if they are capable of oxidizing water in light. Attempts to take the photosynthetic mechanism apart in order to find out how it works have been for a long time as unsuccessful as the proverbial farmer's attempt to get at the source of golden eggs by killing the goose which laid them. Hill's experiments represent the first step forward in this direction, and their continuation appears of great interest. (A confirmation of his results was contributed by French and coworkers, 1942.) They support the conclusion, derived by van Niel and Gaffron from experiments with bacteria and anaerobically treated algae (c/. Chap- ters 5 and 6), that the two partial processes of oxygen evolution and car- bon dioxide reduction are largely independent, and can be investigated separately. While van Niel and Gaffron showed that it is possible to substitute in photosynthesis other redudants for water, Hill's observa- tions indicate that ferric salts can be substituted for carbon dioxide as oxidants in this process. Hill's original experiments with leaf and yeast extracts, as well as the earlier qualitative observations of Friedel and Molisch (in which, too, leaf extracts in glycerol or water were found necessary to bring about the evolution of oxygen by leaf powders) make it seem probable that these extracts contain organic oxidants which can be used to oxidize water, in the presence of illuminated chloroplasts, instead of ferric oxalate. In the case of leaf extracts, the oxidants may well be identical with the intermediate hydrogen acceptors in true photosynthesis. It would be important to identify these oxidants by systematic analysis. One can ask whether Hill's reaction is similar to true photosynthesis from the point of view of conversion of light into chemical energy. Ferric salts are much stronger oxidants than carbon dioxide. At pH 8, where Hill's reaction proceeds most easily, the potential of an oxygen electrode is — 0.75 volt and thus almost equal to the normal potential of the nonassociated ferri-ferro system ( — 0.77 volt). However, ferric oxalate solutions are strongly associated, and their potential is therefore much less negative. The reduction of potassium ferricyanide by ferrous oxa- late, cf. page 64, proves it to be > —0.49 volt. Therefore, the photo- chemical oxidation of water by ferric oxalate must lead to the conversion of a considerable amount of light energy into chemical energy — even if this amount is much smaller than that utilized in true photosynthesis. 2. Experiments with Chlorophyll Preparations Of all chemical components of plants, the only one which is clearly indispensable for photosynthesis is the green pigment chlorophyll. Ingen-Housz, in 1779, established that only green parts of plants improve 68 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 the air in light. Whenever red, brown or blue cells have been found capable of photosynthesis, it could always be proved that they, too, contain the green pigment, even though its color may be masked by carotenoids or anthocyanins. It was natural therefore that attempts to repeat photosynthesis outside the plant have centered on chlorophyll preparations. The re- sults have been disappointing, and we need only devote a few lines to these experiments. Nobody has ever claimed to have achieved photosynthesis by illuminating a chlorophyll solution in an organic solvent in the presence of carbon dioxide; but a few- negative experiments on this subject have been pubUshed by von Euler (1909). Usher and Priestley (19062) have asserted that chlorophyll films on gelatin produce hydrogen peroxide and formaldehyde when exposed to light and carbon dioxide. But this claim, although supported by Schryver (1910), was discredited by Ewart (1908), von Euler (1909), Schiller and Baur (1912), Warner (1914) and Wager (1914), who proved that, although some formaldehyde can be found after the illumination of chlorophyll films in air, it originates in the oxidation of chlorophyll and not in the reduction of carbon dioxide. Willstatter and Stoll (1918) asserted that no formaldehyde is produced at all, if pure chlorophyll preparations are used. Chodat and Schweizer (1915) claimed the formation of formaldehyde and hydrogen peroxide by the illumination of chlorophyll precipitated on calcium carbonate; but Willstatter and Stoll (1918) failed to confirm this claim. It was thought by Willstatter and Stoll that chlorophyll is contained in the leaves in the colloidal state; they therefore carried out some experiments on the photosynthetic activity of colloidal chlorophyll solutions in water, with completely negative results. They also tried the addition of peroxidase (on the assumption that the completion of photosynthesis requires the decomposition of a peroxide), but without success. Knoll, Matthews and Crist (1938) have described an oxygen evolution caused by the addition of catalase to illuminated aqueous solutions of sodium chlorophyllide and carbonate, but details of this experiment have never been pubUshed. We shall find in chapter 14 proofs of the existence in the plant cells of a chlorophyll-protein complex. The preservation of this complex may be necessary to maintain the photosynthetic capacity of chlorophyll. Different methods to extract the chlorophyll-protein complex from leaves have been perfected, and these extracts have been found to possess some of the properties of the chlorophyll in the leaf, e. g., its absorption spectrum, chemical stability and fluorescence. Nevertheless, they, too, lack photosynthetic capacity (c/. Smith 1938). Eisler and Portheim (1923) claimed that artificial chlorophyll protein complexes (prepared by adding horse serum to chlorophyll solutions) were able to reduce carbon dioxide and liberate oxygen in light, but their methods were crude, and the promised detailed publication has not materialized. The incapacity of chlorophyll-protein complexes to bring about photosynthesis appears natural if we remember that even isolated chloroplasts maintain, at the utmost, only a vestige of their normal photosynthetic activity. The question to ask about chlorophyll prepara- PHOTOCHEMICAL OXIDATION OF WATER 69 tions is not whether they are capable of complete photosynthesis, but whether they too, retain some properties reminiscent of the part which chlorophyll plays in photosynthesis. As shown in chapter 3, this part is the utilization of light energy for hydrogen transfer against the gradient of chemical potential. Chlorophyll may achieve this either by a purely physical transfer of energy to a cellular oxidation-reduction system, or, more probably, by direct chemical participation in such a system. Consequently, what we ask is whether chlorophyll in vitro forms a reversible oxidation-reduction system and, if it does, whether the oxi- dizing capacity of its oxidized form, or the reducing capacity of its reduced form (or both) are enhanced by the absorption of light. Indications that chlorophyll in vitro actually possesses the properties of a light-activated oxidation-reduction catalyst, have been found by Rabinowitch and Weiss (1937) in experiments which shall be discussed in chapter 18. Some observations of Baur (1935), Baur and Fricker (1937), and Baur, Gloor and Kiinzler (1928), which point in the same direction, mil be described later in the present chapter (page 90). These interesting, but as yet inconclusive results are the only indications that chlorophyll outside the cell does retain certain of the properties which make it "the most important single organic compound on earth" as long as it is contained in living plant cells. B. The Photochemical Oxidation of Water * We now leave the living cell and the products obtained from it and consider nonbiochemical systems whose behavior is of interest from the point of view of artificial photosynthesis. The essence of photosynthesis is the reduction of the oxidant of an oxidation-reduction system of a higher potential (carbon dioxide-carbo- hydrate) by the reductant of a system of a much lower potential (oxygen- water), with light supplying the necessary energy. The differ- ence in total internal energy between the substrates and products of photosynthesis is 112 kcal per gram atom of carbon; the difference in free energy is a few calories larger {of. Table 3.V). Complete artificial photosynthesis should bridge this whole gap at once. However, all experiments which help to narrow it, may be considered as helpful partial solutions. The bridging may begin at either end or in the middle. It may include photochemical or nonphotochemical reactions likely to bring the two reacting systems closer together. Nonphotochemical reactions cannot contribute to the bridging of the energy gap; but they can make the solution easier, by substituting catalytic reactions with low activation barriers for reactions with the same net heat effect, but with a larger energy of activation. * Bibliography, page 95. 70 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 The experiments on isolated chloroplasts, described on page 63 et seq., as well as considerations based on bacterial metabolism, (cf. Chapter 5), make it feasible that one (and perhaps the only) photo- chemical reaction in photosynthesis may be the transfer of hydrogen from water to an intermediate acceptor. Consequently, in our search for a model of photosynthesis in vitro, we are concerned, in the j&rst place, with the photochemical oxidation of water by substances thermodynami- cally incapable of achieving this oxidation in the dark. The liberation of oxygen from water can occur by "self-oxidation" (dismutation), light (4.3) H2O + H2O > 2 H2 + O2 - 137 kcal or light (4.4) H2O + H2O > H2 + (H202)aq. - 91 kcal or (if hydrogen is taken over by an "acceptor") by an oxidation- reduction: light (4.5) H2O > I O2 + {2H } or light (4.6) 2 H2O y H2O2 + { 2H } where brackets indicate acceptor molecules. Reactions (4.3) to (4.6) can be brought about by direct absorption of ultraviolet light by water, or they can be sensitized. If, in (4.5) and (4.6), the acceptor itself is the light-absorbing species, the reaction is a photoxidation of the acceptor, rather than a true photocatalysis, and can only be called "sensitized" in the wider sense defined on page 56. For the sake of simplicity, we do not speak here of the possibility that an acceptor can be provided also for hydroxyl radicals or oxygen atoms, so that the primary products of oxidation will be complexes of the type {OH}, {OH} 2 or {O2}, rather than free molecules of oxygen or hydrogen peroxide (cf. Chapter 11). Any of reactions (4.3) to (4.6) can provide an appropriate first step towards artificial photosynthesis. To complete the process, carbon diox- ide must be reduced by hydrogen or by the hydrogenated acceptor { H } . In addition to the photodismutations, (4.3) and (4.4), and the phot- oxidations, (4.5) and (4.6), we shall also consider here the "photaut- oxidation": light (4.7) H2O + I O2 > (H202)aq. - 23 kcal (The term " autoxidation " for "oxidation by molecular oxygen" is ugly, but has come into general use.) This reaction could be useful as a first step in artificial photosynthesis only if hydrogen peroxide were able to reduce carbon dioxide. This question will be discussed on page 79 and answered in the negative. We have nevertheless added (4.7) to the other forms of photochemical oxidation of water, because this reaction MERCURY-SENSITIZED DECOMPOSITION OF WATER 71 must be considered whenever the oxidation of water takes place in the presence of air. 1. Decomposition of Water in Ultraviolet Light The direct photochemical decomposition of water into hydrogen and oxygen according to equation (4.3) was described by Coehn (1910) and Coehn and Grote (1912). A " photostationary state" is established in ultraviolet-illuminated water vapor, with light accelerating both its decomposition and the recombination of hydrogen and oxygen (Berthelot and Gaudechon 1910). The decomposition according to (4.4), i. e., with the formation of hydrogen peroxide, was discovered by Thiele (1908) and Kernbaum (1909). Tian (1916) suggested the existence, in ultraviolet-illuminated liquid water, of a stationary state involving photochemical formation and decomposition of hydrogen peroxide. If oxygen is present, all hydrogen formed by (4.4) is taken away, making it possible for hydrogen peroxide to accumulate, and the net effect is a peroxide formation according to equation (4.6). In these investigations, a mercury arc was used, and the decomposition was caused mainly by the first resonance line of mercury (189 mp), which is strongly absorbed by quartz walls and only weakly absorbed by water. A more rapid decomposition can be achieved by means of a hydrogen discharge tube with a fluorite window, as used by Terenin and Neujmin (1934, 1935, 1936). The active wave lengths are 130-140 m/i, which fall into the second absorption band of water. In this region, water decomposes into OH* + H (the asterisk indicating electronic excitation), as proved by the emission of OH bands in fluorescence. The primary process in the first absorption band of water, situated below 178-179 m^, may be either: H2O* -^ OH + H, or H2O* ^ H2 + O (Goodeve and Stein, 1931) or (in the liquid state) : H2O* + H2O -* H2O+ + H2O- The direct photochemical decomposition of water solves a large part of the diflficulties involved in artificial photosynthesis (it accumulates energy and liberates oxygen), but it does not solve all of them, because neither hydrogen molecules nor hydrogen atoms prove capable of reducing carbon dioxide. The reaction between molecular hydrogen and carbon dioxide will be discussed in more detail on page 83; as to atomic hydrogen, Harteck (1933) found that the admission of hydrogen atoms to carbon dioxide gas does not produce more than traces of formaldehyde. 2. Mercury-Sensitized Decomposition of Water We now go over to sensitized photodecompositions of water. The first extension of the photochemically active range towards the visible can be achieved by using mercury vapor as a sensitizer. 72 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 Wood (1925) found that water vapor containing mercury decomposes when illumi- nated with the resonance Une (253.6 mn) of mercury. Senftleben and Rehren (1926) investigated the reaction more closely by measuring the heat conductivity of the illumi- nated mixture. In a closed vessel, the illumination leads to a photostationary state. Ga viola and Wood (1928) have given spectroscopic proofs of the presence in this state of free hydroxyl radicals (OH) and of mercury hydride molecules (HgH). According to Beutler and Rabinowitch (1930) these are the primary products of the reaction: light (4.8) (Hg)g + (H20)g V (HgH)g + (OH)g - 95 kcal dark However, through the recombination of OH radicals, the dissociation of the unstable HgH molecules, and other processes competing with (4.8), numerous other products are formed, among others, H2, O2, H2O2, HgO and free H atoms. Melville (1936) found that the uncondensable fraction of the illuminated Hg/H20 mixture consists mainly of hydrogen, and suggested that the equivalent quantity of oxygen is bound in mercurous oxide. Reaction (4.8) is of the type (4.6), a photoxidation of water by mercury, with the transfer of only one hydrogen atom. This reaction is possible, despite the high energy of dissociation of water into H and OH (109 kcal), because the absorption of the line 253.6 m/j. brings mercury into a state ('Pi) with an excess energy of 112 kcal per mole. This is the first example of how hght energy can be utilized for hydrogen transfer against the gradient of chemical potential. The efficiency of conversion of light energy into chemical energy in reaction (4.8) is 90%. However, most of this energy is dissipated by secondary processes. If the hydroxyl radicals decompose into water and oxygen, while the mercury hydride decom- poses into mercury and hydrogen, the ultimate result is that a quantum equivalent to 112 kcal per mole, has produced the reaction § H20^ ^ H2 -I- i O2, with a heat effect of 27 kcal, corresponding to the conversion of only 25% of hght energy into chemical energy. This result is actually achieved when the reaction between Hg and H2O is carried out in a streaming system. Bates and Taylor (1927) found that the decom- position products contain 73% H2 and 27% O2. The deficiency of oxygen may be due to the incomplete decomposition of hydrogen peroxide. 3. Sensitization of Water Decomposition by Solids (ZnO and AgCl) The photochemical decomposition of water can be extended further towards longer waves by the use of solid sensitizers, e. g., zinc oxide and silver chloride. However, it is not certain whether the sensitization by zinc oxide goes beyond sensitized photautoxidation of water, according to equation (4.7). This reaction was discovered by Baur and Neuweiler (1927). After oxygen-containing water has been shaken with zinc oxide for 10-15 hours in full sunlight, the liquid is found to contain about 1 X 10~^ mole per liter of hydrogen peroxide. According to Baur and Neuweiler, no peroxide is formed in air-free solutions; the oxidant is thus apparently molecular oxygen. The active light belongs to the near ultraviolet (the absorption limit of ZnO lies at 380 m/x), and zinc oxide seems to act as a true photocatalyst, promoting reaction (4.7) without participating in it. SENSITIZATION OF WATER DECOMPOSITION BY SOLIDS 73 Richardson (1939) found that the quantum yield of this reaction is of the order of 0.1 in weak light, and less at the higher light intensity. The rate of peroxide formation shows a "light saturation" similar to that occurring in photosynthesis, proving that the photochemical process is coupled with a thermal process of limited velocity. (For example, the photochemical decomposition of water adsorbed at the surface of ZnO may be followed by the desorption of the reaction products.) The mechanism of this reaction is unknown, but we may assume that the primary process is the decomposition of adsorbed water into OH and H, made possible by the large heats of adsorption of OH and H on zinc oxide. light (4.9) H2O (adsorbed) ^ H (adsorbed) + OH (adsorbed) dark In order to enable reaction (4.9) to occur in the near ultraviolet (that is, with light quanta of about 78 kcal per einstein, one einstein being 6 X 10^^ quanta), the combined heat of adsorption of the radicals must be at least 35 kcal larger than that of water. The assumed primary reaction (4.9) is of the type postulated by van Niel for photosynthesis (c/. Eq. 7.1) — photochemical decomposition of water, with zinc oxide serving as acceptor for both hydrogen atoms and hydroxyl radicals. To explain why hydrogen peroxide is formed only in presence of oxygen, we may assume that oxygen molecules snatch away the adsorbed hydrogen atoms, thus preventing the reversal of reaction (4.9), and leaving to the hydroxyl radicals no other way but to recombine to "biradicals" H2O2. In this way, the primary photo- chemical decomposition of water is again reduced to a " photautoxida- tion," according to equation (4.7), with its comparatively small energy conversion. The question arises as to whether the back reaction in (4.9) is com- pletely effective in absence of oxygen, or whether some hydrogen atoms succeed in recombining to hydrogen molecules, causing an equal num- ber of hydroxyl radicals to recombine to H2O2 and giving the net effect of sensitized water decomposition (4.4), a result much more significant from the point of view of artificial photosynthesis than the photautoxi- dation (4.7). Successful achievement of reaction (4.4) would leave us with the problem of carbon dioxide reduction by molecular hydrogen as the final stage of artificial photosynthesis — a reaction which re- quires no additional conversion of energy. True, we do not yet know how to conduct it in a reversible way, without spending considerable energy on activation; but we shall see in chapter 5, that the so-called "Knallgas bacteria" reduce carbon dioxide to carbohydrates in the dark, by means of molecular hydrogen, with up to 40% of the theoretical yield. 74 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 It is thus an interesting question whether hydrogen peroxide can be obtained by the illumination of zinc oxide suspensions in the absence of oxygen. Baur and Neuweiler (1927) gave a negative answer to this question, while Yamafuji, Nishioeda, and Imagawa (1939) asserted that a small quantity of hydrogen peroxide is formed even if all oxygen had been removed. Vogel (1863), Eder (1906) and Sichling (1911), found that water over a silver chloride precipitate evolves oxygen if exposed to daylight, and this was confirmed by Baur (1908) and Baur and Rebmann (1921). However, the amount of gas evolved is small, and the reaction soon stops. In contrast to water decomposition by zinc oxide, this reaction is probably not a true photocatalysis, but a photoxidation of water hy silver chloride: light (4.10a) Ag+Cl- > Ag + CI (4.10b) CI + H2O > Cl-aq. + H+aq. + OH (4.10c) OH > I H2O + i O2 light (4.10) Ag+Cl- + h H2O > Ag + Cl-aq. + i O2 + H+aq. - 25 kcal This reaction deserves attention because of the apparent conversion of a large part of light energy into chemical energy. If brought about by a single quantum at 400 mn, it would lead to the conversion of about 35% of absorbed light energy, a yield not attained by any other known photochemical reaction in visible light. It is, however, not certain whether the observations of Vogel, Baur, and Rebmann are correct, whether interpretation (4.10) applies to them and, if it does, what the quantum yield of this reaction may be. 4. Photoxidation of Water by Cations We have considered mercury vapor and ionic crystal powders as sensitizers which enable the photochemical liberation of oxygen from water to occur in the medium or near ultraviolet. A third group of such sensitizers is found in dissolved cations. The only cation whose capacity for photochemical water oxidation has been demonstrated by experiments, is the eerie ion, Ce++''"+. The normal oxidation-reduction potential of the system Ce+~^+-Ce"''+"'' is close to — 1.5 volt, that is, far below that of the oxygen electrode. Consequently, Ce++++ ions liberate oxygen from water even in the dark; but this process is slow. Baur (1908) noticed that this oxidation can be acceler- ated by light, and Weiss and Porret (1937) found that oxygen is produced with a quantum yield as high as 0.5. The absorption bands of the eerie ions extend from the far ultraviolet to the blue-violet region of the visible spectrum. (The molar extinction coefficient is about 150 at 400 mn.) It is likely that absorption every- PHOTOXIDATION OF WATER BY CATIONS 75 where in this region is effective in bringing about the reaction Ught (4.11) Ce++++ + h H2O > Ce+++ + H+ + i O2 Since the oxidation potential of eerie ions is more negative than that of molecular oxygen, reaction (4.11) does not convert light into chemical energy. It is known or suspected (see Rabinowitch 1942) that the light absorption by many other cations also leads to a primary oxidation of water, probably, according to the equation: Ught (4.12) M+-H20 )-M-H20+ A. and L. Farkas (1938), who suggested this primary process, pointed out that the final state in (4.12) is unstable, and is terminated either by a return of the electron to water: (4.13) M -1120+ > M+-H20 ("primary back reaction") or by a chain of transformations of the ion H2O+, e. g.: f H2O+ + OH- > H2O + OH (4.14) OH 4- OH >H202 I H2O2 > H2O + \ O2 At any stage of (4.14), the reaction may be reversed by a "secondary" back reaction, that is, the reoxidation of M by hydroxyl, peroxide or oxygen. In the case of the eerie ions, reaction sequence (4.14) has a good chance of occurring in preference to a primary or secondary back reaction. With other cations, no oxygen evolution has been observed upon illumi- nation (at least, as far as attention has been paid to this point) and this can be taken as an indication that the back reactions are more probable than the oxidation chain (4.14). The cause most probably lies in the relative energies of the different states involved in the process. For eerie ions, oxidation releases more energy than the return into the initial state, and the probability of the metastable state Ce+++- H2O+ undergoing a development according to (4.14) is correspondingly high. In the case of other ions (ferric ions, for example) much more energy can be gained by the transformation of the metastable complex (Fe++-H20+) back into Fe+++-H20, than by the completion of oxidation according to (4.15) Fe++-H20+ + (OH-)aq. > Fe++-H20 + \ H2O2 > • • • (as in 4.14) The reaction with ferrous oxalate observed by Hill and described on page 63 is probably of type (4.12) - (4.14), although it requires sensi- tization (by chloroplasts). In this case, the oxygen liberation occurs with a considerable yield, despite the unfavorable position of the energy 76 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 levels. This must be due to an enzymatic mechanism preventing a primary back reaction of type (4.13), and accelerating the completion of the oxidation process. A secondary back reaction (reoxidation of ferrous oxalate by oxygen) actually was observed by Hill, but this reaction is comparatively slow and does not prevent a partial escape of oxygen into the atmosphere, or the fixation of oxygen by hemoglobin. A "hidden," i. e., instantaneously reversible, photochemical oxidation* of water by illuminated cations has also been postulated in the explanation of two forms of the "Becquerel effect," the "photovoltaic" effect, in which the illumination of an oxide-, halide-, or dyestuff-coated electrode causes a change of potential, and the " photogalvanic effect," in which a similar change is induced by the illumination of an electrolyte in contact with an inert electrode. Both effects must be caused by short-lived chemical changes in the surface layer of the electrode, or of the electrolyte. The most probable change is the displacement of an oxidation-reduction equilibrium (c/. Rabinowitch 1940). It has been suggested by Baur (1918, 1919) for the photogalvanic effect, and by Audubert (1934) for the photovoltaic effect, that one partner in the oxidation-reduction equi- librium in aqueous electrolytes is water (the other being the specific photosensitive component — dyestuff, salt, or oxide). However, Svensson (1919) found no evolution of oxygen or hydrogen in illuminated photo- galvanic systems; neither was he able to observe the formation of hydro- gen peroxide or ozone. Baur suggested therefore, that the decomposition of water remains in a "hidden stage," that is, stops short of an actual evolution of oxygen or hydrogen, because of the efficiency of the back reactions. 5. Photoxidation of Water by Dyestuffs In connection with the problem of photosynthesis, the photochemical oxidation of water by dyestuffs, either in consequence of a reduction of the dyestuff itself, or by true sensitization as in Hill's experiments, is of great interest. Many cationic dyestuffs are oxidants, capable of being reduced to so-called leuco dyes. Their oxidation-reduction potentials are much too low to enable them to oxidize water in the dark— the strongest known organic oxidants have potentials of -0.4 volt at pH 7, which is still 0.4 volt above the potential of the oxygen electrode at the same pH. How- ever, the absorption of a light quantum, even of a "red" light quantum of about 600 mM, corresponding to 45 kcal per einstein, should make the free energy of reaction (4.16) negative. (4.16) D* • H2O > DH2 + ^ O2 In other words, light-excited dyestuff molecules contain enough energy PHOTOXIDATION OF WATER BY DYESTUFFS 77 to bring about the oxidation of water. However, the primary process in dyestuff solutions is excitation, and not an electron transfer from water to dyestuff (as assumed above for the inorganic cations, Ce++++ and Fe+++). In this case, the oxidation of water must be brought about by a secondary electron transfer from water to the excited dyestuff ion. Processes of this kind are known to occur between excited dyestuff ions and other electron donors, as, for example, ferrous ions. As shown by Weber (1931), Weiss (1935) and Rabinowitch (1940), excited thionine or methylene blue cations oxidize ferrous ions, even though in the dark the reaction proceeds in the opposite direction, in accordance with the posi- tions of the normal oxidation-reduction potentials: light (4.17) Thionine + 2 Fe++ , leucothionine + 2 Fe+++ For example, at pH 3, the normal potential of the system thionme- leucothionine is approximately - 0.3 volt, while that of the system Fe+++-Fe++ is approximately - 0.75 volt. Nevertheless, in light, ferrous ions are oxidized by thionine ions, and it takes the system several seconds to come back to equilibrium in the dark. The slowness of the back reaction may be attributed to a peculiar relation between AH and AF in reaction (4.17). The normal potentials indicate that the free energy of this reaction is strongly positive; but its heat effect probably is negative. The free energies of hydrogenation of most organic systems, including thionine, are less negative than the total energies— that is, the reduced state has a smaller entropy. The relation is reversed in the case of the reduction of ferric ions by hydrogen. Consequently, the reversal of reaction (4.17) is an endothermal reaction, and as such cannot proceed with a high velocity. This more or less accidental circumstance is the explanation why, in the thionine-iron system, the shift in the oxidation-reduction equilibrium by light, which usually is hidden by rapid back reactions, becomes easily observable, even though it remains transient. It can be asked whether, in the absence of ferrous ions, a reversible reaction does not occur between dye and the solvent (even if with a smaller quantum yield and with a more rapid back reaction). Such "hidden" oxidation-reductions have been held responsible for the photo- voltaic effect of dyestuff-coated electrodes by Audubert and coworkers, Hoang Thi Nga (1935) and Stora (1935, 1936, 1937). In the same way, the directly observable reversible reduction of thionine by ferrous ions has been shown by Rabinowitch (1940^) to produce a strong photo- galvanic effect. If oxygen is present in aqueous dyestuff solutions, one could expect some of the leuco dye formed by the oxidation of water to be reoxidized 78 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 by oxygen, thus leaving an equivalent quantity of oxidized water. In other words, we could expect the occurrence of a dyestuff-sensitized formation of hydrogen peroxide, according to equation (4.14), by the mechanism which was contemplated above in the case of the zinc oxide sensitization. Blum and Spealman (1933) have in fact claimed the formation of hydrogen peroxide in illuminated fluorescein solutions, and Yamafuji and coworkers (1938, 1939) have also obtained positive hydro- gen peroxide tests with illuminated solutions of chlorophyll, eosin and hematoporphyrin. While the dyestuff-sensitized " photautoxidation " of water, indicated (but by no means proved) by the experiments of Blum and Spealman and Yamafuji, appears explicable according to the hypothesis of "hidden" photochemical oxidation-reduction reactions, the alleged formation of hydrogen peroxide in oxygen-free dyestuff solutions is, if true, a much more remarkable phenomenon. Yamafuji and coworkers (1938-1939) have asserted that illuminated tissue extracts, dyestuff solutions, and zinc oxide suspensions also produce hydrogen peroxide in the absence of oxygen, although much less than under aerobic conditions. As discussed on page 73, this result (if true) would indicate a sensitized decomposition of water into hydrogen and hydrogen peroxide, according to equation (4.4). In the case of zinc oxide, the energy of two ultraviolet quanta (about 75 kcal per einstein) is sufl[icient to bring about reaction (4.4); but two quanta of visible, particularly red light (60-40 kcal per einstein), are insufficient for this purpose. This makes us doubt whether oxygen (or other oxidants) have actually been eliminated in the experiments of Yamafuji and coworkers. The problem is too important to allow the acceptance of their results without further confirmation. Baur and Reb- mann (1921) have attempted to achieve a sensitized photolysis of water in visible light, using uranyl salts, quinine, eosin, rhodamine and other sensitizers, but could obtain no traces of oxygen. C. The Chemical and Photochemical Reduction OF Carbon Dioxide * If one primary photochemical process in photosynthesis is the hydro- gen transfer from water to an intermediate acceptor, the reduction of carbon dioxide may be brought about either by a second photochemical reaction, or by a "dark" enzymatic reaction with the reduced primary hydrogen acceptor (cf. Chapter 7). We are therefore interested, in the present chapter, both in photochemical and nonphotochemical reduction of carbon dioxide in vitro. * Bibliography, page 96. CHEMICAL REDUCTION OF CARBON DIOXIDE 79 1. Chemical Reduction of Carbon Dioxide So far, carbon dioxide has been reduced in vitro only by means of the strongest available reductants, or at high temperatures. Fenton (1907) described the reduction of carbon dioxide to formaldehyde by magnesium, while Bredig and Carter (1914) have achieved the reduction of carbon dioxide to formic acid by means of hydrogen and palladium. Reactions of this type are of no use in artificial photosynthesis. Imagine, for example, that we would begin by reducing carbon dioxide with magnesium, and — to provide a similar start at the other end of the reaction chain — oxidize water with fluorine. We would thus obtain magnesium oxide and hydrogen fluoride as the first reaction products. Bringing these two compounds together wall lead to the formation of magnesium fluoride, and the completion of the reaction cycle would now require the photochemical dissociation of this salt into metal and halogen, a task considerably more difficult than photosjm thesis itself. One comparatively mild reductant which has been credited with the capacity to reduce carbon dioxide, was hydrogen peroxide. Kleinstuck (1918) working in Wislicenus' laboratory, found that phosgene, diphenyl carbonate and carbonate ions, can be reduced to formaldehyde by heating with hydrogen peroxide under pressure. Wislicenus (1918) corrected the results, stating that the reduction product obtained from alkah carbonates (or bicarbonates) and hydrogen peroxide is formic acid, and suggested that the process involves the formation of percarbonic acid as an intermediate: OOH (4.18a) H2O2 + H2CO2 > OC + H2O OH (4.18b) H2CO4 > H2CO2 + O2 (4.18) H2O2 + H2CO3 > H2O + O2 + H2CO2 - 44 kcal Thunberg (1923), while faihng to confirm most of Kleinstiick's results, claimed that formaldehyde can be obtained by boiUng lead carbonate with hydrogen peroxide. Thunberg and Weigert (cf. page 70) have used these results as basis for a theory according to which photosynthesis consists of a photochemical decomposition of water into hydrogen and hydrogen peroxide, and a nonphotochemical reduction of carbon dioxide by the latter two compounds: light (4.19a) 2 H2O > (H202)aq. + H2 - 79 kcal (4.19b) CO2 + H2 + (H202)aq. > {CH2O} + O2 + H2O - 33 kcal (4.19) CO2 + 2 H2O > {CH2O! + H2O + O2 - 112 kcal However, not only reaction (4.18), in which hydrogen peroxide alone reduces carbon dioxide, but even reaction (4.19b), in which hydrogen peroxide is assisted by hydrogen, is endothermal to such an extent that it cannot occur spontaneously at low temperatures. Some doubts may be entertained as to the reliability of Thunberg's experiments; but, even if they are correct, they do not point a way by which carbon dioxide can be reduced at low temperatures. The energy accumulated in the oxidation of water to peroxide by oxygen (23 kcal per mole) is much too small to enable the product to reduce carbon dioxide without an external supply of energy. The oxidation-reduction potential of the system O2-H2O2 (- 0.27 volt at pH 7) is much too negative to bring about the reduction of the system H2CO3-H2CO2, or H2CO3-H2CO, whose potentials (at the same pH) are above + 0.4 volt (cf. Table 9,1 V). 80 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 The first reaction in Wislicenus' scheme (4.18), is feasible, because all peroxides have approximately the same energy content; but the decomposition of percarbonic acid into formic acid and oxygen is as impossible as a spontaneous monomolecular decomposition of hydrogen peroxide into hydrogen and oxygen. Peroxides decompose spontaneously only by himolecular dismutation into oxide and oxygen {cf. Chapter 11). If we replace (4.18b) by such a decomposition, the net result will be merely a carbonate- catalyzed decomposition of hydrogen peroxide. Thus, none of the known chemical methods of reduction of carbon dioxide appears significant from the point of view of artificial photo- synthesis. However, it seems probable (c/. Chapter 8) that the immediate substrate of reduction in nature is not free carbon dioxide at all, but carbon dioxide incorporated, by enzymatic catalysis, into a large organic molecule, probably with the formation of a carboxyl group: (4.20) RH + CO2 > RCOOH Once association (4.20) has taken place, the reduction of carbon dioxide to carbohydrate can be replaced by the reduction of the carboxyl group, RCOOH, to the carbinol group RCH2OH: (4.21a) CO2 + RH > RCOOH (4.21b) RCOOH + 4 H > RCH2OH + H2O (4.21c) RCH2OH > RH + { CH2O 1 (4.21) CO2 + 4 H ^ {CH2O) + H2O Furthermore, the reduction of one molecule of acid to one molecule of carbinol can be replaced by the reduction of two molecules of acid to two molecules of aldehyde, and the dismutation of the latter compound (Cannizzaro reaction) : (4.22a) 2 RCOOH + 4 H > 2 RCHO + 2 H2O (4.22b) 2 RCHO + H2O > RCOOH + RCH2OH (4.22) RCOOH + 4 H > RCH2OH + H2O Thus, the chemical problem of carbon dioxide reduction to a carbo- hydrate, can be replaced by the problem of the reduction of a carboxylic acid to an aldehyde. The methods by which this reduction is achieved in organic chemistry are, however, as violent as those used for the reduction of carbon dioxide, i. e., they involve either very strong reduc- tants (sodium amalgam, or hydrogen and palladium under high pressure), or high temperatures (dry distillation of calcium salts). Thermody- namical constants show that there is not much difference between the energies of reduction of carbon dioxide and carboxyl (cf. Table 9. IV); but the substitution of a large molecule of a carboxylic acid for the small molecule of carbon dioxide may decrease the activation energy, and thus make the reduction easier. The free radicals, HCO2 and H3CO2, which must arise as intermediates in the reduction of carbon dioxide if the CARBON DIOXIDE REDUCTION IN ULTRAVIOLET LIGHT 81 hydrogen atoms are transferred one by one, have the full energy of their unsaturated bonds; while the corresponding radicals derived from large organic molecules can often be stabilized by resonance, and there- fore present much less of a barrier to a reversible reduction. We will revert to this function of free radicals in chapter 9 (page 233). At this point, we must state that we do not know of any reaction of carbon dioxide or of the carboxyl group in vitro, which could be called a reversible (or almost reversible) reduction of the C=0 double bond to a CH — OH single bond, and that a closer inquiry into the possibilities of such a reduction would be important for the study of artificial photosynthesis. 