Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
THE SECONDARY PROCESSES IN THE PHOTODECOMPO- SITION OF AMMONIA AND HYDRAZINEt HUGH S. TAYLOR Department of Chemistry, Princeton University, Princeton, New Jersey Received May 25, 1988 Since the diffuse bands of the ammonia spectrum were interpreted by Bonhoeffer and Farkas (1) as due to a dissociation of the molecule, the direction of further research on the photodissociatior~ process has been towards an elucidation of the secondary processes in such a manner that the low quantum yield (A 0.2) might be satisfactorily explained. For the primary process, it has been generally assumed that NH3 + he = NH2 + H (1) though the reaction, NH3 + he = NH + H2 (la) cannot be excluded on energetic grounds. It would, however, suggest an influence of molecular hydrogen which is not found experimentally in the photoreaction, although there is an effect in the reaction initiated by a-par- ticles and ions, in which case, at higher pressures, NH+ is certainly formed by secondary processes. The first reaction accounts for the presence of atomic hydrogen, made certain by the measurements of Geib and Harteck (6) and of Farkas and Harteck (4) on pare-hydrogen conversion in am- monia undergoing decomposition. Gedye and Rideal (5), using a streaming system, obtained yields of hydrazine from the photodecomposition of ammonia of as high as 57 per cent of the stoichiometric yield by the equation 2NH3 > N2H4 + H2 The hydrazine yield decreased rapidly as the temperature of the system was increased. Koenig and Brings (7), and more recently Welge and Beckman (16), have obtained positive confirmation of the formation of hydrazine by photolysis of ammonia in a static system. The amounts obtained, however, were very small. Indirect evidence of the production of amine radicals may be obtained from the photodecomposition of solutions of alkali metals in liquid am- ~ Contribution No. 8 to the Third Report of the Committee on Photochemistry, National Research Council. 85
86 HUGH S. TAYLOR monia (13~. Liquid ammonia does not photodecompose when pure, but does decompose on the addition of alkali metals. The sole products are hydrogen and amide ion, NH2-, in equivalent amounts, and the quantum yield is well under unity. Alkali metals virtually completely dissociate in liquid ammonia into positive metal ions and free electrons which become strongly solvated with the ammonia. It would appear, therefore, that in pure liquid ammonia deactivation or primary recombination effectively retards the decomposition, while in the presence of solvated electrons the NH2 radicals are stabilized by the formation of amide ion, thus permitting the decomposition to proceed. The mechanism, according to this hypoth- esis, would be NHe + he ~ NH2 + H NH' + H + M > NH3 + M (primary recombination) NH2 + ~~ > NH2 H+H+M >H2+M The majority of workers have accounted for the low quantum yield in ammonia decomposition by a recombination process, NH2 + H = NH3 (3) though Farkas and Harteck (4) pointed out certain difficulties in accepting this, due to the low stationary state concentration of atomic hydrogen, especially if reaction 3 requires a third body. That ammonia is regener- ated during the secondary processes was made definite by Taylor and Jungers (15), who showed the formation of deuteroammonias in mixtures of decomposing NH3 and deuterium atoms produced from deuterium mole- cules by excited mercury. No exchange occurred unless the ammonia was undergoing photodecomposition. By investigating the photochemical decomposition in the region of very small decompositions Welge and Beck- mann (16) found that the quantum yield measured by hydrogen produced approached unity; hence they concluded that the recombination reaction (reaction 3) does not play an important role. The products in such cir- cumstances would be substantially hydrogen and hydrazine. They further assumed the secondary processes to be substantially heterogeneous, with a little of the hydrazine reacting with the atomic hydrogen to give am- monia, as shown by Dixon (2~. With larger amounts of products this eject would be exaggerated and the non-condensable products would approach the stoichiometric ratio due to hydrazine decomposition. In experiments of Elgin and Taylor (3) it was shown that the total process in both the photochemical and photosensitized decomposition of hydrazine can be with fair accuracy expressed by the stoichiometric equation 2N2H4 = 2NH3 + N2 + H2 (4)
PHOTODECOMPOSITION OF ~ON! ~ ED HYDRAZINE 87 With hYdrazine in the presence of ammonia, even when only a small frac- tion of the light was absorbed by the hydrazine, it was shown, by Ogg, Leighton, and Bergstrom (14), that hydrazine was the compound which disappeared, with a quantum yield of 1.28 on total quanta absorbed. Such facts provide considerable support for their conclusion, recently restated by Mund and van Tiggelen (12), that the process which regen- erates ammonia and causes the low quantum yield is not reaction 3 but the reaction N2H~ + H = NH3 + NH2 (5) a reaction which would also explain the work of Taylor and Jungers with deuterium atoms. The nitrogen which is formed ultimately in the photodecomposition of ammonia must arise' therefore, from further decomposition of the hydra- zine or NH2 radicals or both. The assumption of Ogg, Leighton, and Bergstrom that this occurs by the reaction N2H4 + 2NH2 = 2NH3 + N2 + H2 (6) has been modified by Mund and van Tiggelen in a sequence of secondary processes which more satisfactorily than any alternatives, it would appear, accounts for the kinetics of the reaction and its quantum yield. The sequence IS: 20(NH3 + by = NH2 + H) 17(2NH2 + M = N2H4) 16(N2H~ + H = NH3 + NH2) 1(N2H4 + 2NH2 = 2N2 + 4H2) 2(H + H + M = H2) 20h~ + 4NH3 = 2N2 + 6H2 (1) (2) (5) (6a) (7) This sequence is based on a quantum yield of 0.2 at 1 atm. pressure, and the numerical coefficients by which the equations are multiplied must evidently be proportional to the velocities of the corresponding reactions. It leads to the following expression of Mund and van Tiggelen for the quantum yield ~ = 1 0.875 + ~/0.0156 + K~/v where v (C~I~.bs.) is the velocity of reaction expressed as the number of mole- cules of ammonia decomposed to nitrogen and hydrogen per cubic centi- meter per second, P is the total pressure, and ~ the quantum yield equal to Q/v, where Q is the number of ammonia molecules per cubic centimeter per second primarily decomposed.