2. Decomposition and Reduction of Carbon Dioxide in Ultraviolet Light In describing the photochemical oxidation of water, we started w^ith the direct effects of ultraviolet light; similarl}^, we begin now with the nonsensitized photochemical decom- position of carbon dioxide by ultra- violet light. The spectrum of the molecule CO2 con- sists of discrete bands, from 200 m/x to 103 m/u; thus, the primary process is electronic exci- tation rather than photochemical decompo- sition. Figure 7 shows the extinction curves of the ions, CO3 and HCOs" in water. In this case, the primary process probably is an electron transfer from the ion to water: hi- (4.23) 3.0 2.5 2.0 1.5 ctI.O o HCOr-HzO ->nco3H20- 0.5 -0 5 that is, an oxidation of the carbonate and reduction of water. This is hardly an appro- priate initial step towards the reduction of the carbonate and oxidation of water. The effect of ultraviolet light on carbon dioxide was first observed by Chapman, Chadwick and Rams- bottom (1907) and Herschfinkel (1909). They found that carbon dioxide gas decomposes in light with an increase in pressure (i. e., probably into car- bon monoxide and oxygen), until a stationary state is reached. Berthelot and Gaudechon (1910) found that ultraviolet light accelerates both the dissociation of carbon dioxide and ■1.0 \ \ \ \ ^ \cor- HCGjN \ \ \ \ 160 180 240 260 200 220 Fig. 7. — Molar extinction curves of CO3 — and HCOa" in water (after Ley and Arends) . 82 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 the recombination of carbon monoxide and oxygen. The photostation- ary state: light (4.24) COo . CO + ^ O2 light was investigated by Coehn and Sieper (1916), Tramm (1923), Coehn and Spitta (1930) and Coehn and Maj^ (1934), with particular emphasis on the effect of moisture. They found that the decomposition has a maxi- mum at a certain moderate degree of drying, and is inhibited both by an excess and by a complete absence of water. The mechanism of this twofold effect of water remains unexplained. From the point of view of photosynthesis, it would be interesting if this effect were due to a photochemical reaction between carbon dioxide and water; but Thiele (1908) and Coehn and Sieper (1916) have asserted that this is not the case, and that the influence of water is a purely catalytic one. Also, Stoklasa and Zdobnicky (1910, 1911) and Stoklasa, Sebor and Zdobnicky (1912, 1913) found that no formaldehyde is formed by the illumination of carbonic acid in solution, while Baly, Heilbron and Barker (1921), Dhar and Sanyal (1925) and Mezzadroli and Gardano (1927) asserted that small quantities of formaldehyde can be obtained in this way. This was disputed by Baur and Rebmann (1922), Spoehr (1923) and Porter and Ramsperger (1925), who found the yield of formaldehyde to decrease with increasing purity of the reactants. How- ever, Baly, Davies, Johnson and Shanassy (1927) reiterated this organic matter is formed by illumination of carbonate solutions with ultraviolet light, claiming this time that it was not formaldehyde, but an undefined higher aldehyde. Mezzadroh and Vareton (1931) claimed that the yield of formaldehyde can be increased by preliminary ionization of carbon dioxide by emanation or electric discharges. It is difficult to judge whether the newer claims of Baly, Dhar, and Mezzadroli and coworkers deserve more confidence than the older ones. The positive test for formalde- hyde in illuminated carbon dioxide solutions, described by Joo and Wingard (1933), is anything but convincing, since it was obtained by Allison's "magneto-optic analysis," whose reliability is open to serious doubts. If it should be confirmed that traces of formaldehyde are formed by ultraviolet illumination of carbonate solutions, this phenomenon may be of some interest from the point of view of the origin of organic matter on earth. The first carbohydrates may have been formed, from carbonic acid and water, by the ultraviolet light, which reaches the higher layers of the atmosphere. Dhar and Ram (1933) have analyzed rain water and found (by iodine titration) 1.5 X 10"^ to 1 X 10"^% of formaldehyde, the larger values being obtained after long periods of sunshine. They suggested that the photochenaical formation of formaldehyde occurs at, CARBON DIOXIDE REDUCTION IN ULTRAVIOLET LIGHT 83 or even above the level where ozone is formed (about 50 km. above the surface), since no rays with wave lengths < 290 mn are available below this layer. From the point of view of artificial photosynthesis, or of natural photosynthesis under the present terrestrial conditions, it appears entirely irrelevant whether traces of formaldehyde can be formed by ultraviolet illumination of carbonate solutions or not. In dealing with photochemical reactions, it must be kept in mind that the energy available in one quantum, particularly a quantum of ultraviolet light, is much larger than the activation energy required for most, if not all chemical reactions. Thermal reactions take place when the energy of molecular vibrations, together with the collision energy, are just sufficient to bring the reacting molecules over the top of an "activation pass" in the many- dimensional relief map representing the potential energy of the reacting system as a function of its various configuration co-ordinates. This favors a uniform fate for all these molecules, which all drop into the same "potential valley." Activation by light absorption, on the other hand, often breaks molecules into free atoms and radicals, thus lifting the system onto a high energy plateau from which it can descend into many different valleys, corresponding to more than one set of reaction products. In the rearrangement of radicals formed from carbon dioxide and water by the absorption of quanta with an energy of 150 kcal per einstein, a few may err into the shallow potential trough of formaldehyde, and miss the opportunity to drop into deeper valleys, representing more stable configurations. In this way, traces of formaldehyde may be formed by an entirely accidental side reaction, which can have nothing in common with the highly specific and purposeful mechanism of photosynthesis. The same consideration applies to experiments in electric discharge tubes, irradiation with x-rays, and other treatments which break the molecules and afford the opportunity for the rearrangement of the broken pieces into all kinds of new patterns. The formation of formaldehyde by silent electric discharges and corona discharges in carbon dioxide — described by Losanitsch and Jovitschitsch (1897), Berthelot (1898, 1900), Lob (1905, 1906), Gibson (1908), Holt (1909), Moser and Isgarishev (1910) and Lunt (1925)— had for a wliile aroused much interest as an approach to the problem of artificial photo- synthesis. In our opinion, the only aspect of these observations which may conceivably be of importance for the understanding of photosynthesis, is the contribution of atmos- pheric discharges to the first synthesis of organic matter on earth. The results discussed above show that, even when the molecules of carbon dioxide and water are broken to pieces and allowed to recombine at random, the chance that they will form formaldehyde and oxygen is very small. The same also seems to be true for mixtures of carbon dioxide and hydrogen, despite the fact that in tWs case, the system CH2O and H2O represents as deep a potential trough as did the system CO2 + 2 H2 in its initial state. Thiele (1908) found no formaldehyde after the ultra- violet illumination of hydrogen-carbon dioxide mixtures, while Berthelot and Gaudechon (1910), Coehn and Sieper (1916) and Mezzadroli and Babes (1929) asserted that some 84 PEOCESSES OUTSIDE THE LIVING CELL CHAP. 4 formaldehyde is formed under these conditions. Stoklasa and Zdobnicky (1912, 1913) emphasized particularly the alleged photochemical production of formaldehyde from carbon dioxide in presence of nascent hydrogen. They suggested that this is the actual photochemical reaction in photosynthesis, and that nascent hydrogen can be formed in the plant by an enzymatic reaction. Since the reduction of carbon dioxide by hydrogen liberates a small amount of energy, while the dissociation of water into hydrogen and oxygen requires even more energy than photosynthesis itself; (4.25a) 2 H2O > 2 H2 + O2 - 137 kcal (4.25b) 2 H2 + CO2 > {CH2OI + H2O + 25 kcal (4.25) H2O + CO2 > {CH2O) + O2 - 112 kcal — it is obviously absurd to suggest that the first reaction is thermal and the second photochemical. 3. Sensitized Reduction of Carbon Dioxide This section describes the investigations in which "artificial photo- synthesis" has been claimed as an accomplished fact. Their usual technique was to illuminate carbon dioxide solutions in presence of different "sensitizers" and then search for traces of formaldehyde or other organic compounds as ardently as alchemists have searched for a grain of gold in the bottom of their crucibles. More often than not, no specific reductant was provided for the reduction of carbon dioxide, and the assumption that water acted as such was made without any attempt to confirm it by proving the liberation of oxygen. In our discussion of these experiments, we will endeavor to keep apart the two phenomena defined in the discussion of the sensitized oxi- dation of water: true photocatalysis (either of the decomposition of carbon dioxide, or of the reduction of carbon dioxide by a specific reductant) and photoreduction of carbon dioxide (or its derivatives) by the "sensi- tizer" itself. In 1893, Bach found that a solution of carbon dioxide and uranyl acetate reacts in light; uranium oxides are precipitated, and Bach thought that carbon dioxide might be reduced to formaldehyde. The same "sensitizers" (uranyl salts) were used by Usher and Priestley (1906) and Moore and Webster (1918). The last named authors thought the colloidal state of the sensitizer to be of particular importance; they obtained positive formaldehyde tests in illuminated carbonate solutions containing colloidal uranium and iron salts, and thought that these results afford an explanation of natural photosynthesis, since colloidal iron compounds are present in the chloroplasts (c/. Chapter 14, page 376). Apart from doubts concerning the correctness of the experimental results of Moore and Webster (c/. page 89), we must ask what happened in these experiments to the "sensitizers." Did they remain unchanged, SENSITIZED REDUCTION OF CARBON DIOXIDE 85 thus playing the part of true photocatalysts, or did they also serve as re- ductants? Of course, to reduce carbon dioxide by uranyl salts or ferrous salts would be an important success, since the oxidation-reduction po- tentials of these substances are far below the potential of the system CO2-H2CO. Still, this reduction would represent only one-half of photo- synthesis, the remaining half being the reduction of the oxidized cata- lyst (e. g., ferric iron) by water, leading to the liberation of oxygen (as in Hill's experiments with isolated chloroplasts). It is, however, improbable that Moore and Webster have achieved even that much, since Baur and Rebmann (1922), in attempts to repeat their experiments, have failed to observe any formation of formaldehyde, oxalic, glyoxalic or formic acid, not to speak of evolution of oxygen. (a) The Experiments of Baly and Dhar The subsequent development of the subject, in two long series of publications, one by Baly and coworkers in Liverpool (1927-1940) and the other by Dhar and coworkers in Allahabad (India) (1925-1933), brought many spectacular claims, but no convincing results. One is therefore tempted to dispense with their presentation altogether; but wishing the reader to be able to form his own opinion as to the validity of claims which have been repeated so persistently (lastly, in Baly's monograph. Photosynthesis, published in 1940), we will review in some detail the experiments on which these claims were based. Baly's experiments have attracted most attention, because of the reputation of the author, and of the comparatively large yields of organic matter which he and his coworkers claimed to have obtained in their first investigations. Baly, Davies, Johnson and Shanassy (1927), employed white powders (barium sul- fate or alumina) as sensitizers, and used ultraviolet light. Baly, Stephan and Hood (1927) went over to the use of colored powders (basic carbonates of nickel and cobalt), illuminated by incandescent lamps. Carbon dioxide was bubbled for several hours through illuminated vessels containing these powders suspended in water; the solution was then separated from the powder and evaporated. A gummy residue was obtained which gave some aldehyde and sugar reactions (reduction of Benedict's solution, MoUsch test; Rubner test; osazone formation). The carbonates soon lost their "catalytic" capacity; Baly attributed this to their oxidation by the oxygen produced by photo- synthesis (no direct test for oxygen production was ever attempted). The yield of "artificial carbohydrates," obtained by Baly and Davies (1927) was up to 75 mg. in two hours, in a vessel with a surface of 300 sq. cm., i. e. "about equal to the yield of natural photosynthesis on an equal area covered by vegetation." Baly and Hood (1929) found that the rate of "artificial photosynthesis" increased between 5° C. and 31°, and declined between 31° and 41°, Uke that of natural photosynthesis. Difficulties in reproducing these first promising results soon arose, and the next ten years were spent on attempts to prepare reliable catalysts. 86 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 In 1931, Baly reported two new methods for obtaining catalytically active prepara- tions: electrolytic precipitation of basic nickel or cobalt carbonates, and deposition of thorium oxide-" promoted " ferric or chromic oxide on alumina-coated kieselguhr (diatomaceous earth). However, Bell (1931) was unable to repeat these experiments, and the same disappointment was later experienced by Baly himself. In 1937, Baly announced that "two methods of preparing active catalysts have been now standard- ized." In 1939, he described the results obtained with these new catalysts; the whole development, from the "early investigations" to the "final achievement of photo- synthesis of carbohydrates," was reviewed in Baly's monograph, Photosijnthesis, in 1940. The two new methods of preparation of the sensitizers were: (a) the deposition of nickel oxide or cobalt oxide on kieselguhr; and (6) the precipitation of (unsupported) nickel oxide by the addition of potassium bicarbonate to a solution of nickel nitrate, and heating of the precipitate in vacuo. The first method, although more compMcated, had the advantage that the supported oxide layers did not dissolve in carbon dioxide- saturated water. In the investigation of Baly, Pepper and Vernon (1939), the surface potentials (f potentials) of the oxide-coated kieselguhr powders were measured by cataphoresis, and it was concluded that the coating consisted of three monomolecular layers. One molecule of thorium oxide was incorporated into the surface layer for each 24 molecules of nickel oxide. In the preparation of unsupported oxide, attention was directed to the avoidance of the adsorption of alkah, which, according to Baly, is the main source of trouble with oxide catalysts. The unsupported catalysts, too, were prepared with one molecule of Th02 as "promoter" for each 24 molecules of NiO. The successes obtained with the new preparations did not go beyond those achieved in 1927. Ten to 20 g. of the catalyst were suspended in 1.5 liters of air-free, carbon dioxide-saturated water, at 30° C, and illuminated for two hours by two 250-watt lamps. The irradiated liquid, separated by filtration, gave a positive Molisch test upon satura- tion with sulfur dioxide. If the solution was left standing for two hours, or heated to 60°, the test became negative; this was taken as proof that the first product of artificial photosynthesis was unstable. Upon evaporation of the liquid, a whitish precipitate was obtained, which was shown by charring to contain some organic matter; but 30 mg. of this matter, collected from several irradiations, gave only traces of carbon dioxide and water in a microcombustion. This was attributed by Baly to the content of the precipitate in silica and was not considered by him as a decisive argument against its predominantly organic nature. The precipitate was taken up in a little water, and treated for two hours with takadiastase at 37°. The product, tested with Fehling's solution, gave 7-8 rag. of cuprous oxide. This was taken as a proof that the white precipitate contained "a kind of starch," which the diastase had con- verted into a reducing sugar. The yield of cuprous oxide could not be increased by prolonged irradiation; this was attributed by Baly to a poisoning of the catalysts by the products of photosynthesis. This is what Baly called the "final achievement" of photosynthesis in vitro! It contained no proof of oxygen liberation; no proof of carbon SENSITIZED REDUCTION OE CARBON DIOXIDE S7 dioxide consumption; and only an unsuccessful attempt to prove the formation of organic matter by combustion. The obvious incompleteness of experimental evidence did not prevent Baly from giving a detailed picture of how six-membered inositol rings grow on the surface of nickel oxide around "hubs" provided by thorium oxide molecules; and how a similar growth occurs in nature on the surface of chlorophyll crystals, also provided with an appropriate number of "anchor points" consisting of "impurities." Baly postulated — without any proof — that nickelous oxide is oxidized in light by carbon dioxide to nickeUc oxide, and the latter decomposes into nickelous oxide and oxygen, and that, in nature, chlorophyll a is oxidized by carbon dioxide to chlorophyll b, and reduced back to chloro- phyll a by carotene (c/. page 554). Practically all attempts to repeat Baly's experiments elsewhere have given negative results. Only Yainik and Trehuna (1931) have obtained positive formaldehyde tests with nickel and cobalt carbonate sensitizers (but also with other colored inorganic salts, as well as with "white powders colored blue, green or red by different dyes"). In attempting to repeat Baly's early work, Emerson (1929) found that a suspension of nickel carbonate absorbs carbon dioxide (probably by bicarbonate formation) in a completely reversible manner; this absorption is unaffected by Ught, and not accom- panied by oxygen evolution. The negative outcome of Bell's (1931) attempts to repeat the experiments with electrolytically deposited carbonates and with kieselguhr-supported ferric oxide was mentioned before. Zscheile (1932) and Qureshi and Mohammad (1932, 1933) repeated Baly's experiments with precipitated basic nickel and cobalt carbonates, "activated" (according to Baly's preception) by illumination with a mercury arc. The same tests for sugars and aldehydes as used by Baly failed to reveal the presence of any carbohydrates. No attempts to repeat Baly's latest experiments (1939) have as yet been pubhshed. We have no specific reasons to deny that the 5-7 mg. of cuprous oxide, which were precipitated by FehUng's solution in these experiments, were due to the presence of a reducing sugar; or that this sugar was formed by the action of diastase on "a kind of starch" (although the specificity of enzymes and the optical activity of the natural carbohydrates raises a difficult problem); or that this "starch" was formed by the reduction of carbonic acid by fight and nickelous oxide. However, the inadequacy of experimental evidence and the results of previous controls outside Baly's laboratory do not encourage us to give credence to these interpretations. It may be worth stressing the fact that, even if the sugar formation should be confirmed, the assertion that it represents the result of true photosynthesis would remain arbitrary as long as no oxygen evolution has been demonstrated. The rapid cessation of the reaction certainly does not speak in favor of true catalysis. Baly thought that he has achieved not only the photosynthesis of carbohydrates from carbon dioxide and water but also the photosynthesis of organic nitrogen com- pounds. Baudisch (1911, 1916), Baudisch and Mayer (1913) and Baudisch and KUnger (1916) have found that nitrate and nitrite solutions in aqueous formaldehyde or methanol are converted in day fight, first into formhydroxamic acid, (OH)CH=NOH, and then into a large variety of complex nitrogen compounds. Baudisch suggested that, while carbon dioxide is photochemically reduced in the plants to formaldehyde (H2C=0), nitrate could be reduced to a similar compound — free nitrosyl, HN=0, after which the two products may imite and give formhydroxamic acid. Baly, Heilbron and Hudson (1922) and Baly, Heilbron and Stern (1923) extended these experiments; and Baly, 88 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 Saunders and Morrison {cf. Baly 1940) claimed to have photosynthesized an amino acid, an alkaloid, and even a protein, by the ultraviolet illumination of mixtures of nitrite and formaldehyde. Baly suggested that the "co-assimilation" of carbonates and nitrites is the source of organic nitrogen compounds in plants: OH light y (4.26) HNO2 + CO2 + 6 H > HC +2 H2O \ NOH These speculations have even less of an experimental foundation than Baly's hypotheses concerning the synthesis of carbohydrates. In another series of papers on photosynthesis in vitro — those of Dhar and coworkers, it was claimed that positive tests for formaldehyde were obtained after the exposure of carbon dioxide or carbonate solutions in open vessels for several hours to "tropical sunlight." (In what respect this light is different from sunlight elsewhere, or from the light of a strong artificial source, was not made clear.) According to Dhar and Sanyal (1925), Rao and Dhar (1931''*), Rajwanshi and Dhar (1932) and Dhar and Ram (1932), formaldehyde can be detected after irradiation, even in pure solutions of carbon dioxide or sodium bicarbonate; the yield can be increased by the addition of the following sensitizers: In carbon dioxide solutions: FeCls, Fe(0H)3, FeS04, NiS04, C0CO3, CUSO4, CuCOa, copper acetate, Cr2(S04)3, Cr203, Cr(0H)3, MnCl2, V2O6, NH4-Ce-nitrate, Pd(N03)2, U02(N03)2, methylene blue, methyl orange and malachite green; in sodium bicarbonate solutions: Fe(0H)3, C0CO3, NiC03, ZnO, Mg and FeCOs. No formaldehyde was found in solutions "sensitized" by cerous oxide, molybdic acid, rhodamine and safranine. The largest concentration of form- aldehyde, obtained in carbon dioxide solutions after four hours of irradiation (in the presence of manganous chloride) was 8 X 10~^%, while in bicarbonate solutions yields up to 4 X 10"'% have been obtained. In the presence of zinc oxide or magnesium, Dhar and Ram (1932) obtained, in four hours, 1-3 mg. of formaldehyde. Results similar to those of Dhar were obtained by Mezzadroli and Vareton (1928), Mezzadroli and Babes (1929) and Gore (1934); but Burk (1927), Reggiani (1932), Qureshi and Mohammad (1932) and Mackinney (1933) were unable to confirm them. Some formaldehyde was found in experiments with dyestuffs. Since its occurrence, however, did not depend on the presence of carbon dioxide, it must have originated in the decomposition of the dyestuff {cf. page 68). In refusing to accept as significant, from the point of view of photo- synthesis, the results of Dhar's experiments, we take into consideration not only the danger of contamination and the generally unsatisfactory experimental technique, but also the general proposition, formulated on page 83, that as long as quantum yields remain extremely small (of the order of 10~^ or 10~^) "everything is possible in photochemistry." This applies not only to direct effects of ultraviolet light but even to sensitized reactions brought about by the comparatively small quanta of visible SENSITIZED REDUCTION OF CARBON DIOXIDE 89 light. Once in a million absorption acts two photons may strike the same molecule or two excited molecules may collide and exchange energy, accumulating a quantum sufficiently large to cause the formation of a free atom or radical. Accidents of this kind may lead to the formation of a few molecules of formaldehyde in carbonate solutions subjected to a prolonged irradiation by visible light. The essential characteristic of natural photosynthesis is that the accumulation of energy occurs w4th an efficiency far in excess of anything explicable by statistical considerations. Unless we are able to imitate nature in this respect, we have no right to speak of having achieved "artificial photo- synthesis" — even if we should succeed in producing traces of formalde- hyde by a prolonged illumination of carbonate solutions. (b) The Experiments of Baur The series of papers by Baur and coworkers, dealing with artificial photosynthesis and related processes in a variety of systems in vitro, remain to be discussed. In many respects, they compare advantageously with the attempts of Baly and Dhar. Unfortunately, Baur's adherence to a strange theory — the reduction of all photochemistry to electro- chemistry {cf. page 90) makes the reading of his papers difficult. The variety of systems investigated by Baur and coworkers was imposing, and the results were always reported in a scrupulous fashion. Neverthe- less, we do not believe that artificial photosynthesis has been achieved by Baur. Aside from the one very complex system (acetate silk- chlorophyll-cetyl alcohol) whose illumination allegedly yielded as much as 20 moles of formaldehyde per mole of chlorophyll present, the essence of all the other experiments was the formation of formaldehyde in quantities roughly equivalent to those of the sensitizing dyes used, and very small compared with the total quantity of the other organic compo- nents of the reacting system. The assumption that this formaldehyde was formed by the reduction of a carboxyl group, or of carbonic acid (and not by oxidation of an alcohol or hydrocarbon) cannot be considered as proved. The formation of oxygen was claimed only in an experiment which was termed by Baur himself as "prefiminary" and in a recent in- vestigation, of which only an abstract could be obtained (c/. p. 93). ffis first papers (Schiller and Baur 1912, Baur and Rebmann 1922, and Baur and Buchi 1923) were concerned with the refutation of the claims by Usher and Priestley (1906), Moore and Webster (1913, 1918) and Baly, Heilbron and Barker (1921). In addition to showing that colloidal ferric oxide, ferric chloride, uranium oxide, sodium uranate, and malachite green do not convert carbon dioxide in hght into formaldehyde or formic acid (as asserted by the above-mentioned authors), Baur and Biichi (1923) also investigated the action of dyestufifs (eosin, phosphine, malachite green) in non- aqueous systems (lecithin emulsions in xylene), as well as in the adsorbed state (on cotton and silk fiber) in the form of resinates, etc. No oxygen evolution was observed, 90 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 and when formaldehyde was found (as in the case of malachite green), it could be detected also in absence of carbon dioxide, and thus must have originated in the decom- position of the dyestuff. In later papers, Baur was not satisfied with the provision of a sensitizer, but at- tempted also the substitution of a stronger redudant (in place of water) or of a less reluctant oxidant (in place of carbonic acid). He was guided in these experiments by a concept of photochemistry as "molecular electrochemistry." He considered a molecule excited by light absorption as "polarized," with a positive and a negative pole, and treated all photochemical reactions as "depolarizations" brought about by the transfer of charges from the " hght-polarized " molecule to appropriate acceptors. Although this picture has little in common with well-founded concepts of molecular excitation, it can be used without much harm as a description of certain facts of sen- sitization. An excited molecule has no "plus" and "minus" pole, but it can have both an increased affinity for an electron (i. c, the properties of an oxidant), and the tendency to lose an electron {i. e., the properties of a reductant). When the excited molecule meets a reductant, it may oxidize it, by taking away an electron (Baur's "cathodic depolarization"). If it meets an oxidant, it can reduce it by donating an electron (Baur's "anodic depolarization"). Baur's conception proved useful in practice in that it induced him to pay attention to the nature of the "depolarizers," that is, to provide complete oxidation-reduction systems, and not to be satisfied with the reduction of carbon dioxide without asking whether the part of the reductant was played by water, by the sensitizer itself, or by some accidental component of the system. In a series of experiments, Baur has attempted to achieve the photochemical reduction of carbon dioxide by providing, in addition to sensitizers, reductants ("anodic depolarizers") which can be expected to donate their electrons more wilhngly than the water molecules. He tried (1928) urea, cyanamide, cyanide, benzidine and sodium sulfite (in benzene), with eosin, resinate dyes or chlorophyll as sensitizers. He also attempted the fixation of the sensitizer by adsorption on carbonates (magnesia alba) and the combination in one molecule of the oxidant (carbonate ion) and sensitizer (uranyl ions, ferrous ions). He also used iron-substituted permutites (for a still stronger fixation of the sensitizers) and colored lacquers, in which tannin was supposed to create a molecular link between carbonate and sensitizer. All these experiments gave negative results; no formaldehyde was produced, and no oxygen was hberated. Similarly negative results were obtained also by Reggiani (1932), who used eosin, quinine sulfate, methylene blue, rhodamine, thionine, and methyl orange as sensitizers, in both artificial light and sunlight, and sodium sulfide, hydrogen, zinc, Dewarda alloy, pyrogallol and hydroquinone as reductants. In another series of experiments, Baur substituted carboxyl groups for carbonate ions as substrates of reduction (c/. page 80). At first (1928), he used /3-resorcyhc acid and other polyphenolcarboxylic acids. Then he tried carboxyl-containing dyestuffs, gallocyanin and pseudopurpurin, \vith different reductants ("anodic depolarizers"), in the hope that uniting sensitizer and oxidant (carboxyl) in one molecule might yield some success. However, no oxygen or formaldehyde were obtained in these experiments as well. In subsequent experiments, (1935), Baur arrived at the conclusion that chlorophyll is capable of producing formaldehyde by the reduction of its two carboxyl groups (c/. Formula 16.III), and, what is more, that this oxidation can be carried out at the cost of water, by the intermediary of an "auxiliary" reversible oxidation-reduction system, e. g., methylene SENSITIZED REDUCTION OF CARBON DIOXIDE 91 blue-leuco methylene blue. This conclusion — which, if correct, would be of extraordinary importance for the theory of photosynthesis, and for the imitation of this process in vitro — was based on the observation that formaldehyde can be detected in water in which collodion films impreg- nated with alcoholic solutions of the two dyestuffs have been exposed to light. We mentioned on page 68 the controversy concerning the formaldehyde formation by gelatin films containing only chlorophyll. Baur found no formaldehyde after the illumination of pure chloro- phyll-collodion films, but obtained positive results with chlorophyll- methylene blue films. One possible explanation of these results is the photoxidation of methyl groups in methylene blue by chlorophyll (or of methyl groups of chlorophyll by methylene blue); but Baur suggested a more complicated mechanism, described by the series of equation (4.27) in which excited chlorophyll molecules are alternatively "depolar- ized" by hydrogen and hydroxyl ions, and whose final result is the oxidation of water by the carboxyl group of chlorophyll, with methylene blue playing the part of a catalyst (XCOOR is chlorophyll, MB is methylene blue and MB — is leuco methylene blue) : (4.27a) s light C=0 + 2 OH- > RO X o \ / c RO O + H2O (4.27b) X ^RO X o o o + MB RO + MB- O o (4.27c) (4.27d) (4.27) C RO + 2H+ light ^ XR++ + H2CO + O2 O XR++ + MB- -> XR + MB XCOOR + H2O > XR + H2CO + O2 The scheme is obviously a very arbitrary one. Baur, however, used it as a starting point for a whole series of experiments. First, he showed (Baur and Fricker 1937) that chlorophyll can be replaced by eosin and that other reversibly reducible dyestuffs or inorganic redox systems can be substituted for methylene blue. Instead of collodion films, he found colophony (rosin) suspensions in water more suitable. The "sensitizer" (chlorophyll or eosin) was contained in the sol particles, the "auxiliary" redox pair in the aqueous phase. The auxiliary systems included thionine, malachite green, safranine, quercetin, Janus red, neutral red, phenosafranine, gallocyanin, hydroquinone, Nile blue and ferric chloride, 92 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 i. e., systems of widely different oxidation-reduction potentials. How- ever, formaldehyde was obtained with all of them, and none was obtained from chlorophyll or eosin sols in the absence of an auxiliary system. The yield was from 16 to 70% of the material available in the two carboxyl groups of chlorophyll. Baur attached a particular importance to experiments with ferric salts and quercetin because both are common components of plants. In this work, no attempt was made to prove the evolution of oxygen. Assuming that oxygen might cause a partial photoxidation of chlorophyll, Baur attempted to improve the yield by adding substances capable of "catching" oxygen, — rubrene and carotene. Rubrene (in benzene) was without effect; carotene (in palm oil suspension) increased the yield by about 30%, which was considered as significant. No formaldehyde was obtained from chlorophyll adsorbed on alumina (suspended in methylene blue solution); from nonfiuorescent chlorophyll, preparations (copper phaeophytin) and from water-soluble dyes (e. g. gallocyanin), which gave no two-phase systems. Baur and Gloor (1937) tested several other dyes as sensitizers and oxidants, and found that only esterified com- pounds can be used, whereas compounds containing free carboxyl groups gave no formaldehyde. Rhodamine derivatives were found to be even better oxidants than eosin. Baur, Gloor and Kiinzler (1938) obtained positive results with rhodamine both in colophony sols and collodion films, and found an increase of the yield with increasing length of the alcohol molecule in the ester; the free acid, rhodamine B, was ineffective. They endeavored further to bring about suitable conditions for the recarboxylation of the (supposedly) decarboxylated dyestuff ; and thought that the use of "ol" phases (higher alcohols), which take carboxylic acids out of the aqueous phase, might help to shift the equilibrium RH 4- CO2 ^ RCOOH towards a more complete carboxylation (c/., however. Chapter 8, page 179). Different "ol" compounds were found useful, particularly geraniol. The authors then used a carbon dioxide atmosphere, to favor still more the recarboxylation of the oxidant. Positive results were obtained, however, only with two very special systems: mashed leaves in geraniol, and acetate silk-chlorophyll-cetyl alcohol. The latter system formed twenty times more formaldehyde than could be accounted for by the carboxyl groups of chlorophyll. This experiment was announced as the first successful photochemical reduction of carbon dioxide in vitro. Baur also tried to give the proof of complete photosynthesis in this system by demonstrating the liberation of oxygen; but the analytical results were not very consistent and the authors themselves termed them "preliminary." Baur, Gloor and Kiinzler (1938) found that positive results can also be obtained with sensitizers not containing esterified carboxyl groups, if SENSITIZED REDUCTION OF CARBON DIOXIDE 93 a higher alcohol is provided which is capable of binding carbon dioxide in a substituted carbonic acid ester: (4.28) CO, + ROH > RHCO, In other words, carbon dioxide must be bound to a "lipophilic" organic molecule, and thus held in the nonaqueous phase. Whether this is O II achieved by the formation of an esterified carboxyl group, Ri — C — OR2, O II or by the formation of a carbonic acid ester, HO — C — OR, is unimportant; in both cases, if R is sufficiently large, the product is lipophilic and does not pass into the aqueous phase. (The absence of free carboxyl reduces the affinity to water.) In the first case, the oxidant can also be the sensitizer; in the second case, a separate sensitizer must be added. Baur, Gloor and Kunzler used acetate silk, colored with "cibacet" or "celliton" dyes, coated with cetyl alcohol and suspended in aqueous methylene blue solution, which also contained suspended calcium carbonate. From all these experiments, Baur concluded that the prerequisite of artificial photosynthesis is a two-phase system, Avith the sensitizer and oxidant in a nonaqueous phase, and the reductant and an "auxiliary oxidation-reduction system'' in the aqueous phase. In 1943 Baur and Niggli announced that two-phase systems contain- ing chlorophyll in geraniol or phjrtol {e. g., 50 mg. chlorophyll in 25 ml. geraniol), and methylene blue in dry glycerol {e. g., 30 mg. in 50 ml.), produced steadily from circulating carbon dioxide gas both oxygen and formaldehyde, at a rate of about 5 mg. per 24 hours, which corresponds to about 5% of the saturation yield produced by the same quantity of pigment in a living plant. Bukatsch (1939) thought that ascorbic acid (cf. Chapter 10) may play the part of the "auxiliary system." He therefore compounded two-phase mixtures containing ascorbic acid (chlorophyll in colophony or lecithin; ascorbic acid in water), and obtained positive formaldehyde tests (with Schiff's reagent) after illuminating these systems for 5-15 hours with 25,000 lux. Our opinion of Baur's work was stated at the beginning of this discussion. Despite the unnecessary complication introduced by the "electrophotochemical" terminology, the general idea of the research was sound. The provision of complete oxidation-reduction systems, the substitution of less reluctant oxidants and reductants for carbon dioxide and water, the attempts to unite the reactants in molecular complexes, and to separate the products by the provision of two phases — all these were reasonable steps towards the reproduction of the essential 94 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 conditions of photosynthesis. However, the technique of Baur's experi- ments was so primitive, and so devoid of the modern quantitative approach to the problems of photochemistry, and conclusions were drawn so hastily, that it is impossible to accept any of them as well founded, let alone proved. The experiments of Baur were hurried excursions along paths which, if more patiently explored, may perhaps one day lead to important results. Bibliography to Chapter 4 Photosynthesis and Related Processes Outside the Living Cell A. Dead Cells, Isolated Chloroplasts, and Chlorophyll Preparations 1881 Engelmann, Th. W., Botan. Z., 39, 441. 1888 Haberlandt, G., Flora, 71, 291. 1896 Ewart, A. J., /. Linnean Soc. London J . Botany, 31, 423. 1901 Friedel, J., Compt. rend., 132, 1138. Friedel, J., ihid., 133, 840. Harroy, M., ihid., 133, 890. 1902 Herzog, R. 0., Z. physiol. Chem., 35, 459. 1903 Macchiati, Bull. soc. botan. ital., 1903. 1904 Molisch, H., Botan. Z., 62, 1. Bernard, Ch., Compt. rend., 140, 509. 1905 Bernard, Ch., Botan. Centr. Beihefte, 16, 1. 1906 Usher, F. L., and Priestley, J. M. V., Proc. Roy. Soc. London B, 78, 318. 1908 Ewart, A. J., ibid., 80, 30. 1909 von Euler, H., Z. physiol. Chem., 59, 122. 1910 Shryver, S. B., Proc. Roy. Soc. London B, 82, 226. 1912 Schiller, H., and Baur, E., Z. physik. Chem., 80, 669. 1914 Warner, C. H., Proc. Roy. Soc. London B, 87, 378. Wager, H., ihid., 87, 386. 1915 Chodat, R., and Schweizer, K., Arch. sci. phys. nat., 39, 334. 1918 Willstatter, R., and StoU, A., Untersuchungen uber die Assimilation der Kohlensaure. Springer, Berlin, 1918. 1923 Eisler, M., and Portheim, L., Biochem. Z., 135, 293. 1925 Molisch, H., Z. Botan., 17, 577. 1935 Inman, 0. L., Cold Spring Harbor Symposia Quant. Biol., 3, 184 Baur, E., Helv. Chim. Acta, 18, 1157. 