88 HUGH S. TAYLOR This equation reproduces the tendency of ~ to vary with the pressure. With ~ = 0.2 at 1 atm. one calculates a value of ~ = 0.314 at 0.1 atm. and 0.126 at 8.5 atm. The former is in agreement with the data of Wiig (18) and the latter with the measurements of Ogg, Leighton, and Bergstrom. At low pressures the influence of the walls begins to predominate and will assist especially reactions 2 and 7. Its effect on the quantum yield will be equivalent to that of high pressures in the gas phase, i.e., will lower the quantum yield as Wiig observed. The influence of the intensity or the velocity of decomposition, v, on the quantum yield is of the same order of magnitude as the equation suggests, as elucidated in a recent research by Mund, Brenard, and Kaertkemeyer (11), and as suggested by the work of Ogg, Leighton, and Bergstrom, by Wiig and Kistiakowsky (20), and by more recent measurements of Wiig (199. The Mund-van Tiggelen nu- merical values for the relative rates of the several reactions do not consti- tute a unique solution, as is evident from an analysis by Leighton (89. :For the ratios 20:17:16:1:2 of the above sequence, Leighton substitutes the sequence lO(NH3 + ho = NH2 + H) 7~2NH2 + M = N2H4 + M) 6(H + N2H4 = NH3 + NH2) 2~2NH2 + N2H4 = 2NH3 + No + H2) 2(H + H + M = H2 + M) It will be noted that in this sequence the nitrogen-producing mechanism simultaneously regenerates ammonia, whereas the Mund-van Tiggelen reaction does not. Both yield similar expressions for the relation between quantum yield, pressure, and intensity. Leighton believes his reaction producing nitrogen to be superior, because it accounts for the 1:1 N2:H2 ratio observed in the ammonia-sensitized decomposition of hydrazine (see later). It has been known since Warburg's original researches on this reaction that neither molecular nitrogen nor hydrogen influenced the quantum yield. The above mechanism is in accord with such inertness except in so far as it might influence the total pressure, P; this influence is small, as ;seen. On the other hand, atomic hydrogen was shown by Melville (9) to inhibit strongly the photodecomposition of ammonia and the same is true of the photosensitized decomposition (10~. This effect, which Mel- ville attributed to the recombination reaction (reaction 3), is rather to be ascribed to the ammonia regenerative reaction with hydrazine (reaction 5), the velocity of which, as is to be seen from the magnitude of the molecula~- coefficient, 16, is very high. Analysis shows that in addition to the effect
PHOTODECOMPOSITION OF AMMONIA ED HYDRAZINE g9 of atomic hydrogen concentration there is also an effect due to the diminu- tion in v, the total decomposition. The cumulative effect of these two factors is evident in the data both of Mitchell and Dickinson and of Melville. Parkas and Harteck found a sharp maximum in the atomic hydrogen concentration from the photodecomposition of ammonia in the presence of hydrogen when the ammonia was only 10 mm. in a total pressure of 30 to 70 cm. This led these authors to the assumption of NH4 radicals in equilibrium with NH3 + H. The maximum should bear some relation to the maximum observed in the quantum yield but this latter, though similar in form, is displaced to higher ammonia pressures. This discrepancy is not yet elucidated. The objection that might be raised to the Mund-van Tiggelen develop- ment, namely, that nitrogen only results from the termolecular process (from 2NH2 + N2H4, reaction 6) and not from such a process as suggested, among others, by Wiig, NH2 + NH2 = N2 + 2H2 (8) is discussed in the original communication of Mund and van Tiggelen and more recently by them in a comment on the newer work of Wiig. They point out that reaction 8 always leads to kinetic expressions from which the observed increase of quantum yield with velocity of decomposition cannot be deduced, and, further, it fails to account for the small variation of yield between 1 and 8.5 atm. found by Ogg, Leighton, and Bergstrom. One might expect that other molecules than NH could act each with a particular efficiency in the recombination-decomposition process (reaction 6) of the Mund-va~ Tiggelen scheme. To introduce such possibilities into the kinetic scheme would further complicate the equation derived' and the experimental data at present are not accurate enough to justify, such further refinements. In the ammonia-sensitized hydrazine decomposition experiments of Ogg, Leighton, and Bergstrom, the two initial stages are quite clear, namely' reactions 1 and 5 of the ammonia scheme above. NH3 + hi = NH2 + H H + N2H4 = NH3 + NH2 The Mund-van Tiggelen mechanism would then give N2H4 + 2NH2 = 2N2 + 4H2 (1) (5) (6a) which would mean a quantum yield of 2 at a maximum diminished by the recombination process, 2NH2 + M = 2N2H4 (2) /
90 HUGH S. TAYLOR The product gas should be 33 per cent nitrogen and 66 per cent hydrogen. The Ogg, Leighton, and Bergstrom mechanism would give N2H4 + 2NH2 = 2NH3 + N2 + H2 (6) again a maximum quantum yield of 2 on the hydrazine Hi.~.nn~.rin~ hilt. a 1:1 ratio of nitrogen and hydrogen. This ratio was actually found by these authors and also by Elgin and Taylor. Wenner and Beckmann (17) found quantum yields ranging from 1 at low pressures to 1.7 at higher pressures and hydrogen concentrations ranging from 58 to 64 per cent in the same pressure range of 2 to 14 mm., decreasing toward 50 per cent with · . ncreasmg pressure. The primary mechanism in hydrazine decomposition is still uncertain. If it be ___w=,< ~ ~ v N2H4 + he = 2NH2 then it would be necessary to fall back on reaction 6 to explain the ammonia formation observed. On the other hand, a primary mechanism NH + he = N2H3 + H could by reaction 5 regenerate one-half the observed ammonia with t.he atomic hydrogen. The NH2 and the N2H3 thus produced could, by several ~ 1 1 ~ 1 _ 1 · ~ 1 ~ ~ . ~ ~ · ~ . . . alternative mechanisms already proposed, but for which there is as yet no experimental test, yield NH3 + N2 + H2 in approximate accord with the observations. The pare-hydrogen conversion reaction might be employed to test these two alternative mechanisms for the primary photoprocess. REFERENCES (l) BONHOEFFER AND FAREAS: z. physik. Chem. 136, 337 (1928). (2) DIXON: J. Am. Chem. soc. 64, 4262 (1932). (3) ELGIN AND TAYLOR: J. Am. Chem. soc. 61, 2059 (1929). (4) AREAS AND HARTECE: z. physik. Chem. B26, 257 (1934). (5) GEDYE AND RIDEAL: J. Chem. soc. 1932, 1160. (6) GEIB AND HARTECE: z. physik. Chem., Bodenstein Festband, p. 849 (1931). (7) KOENIG AND BRINGS: z. physik. Chem., Bodenstein Festband, p. 595 (1931). (8) LEIGHTON: Actualites Scientifiques. Hermann et Pie, Paris (1938). In press. (9) MELVILLE: Trans. Faraday soc. 28, 885 (1932); Proc. Roy. Soc. (London) A162, 323 (1935). (10) MITCHELL AND DICKINSON: J. Am. Chem. soc. 49, 1478 (1927). (ll) MIJND, BRENARD, AND KAERTEEMEYER: Bull. sac. chim. Belg. 46, 211 (1937). (12) MEND AND VAN TIGGELEN: Bull. soc. chim. Belg. 46, 104 (1937). (13) OGG, LEIGHTON, AND BERGSTROM: J. Am. Chem. soc. 66, 1754 (1933). (14) OGG, LEIGHTON, AND BERGSTROM: J. Am. Chem. soc. 66,318 (1934). (15) TAYLOR AND JUNGERS: J. Chem. Phys. a, 373,452 (1934~. (16) WELGE AND BECRMANN: J. Am. Chem. soc. 68, 2462 (1936). (17) WENNER AND BECEMANN: J. Am. Chem. soc. 54,2787 (1932). (18) WI1G: J. Am. Chem. soc. 67, 1559 (1935~. (19) WI]G: J. Am. Chem. soc. 69,827 (1937). (20) WI1G AND KISTIAEOWSEY: J. Am. Chem. soc. 64, 1806 (19321.