1937 Rabinowitch, E., and Weiss, J., Proc. Roy. Soc. London A, 162, 251. Hill, R., Nature, 139, 881. Baur, E., and Fricker, H., Helv. Chim. Acta, 20, 391. 1938 Baur, E., Gloor, K., and Kiinzler, H., ibid., 21, 1038. Smith, E. L., Science, 88, 170. Knoll, A. F., Matthews, Jr. F. L., and Crist, R. H., /. Chem. Phys., 6, 109. Kautsky, H., Naturwissenschaften, 26, 14. Inman, 0. L., Plant Physiol., 13, 859. BIBLIOGRAPHY TO CHAPTER 4 95 1938 Inman, 0. L., Science, 88, 544. 1939 Hill, R., Proc. Roy. Soc. London B, 127, 192. 1940 Hill, R., and Scarisbrick, R., Nature, 146, 61. Hill, R., and Scarisbrick, R., Proc. Roy. Soc. London B, 129, 238. 1942 French, C. S., Newcomb, E., and Anson, M. L., Am. J. Botany, 29, 83. B. Photochemical Oxidation of Water 1863 Vogel, H. W., Ann. Phys., 119, 407, 497. 1906 Eder, J., Photochemie, in Ausfiihrliches Handbuch der Photographic 3rd ed., Knapp, Halle, 1906; p. 208. 1908 Baur, E., Z. physik. Chem., 63, 683. Thiele, H., Ber. deut. chem. Ges., 40, 4914. 1909 Kernbaum, M., Compt. rend., 149, 273. Kernbaum, M., Le Radium, 6, 225. 1910 Coehn, A., Ber. deut. chem. Ges., 43, 880. Berthelot, D., and Gaudechon, H., Compt. rend., 150, 1690. Berthelot, D., and Gaudechon, H., J. pharm. chim., (7), 2, 5. 1911 Sichling, K., Z. physik. Chem., 77, 20. 1912 Coehn, A., and Grote, G., Nernst Festschrift. Knapp, Halle, 1912, p. 136. 1916 Tian, A., Ann. phys., 5, 248. 1918 Baur, E., Helv. Chim. Acta, 1, 186. 1919 Baur, E., Z. Elektrochem., 25, 102. Svensson, T., Arkiv. Kemi Mineral. GeoL, 7, No. 19. 1921 Baur, E., and Rebmann, A., Helv. Chim. Acta, 4, 256. 1925 Wood, R. W., Phil. Mag., 50, 774. 1926 Senftleben, H., and Rehren, I., Z. Physik, 37, 529. 1927 Baur, E., and Neuweiler, C., Helv. Chim. Acta, 10, 901. Bates, J. R., and Taylor, H. S., J. Am. Chem. Soc, 49, 2438. 1928 Gaviola, E., and Wood, R. W., Phil. Mag., 6, 1191. 1930 Beutler, H., and Rabinowitch, E., Z. physik. Chem. B, 8, 403. 1931 Goodeve, C. F., and Stein, N. 0., Trans. Faraday Soc, 27, 393. Weber, K., Z. physik. Chem. B, 15, 18. 1933 Harteck, P., Trans. Faraday Soc, 30, 134. Blum, H. F., and Spealman, C. R., /. Phys. Chem., 37, 1123. 1934 Audubert, R., Phenomenes Photoelectrochimiques, Actualit^s scienti- fiques. No. 91, Gauthier-Villars, Paris, 1934. Terenin, A., and Neujmin, H., Nature, 134, 255. 1935 Terenin, A., and Neujmin, H., /. Chem. Phys., 3, 436. Terenin, A., and Neujmin, H., Acta physicochim. U.R.S.S., 5, 465. Weiss, J., Nature, 136, 794. Hoang Thi Nga, /. chim. phys., 32, 564. Stora, C, Compt. rend., 200, 552. 1936 Stora, C., ibid., 202, 48, 408, 2152. Melville, H. W., Proc Roy. Soc. London A, 157, 621. 1937 Stora, C., /. chim. phys., 34, 536. 96 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 1937 Weiss, J., and Porret, D., Nature, 39, 1019. 1938 Yamafuji, K., Fujii, M., and Nishioeda, M., Biochem. Z., 296, 348. Yamafuji, K., Nishioeda, M., and So, K., ibid., 298, 293. Farkas, A., and Farkas, L., Trans. Faraday Soc, 34, 1113. 1939 Yamafuji, K., So, K., and Tin, H., Biochem. Z., 300, 414. Yamafuji, K., Nishioeda, N., and Imagawa, H., ibid., 304, 404. Richardson, F. D., (unpublished). 1940 Rabinowitch, E., /. Chem. Phys., 8, 551. Rabinowitch, E., ibid., 8, 560. 1942 Rabinowitch, E., Rev. Modern Phys., 14, 112. C. Chemical and Photochemical Reduction of Carbon Dioxide 1893 Bach, A., Compt. rend., 116, 1145. 1897 Losanitsch, S. M., and Jowitschitsch, M. Z., Ber. deut. chem. Ges., 30, 135. 1898 Berthelot, D., Compt. rend., 126, 610. 1900 Berthelot, D., ibid., 131, 772. 1905 Lob, W., Z. Elektrochem., 11, 745. 1906 Lob, W., ibid., 12, 282 (1906); Landw. Jahrb., 35, 541. Usher, F. L., and Priestley, J. M. Y., Proc. Roy. Soc. London B, 77, 369. 1907 Chapman, D. L., Chadwick, S., and Ramsbottom, J. E., J. Chem. Soc, 91, 942. Fenton, H. J. H., ibid., 91, 687. 1908 Gibson, R. J. H., Ann. Botany, 22, 117. Thiele, H., Ber. deut. chem. Ges., 40, 4914. 1909 Holt, A., J. Chem. Soc, 95, 30. Herschfinkel, H., Compt. rend., 149, 395. Herschfinkel, H., Le Radium, 6, 228. 1910 Berthelot, D., and Gaudechon, H., Compt. rend., 150, 1690. Berthelot, D., and Gaudechon, H., /. pharm. chim., (7), 2, 5. Moser, A., and Isgarishev, N., Z. Elektrochem., 16, 613. Stoklasa, J., and Zdobnicky, W., Chem. Z., 34, 945, 1391. 1911 Baudisch, 0., Ber. deut. chem. Ges., 44, 1009. Stoklasa, J., and Zdobnicky, W., Biochem. Z., 30, 433. 1912 Schiller, H., and Baur, E., Z. physik. Chem., 80, 641. Stoklasa, J., Sebor, J., and Zdobnicky, W., Biochem. Z., 41, 333. 1913 Stoklasa, J., Sebor, J. and Zdobnicky, W., ibid., 54, 330. Baudisch, 0., and Mayer, E., Ber. deut. chem. Ges., 46, 115. Moore, B., and Webster, T. A., Proc. Roy. Soc. London B, 87, 163. 1914 Bredig, G., and Carter, G., ibid., 47, 541. 1916 Baudisch, 0., and KUnger, G., ibid., 49, 1167. Baudisch, 0., ibid., 49, 1159. Coehn, A., and Sieper, G., Z. physik. Chem., 91, 347. 1918 Kleinstuck, M., Ber. deut. chem. Ges., 51, 108. Wislicenus, H., ibid. 51, 942. Moore, B., and Webster, T. A., Proc. Roy. Soc. London B, 90, 168. BIBLIOGRAPHY TO CHAPTER 4 97 1921 Baly, E. C. C, Heilbron, I. M., and Barker, W. F., J. Chem. Soc, 119, 1025. Baly, E. C. C, Heilbron, I. M., and Barker, W. F., /. Soc. Chem. Ind., 40, 377. 1922 Baur, E., and Rebmann, A., Helv. Chim. Acta, 5, 828. Baly, E. C. C, Heilbron, I. M., and Hudson, D. P., J. Chem. Soc, 121, 1078. 1923 Baur, E., and Blichi, P., Helv. Chim. Acta, 6, 959. Tramm, H., Z. physik. Chem., 105, 356. Spoehr, H. A., /. Am. Chem. Soc, 45, 1184. Thunberg, T., Z. physik. Chem., 106, 305. Baly, E. C. C, Heilbron, I. M., and Stern, K., /. Chem. Soc, 123, 185. 1925 Lunt, W., Proc Roy. Soc. London A, 108, 172. Porter, C. W., and Ramsperger, H. C, J. Chem. Soc, 47, 79. Dhar, N. R., and Sanyal, R. P., /. Phys. Chem., 29, 926. 1927 Baly, E. C. C, Davies, J. B., Johnson, M. R., and Shanassy, H., Proc Roy. Soc. London A, 116, 197. Baly, E. C. C, Stephan, W. E., and Hood, N. R., ibid., 116, 212. Baly, E. C. C, and Davies, J. B., ibid., 116, 219. Burk, D., J. Phys. Chem., 31, 1338. Mezzadroli, G., and Gardano, G., Atti accad. Lincei Classe sci. fis. mat. nat. Rend., 6, 160. 1928 Mezzadroli, G., and Vareton, B., ibid., 8, 511. Baur, E., Z. physik. Chem., 131, 143. 1929 Emerson, R., /. Gen. Physiol, 13, 163. Baly, E. C. C., and Hood, N. R., Proc Roy. Soc London A, 122, 393. Mezzadroli, G., and Babes, G., Gazz. chim. ital., 59, 305. 1930 Coehn, A., and Spitta, Th., Z. physik. Chem. B, 9, 401. 1931 Baly, E. C. C., Trans. Faraday Soc, 27, 545. Bell, J., ibid., 27, 771. Rao, G. G., and Dhar, N. R., /. Phys. Chem., 35, 1418. Rao, G. G., and Dhar, N. R., ibid., 35, 1425. Yainik, N. A., and Trehuna, F. G., /. chim. phys., 28, 517. Mezzadroli, G., and Vareton, B., Atti accad. Lincei Classe sci. fis. mat. nat. Rend., 14, 347; Chimie & Industrie, Spec. No. 778. 1932 Rajwanshi, A. R., and Dhar, N. R., /. Phys. Chem., 36, 567. Dhar, N. R., and Ram, A., Nature, 129, 205. Dhar, N. R., and Ram, A., Z. anorg. Chem., 206, 171. Reggiani, M., Helv. Chim. Acta, 15, 1383. Zscheile, F. P., Jr., /. Am. Chem. Soc, 54, 973. Qureshi, M., and Mohammad, S. S., /. Phys. Chem., 36, 2205. 1933 Qureshi, M., and Mohammad, S. S., J. Osmania Univ., 1, 48. Dhar, N. R., and Ram, A., /. Phys. Chem., 37, 525. Joo, J. H., and Wingard, R. H., /. Chem. Phys., 1, 886. Mackinney, G., J. Am. Chem. Soc, 54, 1688. 98 PROCESSES OUTSIDE THE LIVING CELL CHAP. 4 1934 1935 1937 1938 1939 1940 1943 Coehn, A., and May, B. W., Z. physik. Chem. B, 26, 117. Gore, v., /. Phys. Chem., 39, 399. Baur, E., Helv. Chim. Acta, 18, 1157. Baur, E., and Fricker, H., ibid., 20, 391. Baur, E., and Gloor, K., ibid., 20, 970. Baly, E. C. C, Nature, 140, 930. Baur, E., Gloor, K., and Kiinzler, H., Helv. Chim. Acta, 21, 1038. Bukatsch, F., Planta, 30, 118. Baly, E. C. C., Pepper, W. P., and Vernon, C. E., Trans. Faraday Soc, 35, 1165. Baly, E. C. C., Proc. Roy. Soc. London A, 172, 445. Baly, E. C. C., Photosynthesis. Van Nostrand, New York, 1940. Baur, E., and F., Niggli, Helv. Chim. Acta, 26, 251, 999. # ■1 Chapter 5 PHOTOSYNTHESIS AND CHEMOSYNTHESIS OF BACTERIA A. Bacterial Photosynthesis* 1. Types of Autotrophic Bacteria Most bacteria are heterotrophic organisms, that is, unlike the auto- trophic green plants, they are unable to synthesize their organic matter directly from carbon dioxide, but require organic nutrients, in common with animals and fungi. They live either in host organisms as parasites or in media containing organic decay products. However, there are two groups of bacteria which are exceptions to this rule. The first group, which consists of photautotrophic bacteria, is capable of reducing carbon dioxide to organic matter in light, using hydrogen sulfide, thiosulfate, hydrogen or other inorganic or organic reductants (but not water, as do the higher green plants). Green and purple sulfur bacteria are representatives of this group; they thrive in sulfide-containing media, and most of them are more or less strictly anaerobic. A second group is formed by chemautotrophic bacteria, colorless organisms which reduce carbon dioxide to organic matter in the dark, by coupling this reaction with different energy-releasing chemical proc- esses. They live in media containing oxidizable substances (sulfide, ferrous iron, methane, etc.) and generally need oxygen (although some can use nitrate instead). The photosynthetic activity of purple bacteria was discovered by Engelmann in 1883. At first, he merely noticed their phototropism which is similar to that of the motile green algae. Under the microscope, the purple bacteria could be observed gathering in a beam of light. Later (1888), Engelmann proved that these bacteria cannot develop in absence of light. If a spectrum is thrown on a culture of purple bacteria, they grow only in the absorption bands of the green ''bacteriochloro- phyll" which is found in all of them, together with a variable assortment of carotenoids (c/. Eymers and Wassink 1938). The strongest absorption band of bacteriochlorophyll lies in the near infrared. Engelmann (1888) pointed out that here for the first time, invisible light appeared to be active in photosynthesis; later, Dangeard (1921, 1927) confirmed this * Bibliography, page 125. 99 100 PHOTO- AND CHEMOSYNTHESIS OF BACTERIA CHAP. 5 and demonstrated that purple bacteria can develop in complete darkness if they are exposed to infrared radiation. Engelmann thought that purple bacteria are normal photosynthesizing organisms, although he was unable to prove, even by means of the extremely oxygen-sensitive motile bacteria, that they produce oxygen in light. He suggested that all oxygen formed by purple bacteria is immediately utilized for the oxidation of sulfide to sulfur. The inability of purple bacteria to produce oxygen was confirmed by Molisch (1907) and van Niel (1931), by means of the even more sensitive luminous bacteria of Beijerinck. On the other hand, Czurda (1936), who observed the oxidation of leuco dyes by purple bacteria in light, and Nakamura (1937), who noticed the decrease in their oxygen consumption in light, interpreted these results as indirect evidence of a photochemical production of oxygen. Van Niel (1941) suggested that the first observa- tion can be explained by the utilization of the leuco dye as reductant in photosynthesis, while the second one proves merely that the respiration of purple bacteria is inhibited by hght (c/. page 111). Van Niel exposed dense suspensions of purple bacteria, mixed with luminous bacteria, to prolonged illumination in closed bottles, without ever being able to detect the slightest traces of oxygen. The indirect arguments of Czurda and Nakamura do not avail against these direct proofs, as was later conceded by Czurda (1937). While Engelmann thought that purple bacteria are normal photo- synthesizing organisms, whose oxygen output is used up by a secondary dark metabolic process, Vinogradsky (1887, 1888) saw in this dark metabolism the main source of organic matter in the bacteria. He based this view on analogies with the colorless chemautotrophic sulfur bacteria, which derive the energy required for organic synthesis, from the chemical oxidation of sulfide by oxygen. If Vinogradsky's conception was correct, why should light be at all necessary for the development of purple bacteria? Vinogradsky, and Skene (1914) offered the following explanation. Purple bacteria thrive only under anaerobic conditions ; they are thus unable to use atmospheric oxygen for the oxidation of sulfide. Vinogradsky and Skene surmised that purple bacteria live in symbiosis with green photosynthesizing bacteria, the latter supplying them with oxygen of such low partial pressure as not to disturb their anaerobic metabolism. However, this hypothesis had to be abandoned when pure cultures of purple bacteria were obtained and found capable of independent growth in light. Buder (1919, 1920) suggested that photosynthesis is carried out by the purple bacteria themselves, to supply the small quantities of oxygen they require for the oxidation of sulfide. He maintained that the latter is their main source of metabolic energy. TYPES OF AUTOTROPHIC BACTERIA 101 Molisch (1907) disagreed with both Engelmann and Vinogradsky, He thought that purple bacteria are not autotrophic at all, but photo- heterotrophic, i. e., that they require organic nutrients but are capable of assimilating them only in light. This confused situation was clarified by van Niel and coworkers in several important papers (van Niel 1930, 1931 ; van Niel and Muller 1931; Muller 1933, Roelefson 1934, van Niel 1935, 19361' 2, 1937; Foster 1940; review by van Niel 1941). Significant con- tributions to this field also were made by Gaffron (1933, 1934, 1935i'2) as well as by French (1936, 19371-2), Wessler and French (1939), Eymers and Wassink (1938), Nakamura (1937i'2, 1938i'2, 1939) and Sapozhnikov (1937). The two main results of van Niel's investigations were as follows : (1) Engelmann, Vinogradsky and Molisch all observed correctly, but used different organisms. There are two kinds of sulfur bacteria: pig- mented, pJiotmdotrophic sulfur bacteria (Engelmann) ; and nonpigmented, chemautotrophic sulfur bacteria (Vinogradsky). In addition, there is a second kind of pigmented bacteria, the heterotrophic purple bacteria (Molisch). {2) In the photosynthesizing sulfur bacteria the oxidation of hydrogen sulfide is not an independent process, coupled with normal photosynthesis through the intermediary of free oxygen, but is a part of the photo- synthetic mechanism itself. The photosynthesis of these bacteria differs from that of the higher plants, in that hydrogen sulfide takes the place of water as reductant. (Gaffron suggested that this type of photochemical metabolism be designated as photoreduction rather than photosynthesis.) However, hydrogen sulfide is not the only reductant which the purple bac- teria can use; in contrast to the photosynthesis of the higher plants, their metabolism is highly adaptable. Some species prefer certain specific re- ductants; but others can use indiscriminantly a large variety of hydrogen donors. Therefore, only a tentative classification of pigmented bacteria according to their normal photosynthetic function is possible. Table 5.1, taken from van Niel (1941), shows the three main classes: green sulfur bacteria, purple sulfur bacteria (Thiorhodaceae) , and purple "nonsulfur" bacteira (Athiorhodaceae) . (Franck and Gaffron designated them as green, red and purple bacteria, respectively; actually the color of both Thio- rhodaceae and Athiorhodaceae can be purple, bright red or brown, de- pending on the nature of the carotenoids associated with the green "bacteriochlorophyll.") The pigment of green bacteria is the so-called " bacterioviridin " (page 445), whose structure probably is intermediate between those of bacteriochlorophyll and ordinary chlorophyll. Table 5.1 shows the variety of compounds which purple sulfur bacteria can use for the reduction of carbon dioxide. Sapozhnikov (1937) found that selenium can be substituted for sulfur. The purple nonsulfur 102 PHOTO- AND CHEMOSYNTHESIS OF BACTERIA CHAP. 5 Table 5.1 Characteristics op the Three Groups of Photostnthesizing Bacteria (after van Niel) Green bacteria: Green-colored bacteria, occurring in hydrogen sulfide media. Photosynthetic activity seems restricted to photoreduction of carbon dioxide with hydrogen sulfide as hydrogen donor. Oxidation proceeds only to elementary sulfur. Other sulfur compounds and organic substances not used as hydrogen donors. Organic growth factors not required. Purple sulfur Purple to red-colored bacteria, also occurring primarily in BACTERIA sulfide-containing media. Capable of oxidizing various inor- (Thiorhodaceae ganic sulfur compounds to sulfate with the simultaneous MoUsch): photoreduction of carbon dioxide. Various organic sub- stances, particularly the lower fatty acids, and some hydroxy and dibasic acids, can be used as hydrogen donors instead of sulfide. Some species can also use molecular hydrogen. Or- ganic growth factors not required. Purple Purple, red or brown-colored bacteria, occurring principally NONSULFUR in media containing organic compounds. Capable of photo- bacteria chemical reduction of carbon dioxide with a large number of (Athiorhodaceae different organic reductanls; some species can use molecular Mohsch) : hydrogen. Although some species are also capable of oxidizing inorganic sulfur compounds to sulfate, growth depends on the presence of small amounts of complex organic materials, such as yeast extract, which presumably furnish necessary organic growth factors. bacteria normally require an organic source of hydrogen. They thrive on a large variety of organic compounds, including acids, alcohols, hydroxy acids, etc. 2. The Over-All Photosynthetic Reactions of Autotrophic Bacteria Over-all chemical equations have not yet been established for all the forms of bacterial photosynthesis by analyses comparable in precision to those which lead to equation (3.6) and (3.7) for the over-all reaction of normal photosynthesis. Since the oxidation products of bacterial photosynthesis are solids (e. g., sulfur) or solutes {e. g.; sulfuric acid), they are unsuitable for manometric assay, which is so convenient for the determination of the "photosynthetic quotient" of the higher plants. On the other hand, bacterial photosyntheses offer the possibility of determining the consumption of the redudant (e. g., hydrogen sulfide or hydrogen) simultaneously with that of the oxidant (carbon dioxide), whereas a determination of water consumption in normal photosynthesis is practically impossible. PHOTOSYNTHETIC REACTIONS OF AUTOTROPHIC BACTERIA 103 The relation between the quantities of carbon dioxide and hydrogen sulfide consumed by bacteria was determined by van Niel (1930, 1931) for a species of purple bacteria which oxidizes sulfide to sulfate. The over-all reaction deduced from these experiments, was: (5.1) CO2 + I (HS)aq. + H2O -> ICHjOi + h (HS04)iq. - 7 kcal This and the following equations of bacterial photosynthesis have been rewritten in the ionic form most suitable for reactions in aqueous phases. For the sake of uni- formity all equations have been reduced to the assimilation of one molecule of carbon dioxide, even if this necessitated the use of fractional coefficients. Formula (5.1) implies the following " photosynthetic quotients": (5.2) - ACO2 : - AH2S : AH2SO4 = 2:1:1 The observed ratios are shown in table 5. II. These ratios are close to Table 5.II The Photosynthetic Quotients for Purple Sulfur Bacteria (after van Niel) After days AH2S0i - AH2S ACO2 AH2S I. 15 0.98 1.86 27 0.97 1.80 34 0.88 1.78 42 0.95 1.94 II. 14 0.97 1.94 24 0.99 1.98 the stoichiometric values, thus proving that carbon dioxide and hydrogen sulfide take part in a common reaction, and not in two "coupled" meta- bolic processes, as assumed by Buder. A further proof of the absence of a separate photosynthetic reaction not involving the sulfide is the fact that the consumption of carbon dioxide ceases as soon as the supply of sulfide has been exhausted (instead of continuing with the liberation of oxygen, as one would expect according to the theories of Vinogradsky and Buder). Van Niel also measured the gas exchange of green sulfur bacteria, and found it to be in agreement with the following equation: (5.2) CO2 + 2 H2S -> {CH2O} -I- H2O 4- 2 S -h 5.1 kcal As mentioned above, various species of purple sulfur bacteria are capable of reducing carbon dioxide by means of compounds intermediate between sulfide and sulfate, e. g., free sulfur, thiosulfate or sulfite. The following over-all equations were suggested by van Niel for these forms of bacterial 104 PHOTO- AND CHEMOSYNTHESIS OF BACTERIA CHAP. 5 photosynthesis : (5.3) CO2 + I H2O + i S (5.4) (5.5) 3- ...V. , 3 > {CHaO! 4- I H+,. + I (S04)aq- - 14 kcal CO2 + I H2O + I (S203)aq-. > {CH2OI + (HS04)aq. " 6 kcal CO2 + H2O + 2 (HS03)aq. • ^ ICH2OI + 2 (HS04)aq. + 17 kcal A form of bacterial photosynthesis particularly suitable for quanti- tative study, is the carbon dioxide-hydrogen assimilation which produces only organic matter and water. According to the equation: (5.6) CO2 + 2H2 > {CHsOl + H2O + 25.1 kcal the quotient AH2/ACO2 should be equal to 2. Reaction (5.6) was discovered by Roelefson (1934) in the study of Thiorhodaceae. In the same year, Gaffron found, that certain Athio- rhodaceae also can reduce carbon dioxide by means of molecular hydrogen in light. Table 5. Ill contains several determinations of the "photo- Table 5.III Photosynthetic Quotients of Hydrogen-Consuming Bacteria Organism Athiorhodaceae Rhodovibrio parvus Streptococcus varians Streptococcus varians Thiorhodaceae: Chromatium sp. AH2/ACO2 1.85-2.25 2.2 -2.6 2.6 2.4 Observer Gaffron (1935) van Niel (1941) Wessler and French (1939) van Niel (1936) synthetic quotient" AH2/ACO2. Most values in the table are somewhat larger than 2, indicating a possible formation of products reduced beyond the carbohydrate stage. We have given, in equations (5.2) to (5.6), the heats of the photo- reduction of one mole of carbon dioxide by bacteria (calculated from the data of Bichowsky and Rossini, assuming 51 kcal for the heat of formation of the {CH2O} group). They vary between Ai7 =4-13 kcal for the oxi- dation of sulfur to sulfuric acid, and AH = — 25 kcal for the reduction of carbon dioxide by molecular hydrogen. Thus, the photochemical reactions of autotrophic bacteria are either exothermal, or only weakly endothermal (as compared with the photosynthesis of the higher plants, AH =112 kcal). However, the reduction of carbon dioxide by ele- mentary selenium (Sapozhnikov) should involve the accumulation of as much as 60 kcal per mole (if selenium is oxidized to selenic acid). Fur- thermore, according to Eymers and Wassink (1938) the carbon dioxide reduction by thiosulfate in Chromatium D leads to the oxidation of the latter to tetrathionate, a strongly endothermal reaction. Using the heats of formation of the ions given by Bichowsky and Rossini, we obtain: (5.7) CO2 + 4 (S203)aq- + 3 H2O > ICH20} + 2 (S406)aq-. + 4 (OH)aq. - 68 kcal PHOTOSYNTHETIC REACTIONS OP AUTOTROPHIC BACTERIA 105 In confirmation of this hypothesis, Eymers and Wassink quoted the observation that, for each molecule of carbon dioxide assimilated by the bacteria, four were taken up by the medium — obviously to neutralize the four hydroxyl ions. However, most of the energy accumulation in (5.7) is associated with the formation of the four hydroxyl ions. If this re- action occurs in an acid medium, and the hydroxyl ions are neutralized, the heat effect is only —12.5 kcal. Altogether, it appears that bacterial photosynthesis is not necessarily inefficient as far as energy conversion is concerned, but can lead to the conversion into chemical energy of up to one-third or one-half the amount which is accumulated in the photosynthesis of the higher plants. Because electrolytes are involved in many forms of bacterial photo- synthesis, the free energies of these reactions often differ considerably from their total energies (while the free energy of normal photosynthesis is almost equal to its total energy; c/. Table 3.V). Calculated for standard conditions (atmospheric pressures of gases and one-molar solutions of the solutes), the gains in free energy in different forms of bacterial photosynthesis, generally are larger than those in total energy, by as much as 20 or 30 kcal per mole. For example, the free energy of reaction (5.7) is AF = 97 kcal per mole in alkaline, and 20 kcal in acid solution. In reactions (5.1) to (5.6), carbon dioxide is reduced to carbohydrate by means of different inorganic reductants. In chapter 3, we have interpreted normal photosynthesis as a transfer of hydrogen atoms from water to carbon dioxide. Van Niel (1931, 1935) generalized this concept by describing all forms of bacterial photosynthesis as hydrogen transfers from various hydrogen donors (reductants) to carbon dioxide as the common hydrogen acceptor. In analogy to the two alternatives, (3.13) and (3.14), in normal photosynthesis, the generalized equation of photosynthesis can be written in two forms: light (5.8a) CO2 + 4 R'H > {CH2O) + HoO + 4 R' or light (5.8b) CO2 + 2 R"H2 > i CH2O I + H2O + 2 R" In the first formulation, each molecule of the reductant contributes one, and in the second formulation two, hydrogen atoms, towards the reduction of one molecule of carbon dioxide. In the case of normal photosynthesis, R' is OH, or — if formulation (5.8b) is preferred — R" is 0. In the case of photoreduction with sulfide, R' is SH or R" is S, and in that of photoreduction with hydrogen, R' is H or R" is nothing. The application of equations (5.8) when the reductant contains no hydrogen at all (as in the case of elementary 106 PHOTO- AND CHEMOSYNTHESIS OF BACTERIA CHAP. 5 sulfur), or is unlikely to yield it (as in the case of the bisulfite ion, HSOs"), will be demonstrated in chapter 9 (page 220). Van Niel's generalized concept of bacterial photosynthesis adds an important argument in favor of the "intermolecular oxidation-reduction" theory of normal photosynthesis, and against the Willstatter-Stoll "internal rearrangement" theory. The easily verifiable fact that in the photosynthesis of sulfur bacteria, the reduction of carbon dioxide leads to the production of sulfur, and not of one part of sulfur and two parts of oxygen (or of sulfur dioxide) is analogous to the fact — proved with much more difficulty by the radioactive isotope experiments of Ruben, Randall, Kamen and Hyde (page 55) — that all oxygen in ordinary photo- synthesis originates in water. 3. Combined Photosynthesis and Heterotrophic Assimilation of Photoheterotrophic Bacteria The metabolism of the "photoheterotrophic" bacteria — that is, bac- teria which require light for the assimilation of organic nutrients, seemed at first to be quite different from that of the " photautotrophic " bacteria discussed above. However, van Niel made it plausible that the organic nutrients serve primarily (although not exclusively) as hydrogen donors, so that the generalized equations (5.8), with R now standing for an organic radical, apply to these organisms as well. The study of the photosynthesis of heterotrophic bacteria had at first encountered difficulties, because the organisms employed were found to require yeast extracts or peptones, and would not thrive in solutions of pure organic compounds. However, this difficulty was overcome by Gaffron (1933) and Muller (1933). Muller used Thiorhodaceae, which were found capable of subsisting not only in sulfide media but also in fatty acid solutions; while Gaffron (1933, 1935) found that the Athiorhodacea, Rhodovibrio parvus, grown in a yeast extract, can be transferred into a simple organic solution for the study of its photosynthetic activity. Similar observations were made by van Niel (1941) with Spirillum rubrum. In these investigations, fatty acids were used as organic hydro- gen donors. However, the utilization of these compounds goes far beyond the contribution of one or two hydrogen atoms. In fact, Muller and Gaffron found that the acids are completely used up, leaving no organic residue at all. It is formally possible to explain this complete assimilation in terms of photochemical dehydrogenation, by assuming the transfer of all hydrogen atoms to carbon dioxide, and a conversion of all carbon atoms into carbon dioxide (c/. Eq. 5.15). However, this explanation is speculative; at least a part of the carbon atoms could be assimilated directly, without taking the roundabout way through carbon dioxide. The fact that the assimilation occurs only in light, and often PHOTOHETEROTROPHIC BACTERIA 107 requires the presence of carbon dioxide (which is " coassimilated " with the acid), makes it probable that the initial step is a photochemical reaction between the organic substrate and carbon dioxide; but the subsequent stages of the process could include the direct "heterotrophic" assimilation of the organic material. Under these conditions, it is important that Foster (1940) found organic substrates capable of yielding only two hydrogen atoms to carbon dioxide in light, and resisting any further assimilation. These were secondary alcohols. The carbon chain in these compounds is not attacked by purple bacteria, and the reaction in light is restricted to the transfer of two hydrogen atoms from the alcohol to carbon dioxide, according to the equation: (5.9) Ri Ri \ light \ : CHOH + CO2 > 2 C=0 + {CH2O) + H2O Ra R2 For example, isopropanol is quantitatively converted into acetone: (5.10) 2 (CH3)2CHOH + CO2 > 2 (CH3)2CO + {CH2O} + H2O - 14 kcal Foster found for this reaction, the " photosynthetic quotients" given in table 5.IV. Table 5.IV Photosynthetic Quotients for Bacterial Photosynthesis with Isopropanol Time, A Isopropanol ACO2 A Acetone days — A Isopropanol 6 2.22 1.02 9 2.02 1.06 12 2.00 1.02 15 2.12 1.00 19 1.97 1.02 25 1.97 1.02 Theoretical 2.0 1.0 We shall now return to the assimilation of fatty acids and attempt to analyze it in terms of photoreduction combined with direct assimila- tion. A "pure" photoreduction of fatty acids, if it is to lead to carbo- hydrates as the only products, must involve either "coassimilation" or hberation of carbon dioxide. When the substrate is " overreduced " (compared to the carbohydrates), it must be diluted with carbon dioxide; when it is "underreduced," some carbon dioxide must be eliminated. 108 PHOTO- AND CHEMOSYNTHESIS OF BACTERIA CHAP. 5 The following general equations can be derived for monobasic acids (5.11) and for dibasic acids: C„H2.+iC00H + ^^5^ CO2 + ^Vi H2O _^3«_+J- {CH2O} (5.12) C„H2,.(COOH)2 + ^^—^ CO2 + ^^-y^ H2O - ~ (12?i + 1) kcal 3n + 1 (CH2O} - ~ {ISn + 4) kcal The heats of the reactions (5.11) and (5.12) have been estimated by assuming 112 kcal for the heat of combustion of iCH20), (156ri + 55) kcal for that of monobasic acids, and il50n + 60) kcal for that of dibasic acids. Equation (5.11) indicates consumption of carbon dioxide for n > 1 and liberation for n < 1, while equation (5.12) requires absorption of carbon dioxide for n > 3 and its Hberation for n < 3. The theoretical "photosynthetic quotients," ACO2/AFA (FA = fatty acid) are {n - l)/2 for monobasic and (n — 3)/2 for dibasic acids. On the whole, these deductions are confirmed by experiments, although the agreement is only qualitative, and individual results scatter considerably. Muller (1933) found that Thiorhodaceae liberate carbon dioxide in the photoreduction of lactate and malate (as they should according to stoichiometric equations, analogous to 5.11 and 5.12, which can be set up for hydroxy acids), but consume it in that of butyrate (in accordance with equation 5.11). Significantly, no butyrate assimilation was observed unless bicarbonate was also provided. More detailed results have been obtained by Gaffron (1933, 1935) with the Athio- rhodaceae (Rhodovibrio) . He found the following values of the "photo- synthetic quotient" Qp = ACO2/AFA. Table 5.V Qp = ACO2/AFA FOR Rhodovibrio (Gaffron) and Spirillum rubrum (van Niel) Qp Qp Acid ?i Acid n Observed Calc. Observed Calc. Acetic 1 1 -0.25 to 0.11 -0.21 (v.N.) Methyl ethyl acetic 4 0.68-0.91 1.5 Propionic 2 0.29-0.42 0.31 (v.N.) 0.5 n-Caproic 5 0.90-1.34 2 n-Butyric 3 0.30-0.43 0.65 (v.N.) 1 z-Caproic 5 1.10-1.30 2 i-Butyric 3 0.61 1 Heptylic 6 1.03-1.54 2.5 n-Valeric 4 0.62-0.90 1.5 Caprylic 7 1.60-1.96 3 i-Valeric 4 0.83 1.5 Nonylic 8 1.90-2.90 3.5 PHOTOHETEROTROPHIC BACTERIA 109 Gaffron's results showed the expected general increase of Qp with n, with the notable exception of n-butyric acid. He saw in the different vahies for the two butyric acids an indication that the mechanism of assimilation may depend on the structure of the molecule. Van Niel (1941) made similar experiments with Spirillum rubrum (another Athio- rhodacea). The individual values again scattered over a considerable range; for example, for acetate, in 48 single experiments, they varied from — 0.15 to + 0.28, but the averages, as given in table 5. VI, showed the expected regular increase from acetate through propionate to n-butyrate. A comparison of the observed values with the calculated ones in table 5.V shows a regular deficiency of carbon dioxide consumption. In the case of acetate, some carbon dioxide is liberated, although the formula of acetic acid, C2H4O2, allows of a quantitative conversion into a carbohydrate (corresponding to Qp = 0). The average "coassimila- tion" of carbon dioxide with the higher fatty acids, amounts to only 60% of the theoretical value. One possible explanation of this fact is that assimilation produces compounds which are more reduced than the carbohydrates. For compounds consisting only of C, and H atoms (and not con- taining peroxide bonds), an appropriate measure of the reduction level is provided by the respiratory quotient, Qr (which is defined on page 32 as the ratio ACO2/— AO2) or, more conveniently, by its inverse value, the reduction level L, which is equal to the number of molecules of oxygen required for the complete combustion of a molecule, divided by the number of carbon atoms in it. It can be calculated by means of the equation ,. ,„, y 1 2nc + Inn - no ^ ^ Qr 2nc where nc, wh and no are the numbers of carbon, hydrogen and oxygen atoms in the molecule, respectively. (We shall see in chapter 9 how closely L determines the energy content of compounds of this type.) The value of L for carbohydrates is 1. A simple calculation shows that products obtained by the coassimilation of fatty acids and 60% of the quantity of carbon dioxide required for the formation of a carbo- hydrate, must have L values between 1.43 {n = 0) and 1.15 (71 = =»). Analyses of the dry matter of purple bacteria are in good agreement with this calculation. Van Niel (1936) found in Spirillum rubrum, Rhodomonas, Streptococcus varians and Chromatium (sulfur-free) an average of 55.7% C, 7.4% H, 15.1% O and 11.8% N. Assuming that all nitrogen is present in the form of amino groups, and substituting an equivalent quantity of hydroxyl groups for them, we arrive at a compo- 110 PHOTO- AND CHEMOSYNTHESIS OF BACTERIA CHAP. 5 sition approximately represented by the formula C5H7O2, corresponding to a L value of 1.15. Gaffron (1933) isolated from the purple bacteria a substance which could be depolymerized to crotonic acid, C4H6O2, L = 1.18, and considered it as a direct product of photosynthesis. However, the results obtained by Foster with isopropanol indicate that the true photosynthetic quotient of purple bacteria probably corresponds to the formation of carbohydrates, rather than to the production of any more completely reduced substances. If this is true, deviations from equations (5.11) and (5.12) in the assimilation of fatty acids are due to a direct assimilation of intermediates, whose reduction level is higher than that of the carbohydrates, rather than to the 'photosynthesis of " overreduced " substances. (Dismutations, which often occur in enzy- matic oxidation, can produce such high-energy intermediates even if the original oxidation substrate itself is not overreduced.) This is not the only argument in favor of a partial heterotrophic nutrition of purple bacteria. Another argument can be derived from the consideration of the metabolism of these organisms in the dark. Barker (1936), Giesberger (1936), Clifton (1937), CUfton and Logan (1939), Winzler and Bamberger (1938) and Winzler (1940) showed that the oxidation of organic sub- strates by respiring bacteria often is coupled with their partial assimilation, e. g., in the case of acetate, according to one of the equations: (5.14a) C2H3O2- + O2 > {CH2O} + HCO3- or (5.14b) 2 C2H3O2- + 3 O2 > {CH2O} + 2 HCO3- + O2 + H2O In the case of the purple bacteria, the mechanism of respiration has a particularly close bearing on that of photosynthesis, because van Niel demonstrated that the first stages of both processes are probably brought about by the same enzymatic system. A few words may be said here about this pecuhar relationship. All Thiorhodaceae (as well as some Athiorhodaceae) are anaerobic, i. e., their dark me- tabolism is of the nature of fermentation. This metabohsm was studied by Gaffron (1 934, 1935), Roelefson (1935), French (1937'' 2) and Nakamura (1937); but its chemistry has not yet been clarified, mainly because it is preponderantly " auto fermentative " (although Nakamura observed the dismutation of formate into hydrogen and carbonate by purple bacteria). The relationship, if any, between this dark metabolism and the metabolism of the same bacteria in light, is as yet not clear. However, some Athiorhodaceae are aerobic (or not strictly anaerobic), and their dark oxidative metabolism was found by van Niel to bear a remarkable relation to their photosynthesis. Respiration and photosynthesis, which are independent (al- though contra-acting), in green plants, appear to be competitive in aerobic Athiorhodaceae. The investigations of Nakamura (1937'''', 1938^'^) with Rhodobacillus palustris showed that the same substances which are readily used as substrates of photosynthesis, also are eagerly consumed as respiration substrates in the dark. Van Niel (1941) found that the uptake of different fatty acids by Spirillum rubrum occurs at the same rate in the dark and in fight — although only oxygen is consumed in the dark, while both oxygen and carbon dioxide are taken up in moderate fight, and carbon dioxide alone is consumed in strong fight. As mentioned on page 100, Nakamura interpreted the decrease in oxygen con- sumption by purple bacteria in fight as evidence of a photochemical production of BACTERIAL CHEMOSYNTHESIS 111 oxygen. However, we now see a more plausible explanation: If the rates of photo- synthesis and respiration are limited by the supply of hydrogen through a common enzyme system, every increase in photosynthesis must lead to a decrease in respiration. The fact that under no circumstance does the organism switch over from oxygen con- sumption to oxygen Uberation, agrees well with this picture, whereas it would be diffi- cult to explain if photosynthesis and respiration were two independent processes, as in the higher plants. Each fatty acid is decomposed by purple bacteria at a different characteristic rate — the same in respiration and photosynthesis; and if a mixture of several acids is provided, their total decomposition rate is additive. This proves that a specific enzyme is available for each of the acids. We made this digression to the subject of the metabolism of purple bacteria in the dark because the parallelism of respiration and photo- synthesis provides an additional argument in favor of a partial direct assimilation of the reductant: Since such an assimilation is known to occur in the dark metabolism, it is also likely to occur in light. Thus, in addition to the direct assimilation of overreduced intermediates, which was suggested as an explanation of deviations from equations (5.11) and (5.12), a part of the carbohydrates formed in the " photoassimilation " of fatty acids, can also be due to a direct "heterotrophic" assimilation, and not to photosynthesis. Van Niel (1941) considered, for the case of acetate assimilation, the three possibilities (5.15), (5.16) and (5.17), to which we may add (5.18) and (5.19) for the sake of completeness: To To photoreduction direct assimilation (5.15) H4C2O2 -t- 2 H2O >[8H + 2C02] (5.16) H4C2O2 -I- U H2O > [6 H + 1| CO2] + h ICH2O} (5.17) H4C2O2 + H2O > [4 H + CO2] + ICH2OI (5.18) H4C202+§H20 >[2H + ^C02] + U {CH2O) (5.19) H4C2O2 > 2 {CH2O) The first equation (5.15), represents pure photoreduction; the next three represent photoreduction coupled with an increasing proportion of direct carbohydrate assimilation; and the last one, direct assimilation without photoreduction. B. Bacterial Chemosynthesis * From the pigmented photosynthesizing bacteria, there is but one step to the nonpigmented " chemosynthesizing " bacteria, which often use the same oxidation substrates (e. g., sulfide, thiosulfate or hydrogen), but work with chemical energy instead of light energy. The discovery of these organisms by Vinogradsky was mentioned on page 100. Some of them can live in (and sometimes only in) purely inorganic media; their autotrophic mode of life is thus easy to prove. These organisms are * Bibliography, page 126. 112 PHOTO- AND CHEMOSYNTHESIS OF BACTERIA CHAP. 5 analogous to the colored sulfur bacteria. Other species require the presence of simple organic compounds, as methane, carbon monoxide, formate, or methanol. They have thus the appearance of hetero- trophants. However, evidence speaks in favor of assigning to the organic substrates of these bacteria, the function of fuels rather than of nutrients. There seems to be an analogy between the colorless bacteria which use these substrates, and the purple bacteria which require organic hydrogen donors for the reduction of carbon dioxide, except that, in the case of the colorless bacteria, the organic substrate has to supply not only hydrogen but also chemical energy. As in the case of the purple Athiorhodaceae, conditions become complicated when colorless bacteria use comparatively complex organic substrates, instead of methane or similar "Ci compounds"; in this case, heterotrophic nutrition can be superimposed upon the chemosynthesis. 1. Types of Chemautotrophic Bacteria The main representatives of this class are the nitrifiers, the (colorless) sulfur bacteria, the iron bacteria, the hydrogen (or "Knallgas") bacteria, the carbon monoxide, methane and carbon bacteria. This list shows that some kind of chemautotrophic bacteria has become associated with prac- tically every oxidizable inorganic compound found on the surface of the earth— ammonia, hydrogen sulfide, sulfur, ferrous iron, methane and coal. Hydrogen and carbon monoxide are not found in natural habitats of the bacteria. These gases are facultative components of the metabo- lism of bacteria, whose mode of life under natural conditions is hetero- trophic. Thiosulfate has occasionally been found in black mud, but most thiosulfate bacteria also can live on organic substrates (that is, they, too, are only facultative autotrophants). The following short description of the chemical activity of the chem- autotrophic bacteria is based mainly on the review by Stephenson (1939), although the thermochemical figures have been revised on the basis of compilations by Kharash, Bichowski and Rossini, and Roth (c/., bibli- ography to Chapter 3), and the over-all equations have been simplified by the consequent use of ionic formulations. All equations are reduced to the consumption of one mole of oxygen, to facilitate comparison with the equation of carbon dioxide reduction, in which one mole of oxygen is produced for each gram atom of assimilated carbon. (a) The Nitrifiers (Vinogradsky 1890) These include Nitrosomonas, which oxidizes ammonia to nitrate, and Nitrobacter, which carries the oxidation from nitrite to nitrate: (5.20a) O2 + ! (NH3)aq. > § (N02)aq. + f H2O + I Hjq. + 49 kcal (5.20b) O2 + 2 (N02)aq. > 2 (N03)lq. + 48 kcal THE SULFUR BACTERIA 113 (h) The Sulfur Bacteria We include in this classification the autotrophic bacteria which utilize hydrogen sulfide, elementary sulfur, thiosulfate, and thiocyanate. Hydrogen Sulfide Oxidizers, e. g., Beggiatoa (Vinogradsky 1887), transform hydrogen sulfide into sulfur globules deposited inside the cell. When the sulfide is exhausted, the sulfur globules are consumed by further oxidation to sulfate: (5.21) O2 + 2 (H2S)aq. > 2 H2O + 2 S + 126 kcal (5.22) O2 + f S + ! HoO > I (S04)aq-. + f H+q. + 98 kcal Sulfur Oxidizers (e. g., Thiohacillus thiooxidans, Waksman and Joffe, 1922). — These bacteria oxidize externally supplied sulfur to sulfate, in accordance with equation (5.22), and are characterized by extreme tolerance to acid. Their optimum pH hes between 3 and 4, and they survive even in 5% sulfuric acid. The metabolism of these organisms recently was investigated by Vogler and Umbreit. Vogler and Umbreit (1941) and Umbreit, Vogel and Vogler (1942) proved that sulfur is first dissolved in fat globules in the ends of the cell, and saw in this fact a proof that only oxidations which take place inside the cell can provide energy for chemosynthesis. According to Vogler (1942), despite its exclusively inorganic nutrition, Thiohacillus possesses an organic metabolism based on storage materials formed by chemosynthesis. Vogler, LePage and Umbreit (1942) showed that the rate of sulfur oxidation is independent of pH (between 2 and 4.8), and of the oxygen pressure; it is inhibited by cyanide (50% inhibition at 10-4 mole/1.), dinitrophenol (50% inhibition at 1.3 X 10"^ mole/1.), azide, iodoacetate, arsenite, indole and phthalate. It is affected by urethane only at comparatively high concentrations (35% inhibition in 0.1 molar solutions). It is 50% inhibited by carbon monoxide in a concentration of 80%, an inhibition which is removed by illumination. All these results indicate that sulfur oxidation proceeds through the intermediary of a heavy-metal enzymatic system (of the hemin type). Since enzymes of this type transfer only electrons, and not oxygen atoms, the oxygen in the SO4 — ions formed by B. thiooxidans must originate in water and not in air, a consequence which could be checked by isotope tracers. The relation between sulfur oxidation and carbon dioxide reduction by B. thiooxodans was studied by Vogler (1942-). Young cultures took up a limited quantity of carbon dioxide even in the absence of sulfur; in older cultures, this uptake was overbalanced by the carbon dioxide production by endogenous respiration. The sulfur-free carbon dioxide uptake could, however, be observed in all cultures, if respiration was suspended by depriving the cells of oxygen. The maximum total uptake, reached in about two hours, was of the order of 0.4 ml. of carbon dioxide 114 PHOTO- AND CHEMOSYNTHESIS OF BACTERIA CHAP. 5 per mg. bacterial nitrogen. This uptake seemed to be reversible and dependent on the concentration of carbon dioxide in the medium (c/. Chapter 8, page 201). Cells which have been allowed to chemosynthesize intensely, subsequently showed a greater capacity for carbon dioxide uptake in absence of sulfur or oxygen than "starved" cells. A short period of "sulfur respiration" restored the capacity for carbon dioxide fixation in "carbon dioxide saturated" cells; while endogenous respira- tion had no such effect. The rate of oxygen consumption decreased in the presence of carbon dioxide, but the sum ACO2 + AO2 remained approximately constant. This is an obvious parallel to the relation between respiration and photosynthesis in purple bacteria, described on page 110. The initial carbon dioxide uptake was unaffected by 0.01 mole per liter of sodium azide or arsenite (which completely inhibit the uptake of oxygen), but was completely inhibited by 10~* mole per liter of iodo- acetate, which caused only 10% inhibition of the oxygen uptake. (Of course, inhibition of sulfur oxidation must cause a corresponding inhibi- tion of carbon dioxide absorption after the initial saturation period.) A concentration of 0.006 mole per liter of sodium pyruvate inhibited both reactions completely, and similar effects were caused by lactic, fumaric and succinic acids, while citric acid had a weaker influence and malic acid none at all. Vogler and Umbreit (1942) inquired into the way in which sulfur oxidation can cause carbon dioxide fixation in a subsequent period of anaerobiosis. They found that during the oxidation period, inorganic phosphate is transferred from the medium into the cells, to be released again during the period of carbon dioxide fixation. Seventy to 80 molecules of oxygen are used up for oxidation while one molecule of phosphate is transferred into the cells; 40-50 molecules of carbon dioxide are taken up concomitantly with the release of one phosphate molecule. This seems to indicate a AO2/ACO2 ratio of about 1.5. Table 5. VII shows that this value corresponds to an almost 100% utilization of the free energy of oxidation, if one assumes that all absorbed carbon dioxide is reduced to carbohydrate. This proves that the question (which Vogler and Umbreit considered as "open") of whether the "delayed" carbon dioxide fixation is a reduction to carbohydrate or not, must be answered in the negative. Probably, it is not a reduction at all, but a carhoxylation (or another reversible carbon dioxide absorption), analogous to that which forms the first stage of photosynthesis (c/. Chapter 8). Remark- able, however, is the large amount of carbon dioxide taken up in this way — it seems to be at least ten and perhaps a hundred times larger than the reversible carbon dioxide fixation by green plant cells (if one excludes from the latter the rather incidental alkali-acid buffer equilibrium). THE SULFUR BACTERIA 115 Vogler and Umbreit considered the phosphate transfer, coupled with sulfur oxidation and carbon dioxide fixation by B. thiooxidans, as a proof that the combustion energy of sulfur is stored in the cells in the form of phosphate bond energy. Of course, the observed transformation of one phosphate molecule per 40 or 50 molecules of carbon dioxide cannot pro- vide more than a small fraction of the energy required for chemosynthesis; but Vogler and Umbreit considered it merely as an index of the inter- cellular formation of high-energy phosphoric acid esters on a much larger scale. If one refuses to consider the "delayed" carbon dioxide fixation by sulfur bacteria as a carbohydrate synthesis, the energy calculations based on this assumption lose their meaning. It is rather improbable that sufficient energy for true chemosynthesis can be stored in the form of "phosphate bond quanta" of 10 kcal per mole each (c/. Chapter 9, page 226). One may suggest that the phosphate transfer and phosphorylation have something to do with the primary reversible carbon dioxide fixation (in chemosynthesis as well as in photosynthesis), rather than with the reduction of carbon dioxide to carbohydrate (c/. Ruben's hypothesis, page 201). The investigations of Vogler and Umbreit show how much information, which may help in the understanding of the closely related phenomena of chemosynthesis and photosynthesis, can be expected from a quanti- tative study of the metabolism of autotrophic bacteria. Bacteria which oxidize sulfur by means of nitrate, instead of oxygen (Thiobacillus denitrificans) , were discovered by Beijerinck in 1904. Since 0.8 mole of NOs" ions, reduced to nitrogen, are equivalent to one mole of oxygen, we write the over-all equation as follows: (5.23) I (NOslaq. + § S + t\ HoO > ! (S04)aq. + A Haq. + f N2 + 86 kcal Thiosulfate Oxidizers. — Thiohacillus thioparus (Natansohn 1902) oxidizes thiosulfate with the deposition of sulfur outside the cell. Natansohn assumed an intermediate formation of tetrathionate inside the cell, and a subsequent external dismutation of tetrathionate into sulfur and sulfate; but Starkey (1935) found no evidence of tetrathionate formation, and gave the following formulation of the over-all reaction: (5.24) O2 + I (S203)a^. + i H2O )■ i (S04)i: (SOO^. + Hii. + 109 kcal 116 PHOTO- AND CHEMOSYNTHESIS OF BACTERIA CHAP. 5 A third kind of thiosulfate-oxidizing bacteria uses nitrate instead of oxygen (Lieske 1912): (5.26) i (N03)aq. + h (S203)a (S04)rq-. + f N2 + i Hjq. + 97 kcal Thiocyanate Oxidizers. — Happold and Key (1937) discovered the Bacillus thiocyan-oxidans, which catalyzes the reaction: (5.27) O2 + h (CNS)rq. + H2O > I (S04)aq- + i (NH4):q. + I CO2 + 112 kcal (c) The Iron Bacteria (Vinogradsky 1888) These organisms precipitate ferric hydroxide from waters containing ferrous salts, and are responsible for the red color of many natural waters. Their chemical activity can be represented by the equation: (5.28) O2 + 4 Fe+qt + 2 H2O > 4 Fe^qt^ + 4 (OH)aq. + 37 kcal The gain in energy becomes larger if we include in the equation the precipitation of ferric hydroxide (c/. Lieske 1911, 1919): (5.29) O2 + 4 Fe+qt + 10 H2O > 4 Fe(0H)3 + 8 H+q. + 63 kcal (d) The Hydrogen Bacteria Bacillus pantotrophus, discovered by Kaserer in 1906, and a number of similar microorganisms of the soil, are heterotrophants which are, however, capable of survival and growth in purely inorganic media, if they are provided with molecular hydrogen, in addition to oxygen and carbon dioxide. Their metabolism is based, under these conditions, on the energy of oxidation of hydrogen to water ("oxyhydrogen reaction"): (5.30) O2 + 2 H2 > 2 H2O + 137 kcal — hence the name "Knallgas bacteria" suggested by Ruhland (1924). Reaction (5.30) is coupled with the reduction of carbon dioxide to organic matter, and further complicated by the simultaneous respiration, i. e., autoxidation of cell material. (Some autotrophic bacteria, the Nitro- somonas, for example, apparently dispense with ordinary respiration altogether, their energy requirements being covered entirely by the oxidation of the inorganic substrate.) The investigations of Kaserer (1906), Nabokich and Lebedev (1907), Lebedev (1908, 1909) and particularly Niklevsky (1908, 1910) have shown a wide distribution of normally heterotrophic but potentially hydrogen-oxidizing bacteria in all soils. Some of them appear to be capable of using nitrate, nitrous oxide or even sulfate as oxidants instead of oxygen, but not much is known about these reactions. Only one species. Bacillus picnoticus, has been thoroughly investigated by Ruhland (1924); and because these hydrogen bacteria appear to be the simplest THE HYDROGEN BACTERIA 117 and most efficient carbon dioxide-assimilating biological systems known, a few words must be said here on his results. Bacillus picnoticus thrives best in inorganic media at pH 6.8 to 8.7; the decline in its activity in alkaline solutions seems to be caused by the precipitation of ferric hydroxide. (It requires a minimum concentration of ferrous iron > 10"^ mole per 1.) It is poisoned by cyanide (5 X 10"^ mole/1.) as well as by urethans (50% reduction in hydrogen consumption by 1.2 moles/1, methylurethan, 2 X 10"^ ethylurethan, 1 X 10"^ propyl- urethan, 5 X 10"^ isobutylurethan, and 1 X 10"* phenylurethan ; com- pare table 12. VIII). While some hydrogen bacteria can use nitrate as oxidant (in absence of oxygen), no such substitution is possible with B. picnoticus. Its rate of consumption of hydrogen is independent of the partial pressure of oxygen (1.5 to 72%) as well as hydrogen. It can operate in "electrolytic gas" (I H2 + ^02), and in nitrogen containing mere traces of oxygen and hydrogen (these traces being completely removed by the activity of the bacteria). The concentration of carbon dioxide is also without specific effect except by its indirect influence on acidity. The rate of hydrogen absorption increases rapidly with temperature (Qio = 3.5 between 20° and 32.5" C). Since the combustion is coupled with the reduction of carbon dioxide, the net ratio AH2/AO2 is larger than 2 (Table 5. VI), the excess hydrogen consumption reaching 40% Table 5.VI Gas Exchange of Bacillus Picnoticus (after Ruhland) AH2 ACO2 - AH2 - AO2 — ACO2 AO2 AH2 - 2AO2 138 53 16.9 2.6 0.53 112 45 10.4 2.5 0.48 90 39 6.2 2.3 0.56 88 41 3.0 2.1 0.53 91 41 5.3 2.2 0.53 103 39 13.1 2.6 0.53 29.5 10.6 4.3 2.8 0.53 85 31 10.9 2.8 0.45 113 41 17.0 2.8 0.53 95 34 13.9 2.8 0.53 225 81 33 2.8 0.53 Average : 2.56 0.52 under the most favorable conditions. The consumption of carbon dioxide, - ACO2, is in exact stoichiometric relationship with the excess consumption of hydrogen, - (zlHg - 2AO2). This " photosynthetic 118 PHOTO- AND CHEMOSYNTHESIS OF BACTERIA CHAP. 5 quotient" is equal to 0.5, indicating the formation of a carbohydrate as the first product of chemosynthesis. Some heterotrophic bacteria (e. g., Acetobader peroxidans, cf. Wieland and Pistor 1936, 1938) also catalyze the " oxyhy drogen " reaction (5.30) but apparently without profiting from its energy for organic synthesis. (e) The Carbon Bacteria The Carbon Monoxide Oxidizers (Bacillus oligocarbophilus) (Beijer- inck and van Delden 1903). The reaction which these bacteria catalyze is: (5.32) O2 + 2 CO > 2 CO2 + 136 kcal which produces no less energy than does the oxidation of hydrogen. Methane Oxidizers. — (Bacillus metanicus), (Sohngen 1906, Miinz 1915). Although methane is usually considered an "organic" carbon compound, we inlcude the methane-burning bacteria in the list of chemo- autotrophic organisms because there is no doubt that methane serves exclusively as a source of energy and as a hydrogen donor, and not as an organic nutrient. The combustion of methane liberates less energy than that of hydrogen : (5.32) 02 + ^ CH4 > h CO2 + H2O + 106 kcal but considerably more than the oxidation of ammonia or ferrous iron. The Benzene and Toluene Oxidizers. — These bacteria, discovered by Tausson (1929), seem to be similar to Sohngen's methane bacteria in that they too use the energy of oxidation of a hydrocarbon for the synthesis of organic matter from carbon dioxide. Carbon Oxidizers. — We mention lastly the carbon bacteria of Potter (1908) which can live autotrophically by oxidizing solid carbon to carbon dioxide : (5.33) C + O2 > CO2 + 94 kcal 2. Efficiency of Chemauto trophic Bacteria The efficiency of the autotrophic bacteria can be expressed in three different ways: by the molecular ratio (AO2 consumed by oxidation divided by ACO2 reduced to organic matter; the latter quantity being determined either directly, or from the amount of synthesized organic material) ; by the ratio of energies (AHr accumulated in synthesis divided by AHo liberated by oxidation); and by the corresponding ratio of the free energies (— AFr accumulated to AFo dissipated). Only the first ratio is derived directly from experiments. The calculation of the last two is based on the fiction that all of the oxidation substrate is completely EFFICIENCY OF CHEMAUTOTROPHIC BACTERIA 119 oxidized by oxygen, while carbon dioxide is reduced independently by water (c/. page 235). The efficiency has been determined for several species, and the results have been discussed, among others, by Baas- Becking and Parks (1927), Burk (1931), and Stern (1933). Many measurements have not been very reHable, so that most of the figures collected in table 5. VII should be considered as preliminary. Table 5.VII Efficiency of Autotrophic Bacteria Type Species Re- action AG 2 AGO 2 Ai/R ^Ho (%) AFr " AFo (%) Observer NlTRIFIEBS Nitrosomonas Nitrobacter (5.20a) (5.20b) 47 66 46 5.4 4.0 8.5 5.9!' 7.4= Vinogradsky (1890) Vinogradsky (1890) Meyerhof (1916, 1917) Sulfur bacteria Thiobacillus thiooxidans Thiobacillus denitrificans Thiobacillus thioparus Thiobacillus novellus Thiobacillus denitrificans (5.22) (5.23) (5.24) (5.25) (5.26) 18 (1.5)'* 43 19 21 9 6.5 (78)<* 6 4.8 4.9 ~13 7.9 (96)" 6 6.5 6.5 ~13' Waksman, Starkey (1922) Vogler, Umbreit (1942) Beijerinck (1920) Starkey (1935) Starkey (1935) Lieske (1912) Hydrogen bacteria Bacillus picnoticus (5.31) 2.5 4 32 20 42/ 26/ Ruhland (1924) Methane bacteria Bacillus methanicus (5.32) (3.5-4.5) » ~20 (22-29)1' ~5 (26-34)0 ~6 Sohngen (1906) " AF of {CH2O) synthesis, for p(C02) = 3 X 10-« atm. and p(02) = 0.2 atm., is 118 kcal (c/. Table. 3 V) ' Calculated for [NH4+] = 5 X 10"= moleA- and [H+] = lO-s mole/1. ' Calculated for [NO2-] = 3 X lO"' mole/I. , . • "* We put these values in parentheses because we believe (c/. page 226) that they apply to primary carbon dioxide fixation rather than to carbon dioxide reduction. The same may be true of the low ratios AO2/ACO2 observed by Sohngen for methane bacteria. i>t o /a 'Baas-Becking and Parks calculated 11%. They thought that Lieske's yields refer to 1 g. Na2S203 instead of 1 g. Na2S!03-5 H2O. / 42% is maximum, and 26% the average. ... o-i, The larger values were calculated by Baas-Becking and Parks from the gasometnc data of bohngen, while the smaller were obtained from the permanganate titration of organic matter. Cf. footnote "■ above. For the majority of organisms in table 5. VII, the "energy yields" are of the order of 5-6%, and the "free energy yields" of the order of 6-8%. The difference is due to the fact that the free energies of oxi- dations leading to the formation of electrolytes often are considerably less negative than the corresponding total energies. Although low, the efficiencies in table 5. VI I are similar to the best efficiencies of the utilization of light energy by green plants under natural conditions. Thus, if chemautotrophic organisms did not succeed in spreading over the whole surface of the earth, as did the green plants, it was not for lack of efficiency, but merely because chemical energy is available only in a few nonequilibrated spots — sulfur springs, coal mines, iron carbonate waters, marsh gases, etc., while sunlight flows abundantly everywhere. 120 PHOTO- AND CHEMOSYNTHESIS OF BACTERIA CHAP. 5 Hydrogen bacteria occupy a unique position in table 5. VII because of their high efficiency. (The high values given for methane bacteria are unreHable, because the gas balance does not agree with any reasonable equation, and the amount of organic matter, as determined by the permanganate method, indicates a much smaller yield. The value given for the delayed chemosynthesis of T. thiooxydans also is of doubtful meaning, as discussed on page 114.) The figures for B. picnoticus are the most reliable of all, having been derived from a series of complete gas analyses by Ruhland (Table 5. VI). The values for the ratio, AH2/AO2, found in these experiments, varied between 2.1 and 2.8 de- pending on the state of the culture. If one neglects the oxygen con- sumption by respiration (a correction for this process would raise the efficiency still higher) a ratio of 2.8 means that for every two hydrogen molecules burned to water, 0.8 molecules of hydrogen are used for the reduction of carbon dioxide. This corresponds to the reduction of 0.4 molecules of carbon dioxide, and leads to the maximum ratio, AO2/ACO2 = 2.5, given in table 5. VII, corresponding to the utilization of 32% of available energy and 42% of available free energy. Even if one uses the average value of AH2/AO2 in table 5. VI (~ 2.5), one calcu- lates that only four oxygen molecules are required for the reduction of one molecule of carbon dioxide, and obtains an energy utilization of 20%, and a free energy utilization of 26%. Gaffron found (cf. page 140), for hydrogen-adapted green algae, a maximum ratio of AH2/AO2 = 3, and considered this value as the theoretical maximum, probably valid also for the hydrogen bacteria. Burk's (1931) calculation which gave a 100% efficiency for the chemosynthesis of hydrogen bacteria, was based on the assumption that the average ratio (0.52) in the last column of table 5.VI represents, not a confirmation of the theoretical stoichiometric value (0.5) — cf. equation (5.6) — but the quantity of hydrogen actually combusted to water to provide energy for the reduction of one mole carbon dioxide by one mole of water (as if all the rest of hydrogen oxidized by the bacteria had nothing to do with the reduction process!). With this assumption, the result of Burk's calculation became a simple consequence of the fact that the free energy of the reaction, 2 H2 + CO2 > {CH20i + H2O, is approximately zero. We can see no relation between his elaborate calculations and the problem of the true thermodynamic efficiency of the hydrogen bacteria. (This was noted also by van Niel, 1943.) 3. Methane-Producing Bacteria and other Cases of Carbon Dioxide Absorption by Heterotrophants We have described the methane-oxidizing bacteria together with other chemautotrophic species, because their use of methane is similar to the use of hydrogen sulfide, sulfur, thiosulfate, or ammonia by "true" autotrophic bacteria. Many other allegedly "heterotrophic" bacteria can live on one chemicall}^ pure organic substrate, and it is quite possible METHANE-PRODUCING BACTERIA 121 that they, too, use it mainly or exclusively as a source of hydrogen and energy, but prepare their cell matter by the reduction of carbon dioxide. However, the metabolism of most of these bacteria is not sufficiently known to allow one to assert that they do not use at least a part of the organic substrate for direct heterotrophic assimilation — especially since we know, from the example of the purple Athiorhodaceae, that synthesis of carbohydrates by the reduction of carbon dioxide can often be coupled with heterotrophic assimilation of one part of the reductant. Only one type of bacteria which uses organic substrates for the reduction of carbon dioxide shall be described here, the methane-producing bacteria, which were discovered in 1910 by Sohngen (who had previously discovered the methane-burning bacteria). IMethane is produced by the fermentation of many organic substrates; these processes have been investigated, e. g., by Neave and Buswell (1930), Fischer, Lieske and AVinzer (1931, 1932), Stephenson and Stickland (1933), Barker (1936i'2, 1937), and Barker, Ruben and Kamen (1940). One thinks, at first, that methane must be the product of dismutation of an organic substrate, as, for example, in the simplest case of the acetate fermentation: (5.34) CH3COOH > CH4 + CO2 - 6 kcal (Decarboxylation can be considered as a special case of dismutation in which one part of the molecule is oxidized to carbon dioxide.) However, Sohngen noticed that the same bacteria which cause methane fermenta- tion of organic substrates, also reduce carbon dioxide to methane in the presence of molecular hydrogen : (5.35) CO2 + 4 H2 > CH4 + 2 H2O + 62 kcal More recently, Barker (1936-) found a species of methane bacteria which reduces carbon dioxide to methane by means of ethanol: (5.36) CO2 + 2 CaHsOH > CH4 + 2 CH3COOH + 21 kcal These examples make it probable that even in methane fermentations which proceed with a net liberation of carbon dioxide, as in (5.34), the way to methane leads through carbon dioxide. Arguments in favor of this hypothesis were adduced by Barker (1936, 1937). He found, for example, that in the gradual decomposition of butanol by methane bacteria, the first stage conforms to the equation: (5.37) 2 C4H9OH + CO2 > 2 C3H7COOH + CH4 + 21 kcal and the second stage to the eciuation: (5.38) 2 C3H7COOH + CO2 + 2 H2O > 4 CH3COOH + CH4 + 9 kcal while, in the third stage, four molecules of carbon dioxide are liberated, according to the over-all equation (5.34), thus giving a net production 122 PHOTO- AND CHEMOSYNTHESIS OF BACTERIA CHAP. 5 of one molecule of carbon dioxide for each fermented molecule of butanol : (5.39) CiHsOH + H2O > 3 CH4 + CO2 + 7 kcal It seems plausible to assume, in analogy to steps (5.37) and (5.38), that the last step in butanol fermentation, the decomposition of acetic acid, also involves the participation of carbon dioxide, i. e., proceeds not according to the "abbreviated" equation (5.34), but by a true oxidation- reduction: (5.40) CH3COOH + CO2 > 2 CO2 + CH4 - 6 kcal A direct proof of the participation of free carbon dioxide in the for- mation of methane in reactions which proceed with the net liberation of carbon dioxide was achieved by means of radioactive carbon, C*. Barker, Ruben and Kamen (1940) showed that the methane fermentation of inactive ethanol by Methanosarcina methanica, in the presence of radioactive carbon dioxide, gives active methane, thus estabhshing the correctness of the equation: (5.41) 4 CH3OH + 3 C*02 > 3 C*H4 + 2 H2O + 4 CO2 + 51 kcal and precluding the "cancelling out" of three carbon dioxide molecules on each side of this equation. A similar interpretation of acetate fermentation, assumed in (5.40), thus becomes increasingly plausible. At first sight, the reduction of carbon dioxide to methane by the methane bacteria appears as a biochemical "art for art's sake," since the product escapes as a gas, carrying with it the accumulated energy. However, Barker, Ruben and Kamen noticed that about 10% of radio- activity supplied in the form of C*02 is found afterwards in the cell material. This shows that, while a large part of reduced carbon dioxide is wasted, a small proportion is utilized for the synthesis of the cell material. This reminds one of the autotrophic bacteria which dissipate most of the available oxidation energy, in order to reduce a small quantity of carbon dioxide to carbohydrate. It seems possible that the methane bacteria have solved a similar problem in a different way (the usual solution being precluded by their anaerobic mode of life). Deprived of oxygen, they cannot derive energy from the autoxidation of the available substrate. Their solution is to use carbon dioxide as an oxidant. We know that none of the available oxidation substrates — not acetate, or methanol, or even hydrogen — has sufficient reducing power to bring about the stoichiometric reduction of carbon dioxide to carbohydrate. However, the reduction of carbon dioxide to methane requires less energy per transferred hydrogen atom than the "halfway" reduction to carbo- hydrate. This is why the reactions (5.35) - (5.41) are exothermal — with the exception of reaction (5.40), which, however, has a positive free energy of about 10 kcal. ROLE OF AUTOTROPHIC BACTERIA IN NATURE 123 The fact that the methane bacteria are capable of reducing carbon dioxide to methane, shows that they have developed a mechanism which avoids the intermediate formation of a carbohydrate (because the latter would present an "energy barrier" which is insurmountable at ordinary temperatures). Under these conditions, it seems possible that the large-scale, exothermal reduction of carbon dioxide to methane may be used by the methane-liberating bacteria to the same purpose as the large-scale, exothermal oxidation of autoxidizable substrates is used by the autotrophic bacteria, namely, to provide energy for the reduction of a relatively small proportion of carbon dioxide to a carbohydrate. In the same class with the methane-producing bacteria may perhaps be placed the species of Clostridium, which reduce carbon dioxide to acetic acid by means of hydrogen (Wieringa 1936): (5.42) 2 H2 + 2 CO2 > 2 H2O -H CH3COOH + 68 kcal or by means of various purines {Clostridium acidi urici, Barker, Ruben and Beck 1940). We shall not continue with the enumeration of bacterial and other biological systems which have been found capable of absorbing carbon dioxide and incorporating it into organic matter. Although some of them probably carry out a " chemosynthetic " reduction of carbon dioxide, similar to that achieved by the microorganisms described above, the most important examples worked out so far appear to belong to a different type, that of enzymatic carboxylations. It is customary to speak of "reduction of carbon dioxide" whenever this compound is bound in an organic molecule. However, it is advisable to distinguish clearly between true reduction of carbon dioxide and carboxylation, carhamination and similar "additive" reactions of carbon dioxide with organic molecules. Whether carboxylations should be called reductions at all, is a matter of definition. In chapter 8 arguments will be presented in favor of not using this designation. This convention would prevent misunder- standings which have led to the use of expressions like "dark assimila- tion," or even "dark photosynthesis" for processes in which carbon dioxide was merely added to existing organic compounds. Carboxylation is important from the point of view of photosynthesis, not as an analogy to the main photosynthetic process, but as a possible way of entry of carbon dioxide into the photosynthetic apparatus. It will therefore be considered in detail in chapter 8, which deals with the immediate fate of carbon dioxide in photosynthesis. C. The Role of Autotrophic Bacteria in Nature Bacterial metabolism is of great importance for the elucidation of the chemical mechanism of photosynthesis. It indicates that photosynthesis consists of two distinct stages, the reduction of carbon dioxide, and the 124 PHOTO- AND CHEMOSYNTHESIS OF BACTERIA CHAP. 5 oxidation of water, and that the second stage can be changed and other reductants substituted for water without affecting the first one. Chemo- synthesis by autotrophic bacteria makes it plausible that the reduction of carbon dioxide is a nonphotochemical process, which can be brought about by the intermediates of the photochemical oxidation of water (or other reductants), as well as by products of exothermal chemical reactions. These problems will be discussed more extensively in chapters 7 and 9. Another interesting question which arises from the study of the photosynthesis and chemosynthesis of autotrophic bacteria, concerns the role which these processes may have played in the development of life on earth. Prior to van Niel's interpretation of the mechanism of bacterial photosynthesis, the synthesis of organic matter by green plants appeared as a unique process, unrelated to all other biochemical reactions in living organisms. Van Niel's investigations have established the long-missing link between the world of green plants and that of the lower micro- organisms. Green plants reduce carbon dioxide in light by means of water; green and purple sulfur bacteria reduce carbon dioxide, also in light, by means of hydrogen sulfide; colorless sulfur bacteria reduce carbon dioxide, by means of hydrogen sulfide, without light. This comparison shows the existence of a hierarchy of autotrophic organisms, and en- courages speculations as to the genetic relationships between them. In considering the present state of life on earth, one is struck by the paradox "no life without chlorophyll — no chlorophyll without life." The large-scale formation of organic matter from inorganic materials has as its prerequisite the existence of complex organic molecules, such as chlorophyll and various enzymes, without which photosynthesis appears impossible, but which themselves cannot be synthesized in nature outside the living cell. Obviously, photosynthesis could not have started on earth without the previous existence of living matter. The existence of chemauto- trophic and photautotrophic bacteria shows a possible development. It was mentioned on page 82 that the first organic molecules may have arisen on earth by photochemical reactions of inorganic compounds in ultraviolet light, or by the action of electric discharges in the atmosphere. Which of these molecules first acquired the capacity of propagation by self -duplication, which is the first sign of life, we cannot surmise; but we can imagine a continuous " chemosynthetic " development leading from this molecule to autotrophic bacteria. At that time, the earth was less settled in its chemical ways than now, and not only hydrogen sulfide, but also free hydrogen might have been available in the atmosphere. From colorless autotrophic bacteria, the development might have BIBLIOGRAPHY TO CHAPTER 5 125 progressed to purple bacteria, and hence to green plants. The transition from bacteria to algae, which liberated the plants from the dependence on uncertain and dwindling supplies of unstable hydrogen donors, has allowed life to spread over the whole surface of the globe. The capacity of certain green algae for adaptation to hydrogen (c/. Chapter 6) may be a reminiscence of their genetic relationship to photoreducing bacteria. Bibliography to Chapter 5 Photosynthesis and Chemosynthesis of Bacteria A. Phoiosynihesizing Bacteria 1883 Engelmann, Th. W., Arch. ges. Physiol. Pfliigers, 30, 90. 1887 Vinogradsky (Winogradsky), S., Botan. Z., 45, 489. 1888 Vinogradsky, S., Zur Morphologie und Physiologic der Schwefelbacterien. O. Felix, Leipzig, 1888. Engelmann, Th. W., Arch. ges. Physiol. Pfliigers, 42, 183. Engelmann, Th. W., Botan. Z., 46, 661 Engelmann, Th. W., Arch, neerland. sci., 23, 151 1907 Molisch, H., Die Purpurbacterien. Fischer, Jena, 1907. 1914 Skene, M. G., New Phytologist, 13, 1. 1919 Buder, J., Jahrb. wiss. Botan., 58, 525. 1920 Buder, J., Naturwissenschaften, 8, 261 1921 Dangeard, P. A., Botaniste, 14, 1. 1927 Dangeard, P. A., ibid., 19, 1. 1930 van Niel, C. B., Contrib. Hopkins Marine Biol. Sta., 1930, 161. 1931 van Neil, C. B., Arch. Mikrobiol., 3, 1. van Niel, C. B., and MuUer, F. M., Rec. irav. botan. neerland., 28, 245. 1933 Muller, F. M., Arch. Mikrobiol., 4, 137. Gaffron, H., Biochem. Z., 260, 1. 1934 Roelefson, P. A., Proc. Acad. Sci. Amsterdam, 37, 660. Gaffron, H., Biochem. Z., 269, 447. 1935 Gaffron, H., ibid., 279, 1. Gaffron, H., ibid., 275, 301. van Niel, C. B., Cold Spring Harbor Symposia Quant. Biol., 3, 138. 1936 Czurda, V., Arch. Mikrobiol, 7, 110. van Niel, C. B., Bull, assoc. diplomes microbial, faculte pharm. Nancy, No. 13. van Niel, C. B., Arch. Mikrobiol, 7, 323. French, C. S., Science, 84, 575. Barker, H. A., /. Cellular Comp. Physiol, 8, 231. Giesberger, G., Dissertation, Univ. Utrecht, 1936. 1937 Czurda, V., Repl Proc. 2nd Intern. Congr. Microbiol London, 1937, 470. Clifton, C. E., Enzymologia, 4, 246. Nakamura, H., Acta Phytochim. Japan, 9, 189. 126 PHOTO- AND CHEMOSYNTHESIS OF BACTERIA CHAP. 5 1937 Nakamura, H., Acta Phytochim. Japan, 10, 211. French, C. S., J. Gen. Physiol, 20, 711. French, C. S., ibid., 21, 71. Sapozhnikov, D. I., Microbiology USSR, 6, 643. van Niel, C. B., Ann. Rev. Biochem., 6, 595. 1938 Winzler, R. J., and Baumberger, J. P., J. Cellular Comp. Physiol, 12, 183. Eymers, J. G., and Wassink, E. C, Enzymologia, 2, 258. Nakamura, H., Acta Phytochim. Japan, 10, 259. Nakamura, H., ibid., 10, 297. 1939 Clifton, C. E., and Logan, W. A., /. Bad, 37, 523. Wessler, S., and French, C. S., J. Cellular Comp. Physiol, 13, 327. Nakamura, H., Acta Phytochim. Japan, 11, 109. 1940 Winzler, R. J., /. Cellular Comp. Physiol, 15, 343. Foster, J. W., J. Gen. Physiol, 24, 123. 1941 van Niel, C. B., in Advances in Enzymology, Vol. 1. Interscience, New York, 1941, p. 263. B. Chemosynthesizing Bacteria 1887 Vinogradsky, S., Ann. inst. Pasteur, 1, 548. 1888 Vinogradsky, S., Botan. Z., 46, 262. 1890 Vinogradsky, S., Ann. inst. Pasteur, 4, 213. 1902 Natansohn, A., Mitt. Zool Stat. Neapel, 15, 665. 1903 Beijerinck, M. W., and van Delden, A., Zentr. Bakt. Parasitenk., II, 2, 33. 1904 Beijerinck, M. W., ibid., 16, 681. 1906 Kaserer, H., ibid., 16, 681. Sohngen, N. L., ibid., 15, 513. 1907 Nabokich, A. J., and Lebedev, A. J., ibid., 17, 350. 1908 Potter, M., Proc. Roy. Soc. London B, 80, 739. Niklewski, B., Zentr. Bakl Parasitenk., II, 20, 469. Lebedev, A. J., Biochem. Z., 7, 1. 1909 Lebedev, A. J., Ber. deut. botan. Ges., 27, 598. 1910 Niklewski, B., Jahrb. wiss. Botan., 48, 113. Sohngen, N. L., Rec. trav. chim., 29, 238. 1911 Lieske, R., Jahrb. wiss. Botan., 49, 91. 1912 Lieske, R., Ber. deut. botan. Ges., 30, 12. 1915 Miinz, E., Dissertation, Univ. Halle, 1915. 1916 Meyerhof, 0., Arch. ges. Physiol. Pfliigers, 164, 353. 1917 Meyerhof, 0., ibid., 166, 240. 1919 Lieske, R., Zentr. Bakt. Parasitenk., II, 49, 413. 1920 Beijerinck, M. W., Proc. Acad. Sci. Amsterdam, 22, 899. 1922 Waksman, S. A., and Starkey, R, L., J. Gen. Physiol, 5, 285. Waksman, S. A., and Joffe, J. S., /. Bad, 7, 239. 1924 Ruhland, W., Jahrb. wiss. Botan., 63, 321. 1927 Baas-Becking, L. G. M., and Parks, G. S., Physiol Revs., 7, 85. 1929 Tausson, W., Planta, 7, 735. BIBLIOGRAPHY TO CHAPTER 5 127 1930 Neave, S. L., and Buswell, A. M., J. Am. Chem. Soc, 52, 3308. 1931 Fischer, F., Lieske, R., and Winzer, K., Biochem. Z., 236, 247. Burk, D., /. Phys. Chem., 35, 432. 1932 Fischer, F., Lieske, R., and Winzer, K., Biochem. Z., 245, 2. 1933 Stephenson, M., and Stickland, L. H., Biochem. J., 27, 1517. Stern, K., Pflanzenthermodynamik. Springer, Berlin, 1933. 1935 Starkey, R. U., Soil Sci., 39, 97. Starkey, R. U., J. Gen. Physiol, 18, 325. 1936 Barker, H. A., Arch. MikrobioL, 7, 404. Barker, H. A., ibid., 7, 420. Wieringa, K. T., Antonie van Leeuwenhoek J. Microbiol. SeroL, 3, 1. Wieland, H., and Pistor, H. T., Ann., 522, 116. 1937 Barker, H. A., Arch. MikrobioL, 8, 415. Happold, F. C, and Key, A., Biochem. J., 31, 1323. 1938 Wieland, H., and Pistor, H. T., Ann. Chemie {Liebig), 535, 205. 1939 Stephenson, M., Bacterial Metabolism. 2d ed., Longmans, Green, London, 1939. 1940 Barker, H. A., Ruben, S., and Kamen, M. D., Proc. Natl. Acad. Sci. U. S., 26, 426. Barker, H. A., Ruben, S., and Beck, J. V., ibid., 26, 477. 1941 Vogler, K. G., and Umbreit, W. W., Soil Sci., 51, 331. 1942 Vogler, K. G., J. Gen. Physiol., 25, 617. Umbreit, W. W., Vogel, H. R., and Vogler, K. G., J. Bad., 43, 141. Vogler, K. G., LePage, G. A., and Umbreit, W. W., J. Gen. Physiol, 26, 89. Vogler, K. G., ibid., 26, 103. Vogler, K. G., and Umbreit, W. W., ibid., 26, 157. 1943 van Niel, C. B., Physiol. Revs., 23, 338. Compilations of Thermochemical Data and Free Energies, see Bibliography to Chapter 3 Chapter 6 THE METABOLISM OF ANAEROBICALLY ADAPTED ALGAE * 1. The Adaptation of Algae to Hydrogen and Hydrogen Sulfide In the preceding chapter, we found that the photosynthesis of bacteria is strikingly adaptable. The photosynthesis of green plants, on the other hand, has long been considered as a rigid process, which can be accelerated or retarded by external influences, but whose chemical mechanism is unalterable. This is, however, not universally true. Nakamura (1937, 1938) found that certain diatoms (Pinnularia) and blue-green algae {Oscillatoria) can use hydrogen sulfide for the reduction of carbon dioxide —in other words, can adopt a metabolism similar to that of the purple sulfur bacteria. Ordinarily, the photosynthesis of green plants is inhibited by hydrogen sulfide (c/. page 315); but Nakamura's algae consumed carbon dioxide even in presence of this gas. The evolution of oxygen, however, was replaced by the deposition of sulfur globules in the cells. This interesting phenomenon certainly deserves more than the cursory attention it has received in Nakamura's work. Much more detailed has been the study which Gaffron devoted to certain unicellular green algae which, after a period of anaerobic incubation, become able to utiHze molecular hydrogen or organic hydrogen donors as reductants in photo- synthesis, that is, adopt a metaboHsm reminiscent of the autotrophic or heterotrophic Athiorhodaceae. The adaptation of green algae to molecular hydrogen was discovered by Gaffron in 1939, and investigated in a series of important papers (Gaffron 1939, 19401-2; 194212; Gaffron and Rubin 1942, reviews Franck and Gaffron 1941, Gaffron 1943). In studying "induction effects" in plants after anaerobiosis in the dark, Gaffron found that some unicellular green algae {Scenedesmus, for example) do not react to this "anaerobic incubation" by a temporary inhibition of gas exchange in light, as do the higher plants, but by a more-lively-than-usual liberation of gas. This "inverse induction" was later found to be caused by a liberation of hydrogen, in addition to (or instead of) the usual exchange of carbon dioxide and oxygen. * Bibliography, page 148. 128 HYDROGEN ADAPTATION AND DE-ADAPTATION 129 In studjdng this phenomenon Gaffron (1939, 1940) found that algae which were capable of liberating hydrogen, were also able to absorb it, if placed in an atmosphere containing a high proportion of this gas. Hydrogen evolution and consumption can be observed even in darkness; but both processes are accelerated by light. The hydrogen exchange continues, gradually decreasing, until the available cellular "hydrogen acceptors" are entirely saturated with hydrogen, or until the available "hydrogen donors" are exhausted. In presence of an added hydrogen acceptor, the absorption of hydrogen can continue for a much longer time, and the same is true of hydrogen liberation in presence of an added donor. Appropriate hydrogen acceptors are oxygen (in small quantities, since larger quantities of this gas cause de-adaptation), and carbon dioxide; while glucose and other organic substrates can act as hydrogen donors. Thus, hydrogen-adapted algae are capable of bringing about the following reactions: In the Dark: (I) and (11) .—Absorption of hydrogen from an atmosphere containing a high proportion of this gas, and evolution of hydrogen into an atmosphere of pure nitrogen (so-called "hydrogen fermentation"). (JU) .—Simultaneous absorption of hydrogen and oxygen (so-called " oxyhydrogen " or "Knallgas" reaction). (lY).— Reduction of carbon dioxide coupled with the oxyhydrogen reaction (III), a process analogous to the metabohsm of autotrophic hydrogen bacteria (page 116). In Light: (V) and (VI).— Enhanced hydrogen absorption in an atmosphere of hydro- gen, and enhanced hydrogen evolution in an atmosphere of nitrogen. The first-named process may, however, be identical with reaction (VII), i. e. it may represent the photoreduction of carbon dioxide produced by acid fermentation, rather than the hydrogenation of an organic hydrogen acceptor. (VII) and {Ylll).— Photosynthesis from carbon dioxide and hydrogen or from carbon dioxide and organic hydrogen donors — processes reminiscent of the metabohsm of auto- trophic and heterotrophic purple bacteria respectively. Gaffron designated these reactions as "photoreductions," and although this term is not very specific, it may be used as a short substitute for "photoreduction of carbon dioxide by reductants other than water," while the term "photosynthesis" is retained to mean "photoreduction of carbon dioxide by water." (To be consistent, one should use the term "photoreduction " also when speaking of the metabohsm of purple bacteria, a terminology which was not rigidly adhered to in chapter V.) 2. The Mechanism of Hydrogen Adaptation and De-adaptation Before discussing the metabolic reactions of hydrogen-adapted algae, we shall deal with the processes of adaptation and de-adaptation (the latter called "reversion" by Gaffron). Not all unicellular green algae can be adapted to hydrogen. Experi- ments with Chlorella, as well as with diatoms (two strains of Nitzschia) and blue-green algae (Oscillatoria), gave no positive results. No generic relationships are apparent: Scenedesmus, Ankistrodesmus and Raphidium 130 ANAEROBICALLY ADAPTED ALGAE CHAP. 6 (three genera which have been successfully adapted) are no more closely related between themselves than they are to Chlorella. Adaptation requires at least two hours of anaerobic incubation at 20° C, less at higher temperatures. At 35°, the hydrogen metabolism of Scenedesmus starts almost immediately upon the removal of oxygen; however, this temperature rapidly causes an irreversible injury to the algae. During the adaptation period, the algae ferment, as all green plants do when the oxygen pressure is below that corresponding to the Pasteur effect (c/. Genevois 1927), producing carbon dioxide and non- volatile acids; the rate of this "acid fermentation" is about the same in hydrogen and in nitrogen. When fermentation has proceeded for a certain time, Scenedesmus and similar algae become capable of picking up hydrogen and this rapidly completes their adaptation. Gaffron suggested that an enzyme, which is usually present in the chloroplasts in an inactive, oxidized form (we may designate it by EhO) is reduced by a fermentation product H2F : (6.1) EhO + HjF >Eh + F + H20 and by this acquires the properties of a hydrogenase, i. e., of a reversible acceptor for molecular hydrogen : (6.2) E H + H2 . H2E H capable of transmitting this hydrogen to other substrates: (6.3) H2EH + R > H2R -1- Eh The effect of a hydrogenase on photosynthesis can be understood if one assumes that the role of the oxidant R in (6.3) can be played by the oxidation intermediates of photosynthesis, whose conversion into free oxygen is thus effectively blocked (c/. Scheme 6.1, page 136). If we use the most general formulation of the primary photochemical process (Eq. 7.10a) and designate the primary oxidation product as Z, we may postulate that in hydrogen-adapted algae, reaction (6.4) : (6.4) 2Z + H2EH >2HZ + Eh takes the place of reaction (7.10b) and photosynthesis is thus converted into "photoreduction." To explain the long "incubation period" and the rapid completion of adaptation after the absorption of hydrogen has finally set in, Gaffron suggested that the activation of the hydrogenase is accelerated auto- catalytically by the reaction: (6.5) HjEh + EhO >2Eh + H20 which is the reverse of a dismutation, and may perhaps be designated as a HYDROGEN ADAPTATION AND DE- ADAPTATION 131 "commutation." As soon as some reduced and hydrogenated enzyme, H2EH, has been formed, reaction (6.5) allows the organism to dispense with the slow incubation reaction (6.1). Enzymes capable of introducing molecular hydrogen into cellular metabolism have been discovered by Stephenson and Stickland (1931) in certain colorless bacteria (as B. coli) and later found by Roelefson (1934), Gaffron (1935) and Nakamura (1937, 1938) in the nonsulfur purple bacteria Rhodovibrio, Rhodohacillus palustris and Rhodospirillum giganteum, and in the sulfur bacteria Thiocystis and Chromatium minutis- simum. In each case, a variety of organic and inorganic substrates (methylene blue, fumarate, nitrate, oxygen, etc.) were found to be suitable as hydrogen acceptors. Their assortment is different for different organisms, so that one can conceive of the existence of a number of different, acceptor-specific hydrogenases. However, Yamagata and Nakamura (1938) concluded, from comparative cyanide inhibition experiments with B. coli formicum, Rhodohacillus palustris and B. del- hruckii, that the hydrogenase in all these organisms is the same, and that it donates hydrogen to a common intermediate acceptor, after which specific oxidoreductases (or an oxidase) transfer it to different final acceptors. We may assume that the same hydrogenase and the same intermediate acceptor are present also in anaerobically incubated algae. We designate the primary hydrogen acceptor by Ah, and spht equation (6.3) into the two equations (6.6a) and (6.6b); and equation (6.4) — into the two equations (6.6a) and (6.6c): (6.6a) HzEh + Ah ^ HaAn + Eh (6.6b) H2AH + R > HoR + Ah or (6.6b') H2R' + A H > H2A H + R' (6.6c) H2AH + 2Z > 2HZ + Ah Reaction (6.6a) must be reversible (since adapted algae can either absorb or liberate hydrogen). As for reaction (6.6b), its direction may depend not only on concentrations, but also on the specific nature of the metabolites R present in the cell (as expressed by the alternative equation 6.6b'). De-adaptation occurs, in presence of carbon dioxide, if the intensity of illumination is raised beyond a certain threshold. The further this threshold is exceeded, the more rapid is the return to normal photosyn- thesis (c/. Fig. 14, p. 145). The photochemical de-adaptation is irrevers- ible, i. e., hydrogen absorption is not resumed upon return to low light intensity (Fig. 8). However, the adapted state can be restored much more rapidly immediately after de-adaptation than after a prolonged period of aerobic photosynthesis, probably because the autocatalytic 132 ANAEROBICALLY ADAPTED ALGAE CHAP. 6 mechanism (6.5) provides for rapid re-adaptation whenever small amounts of the hydrogenated enzyme still are present. De-adaptation can be enforced also in the dark, by means of oxygen. While 0.5% oxygen (constantly renewed to prevent exhaustion by respiration) is sufficient to prevent adaptation, the tolerance for oxygen may rise, after adaptation, to 1 or 2%. This is due to the capacity of the adapted cells for the oxyhydrogen reaction (III) ; only when the rate of oxygen fixation by the cells becomes higher than the maximum possible rate of this reaction, does de-activation become inevitable. c o c u » i- a. E E o O O ■^ V 40 - '^O' >>. _i '^o\ O o \ o ■80 %\ X in 3 1 120 - , 1 1 1 1 1^ 1 20 40 60 80 Time, minufcs 100 120 Fig. 8. — The "de-adaptation" of anaerobically adapted Scenedesmus by light increase from 500 to 5000 lux is not reversed by return to 500 lux (after Gaff r on 1941). If the reduction of an enzyme is the basis of adaptation, its reoxidation must be the basis of de-adaptation. This oxidation may be attributed either to free oxygen, or to cellular oxidants, formed as intermediates either in the photochemical reduction of carbon dioxide (in the case of photochemical de-adaptation), or in the oxyhydrogen reaction (in the case of dark de-adaptation). If one assumes that photochemical and dark de-adaptation are both caused by free oxygen, one cannot help noticing the difference between the partial pressure of oxygen at the moment of photochemical de-adapta- tion (which is exceedingly low) and the comparatively high pressure required for de-adaptation in the dark. To explain this difference, one could suggest that when oxygen is produced photochemically within the cell, a high internal pressure has to be built up before any gas can escape into the atmosphere; so that chemical and photochemical de-adaptation could correspond to the same internal oxygen tension in the chloroplasts. Gaffron argued, however, that experiments with the oxygen electrode {cf. Volume II, Chapter 33), as well as observations with luminous bac- HYDROGEN ADAPTATION AND DE-ADAPTATION 133 teria and fluorescent dyestuffs prove that oxygen appears in the sur- rounding medium within 0.01 second after the beginning of photo- synthesis, and that therefore the internal oxygen tension cannot be markedly different from its external pressure. If this is so, then photochemical de-adaptation, at least, must be attributed to an intermediate of photosynthesis, and not to free oxygen. We shall designate this ''oxygen precursor" as {O2}, with the braces indicating an "acceptor" or "carrier" molecule. The oxidant {O2} is not likely to be the direct product of the primary photochemical process. This is indicated by inhibition experiments (Chapter 12), and by the observation of Rieke and Gaffron (1943) that the de-adapta- tion in flashing light occurs at the same average intensity of illumination as does the de-adaptation in continuous light. We therefore assume, with Gaffron, that (at least) two successive enzymatic reactions are required for the conversion of the primary photochemical product Z into O2 (c/. Eqs. 7.10b and c): (6.7a) 2Z + H2O >H02!+2HZ (6.7b) §102) ^^02 We attribute the photochemical de-adaptation to the reaction : (6.8) {02!+ 2 Eh >2EhO The rate at which the intermediate oxidant {O2} is produced in light, must decrease with decreasing concentration of carbon dioxide, at least in a certain range of concentrations (c/. Chapter 27, Vol. II). The tolerance of the adapted state for hght should rise under these conditions. In fact, if the carbon dioxide formed by fermentation is removed by an alkaline absorber, the adapted state can be preserved in light which would otherwise cause a rapid de-adaptation. Another treatment which prevents photochemical de-adaptation, is poisoning with comparatively large quantities of hydroxylamine (c/. Chapter 12). Apparently, this agent prevents the conversion of the primary photochemical oxidation product into the hydrogenase-destroy- ing intermediate (02}, i. e., it inhibits reaction (6.7a). Finally, the tolerance of adapted algae for light can be increased also by the provision of added oxidation substrates, e. g., glucose. They either accelerate the removal of the primary oxidation products, HZ, and thus prevent the formation of the oxidant {O2} , or reduce this oxidant in competition with the hydrogenase. Having thus attributed the photochemical de-adaptation to accumu- lation of an intermediate oxidant {O2}, we ask whether dark de-adapta- tion should be ascribed to a similar agent, or to free oxygen. Gaffron 134 ANAEROBICALLY ADAPTED ALGAE CHAP. 6 pointed out, in support of the second viewpoint, that the maximum rate of hydrogen consumption by the oxyhydrogen reaction, attained when dark de-adaptation sets in, is approximately equal to the maximum rate of hydrogen consumption by photoreduction, reached just prior to photochemical de-adaptation. This equality finds a plausible explanation in the assumption that de-adaptation is caused by oxidation inter- mediates, which in both cases must be removed by the hydrogenase system. As long as the removal keeps pace with the photochemical or enzymatic supply of the oxidants, the adapted state is stable; whenever the supply becomes too rapid, an accumulation of oxidants occurs and brings about the de-activation of the hydrogenase. In other words: the maximum attainable rates of photoreduction (in light) and of the oxyhydrogen reaction (in the dark) are the same, because they are both limited by the quantity of available hydrogenase. One may further ask whether the intermediate oxidant which causes de-activation in the dark is identical with the intermediate {O2} of photosynthesis and photoreduction, or merely similar to it in its capacity to oxidize the hydrogenase. This question is important because if the first alternative were correct, the oxygen evolution in photosynthesis (reaction 6.7b) would have to be considered as a reversible process, its direction depending on the concentration of {O2} and the partial pressure of oxygen. The following considerations speak against this concept. In the first place, almost the only known reversible oxygen acceptor in nature is hemoglobin, and it is doubtful whether a similar catalyst exists in plants (cf. Chapter 11). In the second place, Gaffron concluded from poisoning experiments that the enzyme Eo ("deoxidase," cf. Chapter 11), which catalyzes the oxygen-liberating reaction (6.7b), is de-activated, in the course of anaerobic adaptation, simultaneously with the activation of the hydrogenase. Eh. If this conclusion is correct, a reversal of reaction (6.7b) in adapted algae is impossible, even if this reaction were thermo- dynamically reversible in the first place. Gaffron's argumentation in favor of a de-activation of the enzyme, Eo, in the adapted state was as follows: Cyanide prevents adaptation; if applied after completed adaptation, it causes a slow de-adaptation in light. This is best explained by assuming that an oxidation of the hydrogenase occurs continuously during photoreduction, but does not lead to de-adaptation as long as the autocatalytic re-adaptation (reaction 6.5) holds step with the de-adapting reaction (6.8). Cyanide blocks re-adaptation (by "freezing" the hydrogenase in its oxidized state, EhO), and thus causes a cumulative de-adaptation in light. This hypothesis implies that a small quantity of the intermediate oxidants {O2} is formed even in the adapted state, where the great HYDROGEN ADAPTATION AND DE-ADAPTATION 135 majority of the primary oxidation products HZ is disposed of by reaction (6.6c). If this is true, then the absence of oxygen evolution during photoreduction indicates that the oxygen-liberating enzyme, Eo, is in an inactive state. Another argument in favor of the absence of an active enzyme, Eo, in the adapted state is, according to Gaffron, the fact that the adaptation process shares with the oxygen evolution in photosynthesis a sensitivity to very small quantities of hydroxylamine and phenantroline. To ex- plain this similarity, one can assume that the complex formation with hydroxylamine "freezes" enzyme Eo in the oxidized state, thus inhibiting its function in photosynthesis, but at the same time preventing its de- activation by reduction during anaerobic incubation. Gaffron (1943) suggested that the de-activation of Eo does not eliminate this enzyme altogether as a catalytic agent, but converts it into an "oxidase" Eo' (whose presence is revealed by the oxyhydrogen reaction). However, it is thermodynamically impossible for Eo' to catalyze the formation of the same complex {O2}, whose decomposition was catalyzed by Eo. (A catalyst has the same effect on the velocity of reaction in both directions, because it cannot shift a thermodynamic equilibrium.) Therefore, we must assume that the transformation of Eo into Eo' brings about a change in specificity— in other words, that enzyme Eo catalyzes the formation of a complex {02}' which is different from the complex {O2}, decomposed by enzyme Eo. As a result of this discussion, we attribute the de-adaptation by excess oxygen to the reactions: (6.9) O2 >!02!' (6.10) {021' + 2Eh >2EhO which is analogous to, but not identical with, reaction (6.8), and occurs whenever the removal of {02}' by the hydrogenase system (i. e., the oxyhydrogen reaction) lags behind the formation of this intermediate by reaction (6.9). It will be noted that, according to (6.9), the formation of the oxidant {O2}' in the oxyhydrogen reaction cannot be avoided; but its rapid consumption (by reaction 6.11) can prevent the deactivating reaction (6.10) from destroying the hydrogenase more rapidly than it is restored by reaction (6.5). As mentioned before, Gaffron assumed that, in photo- reduction too, a certain quantity of the oxidant {O2} is formed despite the removal of the preponderant part of the oxidation product, HZ, by reaction with the hydrogenase system. A summary of the reactions associated with the adaptation and de-adaptation phenomena is given in scheme 6.1. 136 ANAEROBICALLY ADAPTED ALGAE CHAP. 6 2(X+HZ) I [(7.10a) ^ Light I 2HX 2Z I ^ E„0 (6.1) H,F (6.5), f >• Eu + H,0 +F Va (6.2) (6.10) •-' + I I I / (6.7 b) 9) Scheme 6.1. — Reactions in anaerobically adapted algae. (Figures in parentheses refer to equations in text.) Normal photosynthesis. Adaptation and photoreduction (reaction 6.6c competes with 6.7a, thus con- verting photosynthesis into photoreduction). De-adaptation by the intermediates {O2! (in light) or j02i' (in excess oxygen). 3. The Dark Reactions of Adapted Algae We now begin with a more detailed description of the metabolic processes in adapted algae, which were enumerated on page 129. (I) and (II): Hydrogen Absorption and Hydrogen Fermentation. — The absorption and evolution of hydrogen in darkness and in light by pure cultures of Scenedesmus Di, D3 and Scenedesmus ohliquus have been studied by Gaffron and Rubin (1942). During the first hour or two of anaerobic incubation, the algae fermented, liberating carbon dioxide and accumulating nonvolatile acids. After this initial period, Scenedesmus or Raphidium cells, while continuing the steady evolution of carbon dioxide, began to pick up hydrogen, if the incubation took place in a hydrogen atmosphere, and to liberate hydrogen if they were placed in an atmosphere of nitrogen (Fig. 9). Algae which have been allowed to photosynthesize vigorously before incubation, evolved the largest quantity of hydrogen, while those which have been made to re- spire in the dark for a considerable length of time, gave little hydrogen. The rate of absorption of hydrogen was very slow, but the total amount absorbed in several days was considerable — 1 ml. of cells took up as much as 2 ml. of hydrogen (simultanelusly with the evolution of 1.3 ml. of carbon dioxide) . This corresponds to an exchange of about one-tenth THE DARK REACTIONS 137 of one mole of hydrogen per liter of cell volume, and shows that the substrate of hydrogenation is a major cell component, whose concentra- tion is considerably larger than that of chlorophyll (the latter being of the order of 0.01 mole per liter, c/. page 411). The liberation of hydrogen can be increased by the addition of an external fermentation substrate, e. g., glucose (Fig. 10). In the presence of 0.07% glucose, (in phosphate buffer of pH 6.2) 1 ml. of Scenedesmus cells produced hydrogen steadily at the rate of 0.2 ml. per hour. 1 uu / » / ^80 / / o / X / u f ■ Q) / i_ 2 60 - O'v/ • 1. a N. 7 "-A <> y y e c o -^1 <» A y y . ^^ c /• ^^ 6 20 u 00 u. Z^la^ /y /y 1 1 2 4 Tim*, hours 8 Fig. 9. — Liberation of molecular hydrogen by fermentation in Scenedesmus (after Gaffron 19420. 0.027 ml. of cells in culture medium with 0.01 M phosphate buffer of pH 6, con- taining 0.2% glucose. 25° C. Curve b: KOH solution in side arm of vessel, absorbing carbon dioxide. Among several carbohydrates investigated by Gaffron and Rubin, glucose alone was found to stimulate fermentation (in all its forms) immediately. All others acted with a lag, indicative of a need for preliminary enzymatic transformation. In contrast to respiration and photosynthesis, the hydrogen fermentation was dependent on pH, with an optimum between pH 6 and 7. The "acid fermentation" continues independently of the hydrogen fermentation. Therefore, the ratio ACO2/AH2 can vary in the widest limits. Because of the simultaneous appearance of hydrogen fermenta- tion and photoreduction in Scenedesmus, it is safe to assume that the 138 ANAEROBICALLY ADAPTED ALGAE CHAP. 6 site of the hydrogen fermentation is in the chloroplasts (while acid fermentation may occur everywhere in the cell). The nonvolatile acids, produced by the autofermentation of algae, contained only a few per cent of lactic acid, while in presence of glucose, this percentage reached 50%. (In some colorless heterotrophic organ- isms, the fermentation of glucose produces up to 95% lactic acid.) The mechanism of the hydrogen evolution and absorption by algae which contain an active hydrogenase, can be represented by equations (6.2) and (6.3), (or 6.6a,b): the hydrogen is transferred from the atmos- phere, through the reversible systems, Eh-H2Eh and Ah-H2Ah (and probably through specific oxidoreductases), to a cellular oxidant, R; or 3 4 Time, hours 6 Fig. 10. — Increase in fermentation and hydrogen production in Scenedesmus by glucose (after Gaffron 1942i). Final sugar concentration, 0.06%. Curve 1: carbon dioxide with acid. Curve la: carbon dioxide and acid after addition of glucose. Curve 2: hydrogen. Curve 2a: hydrogen after addition of glucose. from a cellular reductant R'H2, through a similar catalytic system, back into the atmosphere. The direction of the process should depend on the oxidation-reduction potentials of the cellular reserve substances R and R', on their concentrations, and on the partial pressure of hydrogen. (Ill) and (IV). Oxyhydrogen Reaction and the Reduction of Carbon Dioxide in the Dark (Algae as Chemautotrophic Bacteria). — As men- tioned on page 132, small quantities of oxygen prevent anaerobic adapta- tion, but after adaptation, they are tolerated, without causing a return to normal photosynthesis, because they are used up by the oxyhydrogen reaction. If a few millimeters of oxygen are added to an adapted Scenedesmus culture in a hydrogen atmosphere, in the dark, the algae THE DARK REACTIONS 139 begin to consume both hydrogen and oxygen (Gaffron 1940-). Figure 11 shows the course of gas absorption in the presence of about 3.8 mm. Hg of oxygen (50 mm. Brodie solution). The total gas absorption ap- proaches, but does not reach 150 mm. (the value which corresponds to the complete absorption of oxygen, with double its volume in hydrogen). Figure 11 also shows that the gas consumption increases above the 2 H2 + O2 mark if carbon dioxide is present; and analysis shows that, in this case, carbon dioxide is absorbed together with hydrogen and oxygen. The algae probably now function as chemautotrophic ''hy- drogen bacteria," that is, they couple the combustion of hydrogen with the reduction of carbon dioxide. 345 Time, hours Fig. 11.— Oxy hydrogen reaction in adapted Scenedesmus in presence and absence of carbon dioxide (after Gaffron 1942). Initial oxygen concentration, 50 mm. Brodie solution. In contrast to Bacillus picnoticus, described in chapter V, chemo- synthesizing Scenedesmus cells have only a very limited tolerance for oxygen. If, for example, the partial pressure of oxygen is increased to 23 mm. Hg (300 mm. Brodie), the rate of the oxyhydrogen reaction decHnes rapidly and "de-adaptation" sets in. B. picnoticus, on the other hand, works well even in pure electrolytic gas. However, other bacteria capable of catalyzing the oxyhydrogen reaction, for example, Acetobacter peroxidans (cf. page 118), are also inhibited by excess oxygen. In a renewed study of the oxyhydrogen reaction in adapted algae, Gaffron (1942^) confirmed that in absence of carbon dioxide, the ratio AH2/AO2 often is much smaller than the theoretical value of 2 for water 140 ANAEROBICALLY ADAPTED ALGAE CHAP. 6 synthesis. A similar observation was made in the case of B. picnoticus (page 117); but there Ruhland attributed the excess oxygen consumption to respiration, i. e., autoxidation of cellular substrates, which proceeds simultaneously with the combustion of hydrogen. A similar explanation is impossible in the case of the hydrogen-adapted algae. In the first place, the ratio AH2/AO2 drops in these algae as low as 1.0 (as against a minimum of 1.8 in hydrogen bacteria). In the second place, respiration is practically absent (as shown by determinations of the carbon dioxide production during the oxyhydrogen reaction) . Gaffron suggested, there- fore, that in the absence of carbon dioxide ox3^gen is reduced only to a peroxide. (However, since a continuous accumulation of peroxide appears impossible, one must assume that its reduction is completed by -200 c-150 o -100 -50 120 180 Time, minutes Fig. 12.— Inhibition by glucose of the hydrogen uptake by the oxyhydrogen reaction in Scenedesmus (after Gaffron 1942). cellular hydrogen donors— without the latter's being oxidized to carbon dioxide.) In presence of carbon dioxide, the ratio AH2/AO2 is between 2 and 3 (as shown in Fig. 11), while the ratio AC02/(AH2 - 2 AO2) (compare Table 5. VI) is close to 0.5. This indicates that now all absorbed oxygen is reduced to water, while all absorbed carbon dioxide is converted into carbohydrates. Thus, the reduction of carbon dioxide helps the oxy- hydrogen reaction to run to completion. The efficiency of chemosynthesis was measured, on page 117, by the ratios ACO2/AH2 and AH2/AO2. The minimum value of the second ratio, in the presence of carbon dioxide, is 2.0 (no chemosynthesis, but a complete combustion of hydrogen to water), while the majority of experiments gave values between 2.6 and 3, a result similar to that obtained by Ruhland with Bacillus picnoticus (AH2/AO2 < 2.8, cf. Table THE DARK REACTIONS 141 5. VI). In other words, in the presence of carbon dioxide, for every two molecules of hydrogen transferred to oxygen, up to one molecule finds its way to carbon dioxide. In place of molecular hydrogen, hydrogen from organic donors can be used by adapted algae in reactions with oxygen, as shown by the diminution of hydrogen consumption caused by the addition of 0.05-1% of glucose, or yeast autolysate (c/. curves a and h in Fig. 12). The occur- rence of coupled carbon dioxide reduction remains to be demonstrated in this case. The explanation of the oxyhydrogen reaction in adapted algae requires one new assumption (in addition to the presence of a hydrogenase and of an oxidase, which already have been postulated in the interpretation of the adaptation phenomena). This assumption is that the oxidant {O2}' can react not only with the hydrogenase, Eh (thus causing de- adaptation) but also with the intermediate reductant, H2EH, thus com- pleting the transfer of hydrogen to oxygen. The fact that the ratio AH2/AO2 in the absence of carbon dioxide often is closer to 1 than to 2 indicates that this reaction takes place in two steps: (6.11a) {OsT + HoAh >{H202!+Ah (6.11b) IH2O2! + HoAh > 2 H2O + Ah and that the second step can be replaced by reaction with an internal hydrogen donor: (6.11c) {H2O2I+H2R >2H20 + R thus reducing the hydrogen consumption to one molecule of hydrogen per molecule of oxygen. This mechanism of the oxyhydrogen reaction is represented in scheme 6.IIA, which is merely a partial elaboration of scheme 6.1. In the presence of carbon dioxide, reaction (6.11) runs to completion, and one molecule of carbon dioxide can be reduced simultaneously with the reduction of two molecules of oxygen. This indicates that reaction (6.11c) is replaced by a "coupled" reaction: (6.12) 4 H2AH 1^ j(.Q^j ^ j(^jj,Oi + H2O + 2 Ah represented by scheme 6.IIB. A combination of (6.11) with (6.12) leads to the ratios AH2/AO2 = 3 and AC02/(AH2 - 2AO2) = h, in agreement with the experimental values. The nature of the coupling between the oxyhydrogen reaction and the reduction of carbon dioxide, symbolized by bracketing in (6.12) will be discussed in chapter 9 (page 235). Equation (6.12) shows that the " chemosynthetic " reduction of carbon dioxide in adapted algae requires the existence of an enzymatic 142 ANAEROBICALLY ADAPTED ALGAE CHAP. 6 link between the hydrogenase-oxidase system on the "oxidation side" of the primary photochemical process (c/. Scheme 6.1) and the catalytic system on the "reduction side" of this primary process (which takes care of the reduction of carbon dioxide in photosynthesis). A similar e;|(6.9) {Oa} H, Ah J [(6.lla) "[(6.6o) HzEh J Eh I 2/H2O2} 4H2A„ Cp2 2HjO 1'^"'=) 1 R 4H20 (6'2) 4A„ {CH2O} Scheme 6. 1 1 A. Mechanism of the oxyhydrogen reaction in adapted algae. B. Carbon dioxide reduction coupled with the second stage of the oxyhydrogen reaction. (Figures in parentheses refer to equations in text.) link is required also for the explanation of the photochemical absorption and evolution of hydrogen which will be discussed in the next section. 4. The Photochemical Reactions of Adapted Algae (V) and (VI). Photochemical Absorption and Liberation of Hydro- gen. — As mentioned on page 129, illumination accelerates both evolution and absorption of hydrogen by adapted Scenedesmus cells (cf. Fig. 13). The dependence of both effects on light intensity is so different, that sometimes a change in light intensity can convert hydrogen evolution into hydrogen consumption. For example, on the left side of figure 13, (which corresponds to 0.2% H2 in the air), hydrogen absorption prevails at 800 lux and hydrogen evolution at 3500 lux. On the right side of the same figure (in an atmosphere of pure hydrogen), even an illumination of only 400 lux causes a rapid saturation of the cells with hydrogen. The photochemical liberation of hydrogen can be observed only in absence of carbon dioxide. If the latter gas is admitted, it acts as an acceptor for hydrogen, and photochemical hydrogen liberation is trans- formed into photoreduction. Since acid fermentation liberates carbon dioxide continuously as long as the cells are deprived of oxygen, experi- ments on hydrogen liberation in light have to be carried out with alkali in a side arm of the manometer, and allowance must be made for the amount of carbon dioxide consumed by photoreduction before it had THE PHOTOCHEMICAL REACTIONS 143 time to reach the absorption vessel. Similarly to the hydrogen fermenta- tion in the dark, the hydrogen production in light can be sustained by an added hydrogen donor, e. g., glucose. The hydrogen evolution in light is not ajffected by some poisons (e. g., dinitrophenol) which inhibit the hydrogen fermentation in the dark. It thus seems that light permits the by-passing of an enzymatic step necessary for the hydrogen fermentation in the dark. 60 80 100 Tims, minutes Fig. 13. — Photochemical absorption and evolution of hydrogen by Scenedesmus in absence of carbon dioxide at different light intensities (after Gaffron 1942). 0.074 ml. of cells of Scenedesmus D3 in 0.05 M phosphate buffer of pH 6.2 at 25° C. Side arm contains KOH. Preceding dark periods, 20 hours; during this time the algae have formed hydrogen up to 0.2% of the gas phase. The explanation of the photochemical evolution of hydrogen requires the assumption that the photochemical hydrogen transfer can be inter- posed between the hydrogen donor R'H2 and the hydrogenase system H E H H2A H H2E H, H as represented in scheme 6. Ill, where the photochemical reaction se^ quence (6.14), (7.10a), (6.13), achieves the same result as the dark reaction (6.6b')- The method of representation used in scheme 6. Ill is different from that in schemes 6.1 and 6. II. Instead of giving a complete representation of each partial reaction, we have merely written down the oxidation-reduction systems which participate in the process, and indicated by arrows the direction of hydrogen transfer between them. The photochemical absorption of hydrogen can be interpreted in two ways: either as a light-accelerated hydrogenation of organic acceptors, R, or (as mentioned on page 129) as a photoreduction of carbon dioxide produced by fermentation. The first alternative is represented in scheme 6. Ill by the reaction sequence (6.6c), (7.10a), (6.15), whose 144 ANAEROBICALLY ADAPTED ALGAE CHAP. 6 final result is identical with that of the dark reaction (6.6b). In the second alternative, reaction (6.15) is replaced by (7.10d,e). Comparing scheme 6. Ill with scheme 6.1, we find two new features. In the first place, it provides for an enzymatic link between the catalytic systems on both sides of the primary photochemical hydrogen transfer, by means of the reaction: (6.13) 2HX + Ah >H2Ah + X This means that the primary reduction product in photosynthesis is capable of supplying hydrogen back to the same acceptor which serves as a reductant for the primary photochemical oxidation product, Z, CO, . Hx .^ "g*'^ HoO {C^^aO} TflSd: - - -■ X Scheme 6.III. — Photochemical and dark reactions in adapted algae. Simplified representation. Arrows indicate hydrogen transfers between two oxidation-reduction systems. FuU equations given in text and referred to by figures in parentheses. < Normal photosynthesis (6.7a, b), (7.10a), (7.10d, e). -< Photochemical and dark hydrogen liberation. The first step of the dark reaction (6.6b') is by-passed in light via (6.14), (7.10a) and (6.13). •< Photochemical and dark hydrogen consumption. The last step of the dark reaction (6.6b) is by-passed in light, via (6.6c), (7.10a) and (6.15); alterna- tively hydrogen may go in Ught to CO2 (by reactions 7.10d, e) instead of R (by reaction 6.15). by means of reaction (6.6c). The necessity for a nonphotochemical linkage between the two catalytic systems already was emphasized in connection with the mechanism of the chemosynthetic reduction of carbon dioxide, where an enzymatic link had to be provided from the hydrogenase system to carbon dioxide. The second feature of scheme 6. Ill are the direct links (6.14) and (6.15): (6.14) H2R' + 2Z ).R' + 2HZ (6.15) R -I- 2 HX > H2R + 2 X THE PHOTOCHEMICAL REACTIONS 145 between the cellular hydrogen donors, H2R', and cellular hydrogen acceptors, R, on the one side, and the primary photochemical products, Z and HX, on the other, which provide parallel photochemical channels to dark reactions (6.6b) and (6.6b'). (This relation between the paths of dark and photochemical reaction explains, incidentally, why they are not necessarily inhibited by the same poisons.) The assumption of reaction (6.14) leads to a question which must for the time being be left open: Why are nonadapted green plants — in which 20 10 c o U 3 — S.-IO E E -20 -30 - S/^ > ■^ 4/ X -^ ::^-^^-^ ^^ - \ \ — *-g_ - \ 1 A 1 6 9 Time, minuiss 12 IS Fig. 14. — Time course of gas exchange of anaerobically incubated Scenedesmus (after Gaffron 1942). Downward trend — absorption of hydrogen; upward trend — liberation of oxygen. 1— Hydrogen— 560 lux. 3— Hydrogen— 200 lux. 2— Hydrogen— 1020 lux. 4— Hydrogen— 6000 lux. 5— Air— 6000 lux. allegedly only the system H2Eh/Eh is ehminated by oxidation to EhO — incapable of using organic compounds as hydrogen donors in photosyn- thesis (instead of water)? One may suggest that the enzyme which catalyzes reaction (6.14) is de-activated simultaneously with the hydro- genase, or that the rate of reaction (6.14) is so slow as to make its com- petition with reaction (6.7a,b) impossible, as long as the "deoxidase" which catalyzes the latter reaction is fully active. (The de-activation of this enzyme was postulated, on page 134, to be the second feature of the adaptation process, supplementing the activation of the hydrogenase.) Reaction (6.15) poses a similar question as to why plants are unable 146 ANAEROBICALLY ADAPTED ALGAE CHAP. 6 to use cellular oxidants, R, as hydrogen acceptors in photosynthesis instead of carbon dioxide. However, it was mentioned above that this reaction can be eliminated if one assumes that the hydrogen absorbed in light is conveyed to the fermentation carbon dioxide rather than to R. (VI) and (VII). Photoreduction by Adapted Algae (Algae as Photo- synthesizing Bacteria). — Hydrogen-adapted algae can reduce carbon dioxide either at the cost of molecular hydrogen or of organic hydrogen donors, the latter taking precedence as long as they are available. When adapted algae are illuminated in presence of carbon dioxide, the photo- chemical consumption of hydrogen sets in only after a certain delay, 10 I 20 30 - \ ^ A 1500 LUX > 3 1 > 40 5 10 15 Time, minutes Fig. 15. — ^Time course of photoreduction by adapted Scenedesmus D3 as a function of the partial pressure of hydrogen (after Gaffron 1942). Culture medium. Adaptation 16 hours in hydrogen; 4% carbon dioxide. All mixtures of hydrogen and nitrogen contained 4% carbon dioxide. D — 96% hydrogen; 0—22% hydrogen; A— 8% hydrogen; • — 4% hydrogen. during which the cellular hydrogen donors are used up. In agreement with this explanation, the length of the "induction period" decreases with increased light intensity (while in ordinary photosynthesis the length of the induction period is independent of light intensity). The induction effect is illustrated by the curves for 200 and 560 lux in figure 14; the curves for 1020 and 6000 lux show, in addition to induction, the de-adaptation by excessive light. The "photosynthetic quotient," AH2/ACO2, of the hydrogen-adapted algae was found by Gaffron (1940^) to be 1.97 (average of five measure- ments with Scenedesmus D3, and seven measurements with Scenedesmus obliquus, individual values varying between 1.84 and 2.17). This result can be compared with the corresponding quotients found for purple hydrogen bacteria and assembled in table 5.IIL In that table, some THE PHOTOCHEMICAL REACTIONS 147 figures (particularly those of French and Wessler) were considerably above 2. Gaffron, too, at first found quotients as high as 3, but decided that these high values were due: (a) to hydrogen absorption not connected with photoreduction; and (6) to carbon dioxide liberation by acid fermentation. The consumption of hydrogen increases linearly with light intensity between 200 and 600 lux; but before any "light saturation" can be observed "de-adaptation" sets in, as illustrated by figure 14, and photo- reduction is replaced by normal photosynthesis. De-adaptation can be delayed by hydroxylamine or phenantroline (c/. Chapter 12, page 319); the maximum rate of photoreduction observed under these conditions was three times the rate of dark respiration. 12 16 20 24 Time, minuses 28 30 Fig. 16. — Photoreduction with hydrogen in Scenedesmus (after Gaffron 1940i). Preceding anaerobiosis, 12 hours. 20° C. in Ho. Curve /: cells washed and sus- pended in 0.01 M NaHCOs. Curve //: cells suspended in nutrient medium with 0.01 M NaHCOa and 0.5% glucose. The effect of hydrogen concentration on the rate of photoreduction in nitrogen is shown in figure 15. The reaction is slowed down when the hydrogen concentration is below 20% and de-adaptation starts rapidly below 4%. It was mentioned before that hydrogen-adapted algae also may function like heterotrophic purple bacteria, that is, reduce carbon dioxide at the cost of hydrogen from organic donors. The delay in hydrogen absorption, which was attributed above to the "cleanup" of intercellular hydrogen donors, and illustrated by figure 14, can be interpreted as e\'idence of this type of metabolism. This delay can be extended by a supply of organic reductants. Figure 16 shows the effect of glucose on the consumption of hydrogen. The hydrogen absorption in the dark is inhibited entirely, and that in light is strongly reduced. An induction 148 ANAEROBICALLY ADAPTED ALGAE CHAP. 6 period of four minutes appears in light, during which no hydrogen is consumed at all. Obviously, the supply of hydrogen from glucose or its metabolic derivatives suffices to cover all requirements in the dark, and to eliminate the absorption of hydrogen from outside in the first four minutes of illumination. After that, the more rapidly diffusing molecular hydrogen enters into competition with the slower diffusing glucose. A complete inhibition of hydrogen uptake has been observed when yeast autolysate is used instead of glucose. The explanation of the photoreduction by adapted algae is contained in scheme 6.1. It is due to the "interception" of the photochemical oxidation products by the hydrogenase system, with the cellular hydrogen donors R'H2 and external hydrogen competing as suppliers of hydrogen to the intermediate reductant, H2AH (c/. Schemes 6. II and 6. III). The evolution of oxygen is probably prevented not only by this interception (which, as noted on page 135, is not perfect), but also by the de-activation of the "deoxidase," Eo (page 134). The photosynthesis of green and purple bacteria may proceed by exactly the same mechanism as the photoreduction by adapted algae (except that their enzymatic system is "frozen" and under no circum- stances can switch over to the liberation of oxygen) ; but more probably, the incapacity of purple bacteria to produce oxygen is caused by a different character of the primary oxidation products, Z, which do not contain sufiicient energy for transformation into {O2} and free oxygen, and can only be reduced by the hydrogenase system (c/. Chapter 7, page 169). Despite the analogy between the metabolism of adapted algae and purple bacteria, there is a difference in the role which this metabolism can play in the life of these plants: Gaffron (1943) found that, after several days of "photoreduction," the algae showed no multiplication or increase in chlorophyll concentration comparable to that caused by a similar period of photosynthesis. Bibliography to Chapter 6 The Metabolism of Anaerobically Adapted Algae 1927 G6n6vois, L., Biochem. Z., 186, 461. 1931 Stephenson, M., and Stickland, L. H., Biochem. J., 25, 205, 215. 1934 Roelefson, P. A., Proc. Acad. Sci. Amsterdam, 37, 660. 1935 Gaffron, H., Biochem. Z., 275, 301. 1937 Nakamura, H., Acta Phytochim. Japan, 9, 189. 1938 Nakamura, H., ibid., 10, 259. Nakamura, H., ibid., 10, 271. Yamagata, S., and Nakamura, H., ibid., 10, 297. BIBLIOGRAPHY TO CHAPTER 6 149 1939 Gaffron, H., Nature, 143, 204. 1940 Gaffron, H., Avi. J. Botany, 27, 273. Gaffron, H., Science, 91, 529. 1941 Franck, J., and Gaffron, H., in Advances in Enzymology, Vol. 1. Interscience, New York, 1941, pp. 199-262. 1942 Gaffron, H., /. Gen. Physiol., 26, 195. Gaffron, H., ibid., 26, 241. Gaffron, H., and Rubin, J., ibid., 26, 218, 1943 Rieke, F. F., and Gaffron, H., /. Phys. Chern., 47, 299. 1944 Gaffron, H., BioL Rev. Cambridge Phil. Soc, 19, 1. Chapter 7 THE PRIMARY PHOTOCHEMICAL PROCESS* 1. The Problem of the Primary Process Observations with isolated chloroplasts, bacteria, and hydrogen- adapted algae, described in chapters 4, 5 and 6, as well as kinetic meas- urements (to be described in Volume II), indicate that photosynthesis is not a direct reaction between carbon dioxide and water, but a com- plicated sequence of physical, chemical and photochemical processes. One of the most important problems in the study of the mechanism of photosynthesis is the identification of the primary 'photochemical reaction (or reactions), and its separation from the nonphotochemical processes, which may precede or follow it in the reaction sequence. In the third chapter, the photosynthesis of green plants was described as a hydrogen transfer from water to carbon dioxide; and in the fifth chapter, bacterial photosynthesis was characterized as a transfer of hydrogen to the same acceptor, from reductants other than water. Although these hydrogen transfers may be associated with reactions of a different type, e. g., carboxylations, hydrations, phosphorylations or dismutations, we feel safe to assume that the primary photochemical process is a stage in the main oxidation-reduction process. One suggestion of a different kind (c/. Ruben 1943, and Emerson, Stauffer and Umbreit 1944) was that the absorbed light energy (or, at least, a part of it) is used for the synthesis of high energy phosphate esters, whose subsequent degradation is coupled with endergonic oxidation- reductions. This theory, derived from observations on the mechanism of energy utilization in respiration and fermentation, will be discussed in chapter 9 (page 226), and found improbable. Even less plausible is the hypothesis of Kautsky (1932) that the light energy is first stored in metastable oxygen molecules {cf. Chapter 18, page 514), later modified by the substitution of a dissociable oxygen complex for free oxygen (cf. Kautsky and Franck 1943). Another suggestion, which we consider quite implausible, was made by Seybold (1941). He thought that the Ught energy absorbed by chlorophyll b is used for polym- erization of sugars to starch rather than for the reduction of carbon dioxide. * BibUography, page 170. 150 PROBLEM OF THE PRIMARY PROCESS 151 In the present chapter, we will disregard these hypotheses and consider the primary photochemical process as a stage in the main oxidation- reduction reaction between water (or a substitute reductant) and carbon dioxide. In green plants, and in many bacteria as well, the hydrogen transfer must occur against the gradient of chemical potential, i. e., from an oxidation-reduction system (O2-H2O) with a more negative potential to a system (C02-{CH20}) with a more positive potential. This "uphill flow" is possible only with the assistance of external energy; here, light is called upon to play its part. However, the exact location of this "lift" in the reaction sequence cannot be predicted a priori. When a canal is built between two bodies of water situated at different levels, the provision of locks cannot be avoided; but whether these locks are constructed at the upper or lower end of the waterway is a purely practical problem. Similarly, the photochemical processes, which serve as "locks " in the flow of hydrogen from water to carbon dioxide, can be located either at the beginning of the transfer (in the oxidation of water) or at its end (in the reduction of carbon dioxide), or somewhere in the middle, or even in several different places. The description of the photochemical process in photosynthesis as a transfer of hydrogen atoms, which will be used throughout this chapter, does not exclude the possibiUty that it may be primarily an electron transfer. As described elsewhere (c/. Chapter 9, page 219), electron transfers coupled with acid-base equilibria, are equivalent to hydrogen transfers (and, if coupled with hydrations and dehydrations, may be equivalent to oxygenations). Franck (1935) and Stoll (1936) once made the suggestion that the primary photochemical process in photosynthesis may be an exchange of hydrogen for hydroxyl (cf. Chapter 19, page 555). However, the assump- tion of a transfer of hydroxyl radicals from carbonic acid to water is equivalent to the postulate that one part of the liberated oxygen originates in carbon dioxide, a concept which was found in chapter 3 (page 55) to be in conflict with experimental evidence. Looking for analogies to the postulated primary photochemical reaction in the realm of ordinary photochemistry, we find them in certain phenomena discussed in sections 4 and 5 of chapter 4. It was suggested there, that light absorption by inorganic ions in solution often leads to the oxidation of water, even though this effect remains "hidden" be- cause of the high rate of back reactions. In certain dyestuff solutions, a similar photochemical electron transfer takes place in the presence of added reductants, for example, ferrous ions, and may perhaps occur also in their absence. In the system thionine-ferrous ions, the back reaction is so slow that the mixture can lose all its color in light (as described 152 PRIMARY PHOTOCHEMICAL PROCESS CHAP. 7 in Chapter 4, page 77), despite the fact that the oxidation potential of thionine is several tenths of a volt more positive than that of ferric iron. This is the best-known photochemical reaction in vitro which is funda- mentally similar to the postulated primary photochemical process in photosynthesis — similar in that it, too, is an oxidation-reduction which, with the help of light, proceeds against a considerable gradient of chemi- cal potential. The unique characteristic of photosynthesis probably is not the photochemical transfer of hydrogen from water to an oxidant much weaker than oxygen, brought about by visible light — this may be a common occurrence even in nonbiological systems — but the avoidance of back reactions. The latter prevent a direct demonstration of primary photochemical water oxidation in many simple inorganic systems, and make even the photoxidation of ferrous ions by thionine a transitory phenomenon. The secret of how back reactions are prevented in photosynthesis must be sought in the heterogeneous structure of the photosynthetic appa- ratus and the consequent topochemical mechanism of the whole process — meaning by this term a chemical mechanism in which the participants follow prescribed paths on the catalytic surfaces, without appearing as free intermediates between the successive steps of their catalytic trans- formations. The preservation of at least a part of this structure in isolated chloroplasts may account for the success of Hill's experiments on chloroplast-sensitized photoxidation of water by ferric oxalate. Theoretically, there is no reason why all electronic energy contained in molecules excited by the absorption of light should not be available for oxidation-reductions. A light-excited molecule is both an efficient electron donor (that is, reductant) because it contains a "loose" electron; and a potential electron acceptor (that is, oxidant) because (to use a picture suggested by Weiss) it contains a "hole" in its usual complement of electrons. All electronic excitation energy is "free energy" and thus available for chemical reactions. Therefore, in a true thennodynamic equilibrium, hght-excited molecules can be assigned oxidation-reduction potentials equal to those of the same molecules in the normal state plus (or minus) their electronic excitation energy. Excitation by visible light (X = 700-400 myu) should add (or subtract) from 1.7 to 3 volt to the oxidation- reduction potential of the excited molecules, and thus make even the weakest oxidants thermodynamically capable of oxidizing water, and even the weakest reductants able to reducfe carbon dioxide. However, a photochemical reaction is practically never a part of a true thermodynamic equilibrium (unless we consider systems at very high temperatures, as, for example, the cosmic bodies). What is observed under ordinary conditions is a progressive conversion of hght, partly into heat and partly into chemical energy; the high theoretical oxidation or reduction potentials of the light-excited molecules are of no practical avail if the conversion into heat occurs much more rapidly PROBLEM OF THE PRIMARY PROCESS 153 than the energy-storing photochemical reaction. In other words, the problem of oxidation and reduction by light-activated molecules is one of reaction kinetics rather than thermodynamics. We will now describe the different specific interpretations of the photochemical oxidation-reduction process in photosynthesis, using a logical rather than historic approach. It was stated above that the photochemical stage may be located at the "oxidation end" or at the "reduction end," or in the middle, of the sequence of reactions by which hydrogen atoms are transferred from water to oxygen. We can represent this "hydrogen bucket brigade" by the following scheme: CO2 O2 i I !C02i < HX Y < HZ {OHl {HC02I X< HY Z< {HoOl I ! i : {CH2O} H20 Scheme 7.1. — Photosynthesis as an oxidation-reduction, coupled with preparatory and finishing catalytic reactions (represented by dotted and dashed arrows respec- tively). Full arrows symbolize hydrogen transfers (or electron transfers) between adjacent oxidation-reduction systems. Full arrows in scheme 7.1 symbolize hydrogen transfers between adjacent oxidation-reduction systems (e. g., the second full arrow from the left represents the reaction HY + X > Y + HX). For the sake of simplicity, all these systems are assumed to be "monovalent" (c/. Chapter 9). Broken arrows represent "finishing" catalytic reactions (dismutations, polymerizations, etc.) by which the first reduction product of carbon dioxide, {HCO2}, is converted into a carbohj^drate, and the first oxidation product of water, {OHj, into free oxygen, while dotted arrows symbolize the "preparatory" reactions by which the reactants (CO2 and H2O) are "fixed" prior to their participation in the oxidation- reduction proper. Braces in scheme 7.1 — as throughout this book — indicate that the components are supposed to be present, not in the free state, but as parts of larger molecules or complexes. The catalytic system which serves as the immediate hydrogen donor to carbon dioxide (or to a carbon dioxide-acceptor complex, cj. Chapter 8), is designated in scheme 7.1 by X, and the system which serves as the immediate hydrogen acceptor from water (or a water-acceptor complex), by Z, while Y stands for an intermediate catalyst which does not react 154 PRIMARY PHOTOCHEMICAL PROCESS CHAP. 7 directly with either of the two reaction components. Photosynthesis might require several such intermediates (Y', Y" • • •), or none at all. It is even possible (although not very probable) that only a single oxidation-reduction system lies between water and carbon dioxide, i. e., that X and Z are identical (this single intermediary being the chloro- phyll, cf. Chapter 19). Any one (or several) full arrows iji scheme 7.1 may represent the pri- mary photochemical process (or processes) . If one assumes only one such process, it appears that four quanta should be sujEficient to bring about photosynthesis, since four hydrogen atoms are required for the reduction of one molecule of carbon dioxide to the carbohydrate level. Earlier quantum yield determinations seemed to support this conclusion (Vol. II, Chap. 29); and, even though recent experiments have proved it to be incorrect, it is still useful to begin our discussion with the considera- tion of "four quanta theories," since they can be used afterwards as a basis for theories in which a larger number of quanta are assumed to contribute to the reduction of one molecule of carbon dioxide. We shall initiate this discussion in section 2 (page 155) with the four quanta theories which consider the primary photochemical process to be the dehydrogenation of water (cf. Scheme 7. II). In section 3 (page 157), we shall consider a similar theory which identifies this process with the hydrogenation of carbon dioxide (cf. Scheme 7. Ill); and in section 4 (page 159), we shall make the least specific assumption that the primary process is an exchange of hydrogen between two intermediates (cf. Scheme 7. IV). Thermochemical difficulties (Vol. II, Chapter 29) make four quanta theories implausible, and recent redeterminations of the quantum yield of photosynthesis have confirmed that at least eight quanta are required for the reduction of one molecule of carbon dioxide. The next step in our discussion will thus be the transition to "eight quanta theories," by a combination of two different or identical four quanta processes. Two "eight quanta theories" will be discussed in sections 5 and 6 (pages 160 and 164). In the first one, four hydrogen atoms take part in two different photochemical transfers each (cf. Schemes 7.V and 7.VA), while in the second, eight hydrogen atoms are transferred by eight identical photo- chemical reactions, but the energy of four of them is used afterwards for a second activation of the other four (cf. Scheme 7. VI). In these eight quanta schemes, too, the primary photochemical processes may be located either at the "oxidation end" or at the "re- duction end" of the reaction sequence (or in both places), or somewhere in the middle. In schemes 7,V and 7. VI, the last alternative is used as the least specific one. We consider these schemes the most appropriate starting points in the quest for the true chemical mechanism of photo- synthesis. Scheme 7.VA, suggested by Franck and Herzfeld (1941) OXIDATION OF WATER AS PRIMARY PROCESS 155 represents a possible elaboration of 7.V; its advantages and disadvan- tages will be discussed in section 7. In Franck and Herzfeld's theory (as well as in several other theories of the primary process), one of the participants in the photochemical oxidation-reduction was identified with chlorophyll. We have eliminated all reference to chlorophjdl from reaction schemes in this chapter, so as not to prejudice their generality. The chemical function of chlorophyll in photosynthesis, and its possible identification with one of the inter- mediate oxidation-reduction catalysts in scheme 7.1, will be discussed in chapter 19. 2. Oxidation of Water as the Primary Process First Four Quanta Theory In first formulating the oxidation-reduction theory of photosynthesis, van Niel (c/. van Niel and Muller, 1931, Muller 1933, van Niel 1931, 1935, 1941) postulated that photosynthesis involves a single photo- chemical reaction and that this reaction is the decomposition of water (as was suggested earlier by Bredig in 1914, Hofmann and Schumpelt in 1916, Thunberg and Weigert in 1923, and Wurmser in 1930). The reduction of one molecule of carbon dioxide to a carbohydrate requires the transfer of four hydrogen atoms. If we assume that each of these atoms is contributed by a different molecule of water, that is, if we select the alternative reaction (3.14) in preference to (3.13), we can postulate, with van Niel, four identical primary photochemical reactions: (7.1) 4{H20i ^4 {HI +4 {OH} where hp symbolizes, in the usual way, a quantum of light energy. The decomposition of water has also been postulated, as the principal or only photochemical reaction in photosynthesis, by Shibata and Yakushiji (1933), Dhar (1934), and Gaffron (1942). In equation (7.1), braces again indicate that the components and products of this reaction do not occur in the free state. Water is assumed to be attached to a molecular complex, which probably includes the sensitizer (chlorophyll), while the "primary reduction product," {H}, and the "primary oxidation product," {OH}, are taken up by unknown "acceptors." With free molecules, atoms and radicals, reaction (7.1) would require not less than 110 kcal per mole (c/. Table 9. II), that is, more than twice the energy available in one quantum of red light. In order to make the primary reaction (7.1) at all possible, the energy of association of the products, {H| and {OH}, with their acceptors, must be larger than that of water, by at least 70 kcal per mole. (We recall 156 PRIMARY PHOTOCHEMICAL PROCESS CHAP. 7 that, on page 73, the adsorption of hydrogen and hydroxjd radicals by zinc oxide was suggested as an explanation of the zinc oxide-sensitized decomposition of water in ultraviolet light.) Instead of representing the elementary photochemical process as a decomposition of water (as it appears in equation 7.1), it may be useful to emphasize its character as an oxidation-reduction, with water (or a water-acceptor complex) in the part of the reductant and an inter- mediary hydrogen acceptor in the part of the oxidant. In this case, we may write, using the symbols introduced in scheme 7.1: (7.2) 4{H20)4-4Z ^4 {OH) +4 HZ The completion of photosynthesis, initiated by reaction (7.2), calls for a nonphotochemical reduction of carbon dioxide (or, more probably, of an association product, {CO2}), by the primary reduction product HZ, perhaps through the intermediary of other catalysts (Y and X in scheme 7.1): catalysts (7.3) {C021+4HZ > {CH2O! +H2O + 4Z We have further to assume the dismutation of the oxidation product, {OH}, into water and oxygen. The latter can occur either directly, catalysts (7.4) 4 {OH} > 2 H2O + O2 or through the intermediary of biradicals (peroxides { OH } 2 or moloxides {Oohc/. Chapter 11): (7.4a) 4 |0H1 ^2{OHl2 or (7.4a') 4 10H| > 2 H2O + {O2) (7.4b) 2 {0H!2 >2H20 + 02 or (7.4b') {O2I ^02 The alternative formulation of equations (7.4a) and (7.4b) is the one used previously in chapter 6, e. g., in scheme 6.1. By the summation of (7.2) and (7.4), one obtains, for the oxidation of water in photosynthesis, the equation: (7.5) 4{H20}+4Z ^4HZ + 2H20 + 02 By further addition of equation (7.3), the over-all reaction of photosynthesis becomes (7.6) 4 {H2O! + {CO2I ) {CH2OI + 3 H2O + O2 Thus, in van Niel's theory, four water molecules and one carbon dioxide molecule participate in the formation of one {CH2O} group, and three water molecules are recovered in the end — two by the dismutation of the REDUCTION OF CARBON DIOXIDE AS PRIMARY PROCESS 157 first oxidation product, according to (7.4), and one by the dehydration of an intermediate reduction product, as described by equations (3.11) and (3.12). CO2 ' 1 ^{HzO) {CH20}*H20+4Z (7.4a) (74 b) 2H2O + O2 Scheme 7.II.— Photosynthesis, with water oxidation by an intermediate catalyst as the primary photochemical process (first four quanta theory). Scheme 7. II may help to visuaUze the mechanism of photosynthesis according to van Niel. The heavy arrow in this and all subsequent schemes designates the primary photochemical process, while figures in parentheses refer to equations in text. Wislicenus (1918), Thunberg (1923) and Weigert (1923, 1924) suggested that the primary photochemical process in photosynthesis is the decomposition of water into hydrogen and hydrogen peroxide, and that it is followed by a nonphotochemical reduction of carbon dioxide by hydrogen peroxide, either alone (Eq. 4.18), or in cooperation with hydrogen (Eq. 4.19b). As stated in chapter 4 (page 79), the nonphotochemical re- actions postulated in this theory require too much energy to occur spontaneously at the low temperatures prevailing in living organisms. There is therefore no need to consider the Thunberg-Weigert theory in more detail here. Experiments to be described in chapter 11 (page 295) prove that the substitution of heavy water for ordinary water affects the rate of a nonphotochemical reaction in photosynthesis. This does not furnish, however, an argument against the participation of water in the primary photochemical process, because hydrogen (or deuterium) atoms transferred by hght from water to an intermediate acceptor must afterwards take part in a number of catalytic reactions. In fact, the only partial reaction in photosynthesis whose rate is Ukely to be left unaffected by the substitution of heavy water for ordinary water, is the fixation of carbon dioxide in the {CO2I complex. 3. Reduction of Carbon Dioxide as the Primary Process Second Four Quanta Theory The oldest theory, according to which the primary process in photo- synthesis was thought to be the decomposition of carbon dioxide, had to be discarded when hydrogen transfer was proved to be the main mechanism of biochemical oxidation-reductions, and when all oxygen in photosynthe- sis was shown to originate in water. We can nevertheless associate the primary photochemical process in photosynthesis with a transformation of carbon dioxide, if we consider this process as a photochemical hydro- genation rather than a decomposition of this compound (cf. Scheme 7.1). 158 PRIMARY PHOTOCHEMICAL PROCESS CHAP. 7 If water does not participate directly in the primary process, the photo- chemical reduction of carbon dioxide must occur at the cost of an inter- mediate hydrogen donor (designated by HX in Scheme 7.1). It has often been assumed that the reduction of carbon dioxide must of necessity involve several successive photochemical steps, e. g. : (7.7a) {CO2I + HX > {HCO2} + X (7.7b) {HCO2) + HX "—^ {H2CO2} + X (7.7c) {H2CO2I + HX "-^ {H3CO2) + X (7.7d) {H3CO2} + HX "—^ {H4CO2! + X > {CH2OI + H2O + X (7.7) {C02J+4HX ^ {CH2OI +H2O + 4X The first and third stage lead to "odd" molecules (that is, free radicals), while the second one produces an intermediate of the reduction level of formic acid. In equations (7.7) we postulated four differ ejit 'primary 'photochemical reactions, and this may be considered as a setback compared with van Niel's theory. However, the assumption of a single primary process is possible in van Niel's theory only through combination of this primary process with the catalytic dismutation of the primary oxidation product, {OH}, by reactions (7.4). A similar scheme, with a single primary photochemical reaction, also can be substituted for (7.7), with the only difference that, because of the participation of four hydrogen atoms in the reduction of oifie molecule of carbon dioxide, two successive dismuta- tions are required to complete the reaction. In the first one, two radicals, {HCO2}, dismutate into two "even" molecules, {CO2) and {H2CO2I, while in the second one, two molecules, (H2CO2}, dismute into {CO2} and {H4CO2}: (7.8a) 4{C02l+4HX ^4|HC02l+4X catalysts (7.8b) 4 {HCO2I > 2 {CO2! + 2 {H2CO2} catalysts (7.8c) 2 {HaCOz) > {CO2} + {H4CO2} > {002} + {CH2O} + H2O (7.8) 4 {CO2} +4HX "-^ {CH2OI +H2O + 3 {CO2} +4X catalysts If (7.8a) is considered to be the only photochemical reaction in photosynthesis, the process must be completed by a nonphotochemical oxidation of water by the oxidized intermediate X, possibly involving the intermediate catalysts Y and Z (c/. Scheme 7.1) : catalysts (7.8d) 4X + 2{H20} > 2 HX + O2 HYDROGEN EXCHANGE BETWEEN INTERMEDIATES 159 SO that the over-all reaction becomes: (7.8) 4 {CO2I + 2 H2O "-^ {CH2O) + 3 {CO2} + H2O + O2 catalysts The reaction mechanism (7.8) is represented graphically in scheme 7. III. 4C02 SHjO 3CO2 +{CH20J -t-HjO 4HX Og Scheme 7.III. — Photosynthesis, with reduction of carbon dioxide (in the form of a compound {CO2}) by an intermediate catalyst as the primary photochemical process (second four quanta theory). If {CO2I is a carboxyhc acid (c/. Chapter 8), reaction (7.8c) is analo- gous to the "Cannizzaro reaction" (4.22b). As an analogy to (7.8b), we may mention the dismutation of semiquinones into quinones and hydroquinones. For example, the reduction of thionine by ferrous ions in light — a reaction whose first step was described on page 000 as similar to the primary process in photosynthesis — runs to completion by the dismutation of the primary reduction product (semithionine) into thionine and leucothionine: (7.9a) 2 Thionine + 2 Fe+++ > 2 semithionine + 2 Fe++ (7.9b) 2 Semithionine > leucothionine + thionine 2hi' (7.9) Thionine + 2 Fe+++ > leucothionine + 2 Fe++ In van Niel's mechanism (7.6), four water molecules participate in the primary reaction, and three of them are recovered; in mechanism (7.8), four carbon dioxide molecules participate in the primary reaction, and three of them are restored at the end. The dismutation of the radicals, {HCO2}, can take place either directly, as assumed in (7.8b), or through the intermediate formation of "biradicals," {HC02J2, analogous to the peroxides {OHI2, postulated in (7.4a,b). 4. Hydrogen Exchange Between Intermediates as the Primary Process Third Four Quanta Theory In the absence of decisive arguments in favor of a direct association of the primary photochemical process with either carbon dioxide or 160 PRIMARY PHOTOCHEMICAL PROCESS CHAP. 7 water, it may be useful to write down also a less specific scheme, in which both the reduction of carbon dioxide and the oxidation of water are assigned to secondary catalytic reactions, and the primary photo- chemical process is thought of as an oxidation-reduction reaction between two intermediates, e. g., X and Z in scheme 7.1: (7.10a) (7.10b) (7.10b') (7.10c) (7.10c') (7.10d) (7.10e) (7.10) 4HZ + 4X 4 Ay -» 4 Z + 4 HX ^4 HZ + 4 {0H| ).4HZ+ {O2I or 4Z + 4 {H2OI — 4Z + 2 IH2O) — 4 {OH! > 2 H2O + O2 {O2} >02 4 HX + 4 {CO2! > 4 {HCO2! + 4 X or 4 {HCO2 -> {CH2OI +H2O + 3CO2 4{C02l +4{H20i ihy -^ { CH2O) + O2 + 3 HoO + 3 CO2 Reaction system (7.10) is represented graphically in scheme 7.1 V. It also was used, because of its unspecific character, in the construction of schemes 6.1 and 6.III in the preceding chapter. 2H,0 jCHaO} +H2O + 3C0; Scheme 7.IV. — Photosynthesis, with an oxidation-reduction reaction between two inter- mediate catalysts as the primary photochemical process (third four quanta theory). 5. Two Different Primary Four Quanta Processes First Eight Quanta Theory Schemes 7. II, 7. Ill, and 7. IV have in common the assumption oi four identical photochemical reactions for each reduced molecule of carbon dioxide. But after a controversy which lasted for several years (Vol. II, Chapter 29) it now seems probable that the maximum quantum yield of photosynthesis is not 1/4, but 1/8 (perhaps even 1/10 or 1/12), that is, that at least eight quanta must be absorbed for each reduced molecule of carbon dioxide. This larger number of available quanta is welcome, because the energy contained in four quanta of red light (about 160 kcal TWO DIFFERENT PRIMARY FOUR QUANTA PROCESSES 161 per einstein) is scarcely sufficient to cover the net requirements of photosynthesis (112 kcal per mole) and leave a sufficient margin for losses involved in the stabilization of unstable intermediates (c/. Wohl 1935). If we assume, as a working hypothesis, that eight quanta are actually utilized in photosynthesis (while quanta absorbed above this number are lost by energy dissipation), we may ask how eight primary photochemical processes can be utilized for the transfer of four hydrogen atoms. The obvious answer is that each hydrogen atom can be activated twice, thus enhancing its reducing power. This result can be achieved in two ways. One is to activate the same four hydrogen atoms photochemically tvnce in succession, e. g., to combine two of the different four quanta processes discussed in the preceding sections. The other solution is to double the number of identical primary photochemical processes, e. g., (7.2), (7.8a) or (7.10a), and to allow four of the primary products to recombine, transferring their recombination energy to the remaining four intermediates. This kind of secondary reactions can be designated as "energy dismutations," because of their analogy to chemical dismutations repeatedly mentioned in this chapter. We will begin by exploring the first alternative, that is, by discussing hypotheses which postulate two sets of different primary photochemical reactions. We may designate one set — in which hydrogen atoms are taken away from water (or from an intermediate donor), as photoxidations, and those of the second set— in which the same hydrogen atoms are transferred to carbon dioxide (or an intermediate acceptor), as photo- reductions (using this term in a sense different from that assigned to it by Gaffron, cf. Chapter 6). The hypothesis of two primary processes has often been associated with the assumption that the intermediate hydrogen acceptor is chloro- phyll, and that this pigment is capable of taking hydrogen atoms away from water (or another donor), with the help of light, and transferring them to carbon dioxide (or another acceptor), also with the help of light. As announced before, we will postpone the question of the role of chlorophyll in photosynthesis until chapter 19, and use in the following schemes the symbols X or Z where the original papers may have used Chi (= chlorophyll). However, we will retain the assumption that the same catalyst whose oxidized form participates in the photoxidation of water also participates, in the reduced form, in the photoreduction of carbon dioxide. (A less specific assumption would be to consider the photoxidation and photoreduction as separated by an unknown num- ber of intermediate oxidation-reduction catalysts.) In other words, we assume that only one of the intermediate catalysts in scheme 7.1, either X, Y, or Z, is a "photocatalyst." We begin with the second alter- 162 PRIMARY PHOTOCHEMICAL PROCESS CHAP. 7 native (Y as a photocatalyst), since it removes both carbon dioxide and water from the direct participation in the primary photochemical process, and can thus be considered as the least specific of all. The resulting system of reactions, (7.11), represented in scheme 7.V, is a logical generalization of system (7.10) and scheme 7. IV. (7.11a) (7.11b) (7.11c) (7.11d) (7.11) 4Y + 4HZ 4 HY + 4 X 4 A;- -> 4 HY + 4 Z 'ihp > 4 HX + 4 Y 4 {CO2I + 4 HX > iCHjO) + 3 CO2 + H2O + 4 X 4Z + 4 {H2OJ -^ 4 HZ + O2 + 2 H2O 4{C02} +4 IH2O} 8hy -^02+ {CH2OJ + 3 CO2 + 3 H2O Another scheme of the same type was suggested by Franck and Herzfeld (1941) to replace the older "four quanta theories" of Franck 4CO2 4X ^{cOa} 4Y r 4HX _J I r 4-HY I 4HZ -I Ahv\(l.\\a) 4Z L 4{h20) 4hv|(7.ll b) 4Y (7.|ld) 4HZ 1- {Oj} + SHjO (7.11c) 4{hC02} i 4X (7. lid) {CHjOJ + HzO+aCOa Scheme 7.V. — Photosynthesis, with oxidation-reduction reactions between three intermediary catalysts as the two primary photochemical processes. (The central catalyst, which participates in both photochemical reactions, may be chlorophyll.) (First eight quanta theory.) (1935) and Franck and Herzfeld (1937). In this scheme, the "photo- catalyst" was identified with X in scheme 7.1, that is, it was assumed to react directly with the carbon dioxide-acceptor complex in the "photo- reduction," and to be restored by an intermediate hydrogen donor in the "photoxidation." Two additional specific assumptions were made by Franck and Herzfeld. In the first place, they assumed that the reduced photocatalyst, HX, hydrogenates not only the complex, {CO2}, but also its three reduction intermediates, {HCO2}, {H2CO2}, and {H3CO2} — in other words, they assumed four different photoreduction processes (c/. 7.7a, b, c, and d), and combined them with four identical primary TWO DIFFERENT PRIMARY FOUR QUANTA PROCESSES 163 photoxidation processes of the type (7.10a). In the second place, they postulated that the intermediary catalyst, HZ, which reduces X to HX, , is an organic hydroxyl compound, ROH. Upon its dehydrogenation to a radical, RO, this catalyst was assumed to oxidize water, forming an organic hydroperoxide, ROOH, which finally dismutates, restoring ROH and liberating oxygen. This special mechanism of catalytic water oxidation will be compared with the mechanism (7.4) in chapter 11 (page 189). The hypothesis of Franck and Herzfeld is represented by the formulae (7.12) and the reaction scheme 7.VA. (7.12a) 4 X + 4 ROH - (7.12b) HX+ {CO2} — > {HCO2} + X (7.12c) HX + {HCO2I - t (TT CO 1 1 "V (7.12d) HX + {H2CO2I (7.12e) HX + {H3CO2I h,> •, (PTT Ol 1 TT n 1 V (7.12f) 4 RO + 2 H2O - 4 '^ POTT 1 ^ POOTT (7.12g) '1 ROOTT ^ 9 POTT -1- n„ (7.12) 2 H2O + {CO2I "—^ {CH2OI + H2O + O2 A characteristic assumption in scheme 7.VA is that all intermediary products (designated by asterisks) are stabilized by one and the same catalyst (Eb), to protect them from back reactions. The assumption of a common effect of a catalyst on four different intermediates on the "reduction side" of the primary photochemical process, and on one intermediate "on the oxidation side" is not very plausible. The number of different photochemical products requiring catalytic stabilization could be reduced from five to two by combining (7.10a) with (7.8), instead of with (7.7), thus arriving at a scheme similar to 7.V except for the elimination of the intermediate oxidation-reduction system HY-Y — a change which makes (CO2} a direct participant in the primary photo- chemical process: (7.13a) 4X + 4HZ > 4 HX + 4 Z (7.13b) 4HX + 4{C02! ^4X4-4{HC02l (7.13c) 4Z + 4H2O )-4HZ + 2H2O + O2 (7.13d) 4 {HCO2} > 3 CO2 + CH2O + H2O (7.13) 4 {CO2I + 4 H2O "-^ 3 CO2 + 3 H2O + {CH2O) + O2 164 PRIMARY PHOTOCHEMICAL PROCESS CHAP. 7 6. Dismutation of Energy as an Alternative to a Second Primary Process Second Eight Quanta Theory As stated on page 154, an alternative to two sets of different photo- chemical reactions is a single set of eight identical primary reactions, Photoreduction 0»ida+ion- of Carbon Reduction of Catalyst X (^Chlorophyll?) HXRO* "M h 1 X ROH Photoxidation of an Intermediate Catalyst, ROH RO- t Oxidation of Woter ► ROH "^ROOH' ->^2R0H+0a ► ROOH-' ^->ROH Scheme 7.VA. — Photosynthesis according to Franck and Herzfeld. This is a varia- tion of scheme 7.V, with the carbon dioxide complex {C02! and its reduction inter- mediates substituted for the intermediate oxidant. The intermediate reductant is designated as ROH (instead of Z). The central catalyst X (which corresponds to Y in scheme 7.V) may be chlorophyll. coupled with secondary processes by which the energy contained in eight primary intermediate products is transferred to four secondary intermediates. As an example, we assume that the primary reactions are all of the type (7.10a). This assumption leads to the following mechanism : DISMUTATION OF ENERGY 165 8 HZ + 8 X Shf -> 8 Z -f 8 HX 8HX { + 4 {CO2I -^4 (HCOzl +4X -> 4 HZ + 4 X .+ 4Z 4 {HCO2I > 3 CO2 + H2O + {CH2OI 4 Z + 4 H2O -^ 4 HZ + O2 + 2 H2O 4H2O + 4 !C02| Shv (7.14a) (7.14b) (7.14c) (7.14d) (7.14) This mechanism is represented in scheme 7. VI. Its essential part is the "energy dismutation " by the coupled reaction (7.14b), in which the reoxidation of four reduction intermediates HX by four oxidation intermediates Z is supposed to assist four other molecules HX in reducing carbon dioxide. -> {CH2OI + O2 + 3 H2O + 3 CO2 4CO, 4{C0^ 8X 8HZ I , ' (7.14 a) I 8 hi. i — — X 8HX + 4Z + 4Z J L_ a{h^o} (714b) ;7.i4d) 4IHCO2) + 8X + 4HZ 4HZ (7.14c) {CH^o; +3CO2+H2O {o^l + aH^o (7.l4d) O2 Scheme 7. VI.— Photosynthesis according to the concept of energy dismutation (second eight quanta scheme). Primary photochemical process is an oxidation-reduc- tion reaction between eight molecules of an intermediate catalyst X and eight molecules of another catalyst Z (one of them may be chlorophyll). The reduction of carbon dioxide is coupled with the recombination of one-half of the primary photochemical products. A similar scheme can be devised also by assuming an "energy dismu- tation" on the "oxidation side" of the primary photochemical process, that is, by postulating that the recombination of four pairs of primary products gives four other primary oxidation products the power to oxidize water according to reaction (7.14d). Simple energy dismutations are well known in physics. The descent of one weight in a clock lifts the other weight to twice its original height, doubling its potential energy. When two excited mercury atoms collide, the result is often the excitation of one of them to twice its original energy level, and the return of the other into the ground state (c/. Beutler and Rabino witch 1930). (7.15) 2Hg^ -> Hg** + Hg Chemical reactions involving "energy dismutations" undoubtedly occur in chemosynthesizing bacteria, in which the oxidation of several 166 PRIMARY PHOTOCHEMICAL PROCESS CHAP. 7 molecules of a comparatively mild reductant is utilized for the production of one molecule (or radical) able to react with carbon dioxide. This analogy with chemosynthesis is the main reason for the introduction of the concept of "energy dismutation" into the discussion of the mechanism of photosynthesis. This concept enables one to postulate only one kind of primary photochemical processes even if the number of these processes is much larger than the number of elementary oxidation-reduction acts (hydrogen transfers or electron transfers) required for the completion of the overall reaction. A possible mechanism of "energy dismutation" in photosynthesis and chemosynthesis will be discussed in chapter 9, and the results presented in schemes 9. Ill and 9. IV. The assumption on which these schemes are based is that, after a compound, RH2, has been first oxidized by a strong oxidant (oxygen, for example) to a radical, RH, the latter may be able to yield its remaining hydrogen atom to a much weaker second oxidant (carbon dioxide, for example). 7. Comparison of Different Primary Processes Comparing critically the various schemes of photosynthesis presented in this chapter, we can discard the four quantum schemes as contradicting recent quantum yield determinations, as well as straining dangerously the thermochemical possibilities. As between the alternative eight quanta theories, no final decision is possible at present. Two questions remain to be decided: Is the assumption of eight identical photochemical processes (as in 7.14) more probable than that of two different kinds of primary processes (as in 7.11 or 7.13) or of five such processes (as in 7.12)? Does carbon dioxide or water (or both or neither) participate (directly or as complexes) in the primary photochemical process? Although the hypothesis of two sets of primary photochemical processes — photoxidations and photoreductions — does not require the direct chemical participation of chlorophyll in both of them, an experi- mental proof of the existence of two interconvertible colored forms of chlorophyll belonging to different reduction levels and capable of using light energy for photoxidations and photoreductions, respectively, would strengthen this hypothesis almost to the point of certainty. We shall see, in chapter 18, that experiments with extracted chlorophyll make the existence of two such chlorophyll modifications plausible but do not prove it. The main argument in favor of the alternative theory of eight identical photochemical reactions (beside the greater simplicity of this scheme) is the analogy which it enables to establish between the mechan- isms of photosynthesis and chemosynthesis. This appeals to our desire for a unified conception of all forms of organic synthesis, and receives support from the discovery of Gaffron that photochemical and non- COMPARISON OF DIFFERENT PRIMARY PROCESSES 167 photochemical reduction of carbon dioxide can occur in the same organ- isms (anaerobically-adapted Scenedesmus and similar green algae). As to the second question, that of the direct participation of the reaction components in the primary photochemical process, the non- photochemical reduction of carbon dioxide in autotrophic bacteria and hydrogen-adapted algae constitutes a strong, even if indirect, argument against the association of carbon dioxide with the photochemical reaction proper. One is tempted to attribute photosynthesis and chemosynthesis to a reaction of carbon dioxide with the same reducing agents, formed in one case by a photochemical reaction and in the other case by the catalytic oxidation of hydrogen, hydrogen sulfide, or another inorganic or organic reductant. Against this argument, one must weigh several observations which speak in favor of a closer association of carbon dioxide with the photo- chemical apparatus, and which caused Franck and Herzfeld to assume such an association in their scheme 7.VA. One such observation is the light-induced liberation of carbon dioxide, which occurs occasionally (c/. Emerson and Lewis, page 207) during the induction period of photosynthesis, and which ma^^ be attributed to a photochemical decomposition of the complex, {CO2}, into acceptor and free carbon dioxide. HoM'ever, Franck (1942), in a discussion of this "CO2 gush," decided that it is caused, not by a direct photochemical interaction between {CO2} and excited chlorophyll, but by back reactions of the first intermediate ({HCO2} in scheme 7.VA), in which so much energy is released that the regenerated complex, {CO2}, dissociates immediately into free acceptor and carbon dioxide. This mechanism does not require that {HCO2} be formed by a direct photochemical interaction of {CO2I with chlorophyll, but can equally well be fitted into a scheme in which the complex {CO2} is reduced by an intermediate reductant. A second argument in support of a photochemical interaction between chlorophyll and carbon dioxide is the relationship between the photosyn- thesis and chlorophyll fluorescence in vivo. This phenomenon will be dis- cussed in detail in volume II, chapters 24 and 32. The essential point is that the yield of fluorescence sometimes increases at high light intensities and that, according to Franck, French, and Puck (1941), this occurs whenever the stationary concentration of the complexes, {CO2}, is de- pleted. The simplest explanation of the quenching effect, which the complex {CO2} apparently exercises on chlorophyll fluorescence, is the assumption of a direct photochemical interaction of this complex with excited cidorophyll molecules. However, in this case, too, an indirect interaction may suffice to produce the observed results. For example, if chlorophyll reacts photo- chemically with an intermediate oxidant X, and the reduced intermediate 168 PRIMARY PHOTOCHEMICAL PROCESS CHAP. 7 HX is reoxidized by reaction with the complex, {CO2}, an exhaustion of {CO2} may lead to an accumulation of reduced intermediates, HX, and exhaustion of the quenching species, X. On the basis of all these considerations, without pretending to be able to give a final answer to the problem of the primary photochemical process in photosynthesis, it seems that eight primary processes of the type assumed in scheme 7. VI, perhaps, with chlorophyll identified with the reductant, HZ (c/. Chapter 19, page 552), provides the best working hypothesis. Scheme 7.V, which contains two sets of different primary processes but leaves open the possibility of the same intermediate reductants occurring in photosynthesis and chemosynthesis, is our second choice, and would become the first one if the existence of two interconvertible green modifications of chlorophyll — one a photo-oxidant and one a photoreductant — would be definitely confirmed by experiments in vitro. 8. The Primary Process in Bacteria and Adapted Algae In hydrogen-adapted algae and in bacteria, molecular hydrogen, hy- drogen sulfide, or other inorganic or organic hydrogen donors replace water in the role of the ultimate reductant in photosynthesis. Does this substitution mean a change in the primary photochemical process, or merely a different course of secondary catalytic reactions? Nakamura (1938), van Niel (1941), Franck and Gaffron (1941), and Gaffron (1942) all suggested, for different reasons, that the substitute reductants do not participate in the primary photochemical process. One of van Niel's arguments was the observation (c/. page 110) that organic reductants are used up by Spirillum riibrum at the same rate in the dark and in light. This indicates a preliminary enzymatic transfor- mation of these reductants, e. g., hydrogen transfer to the hydrogenase system (cf. Chapter 6, Eq. 6.6b), which they have to undergo both in respiration and in photoreduction. Since van Niel and Gaffron considered the oxidation of water as one (or even the only) primary photochemical reaction in ordinary photosyn- thesis (as in Scheme 7. II), the assumption that substitute reductants do not participate in the photochemical process led them to the logical con- clusion that, in bacterial photoreduction too, the primary photochemical process is the oxidation of water. The fact that purple bacteria do not evolve oxygen in light could then be explained in two ways. One hy- pothesis, suggested by Gaffron, was that the intermediate product of water oxidation, {OH}, can only be reduced in bacteria by substitute reductants — hydrogen, hydrogen sulfide, etc.— (and not by water), because these organisms contain an active hydrogenase system, but not the oxygen-liberating enzyme, Eq. The other hypothesis, proposed by PRIMARY PROCESS IN BACTERIA AND ADAPTED ALGAE 169 van Niel, was that the primary product obtained by the oxidation of water in bacteria, {OH}^, is somewhat different from that formed in green plants, {OH}-^, and therefore incapable of conversion into oxygen. For example, the energy content of {OH}^ could be insufficient for this conversion, perhaps because this product is formed with the help of infrared quanta, supplied by bacteriochlorophyll, which are about 30% smaller than the red quanta made available by ordinary chlorophyll. The difference between {OH}^ and {OHp may be in the nature of the acceptor (symbolized by brackets), the simplest hypothesis being that this acceptor is the sensitizing pigment itself, that is, chlorophyll in green plants and bacteriochlorophyll in purple bacteria. If we accept van Niel's hypothesis, we must conclude that the mechanism of photoreduction is somewhat different in hydrogen-adapted algae and in purple bacteria. The former contain ordinary chlorophyll, apparently unaffected by the adaptation process; the primary oxidation product of water, {OH}^, is thus probably the same in the ordinary and in the adapted state, and the difference in the final stages of oxidation must be attributed to the activation of the hydrogenase system and the simultaneous inactivation of the oxygen-liberating enzyme, Eo, as suggested by Gaffron (c/. Chapter 6, page 134). The identity of the primary processes in adapted and ordinary green algae is supported by the observations of Rieke and Gaffron (1943) that the maximum quantum yield and the saturation rate in flashing light are the same in the photo- reduction by adapted algae as in the photosynthesis in the nonadapted state. In the case of purple bacteria, on the other hand, the primary oxidation product, {OH}^, is naturally incapable of conversion into free oxygen; therefore, aerobic conditions may cause only a complete cessation of synthesis (if they lead to an oxidative deactivation of the hydrogenase) but cannot cause a transition to ordinary photosynthesis (with water as reductant), as this occurs in the "de-adaptation" of green algae. However, an even simpler description of the same facts becomes pos- sible if one assumes, as we have done above, that the primary photo- chemical process is the oxidation of an intermediate reductant, HZ, and that, in the course of normal photosynthesis, the oxidation product, Z, recovers hydrogen from water by a nonphotochemical reaction. In adapted algae, this recovery is blocked, and a reaction with a substitute reductant (e. g., H2) is made possible by a characteristic transformation of the enzymatic system (activation of the hydrogenase, deactivation of the deoxidase). This was the mechanism assumed in schemes 6.1 and 6. III. In purple bacteria, on the other hand, the primary reductant, HZ^, is different from the corresponding compound in green plants, HZ^, and its oxidation product, Z^, is incapable of oxidizing water, but capable 170 PRIMARY PHOTOCHEMICAL PROCESS CHAP. 7 of oxidizing the less stable ''substitute reductants" (H2, H2S, S2O3 , etc.). (Again, the simplest explanation of the difference between Z^ and Z^ is their identification with chlorophyll and bacteriochlorophyll respec- tively.) It may further be asked whether the primary oxidation product Z^ is the same in all purple bacteria, or whether it may further depend on the specific nature of the reductant (H2S, or S2O3 , or H2, etc.). Was- sink, Katz, and Dorrestein (1939) noticed that, in the spectra of living purple bacteria, the single absorption band of free bacteriochlorophyll in the far red is replaced by several red and infrared bands whose pattern varies from species to species (Vol. II, Chapter 22), and suggested that each of these bands corresponds to a different bacteriochlorophyll-bearing complex adapted to the reduction of a specific hydrogen donor. How- ever, this hypothesis disagrees with the assumption, made in chapter 6, that the hydrogenase system, even if it acquires hydrogen from different donors by means of specific "oxidoreductases," transfers it to a common acceptor (designated by Ah in Chapter 6). This seems to leave no place for a specific photocatalyst between Ah and carbon dioxide. Bibliography to Chapter 7 The Primary Photochemical Process 1914 Bredig, G., Umschau, 18, 362. 1916 Hoffmann, K. A., and Schumpelt, K., Ber. deut. chem. Ges., 49, 303. 1918 Wislicenus, C, ibid., 51, 942. 1923 Thunberg, K., Z. physik. Chem., 106, 305. Weigert, F., ibid., 106, 313. 1924 Weigert, F., ibid., 109, 79. 1930 Wurmser, R., Oxidations et reductions. Presses universitaires de France, Paris, 1930. Beutler, H., and Rabinowitch, E., Z. physik. Chem., B6, 233. 1931 van Niel, C. B., and MuUer, F. M., Rec. trav. botan. nierland., 28, 245. van Niel, C. B., Arch. Mikrobiol., 3, 1. 1932 StoU, A., Naturwissenschaften, 20, 955. Kautsky, H., Ber. deut. chem,. Ges., 65, 1762. 1933 Muller, F. M., Arch. Mikrobiol., 4, 131. MuUer, F. M., Chem. Weekblad, 30, 1. Shibata, K., and Yakushiji, E., Naturwissenschaften, 21, 267. 1934 Dhar, N. R., Trans. Faraday Sac, 30, 142. 1935 van Niel, C. B., Cold Spring Harbor Symposia Quant. Biol., 3, 138. Franck, J., Naturwissenschaften, 23, 226. Franck, J., Chem. Revs., 17, 433. Wohl, K., Z. physik. Chem., B31, 152. 1936 Stoll, A., Naturwissenschaften, 24, 53. BIBLIOGRAPHY TO CHAPTER 7 171 1938 Nakamura, H., Acta Phytochim. Japan, 10, 271. 1939 Wassink, E. C, Katz, E., and Dorrestein, R., Enzymologia, 7, 113. 1941 Franck, J., French, C. S., and Puck, T. T., /. Phys. Chem., 45, 1268. Franck, J., and Gaffron, H., in Advances in Enzymology, Vol. 1. Interscience, New York, 1941, p. 199. Franck, J., and Herzfeld, K. F., /. Phys. Chem., 45, 978. van Niel, C. B., in Advances in Enzymology, Vol. 1. Interscience, New York, 1941, p. 263. Seybold, A., Botan. Arch., 42, 254. 1942 Gaffron, H., /. Gen. Physiol., 26, 195. Gaffron, H., ibid., 26, 241. Gaffron, H., and Rubin, J., ibid., 26, 218. Franck, J., Am. J. Botany, 29, 314. 1943 Ruben, S., /. Am. Chem. Soc, 65, 279. Rieke, F. F., and Gaffron, H., /. Phys. Chem., 47, 299. Kautsky, H., and Franck, U., Biochem. Z., 315, 207. 1944 Emerson, R. L., Stauffer, J. F., and Umbreit, W. W., Am. J. Botany, 31, 107. Chapter 8 NONPHOTOCHEMICAL PARTIAL PROCESS IN PHOTOSYNTHESIS I. FIXATION OF CARBON DIOXIDE In the reaction schemes developed in chapter 7, the primary photo- chemical process was coupled with several nonphotochemical reactions. Since these reactions proceed at low temperatures, they probably require catalysts. Some of these undoubtedly are true enzymes, while others may be comparatively simple organic, or even inorganic, compounds. The realization that photosynthesis includes nonphotochemical re- action steps first came from kinetic studies. About 1905, Blackman had established that, under certain conditions, photosynthesis cannot be ac- celerated by further increase in light intensity, or carbon dioxide supply, but only by a raise in temperature; Willstatter and Stoll (1918) and War- burg (1919) interpreted this as evidence that photosynthesis includes a non-photochemical process (which they called "Blackman reaction") whose maximum rate is limited by the available quantity of an enzyme (Vol. II, Chapter 28). Willstatter and Stoll suggested, more specifically, that the Blackman reaction may be the liberation of oxygen from perox- ides. Warburg thought at first that the Blackman reaction consists in a transformation of carbon dioxide preliminary to its participation in the photochemical reaction, but later agreed with the Willstatter-Stoll hypothesis because of the similarity which he found between the effects of poisons on photosynthesis and on catalase activity (c/. page 284), The assumption, which appeared natural at that time, of a single "Black- man reaction" later led to various difficulties. Suggestions that there may be several "Blackman reactions" were made repeatedly, but without much conviction, until Franck postulated, on the basis of an analysis of various kinetic data, that photosynthesis must include (at least) three different catalytic reactions. In addition to the preliminary transfor- mation of carbon dioxide (first postulated by Warburg), and the peroxide decomposition (first suggested by Willstatter and Stoll), Franck assumed a third catalytic reaction, the stabilization of the primary photochemical products, which prevents their destruction by back reactions. He made no suggestion as to the chemical nature of this reaction, but our discus- sions in chapter 7, would indicate that it may possibly be a dismutation, 172 CARBON DIOXIDE-WATER EQUILIBRIUM 173 which converts free radicals into saturated molecules. Franck designated the catalysts involved in these three reactions as "catalyst A" (probably a "carboxylase"), "catalyst B" (the "stabiHzing" catalyst, perhaps a "mutase") and "catalyst C" (possibly a "catalase"); we have desig- nated them, in chapters 6 and 7, as Ea, Eb and Ec, respectively. These three catalysts are only a minimum; and the actual number of nonphoto- chemical reactions in photosynthesis may be larger than three. The evolution of oxygen, for example, may require two successive catalytic reactions (c/. Schemes 6.1, etc.), while the reduction and polymerization of the carbon dioxide-acceptor complex, {CO2}, to glucose, probably involves a whole series of oxidoreductions, dismutations, and polymerizations, requiring a complex catalytic system of which Franck's "catalyst B" may be only the first component. This and the next four chapters will deal with these catalytic proc- esses. We begin with the primary carbon dioxide fixation, represented in chapter 7 by the formula: CO2 >• {CO2! . Evidence pertaining to the nature of the carbon dioxide-acceptor complex in photosynthesis includes kinetic observations, experiments on carbon dioxide absorption by plants in the dark, carbon dioxide fixation by bacteria and other heterotrophic organisms, and the binding of carbon dioxide by different absorbers in vitro. A. The Carbon Dioxide Fixation in vitro* 1. The Carbon Dioxide-Water Equilibrium The primary absorber of carbon dioxide in all organisms is water, which forms 70-80% of the average tissue. The absorption of carbon dioxide by water is partly physical solution, partly chemical hydration, determined by the constants of the following equihbria: ^'s ^H20 ^Di (i = 4.54 X 10"' for the calculation of [H+] and [HCO3-], and neglecting the species HjCOj and COj — because of the small values of the constants KhjO and K^i (see below), one obtains the compositions of carbon dioxide solutions (at 25° C.) given in table 8.II. Table 8.II Distribution of Carbon Dioxide between Air and Distilled Water at 25° C. (calculated with Ks = 0.827 and Kt>^ = 4.54 X 10"^) PcOj- [CO 2] , mole/liter [HCO3-], mole/1. [H^],' atm. Gas" Solution pH 10-5 4.07 X 10-' 3.37 X 10-' 3.78 X 10-' 6.42 10-^ 4.07 X 10-0 3.37 X 10-^ 1.24 X 10-6 5.91 -=3.1 X 10-^ 1.26 X 10-s 0.94 X 10-5 2.07 X 10-« 5.68 10-3 4.07 X 10-5 3.37 X 10-5 3.92 X 10-" 5.41 i io-« 4.07 X 10-" 3.37 X lO-" 1.24 X 10-5 4.91 10-1 4.07 X 10-3 3.37 X 10-3 3.92 X 10-" 4.41 1 4.07 X 10-2 3.37 X lO-i* 1.24 X 10-3 3.91 " Assuming ideal gas laws. >• Cf. Quinn and Jones (1936), p. 119. ' Normal carbon dioxide content of free atmosphere. CARBON DIOXIDE-WATER EQUILIBRIUM 175 The table shows that, in very dilute carbon dioxide solutions, the concentration of bicarbonate ions is close to that of carbon dioxide molecules; in distilled water equi- librated with the atmosphere the ratio [CO2] : [HCOj"] is approximately 5:1. The constant of the hydration equilibrium: (8.2) Kmo = CH2CO,]/[C02] is not known precisely, but can be calculated approximately from the rate constants of hydration (kn) and dehydration (k'n)' (8.3) KH20 = knik'n McBain noticed, in 1912, that the hydration of carbon dioxide is a comparatively slow process. Thiel and Strohecker (1914) and Strohecker (1916) measured its rate in the alkaline region. More recently, Faurholt (1924), Stadie and O'Brien (1933), and Brinkman, Margaria and Roughton (1933) determined A;h ^^^ ^'h over a wide range of hydrogen-ion concentrations. Table 8.III Hydration and Dehydration of Carbon Dioxide" t°C. sec ' sec ' ^HzO AH of hydration, kcal/mole 0° 2.67 (F)*- 1.70 (BMR)^ 1.4 (SO)* 0.0030 (F)* 0.0026 (BMR)* 0.0027 (SO)*- 0.0027 (MU)" 0.0021 (RB) 0.0021 (MU) 0.0012 0.0015 2.80 (R) 18° 16.4 (F)» 11.1 (BNR)* 14 (R)* 0.025 (F)* 0.024 (BMR)* 0.0016 0.0022 1.40 (R) 25° 0.0275 (MU) 27° 31 (R)* 1.05 (R) 37° 77 (R)» 0.38 (R) 38° 0.23 (F)* 0.26 (BMR)* 0.10 (MU) " F— Faurholt (1924); BMR— Brinkman, Margaria and Roughton (1933); SO— Stadie and O'Brien (1933); RB— Roughton and Booth (1938); R— Roughton (1940); MU— MiUs and Urey (1940). * Buffered solutions! The hydration occurs, in addition to the reaction: (8.4) CO2 + H2O ^ HsCOj (equilibrium constant, Kmo) also through the bicarbonate ion: (8.5) CO2 + OH- . HCOr (equilibrium constant, Kqb.) 176 FIXATION OF CARBON DIOXIDE CHAP. 8 Reactions (8.4) and (8.5) can be interpreted as additions of HOH and 0H~ respectively to a C==0 double bond in CO2. According to Olsen and Joule (1940), the activation energy of reaction (8.4) is 19 kcal, and that of reaction (8.5) between 10 and 13 kcal. In the pH range of 8-10, the rates of the two reactions (8.4) and (8.5), are of the same order of magnitude. At pH > 8, the pH-independent reaction (8.4) predomi- nates, while at pH > 10, hydration and dehydration occur practically exclusively by reaction (8.5). At pH > 8, and 18° C, a dissolved carbon dioxide molecule lives, on the average, about one minute before it is hydrated, but remains only about 0.1 second in the hydrated state. Hydration and dehydration are accelerated by the anions of many weak acids, e. g. phosphate, borate, and acetate (Roughton and Booth 1938). This is of importance whenever buffers are used. Particularly strong is the effect of a specific enzyme, carbonic anhydrase, found in red blood corpuscles by Meldrum and Roughton (1932): cf. the reviews by Roughton 1934, 1935. Further results on the rate of hydration of carbon dioxide were obtained by Mills and Urey (1939, 1940) by the use of isotopic indicators. Their results are summarized in table 8.III. The apparent first dissociation constant of carbonic acid was redetermined by Maclnnes and Belcher (1933), who found: Kauko and Carlberg (1935) obtained a smaller value, 3.50 X 10"^. The true dissociation constant is considerably larger: (8.7) gp, = ^^r^?n?'"-^ = ^'^^ ^^T + ^^ ^f^ = 1-8 X 10-^ (25° C.) LJl2l>UjJ /VH2O AH2O The equilibrium constant of reaction (8.5) is: <«■'"> '^°" - [CftKOH-] ° [H^YoH-] - '■' X '"' <'5°C-' Thus, the standard free energies of hydration of CO 2 molecules are: AFhsO = + 3.7 kcal/mole (18° C.) for the hydration to H2CO3 molecules, and: Ah/^oh = - 10.4 kcal/mole (25° C). for the association with hydroxyl ions to HCOj~ ions. The second dissociation constant of carbonic acid (according to Maclnnes and Belcher) is: (8.8) Kv, = '"^HCOr]"' = ^-^^ ^ ^°~" ^^^° ^'^ A knowledge of the equilibrium constants Ks, Kn^o, K'di and Kdj permits the calculation of the equilibrium concentrations of all the molecular species in carbonic acid solutions (cf. Tables 8. IV and 8.V). Table 8. IV and figure 17 show the composition of carbonic acid solutions at 0° C. according to Faurholt (1924), based on the following values of the dissociation constants: Kdi = 2.24 X 10-' and Kv, = 3.2 X IQ-" (0° C.) CARBON DIOXIDE-WATER EQUILIBRIUM 177 Table 8.IV Equilibrium Concentrations in Carbonic Acid Solutions at 0° C. (from Faurholt 1924) PH CO., % HCO3-. 7o CO3-, % 99.888 1 99.888 2 99.886 0.0022 3 99.886 0.022 4 99.664 0.224 6 97.70 2.19 6 81.60 18.3 7 30.81 69.1 0.022 8 4.25 95.4 0.302 9 0.43 96.5 3.05 10 0.034 76.0 24.0 11 0.0011 24.0 76.0 12 3.07 96.9 Table 8. V shows the concentration of the species CO2 in the carbonate- bicarbonate buffer mixtures, which were first used in the study of photo- synthesis by Warburg (1919), and found many appHcations, particularly 100 % 80 60 40 20 - / / - 1 - C02 / 1- COJ / C03- - / / 1 1 ^ 1 \^ 1 1 7 13 Fig. 17. — Distribution of carbonic acid between CO2, HCOs" and CO3 as a function of pH (0° C, ionic strength 0) (after Faurholt 1924^). in kinetic investigations. The presence of a large quantity of bicarbonate ions has the effect of stabilizing the concentration of the carbon dioxide molecules and thus preventing a local exhaustion of carbon dioxide dur- ing intense photosj^nthesis. 178 FIXATION OF CARBON DIOXIDE CHAP. 8 Such exhaustion effects — caused by the slow diffusion of carbon dioxide through water — can falsify the results of kinetic studies; the use of carbonate buffer solutions prevents these errors. Warburg's buffer No. 6 (Table 8.V) contains, for example, 5800 HCOs" ions (and an equal number of CO3 ions) for each CO2 molecule. One can withdraw 10~3 moles of CO2 from a liter of this solution, that is, 120 times more than its initial content in CO2 molecules — and the concentration of CO2 molecules wdll not change by more than 12% (from the initial 8.7 X 10~^ moles/1, to 7.8 X 10-« moles/1.). Table 8.V Carbonate-Bicarbonate Buffer Solutions" Buffer Moles Alter [CO2] moles/liter X 10« (25° C.) no. [K2CO3] [KHCO3] 1. 0.085 0.015 0.481 2 0.080 .020 0.902 3 .075 .025 1.49 4 0.070 .030 2.29 5 .060 .040 4.48 6* .050 .050 8.67» 7 .035 .065 20.5 8 .025 .025 37.5 9 .015 .085 78.7 10 .010 .090 131. 11 .005 .095 290. " Warburg (1919), recalculated by Smith (1937), using the dissociation constants of Maclnnes and Belcher. * This buflfer is closest to the concentration of carbon dioxide in pure water equiHbrated with the free atmosphere. Other carbonate-bicarbonate buffer mixtures are listed in Kolthoff's book (1937), p. 259. Warburg's buffers are strongly alkaline (pH 8.5 to 11) and therefore "unphysiological," which calls for a certain caution in their use (c/. Vol. II, Chapter 27). Pure bicarbonate solutions have, at 25° C, in the concentration range of 0.001 to 0.1 mole per liter, an approximately con- stant pH of 8.37 (Kolthoff 1937, p. 21), and therefore also an approxi- mately constant ratio [HC03~]/[C02] = 90. The solubility of carbon dioxide in water is enhanced by the presence of solid alkaline earth carbonates. Investigators have usually been concerned with another aspect of this phenomenon — the dissolving action of carbonated water on sohd carbonates — because this effect has great practical importance. According to the equations: CARBON DIOXIDE ABSORPTION BY ALCOHOLS 179 (8.9a) MgCO, (or CaCOs) ^f= (8.9b) COr- + CO2 + HoO ^ Mg+^ (or Ca++) + CO3- =^ 2 HCO3- (8.9) MgCOs (or CaCOa) + CO2 + H2O ^ =^ Mg++ (or Ca++) + 2 HCOr the dissolution of one mole of alkaline earth carbonate is coupled with the absorption of one mole carbon dioxide from the air. The equihbrium (8.9) has been studied by Tillmans and Heublein (1912), Auerbach (1912), Tillmans (1919, 1921), Johnston and Wilhamson (1916), Frear and Johnston (1929), and Kline (1929). Table 8. VI contains some results. Table 8.VI Solubility of CO2 in Presence of Alkaline Earth Carbonates at 25° C. PCOj, [HCO3-] in (mole/1.) X 10' atm. Without solid carbonates In presence of CaCOa In presence of MgCOs 3.1 X 10-" 0.0021 1.02 12 1 X 10-3 0.0039 1.54 16 1 X 10-2 0.0124 3.4 26 1 X 10-1 0.0392 7.8 60 1 0.124 18.0 215 The "natural" concentration of bicarbonate ions in solution is increased by the presence of calcium carbonate, by a factor of 500 in air, and a factor of 140 in pure carbon dioxide. The effect of magnesium carbonate is ten times stronger. One-half of this bicarbonate comes from the solid salt and the other half from the atmosphere. The solubility of carbon dioxide in water is decreased by the presence of electrolytes (salting-out effect). For salt concentrations found in plant saps (~ 10~^ mole/liter), this depression may be of the order of 5-10% (cf. Quinn and Jones 1936, pp. 97 and 102). 2. Carbon Dioxide Absorption by Alcohols Plants contain, in addition to aqueous phases (cell sap and cytoplasm), phases of a predominantly "Upoid" character. Chloroplasts, in particular, are rich in hpoids. The carbon dioxide distribution between atmosphere and plant cells can therefore be affected by the solubility of carbon dioxide in hpoids. In general, carbon dioxide is more soluble in organic solvents, than in water — 3 times more soluble in toluene and benzene, 3.5 times more in ethanol, and 7.5 times more in acetone. In true lipids, the solubiUty may well be even higher. The "physical" solution of carbon dioxide in organic solvents is often enhanced by chemical reactions, analogous to hydration. As in water, the effect of chemical solvation is small in pure solvents, but becomes large when occasion is given for the formation of anions, analogous to the bicarbonate ions in water. In the case of alcohols, for example, the molecular solvation equilibrium (8.10) 0C=0 + ROH . OC— OH I OR 180 FIXATION OF CARBON DIOXIDE CHAP. 8 which leads to carbonic acid esters, is overshadowed in the presence of alcoholates, by the ionic equihbrium (8.11) 0C=0 + RO- > OC— O- )R L which is analogous to reaction (8.5) in water. Faurholt (1927) estimated, for example, that the two carbon dioxide methanolation constants are (at 0° C): (8.12) Knon = ^?^?^ - 0.01; AFroh = + 2.7 kcal (as against KuiO — 0.001 in water), and (8.13) Kro = [^cSkRO"] " ^ ^ ■^°'' ^^°" ^ ~ ^'^ ^^^^ (the corresponding value for water is Kqh = 4 X 10^) Since the dissociation of methanol is much weaker than that of water ([RO-][H+] ^ 10-"), neither (8.10) nor (8.11) can contribute more than one per cent to the solubility of carbon dioxide in pure methanol; but in sodium alcoholate solutions, with their high concentration of RO" ions, carbon dioxide is eagerly absorbed with the formation of RCO3- ions. QuaUtatively, the absorption of carbon dioxide by alcohols in the presence of alkali was known for a long time. It was studied by Siegfried and Howwjanz (1909), who observed the absorption of 0.2 to 1 mole of carbon dioxide by one mole of methanol, ethanol, glycol, glycerol, erythrol, quercitol, lactose, sucrose, and lactic acid, in the presence of calcium hydroxyde. In aqueous alcohol, carbon dioxide becomes an object of competition between water and alcohol. The resulting equilibria have been studied by Faurholt il927^^'^) and Faurholt and Jespersen (1933), for methanol, ethanol, propanol and sucrose. They found that the contribution of alcohols to the absorption of carbon dioxide in alcohol-water mixtures is comparatively small and disappears on both side of the "bicarbonate range" (pH around 8.4) because of the decomposition of the carbonic acid esters (into CO2 and alcohol on the acid side, and into CO3 — and alcohol on the alkaline side of this range). The ratio: <«•") ^-'- - [hco.-]Troh] is 0.083 for methanol, 0.044 for ethanol (at 0° C), and even smaller for sucrose. Con- sequently, the proportion of carbon dioxide bound to alkyl in a molar methanol solution is only 7.4% at pH 8-9. Without reference to Faurholt's quantitative results, Baur and Namek (1940) made some rough qualitative observations concerning the carbon dioxide absorption by alcohols. In addition to the rapid gas uptake by dissolution, they noticed a slow absorption which they attributed to esterification. For the extent of this chemical binding, they gave values between 18 ml. of carbon dioxide per mole of ethanol (8 X lO—* mole/1.) which is many times more than one would expect from Faurholt's equihbrium constants, and 470 ml. (0.02 mole/1.) per mole of phytol. Since phytol is a component of chlorophyll, Baur considered this result as significant for the carbon dioxide fixation in photosynthesis. CARBON DIOXIDE ABSORPTION BY AMINES 181 3. Carbon Dioxide Absorption by Amines Carbon dioxide addition to N— H bonds is similar to its addition to O — H bonds in HOH and ROH, except that the basicity of the amines may lead to the formation of carbamates (by the addition of a second molecule of amine to the primary formed carhamic acids), for example: (8.15a) 0C=0 + RNH2 > OC— OH (R-carbamic acid) NHR (8.15b) RNH— COOH + RNH2 > RNH— COONH3R RNH— COO- + RNH3+ (R-carbamate) (8.15) 0C=0 + 2 RNH2 » RNH— COO- + RNH,+ Faurholt (1921, 1922, 1924, 1925) investigated these equilibria in aqueous solutions, and found for the ratio: ^«1R^ TC [RNHCO2-] (8.16) Knhr/oh = ^RNH2] [HCO3-] values of about 2 for ammonia, 165 for methylamine, 46 for dimethylamine and 32 for glycine, that is, considerably larger than the corresponding ratios for alcohols. How- ever, because of the basicity of the amines, the RNH2 molecules constitute (except in very alkaline solutions) only a small proportion of the dissolved amine (most of it being present as RNH3+ ions, which have no affinity for carbon dioxide). This restricts the contribution of amines to the carbon dioxide absorption by aqueous solutions. Nether- theless, in a one-molar solution of methylamine, at 18° C. (pH 8.9), 61% of the absorbed carbon dioxide is present in the form of carbamate while 39% is present as HCOa" or CO3 — ions. In less alkahne solutions, the extent of carbamination is much smaller; solid carbamates decompose in pure water. Siegfried (190512), Siegfried and Neumaim (1908) and Siegfried and Liebermann (1908) have studied quaUtatively the formation of carbamates in aqueous solutions of amines, amino acids, peptones, and proteins, in the presence of alkali (calcium hydroxyde) and Fichter and Becker (1911) have observed the formation of carbamate from gaseous methyl amine and carbon dioxide at low temperatures. According to Siegfried, the proximity of an oxidized group (e. g., carboxyl) favors the addition of carbon dioxide to the amino group; thus, amino acids absorb carbon dioxide more eagerly than alkyl amines. Siegfried suggested that the fixation of carbon dioxide by amino acids may be of importance for photosynthesis. He claimed (1905^) that this reaction can occur even in absence of alkaU. However, this conclusion, based on measurements of the increase in the conductivity of glycocoU solutions by saturation with carbon dioxide, requires confirmation. The carbamination equilibria of simple amino acids were again investigated by Stadie and O'Brien (1936). Theyconfirmedthefact that the dipolar ions NHj^-RCOO-, which predominate near the isoelectric point, do not unite with carbon dioxide at all; this association is restricted to the anions NH2-RC00-, which predominate on the alkaline side of the isoelectric point. The equihbrium: (8.17) NH2RCOO- + CO2 V RCOO- HOOCNHRCOO-;;— ^HN + H+ COO- 182 FIXATION OF CARBON DIOXIDE CHAP. 8 is established so rapidly that it can be studied, in aqueous solution, without interference from the side of the more slowly estabhshed hydration equiUbrium. The constants (8.18) xV Carbarn. — NH coo- RCOO- [H+] [H2NRCOO-] [CO2] of alanine and glycine are, according to Stadie and O'Brien, of the order of 2.5 X IQ-^ at 20° C. Consequently, half-saturation of these amino acids with carbon dioxide is reached at [CO2] = 4 X 10"^ mole per liter {i. e., at a partial pressure of 90 mm.) if pH = 8, and at a ten times smaller pressure if the pH is 9. Carbamination is of particular importance for the carbon dioxide transportation by blood. The entire absorption of carbon dioxide by blood w^as attributed, until 1928, to the carbonate-bicarbonate interconversion, although Bohr had postulated the existence of a hemoglobin-carbon dioxide compound in 1905, and Siegfried had demon- strated in the same year the formation of carbamates in blood serum. Kinetic studies by Henriquez (1928, 1931), Margaria and Green (1933), and Meldrum and Roughton (1933) have proved since that an important proportion of carbon dioxide is present in the form of carbamate. The carbamination of blood was discussed quantitatively also by Roughton (1935) and Stadie and O'Brien (1937). Despite the presence of carbonic anhydrase in blood, the carbamination equilibrium can be studied, independently from the hydration equiUbrium, by poisoning this enzyme with cyanide (0.05-0.1 mole/1.). At 0°, with poisoned enzyme, the half-saturation of oxyhemoglobin is reached at about 10 mm. carbon dioxide in the air, and that of reduced hemoglobin at about 30 mm. The heat of carbamination is considerable, about 17 kcal per mole. Table 8.VII illustrates the role which carbamination plays in the carbon dioxide balance of blood. Table 8.VII BiCAKBONATES AND CARBAMATES IN BlOOD Component Arterial Venous Plasma Cells Plasma Cells pH CO2 (free), ml. /I. CO2 as HCO3-, ml. /I. CO2 as carbamate, ml. /I. 7.45 16 331 10 7.12 8 98 20 7.43 18 352 11 7.11 9 105 26 Up to 20% of carbon dioxide in red blood cells is present as carbamate, and 45% of the difference between the carbon dioxide content of these cells in venous and arterial blood is caused by a shift of the carbamination equilibrium. The possible role of amino acids in the carbon dioxide absorption by plants will be mentioned on pages 191 and 194. We must, at this point, cite a case of carbamination which may conceivably be of importance for photosynthesis. For C — H bonds, the substitution of a metal for hydrogen makes the affinity for carbon dioxide stronger, as shown by the eager carboxylation of Grignard's reagents and other metalloorganic compounds. Is it possible for nitrogen-metal bonds also to be more efficient carbon CARBOXYLATION EQUILIBRIA 183 dioxide acceptors than the nitrogen-hydrogen bonds? In other words, may we expect the reaction: (8.19) 0C=0 + R=N— M > OC— NR OM where M = metal, to proceed more easily and completely than reaction (8.15a)? The importance of this question hes in the fact that chlorophyll contains two nitrogen-magnesium bonds. The interaction of carbon dioxide with chlorophyll in vitro will be discussed in chapter 16. One mole of sohd or colloidal chlorophyll ap- parently can absorb up to two moles of carbon dioxide. In interpreting this uptake (page 454), we shall have to consider a reaction of type (8.19) as one possibility. 4. Carboxylation Equilibria The reaction which has aroused most interest in connection with the primary carbon dioxide fixation in photosynthesis is carboxylation. It can be interpreted as an addition of an organic compound RH to the C=0 double bonds in carbon dioxide; in other words, the C — H bond plays in carboxylation the same part which the N — H bond plays in carbamination and the — H bond in the hydration of carbon dioxide. OH / (8.20) OC =0 + RH > OC R Respiration ends with the elimination of carbon dioxide by decarboxylation of certain keto acids. Since photosynthesis is the reversal of respiration, one is tempted to consider the reversal of this last step in respiration as a possible first step in photosynthesis (Thimann 1938). However, the analogy between the role of decarboxylation in respi- ration and the role of a preliminary carboxylation in photosynthesis is not quite so close as it may appear. In the respiratory process, decar- boxylation is a step in the breakdown of the sugar molecule. Carboxyla- tion would play a corresponding role in photosynthesis only if carbon dioxide were added to an intermediate reduction product, and not to a catalyst which must be restored at the end of the reaction. The car- boxylation of chlorophyll or another temporary carrier may be useful for kinetic purposes, but it does not constitute a first step in the building up of a carbon chain. Carboxylations and decarboxylations do not change the average reduction level of the reacting system and hence have only relatively small heat effects (c/. page 216). Table 8. VIII shows that decarboxyla- tions usually are slightly endothermal (AH > 0). If decarboxylation leads to the disruption of a conjugation between the C=0 double bond in the carboxyl and another C=0 double bond in the molecule (as in 184 FIXATION OF CARBON DIOXIDE CHAP. 8 pyruvic and oxalic acid) the energy is not markedly different; but conju- gation with a C=C double bond apparently has a stabilizing effect, since the decarboxylation energy of benzoic and fumaric acid is as high as 16-17 kcal. Table 8.VIII Heat and Free Energy of Decarboxylation" Acid Reaction AH AF I. No CONJUGATION + 5.5 + 5.7 + 3.6 + 5.0 + 3.3 + 2.2 + 5.3 — 89 Formic ' 1 H+ + HCOO-(aq.) > HoCg.) + COsCaq.) H2O + HCOO-(aq.) > Ho(g.) + HC03-(aq.) -11.5 - 0.8 — 11 8 Acetic Malonic Hexylic H2O + CH3C00-(aq.) > CH^Cg.) + HC03-(aq.) - 5.9 /-I TT nor»TT/'a ^ > p tt n "^ 1 PH.. Ueiin^-^UUlHs.j > ^-^6n.i4UJ 1 ^yj'z — C=-0 II. Conjugation | — C— PTT pnpnnTTH ^ » ptt pttom "i i onjir ^ + 1.4 - 1.8 + 3.1 + 8.6 (-15.2) Pyruvic • TT r\ 1 PTT fTir^rif^—trr. \ \ II2U + Uil3UUUUU (.aq.; > CH2CH0(aq.) + HC03-(aq.) TTnr^r^ pnr\TTCn ^ ^ ttpi^ottm ^ 1 (- 2.9) Oxalic IIUUU — UUUiHs.; ' xlL/UUxH,i.; 1 C02(g.) TTonp ponTT/'o ^ > tt ^0- ^ 1 '^ PO.. -13.4 — 22 3 2 H2O + -OOC-COO- > H2(g.) + 2 HC03-(aq.) - 8.8 Acrylic Fumaric Benzoic Salicylic GalUc 1 III. Conjugation — C — C— PIT PTT Pnnili'c: ^ t Vf (o- \ \ POuCrr ^ + 9.1 + 16.6 + 16 + 9.1 + 5.0 UII2 — *-^ii UUU1H.S.; > 02ii4(,g-J 1 ^*-'2i,g.; TTnOP PTT PTT Pr^riTT/'o 'i k CH2-CH— COOH(s.) + C02(g.) P TT Pr»OTT^r- ^ 4 P TT I'M 1 PO.Yrr "1 — 4 6 C6H4(OH)COOH(s.) > C6H50H(s.) + C02(g.) p TT ^rvTT^ prkr»TT/'n ^ » p tt (r\Vf\ (v \ i - 3.3 C02(g.) " Cf. bibliography to chapter 3, page 56. The free energies of decarboxylations of pure (solid or liquid) organic acids are more negative (by as much as 10 or 15 kcal) than the total energies of these reactions, and thus do not favor back reactions. The AF values are negative even for benzoic and salicyhc acid, despite their stabilization by conjugation. This explains why attempts of Widmer CARBOXYLATION EQUILIBRIA 185 (1929) and Hirsbrunner (1934) to reverse the decarboxylation of salicjdic, gallic, and phloroglucin-carboxjdic acid have been unsuccessful. In alkaline solutions, where carbonic acid is present in the form of anions, the carboxylation reaction becomes: (8.21) HCO,- + RH > RCOO- + H^O The free energy of reaction (8.21) is considerably less positive than that of reaction (8.20), because carbonic acid is weaker than most carboxylic acids. This explains why the decarboxjdation of formic acid in alkaline solution is reversible. This reversibilit}^ was demonstrated by experi- ments with biological catalysts {Escherichia coli, cf. page 208). However, with acids much weaker than formic acid (for example, acetic acid), not even the substitution of bicarbonate ions for carbon dioxide molecules will suffice to make carboxylation thermodj'namically possible at low temperatures and low partial pressures of carbon dioxide. Aromatic compounds (benzene, phenol, polyphenols) as well as noncyclic, unsaturated compounds, whose free energies of carboxylation in the acid range are less positive than those of the saturated aliphatic compounds, can be expected to show negative free energies of carboxyla- tion in alkaline media. It is well known that phenols can be carboxylated, in the presence of alkali, at comparatively low temperatures (100-200° C.) and low carbon dioxide pressures. (The usual method of preparation of salicylic acid is by carboxylation of phenolate.) Ruben and Kamen (1940) suggested that the presence in plants of polyphenols (of the type of tannin and quercetin) may be of importance for the fixation of carbon dioxide. However, it still remains to be demonstrated that carboxyla- tions of this type can occur at the comparatively low pH values prevailing in plant cells. In respiration, the elimination of carbon dioxide involves the decar- boxylation of two a-keto acids, oxalacetic and pyruvic: (8.22) HOOC— CH2— CO— COOH > CH3— CO— COOH + CO2 (8.23) CH3CO— COOH > CH3— CHO + CO2 According to table 8. VIII, the decarboxylation of pyruvic acid is not easily reversible (AF = — 15 kcal), not even in alkaline solution (AF = — 3 kcal). Carson, Ruben, Kamen, and Foster (1941) tried un- successfully to prove, by the use of radioactive carbon dioxide, the rever- sion of this reaction in enzymatic systems. No data are available in standard compilations on the thermochemical properties of oxalacetic acid. However, the decarboxylation of this acid was studied by means of radioactive indicators by Carson, Foster, Ruben, and Barker (1941), Wood, Werkman, Hemingway, and Nier (1940, 1941), Krampitz and Werkman (1941), and Krampitz, Wood, and 186 FIXATION OF CARBON DIOXIDE CHAP. 8 Werkman (1943) ; and good evidence of reversibility obtained. Werkman and coworkers found, for example, that if nonradioactive oxalacetic acid is allowed to lose by enzymatic action in an atmosphere of radioactive carbon dioxide about one-half its carbon dioxide content, and the remain- ing portion is analyzed for radioactivity, a measurable quantity of active carbon is found in the acid (in the carboxyl adjoining the CH2 group). It thus seems that, in the case of oxalacetic acid, the equilibrium lies further on the side of carboxylation than it does in other organic acids. It would be interesting to check this conclusion directly by the car- boxylation of pyruvates. Baur and Namek (1940) suggested that the carboxylation equihbrium can be shifted towards association not only by the formation of salts (as discussed above) but also by the formation of esters: (8.24) 0C=0 + R'OH + R"H > R"COOR' + H2O Experiments, by which the occurrence of reaction (8.24) was allegedly proved, consisted in determining the effect of phloroglucinol, C6H3(OH)3, and of rosohc acid (both rep- resenting R"H) on the carbon dioxide absorption by glycerol (representing R'OH). A certain increase in absorption was observed in the case of phloroglucinol, but no glycerate of the phloroglucinol-carboxylic acid, C6H2(OH)3COOH, could be isolated. The addition of 0.7 g. rosohc acid to 5 ml. of glycerol caused an increase in the carbon dioxide absorption by 2.5 ml.; this, as well as certain color and fluorescence effects, was interpreted as evidence of the formation of a dyestuff derivative of triphenylmethyl- carboxylic acid by the carboxylation of about 5% of added rosohc acid. As in the case of his work on the mechanism of photosynthesis (c/. Chapter 4), the conclusions of Baur run far ahead of the very rough experiments. Organic compounds which are known to absorb carbon dioxide eagerly, with the formation of carboxyl groups, are metal alkyls, e. g. Grignard reagents. These reactions can be interpreted as additions of R — M (M = metal) to 0=0 R / (8.25) 0C--0 + RM > OC OM The instabihty of the carbon-metal bonds and the stabihty conferred on the salts by ionic dissociation offer sufficient explanation of why, in this case, the equihbrium Ues far on the side of synthesis. We have reviewed the reversible addition of carbon dioxide to O — H, N — H, N — M, C — H and C — M bonds. Before applying these results to observations on the carbon dioxide fixation by plants, it might be worth while to mention one important example of reversible carbon dioxide fixation in nature — the equihbrium between carbon dioxide and carbonic anhydrase. According to Roughton and coworkers (1940), the equihbrium constant: where E = enzyme, is of the order of 0.1 atm. at 0° C, and 1 atm. at room temperature. Thus, the energy of formation of the E-C02 complex is of the order of AH = — 15 kcal, while the free energy is about AF = + 1.85 kcal at 25°. The chemical nature of this complex is unknown. DEFINITION OF CARBOXYLATION 187 5. Is Carboxylation a Reduction of Carbon Dioxide? It is customary to speak of "reduction of carbon dioxide" whenever a carbon dioxide molecule is incorporated into an organic compound with the formation of a new C — C bond. This practice leads to mis- understandings when processes of this kind are put on the same level with the reduction of carbon dioxide in photautotrophic and chem- autotrophic organisms. It is necessary to distinguish clearly between the two types of reactions involving carbon dioxide: reversible additions (e. g., the addition of CO2 to RH, leading to the substitution of C — C bonds for C — bonds) ; and true reductions (characterized by the creation of new C — H bonds). Whether carboxylation should be called a "reduc- tion" of carbon dioxide at all is a matter of convention. The definition of the word reduction is unambiguous only in the case of intermolecular oxidation-reductions, in which the reaction partners exchange electrons (or hydrogen atoms) and then separate, one having experienced oxidation and the other reduction. If the reaction partners remain Unked together, the definition becomes vague. To find out whether an oxidation- reduction has occurred, one may consider the positions of electrons or hydrogen atoms before and after the reaction. By this criterion, a carboxylation: (8.27) RH + CO2 > RCOOH could be considered as the oxidation of the organic radical R and reduc- tion of carbon dioxide, since it involves the shift of a hydrogen from RH to CO2. However, by the same token, one could also describe the hydration of carbon dioxide: (8.28) HOH + CO2 > HOCOOH as a reduction of carbon dioxide and oxidation of water. Both in (8.27) and (8.28), the hydrogen is transferred to oxygen (in the carbon dioxide); because of the high aflfinity of oxygen for hydrogen, this requires no supply of energy. A true reduction of carbon dioxide (as defined above) would require the shift of hydrogen to carbon. If, after such a shift (from RH or H2O to CO2) the products remain united in a single mole- cule, the results will be: (8.2 ) RH + CO2 > ROCHO and (8.30) HOH + CO2 > HOOCHO that is, the formation of Sifor7nic acid ester and performic acid, respectively. This would constitute a true reduction of carbon dioxide (and oxidation 188 'fixation of carbon dioxide chap. 8 of the donor, RH, or water). (Reactions of type 8.30 were postulated by Willstatter and Stoll in 1918 as the main photochemical steps in photosynthesis.) To sum up, it seems more logical to reserve the term "carbon dioxide reduction" for reactions in which hydrogen atoms (or electrons) are transferred from "donor" molecules to the carbon atom in carbon diox- ide, and not to apply it to carboxylations and similar additive reactions. True, in dealing with metabolic processes, it is often not clear whether an observed consumption of carbon dioxide is caused by addition or reduction; but the distinction should be kept in mind and applied whenever possible. B. Carbon Dioxide Fixation by Living Cells* The review of different reversible carbon dioxide addition processes in vitro in the preceding section illustrates the variety of reactions which may occur when carbon dioxide comes in contact with living organisms. This interaction has been studied in detail only in the case of blood; observations of the absorption of carbon dioxide by other tissues, animal or vegetable, have for the most part been qualitative. However, the work of Spoehr and Smith on sunflower leaves has opened the way to a more quantitative treatment, which is a prerequisite for the complete understanding of the fate of carbon dioxide in photosynthesis. From the studies of Smith, water, phosphate, and alkaline earth car- bonates emerge as the three main factors determining the carbon dioxide balance of nonilluminated leaves under high partial pressures of carbon dioxide. It was mentioned above that the carbon dioxide balance of blood was originally attributed exclusively to the conversion of carbonates into bicarbonates. Later, it was found that carbamination also plays a limited, but not negligible, part. A similar development may possibly occur in the theory of the carbon dioxide absorption by plants; but the suggestion of Willstatter and Stoll that carbamination is the inain factor in this absorption is not borne out by the analysis of Spoehr and Smith. While dissolution in water and bicarbonate formation (and possibly carbamination) determine the carbon dioxide balance of plants under high partial pressures of this gas, the carbon dioxide binding in the complex, {CO2J — which is probably a carboxylation — comes into greater prominence under low pressures, for instance, in the free atmosphere. Under these conditions, the {CO2} complex may account for carbon dioxide quantities of the same order of magnitude (1 to 5 X 10~^% of the dry weight of the leaves) as those absorbed by conversion into bicarbonate. * Bibliography, page 211. CONVERSION OF CARBON DIOXIDE INTO BICARBONATE 189 1. Solubility of Carbon Dioxide in Plant Sap In all measurements of the carbon dioxide absorption by plant tissues, the solubility in cell water has to be corrected for in order to determine the extent of "chemical" binding. This component is small at the low carbon dioxide concentrations, but grows with increasing pressure, when the chemical absorbers become saturated. According to table 8.1, the cell water, if it wxre pure, would con- tain, in contact with the free atmosphere at 25° C, about 9 X 10"^ mole per liter of CO2 molecules, and in contact with an atmosphere of pure carbon dioxide, approximately 3 X IQ-^ mole per liter. Since an aver- age leaf is about 80% water, the first concentration corresponds to about 2 X 10-''%, and the second to 0.6% dissolved CO2, relative to the dry weight of the leaves. A correction is needed in exact calculations for the effect of salts and nonelectrolytes on the solubility of carbon dioxide. However, this effect cannot exceed a few per cent (c/. page 179). Smith (1940) found that partial drying of leaves affects the absorption of carbon dioxide somewhat more than can be accounted for by the amount of evaporated water, and ascribed this effect to the solubility-depressing influence of sugars and salts. He noticed also that expressed and acidified cell sap absorbs about 10% less carbon dioxide than the same volume of pure water. Leaves of Sedum prealtum, whose sap has a strongly acid reaction (pH 4.08) and in which no bicarbonates can be formed, also absorb less carbon dioxide than calculated from the solubility of this gas in pure water. The possible contribution of lipoids to the reversible absorption of carbon dioxide by plants (c/. page 179) has never yet been taken into consideration. The small volume of the lipoid phase compared with the hydrophilic phases (cytoplasm and cell sap) perhaps makes the omission permissible. 2. Conversion of Carbon Dioxide into Bicarbonate in Plants If the cell water were unbuffered, about 15% of dissolved carbon dioxide would be in the form of bicarbonate ions in ordinary air, and a smaller proportion in an atmosphere enriched in carbon dioxide. The sap, under these conditions, would be acid. The absorption of carbon dioxide in excess of normal solubility, actually observed with almost all investigated plants, must be attributed to a conversion of carbon dioxide into bicarbonate by alkahzing agents. The most important of these are soHd alkaline earth carbonates and dissolved primary phosphates. The presence in plants of solid carbonates was discovered by Berthelot and Andre in 1887. They removed "free" carbon dioxide (that is. 190 FIXATION OF CARBON DIOXIDE CHAP. 8 dissolved carbon dioxide and one-half the carbonic acid of the bicar- bonates) by pumping, then extracted the leaves with water, acidified the extract and the insoluble residue, and measured the evolved carbon dioxide. Table 8. IX contains some of the results obtained in this way. Table 8.IX Carbonates in Plants (after Berthelot and Andre) CO2, % of dry matter Snecies Insoluble carbonates Soluble carbonates Chenopodium quinoa 0.03 Rumex acetosa 0.50 0.14 Oxalis strida 0.36 0.06 Amaranthus caudatus 0.09 Mesembrianthemum cristallinum none 0.13 to 0.72 The quantities of soluble carbonates (presumably, alkaU carbonates) found by Berthelot and Andre appear too high when one considers the comparatively low pH of the cell sap, and do not agree with the quantity of carbon dioxide which the leaves absorb under an increased pressure of carbon dioxide, and liberate in vacuo. Recent investigations of Smith make it probable that divalent cations account for all the carbonate anions found in the leaves. The presence of phosphates in the cell sap of green plants has been demonstrated by Martin (1927). Their concentration is of the order of 10-- mole per liter. Primary phosphate absorbs carbon dioxide according to the equation: (6.31) CO2 + H2O + HPO, :^ HCO3- + H2PO4- The presence in leaves of alkaline earth carbonates and primary phosphates makes it necessary to consider these factors first in the interpretation of the reversible carbon dioxide absorption by plants. The first determinations of the reversible carbon dioxide absorption by leaves were carried out by Willstatter and Stoll (1918), with Urtica dioica (nettle) and Helianthus annuus (sunflower). The absorption isothermals are reproduced in figure 18. Half-saturation is reached in Helianthus at about 40 mm., and in Urtica at a somewhat higher pressure. The maximum absorbed quantities (after correcting for solubility) are, in both cases, of the order of 1 ml. CO2 per 10 g. fresh leaves (5 X 10"^ mole/1., or about 0.1% of the dry weight of the leaves). Willstatter and Stoll mentioned the carbonate-bicarbonate conversion as a possible explanation of the carbon dioxide absorption, but thought CONVERSION OF CARBON DIOXIDE INTO BICARBONATE 191 that the comparatively high pressure required for saturation argues against this hypothesis and in favor of carbamate formation. The com- parison of figure 18 with Smith's figure 19, which contains the calculated absorption curve for a solution of primary phosphate, shows, however, that the Willstatter-Stoll results can be accounted for almost entirely by the phosphate buffer equilibrium. 20 16 12 E O 8 O - ^^ ^ -o - l^ ^ - A -^■ • r\.- ...o- ' 1 1 1 1 1 1 1 1 1 1 1 20 16 .12 E d^8 u - ^ y^ >*- - / y" - y ^--'' ~ A - -O' ^ _^^^ "■'* ~i-\^ ...-■a- 1 1 1 1 1 1 1 1 1 1 1 3 6 9 12 15 18 (A) COi,56 3 6 9 12 15 18 (8) COjV. Fig. 18.— Absorption of carbon dioxide at 5° C. A. Helianthus annuus (about 20 g. fresh leaves); B. Urtica dioica (same quantity) (after Willstatter and Stoll 1918). Leaves Water in the leaves - - Leaves without water (calculated) According to Willstatter and Stoll, sunflower leaves absorb, under 150 mm. partial pressure, twice as much carbon dioxide as can be dissolved in pure cell water and, under a partial pressure of 0.75 mm. (0.1% CO2 in the air), twelve times as much. The ratio between chemically bound and physically dissolved carbon dioxide must become even larger at still lower pressures. Thus, Schafer (1938) found that leaves of Vicia faha (broad beans) may liberate, in vacuo, fifty times as much carbon dioxide as could have been dissolved in the cell water under the partial pressure of carbon dioxide in the air (0.23 mm.). Sj^stematic attempts to elucidate the nature of the carbon dioxide absorbing agents in plants were first undertaken by Spoehr and coworkers. Spoehr and McGee (1923, 1924) proved that dried or frozen leaves retain the capacity for carbon dioxide absorption. They found that the absorb- ing compounds can be extracted from the leaves by ether-saturated water, and considered this at first as a proof of their proteinaceous nature. Later, however, Spoehr and Newton (1925, 1926) found that the ab- sorbing agent (reprecipitated by alcohol from the ether-water extract) does not contain enough nitrogen to account for the carbon dioxide absorption on stoichiometric basis. They therefore abandoned the carbamate hypothesis and turned to the bicarbonate hypothesis. 192 FIXATION OF CARBON DIOXIDE CHAP. 8 Spoehr and Newton observed that leaves of sunflower arid nettle absorb about twenty times more carbon dioxide than those of alfalfa, rhubarb, spinach, and hydrangea. However, the properties of sunflower are not exceptional, as shown by the later data of Smith (1940), some examples of which are given in table 8.X. Table 8.X Carbon Dioxide Absorption by Fresh Leaves CO 2 absorbed by 10 g . at 15° C. under 1 atm pressure of CO2, ml. Leaves of Total Dissolved (calcd.) Excess Helianthus annuus Malva parviflora Libo cedrus EschschoUzia californica Rosa species Quercus douglasii Trifolium repens 10.2-11.6 11.9 7.7 9.0 8.0 6.4 10.2 7.5-8.0 7.8 5.0 7.6 5.9 5.6 7.6 2.2-3.7 4.1 2.7 1.4 2.1 0.8 2.6 In all species listed in table 8.X, "chemical" absorption of carbon dioxide enhances the solubility under atmospheric pressure by 20-50% (in agreement with the 100% increase under 150 mm. pressure found by Willstatter and Stoll). The absorption is fully reversible — in fact, a little more carbon dioxide is usually removed by evacuation than has been absorbed under high pressure (obviously because of respiration). Acidi- fication liberates an additional quantity of carbon dioxide by the de- composition of neutral carbonates. Table 8.XI Carbon Dioxide Content of Helianthus Leaves " Material CO2 content in excess of normal solubility, ml. No. "Reversible" CO2 (.under 1 atm.)'' "Irreversible" Total 002" (under 1 atm.) 1 Living leaves 3.0 15.8 18.8 2 Frozen leaves 8.9 7.8 16.7 3 Water extract from 2 4.8 -0.3 4.5 4 Water-insoluble residue of 2 4.9 4.3 9.2 5 CO2 — water extract from 4 5.9 5.3 11.2 6 CO2 — water-insoluble residue of 4 0.5 0.5 " All figures refer to 10 g. fresh leaves or material derived therefrom. ' Amount of carbon dioxide absorbed when CO2 pressure is increased from to 1 atm. ' Amount of carbon dioxide released by cold dilute acid, minus the "reversible" CO2. CONVERSION OF CARBON DIOXIDE INTO BICARBONATE 193 The total quantity of chemically bound carbon dioxide in Helianthus leaves, equilibrated with an atmosphere of pure carbon dioxide, is 17-19 ml. per 10 g. fresh leaves, corresponding to an average CO2 concentration of 0.1 mole per liter, or 2% CO2 relative to the dry weight of the leaves. This absorption equilibrium can have nothing to do with chlorophyll, whose average concentration in the leaves is only of the order of 2 X 10"' mole per liter. This conclusion is borne out by the observations that yellow leaves absorb the same quantities of carbon dioxide as green leaves (Willstatter and Stoll), that white leaves also yield carbon dioxide in vacuo (Schafer), and that stalks, roots, and petals show^ the same reversible carbon dioxide absorption as leaves (Smith). Schafer found that the quantity of dissociable carbon dioxide increases in light; but this can scarcely be taken as an indication of a direct relationship between the agent absorbing carbon dioxide and the photochemical apparatus of the leaves. Not all figures in table 8. XI can easily be interpreted. The properties of fractions 3, 4, 5 and 6 are understandable, but the carbon dioxide up- take of whole leaves (rows 1 and 2) is considerabl}^ larger than the sum of the volumes taken up by fractions 3 and 4, and its distribution between "reversible" and "irreversible" CO2 is remarkably different for the living and the frozen leaves. Fraction 3 behaves as a buffered solution which takes up carbon dioxide under pressure (in excess of the solubiUty of this gas in pure water) by conversion into bicar- bonate, but releases all of it upon evacuation. It will be shown below (cf. Fig. 19) that this uptake can be attributed practically entirely to the presence of a phosphate buffer. The behavior of fraction 4 is that of an insoluble carbonate, which absorbs reversibly an equivalent quantity of carbon dioxide by conversion into bicarbonate (c/. p. 179), and thus contains, upon saturation, equal amounts of "reversible" and "irreversible" carbon dioxide. This interpretation is confirmed by the properties of fractions 5 and 6, since they show that the carbon dioxide-absorbing component of fraction 4 is completely soluble in carbonated water. Fractions 3 to 6 were prepared from 10 g. of frozen leaves. An ahquot portion of whole frozen leaves (No. 2) took up the expected quantity of "reversible "carbon dioxide (roughly the sum of those absorbed by fractions 3 and 4), but proved to contain considerably more "irreversible" carbon dioxide than did these two fractions together. Fresh leaves showed an even stronger deviation from additive behavior: the amount of "reversible" CO2 was only one-third of that of fractions 3 and 4, while that of "irre- versible" carbon dioxide was four times larger. Since carbonates yield equal quantities of "reversible" and "irreversible" carbon dioxide, while phosphates take up only "reversible" carbon dioxide, the combined action of these two agents should lead to the uptake of more "reversible" than "irre- versible" CO2 — while fresh leaves in table 8.XI show the reverse relation. This can only be explained by assuming the presence of carbonates in such a state or location that they are unable to take part in the absorption of gaseous carbon dioxide, but can be decomposed by acid. Smith suggested that the difference between Uving and frozen leaves can be ex- plained by the rapid carbon dioxide production by respiration in the former ones — a 194 FIXATION OF CARBON DIOXIDE CHAP. 8 production which partially saturates the buffers during the short time between prehmi- nary evacuation and the admission of carbon dioxide. This may explain the smaller CO2 uptake after the admission of this gas, but not the increased quantity of "irrevers- ible" carbon dioxide found in living leaves. Furthermore, this explanation imphes that respiration builds up a large internal pressure of carbon dioxide before the latter escapes into the atmosphere — which is improbable (compare Vol. II, Chapter 33). Despite these difficulties of quantitative interpretation, which show the desirabihty of continued experimentation, Smith's general conclusion that leaves contain two main carbon dioxide-absorbing factors — solid carbonates, and a water soluble buffer — appears plausible. The behavior of the aqueous fraction in particular can be quantita- tively accounted for by the action of a phosphate buffer, as shown by measurements of the CO2 uptake by this fraction under varying partial pressures of carbon dioxide {cf. Table 8. XII and Fig. 19). Table 8.XII Carbon Dioxide Absorption by Water Extract of Sunflower Leaves PCOS- atm. [HCO3-], moleA- (excess CO 2) [HCO3-], moleA- calcd. from pH pH Corresponding CO 2 absorption by 10 g. fresh leaves, ml. 0.05 0.20 0.75 0.99 0.0088 0.0133 0.0180 0.0187 0.0088 0.0138 0.0186 0.0197 6.83 6.42 5.98 5.89 1.6 2.4 3.3 3.6 A comparison of pH values in table 8. XII with those in table 8. II confirms the presence of buffers in the sap. The second column in table 8.XII shows that if the bicarbonate concentration is calculated from the carbon dioxide absorption, by assuming that all absorption in excess of solubility is due to bicarbonate formation, the result is equal to that derived from acidity. This is taken by Smith as a proof that carbamina- tion plays no part in the carbon dioxide absorption by the water-soluble leaf fraction. In figure 19, the experimental absorption values are com- pared with the calculated absorption by the phosphate buffer alone; the comparison shows that the phosphate can account for most, but not quite all, the bicarbonate formation in the extract. The remaining discrep- ancy indicates the presence of some minor buffering components. The assumption that the carbon dioxide absorbing capacity of the insoluble leaf fraction is caused by the presence of alkaline earth car- bonates is supported not only by the solubiHty of the absorbing agent in carbon dioxide-saturated water, but also by the analysis of the ash. It shows the presence of 6.7 X 10"" gram atom of calcium and 1.8 X IQ-" gram atom of magnesium in 10 grams of fresh leaves. In the form of ROLE OF BICARBONATE IONS IN PHOTOSYNTHESIS 195 carbonates, these cations could account for the absorption of 8.5 X 10~* mole, or 19 ml. carbon dioxide. This is a little more than the observed effect; but not all alkahne earths need to be present as carbonates. Insoluble phosphates, as well as manganese (found in the ash), also can 0.020 t^ L 1 0.015 ^ D ^ / o \/^ ♦- 7 c J •> / 3 c y o / u / a, 0.0 10 / , .^ /^ ^rfl c f o ' .o CD 005 0.2 0.4 0.6 0.8 Carbon dioxide pressure, atmospheres 1.0 Fig. 19. — The bicarbonate-ion concentration determined from e.m.f. measurements (A) compared with the total combined carbon dioxide obtained by gas-analytical methods (o) and that calculated from the buffer action of the phosphates (D) in the sunflower-leaf sap (after J. H. C. Smith 1940). contribute to the absorption of carbon dioxide by the water-insoluble fraction. 3. Role of Bicarbonate Ions in Photosynthesis The preceding section showed that plant cells (at least those of the higher plants, since no data are available on algae) usually contain, in equilibrium with the atmosphere, considerably more bicarbonate ions than carbon dioxide molecules. The role of these ions in photosynthesis has been much discussed in the literature, but most arguments used in this discussion are now obsolete; they were based on the effect of the presence of bicarbonate ions in the environment on the photosynthesis of aquatic plants. When Draper (1844) discovered that plants can live in bicarbonate solutions without a carbon dioxide supply, he concluded that bicarbonate ions can be used as such in photosynthesis. Later, it was reaUzed that all bicarbonate solutions contain carbon dioxide molecules; but it was thought that quantitative determinations of the rate of photosynthesis in relation to the concentrations, [CO2] and [HCOj"], can reveal whether the bicarbonate ions participate directly in photosynthesis or not. Natanson (1907, 1910) postulated that CO2 molecules are the only form in which carbonic acid is utilized in photosynthesis, while Angelstein (1911), who had observed 196 FIXATION OF CARBON DIOXIDE CHAP. 8 that at a constant value of [CO2], the rate of photosynthesis is improved by the addition of bicarbonate, beheved in the availabiUty of the latter for the photosynthetic process. Wilmott (1921) found, on the other hand, that the rate of oxygen production by Elodea is the same in acid carbon dioxide solutions and in alkaline bicarbonate solutions with the same concentration of CO2 molecules. Romell (1927) attributed Angelstein's results to the capacity of bicarbonates to renew the supply of carbon dioxide (cf. page 177), an interpretation confirmed by James (1928), who noticed that the improve- ment of photosynthesis, caused by the addition of bicarbonate ions, disappeared with an increase in the rate of circulation of the medium — thus indicating that it was predi- cated upon a local exhaustion of carbon dioxide. Because of the permeability of cell membranes to carbon dioxide molecules, changes in the external concentration of this molecular species produce shifts in the carbonate concentration inside the cell (accom- panied by changes in the acidity of the aqueous cell phases). Whether variations in the external concentration of bicarbonate ions, for which the cell membrane is almost impermeable, also affect the composition of the carbonic acid system in the cell, is a complicated problem of mem- brane equilibrium, and as long as we do not know the answer, experi- ments with varying concentrations of carbonate ions in the external medium do not tell us anything definite about the part which these ions play inside the cell. One thing, however, can be stated with certainty. If the presence of carbonate anions in the medium does have an influence on the composi- tion of the carbonic acid system within the cell (in the equilibrium state or in the steady state of illumination), this influence is at least several orders of magnitude smaller than that of free carbon dioxide molecules. Warburg's buffers contain tens of thousands of HCOs" and CO3 ions for each CO2 molecule (cf. Table 8.V). Nevertheless, the curves show- ing the yield of photosynthesis in relation to the concentration of the species CO2 in these mixtures have approximately the same shape as those obtained in experiments with land plants supplied with free CO2 molecules only. For instance, according to chapter 27 (Vol. II), the sat- uration of the photosynthetic apparatus of Chlorella occurs, in carbonate- bicarbonate buffers, at [CO2] = 5 X IQ-^ mole/1., with 7.5 X IQ-^ mole/1. HCOs" and 2.5 X 10"^ mole/1. CO3 ions also present in solution; while the photosynthetic apparatus of wheat is saturated when the concentration of carbon dioxide is of the order of 2 to 4 X 10~^ mole/1. In other words, the presence of an enormous excess of HCOs" and CO3 ions does not essentially affect the carbon dioxide saturation, which remains determined, in the first approximation, by the concentration of the carbon dioxide molecules alone. The assertion that carbonate ions cannot penetrate into the cells as rapidly as do the carbon dioxide molecules is based not only on the gen- eral experience that ions, with their clusters of water molecules, are much ROLE OF BICARBONATE IONS IN PHOTOSYNTHESIS 197 less capable of penetrating through cell membranes than neutral, par- ticularly lipophilic molecules, but also on direct experiments of Osterhout and Dorcas (1926), who found that the rate of penetration of carbonic acid into the interior of the unicellular alga, Valonia, is proportional to the external concentration of carbon dioxide and unaffected by the addi- tion of a large quantity of carbonate and bicarbonate ions. At first sight, certain results of Arens (1930, 1933, 19361-2) seem to contradict the conclusions of Osterhout and Dorcas. He investigated the well-known fact that aquatic plants, Elodea or Potamogeton, for instance, while carrying out photosynthesis in natural waters, often become covered by a precipitate of calcium carbonate; at the same time, the water in the neighborhood of the leaves becomes alkahne. Both observa- tions are easily explained by shifts in the equihbrium (8.9), caused by the elimination of carbon dioxide by photosynthesis. The interesting aspect of the phenomenon is that the deposition of calcium carbonate often takes place on the upper surface only. This would be natural if the consumption of HCO3" ions also took place only there. Arens found, however (by experiments in which leaves were used as membranes between two water-filled cells), that the bicarbonate— Ca(HC03)2 or KHCO3— is consumed on the lower surface of the leaf, while an equivalent quantity of Ca++ (or K+) ions emerges at the upper surface, accompanied either by CO3 ions (in the case of potassium bicarbonate), or by OH" ions (in the case of calcium bicarbonate). This directed transfer of ions through the leaves takes place only in light and thus appears to be related to photosynthesis. Although the results of Arens indicate a penetration of ions across the leaf, they do not necessarily clash with the conclusions of Osterhout and Dorcas. According to the latter, the flow of carbonic acid into the cells is maintained practically exclusively by the molecules of CO2, even when the medium contains a large excess of carbonate or bicarbonate ions. This makes it probable that the ions, HCO3- and CO3 — , cannot penetrate through the membranes at all. However, carbon- ate solutions contain a small proportion of undissociated salt molecules (KHCO3, K2CO3 etc.). It seems plausible that salt molecules can pass through the membranes as easily as acid molecules. If this is so, the results of Arens could be attributed to the penetration of the cell by these molecules, rather than by free ions. A salt, e. g. KHCO3, would enter the cell in the form of neutral molecules, dissociate there into ions, have a part or all of its HC03~ ions consumed by photosynthesis, and escape on the opposite side of the cell in the form of other neutral molecules, e. g., K2CO3 or KOH. Simultaneously with this comparatively slow flow of carbonates and bicarbonates across the cell, a much larger quantity of free carbon dioxide — unobserved in Arens technique — enters the cell (as shown by Osterhout and Dorcas), to be completely consumed there by photosynthesis. It must be added that the correctness of the results of Arens is not beyond doubt. Gessner (1937) found, for instance, by experiments with vaseline-covered leaves, that both surfaces of Elodea leaves are equally active in supplying carbon dioxide for photosynthesis. A complete interpretation of the transport of ions across the leaves must also take into consideration the possibility of diffusion through the cell walls without actual entrance into the membrane-shielded interior of the cells. It was repeatedly stated that the ratio [HCOs-J/ECOa] in the me- dium cannot be changed without simultaneous change in acidity. The occasionally observed depressing influence of carbonates on the rate 198 FIXATION OF CARBON DIOXIDE CHAP, 8 of photosynthesis at constant [002] may thus have been caused by the