The Photochemical Decomposition of Hydrazine Citation Wenner, Ralph Richter (1931) The Photochemical Decomposition of Hydrazine. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/hjtw-a149. https://resolver.caltech.edu/CaltechTHESIS:09062024-171104247 Abstract No abstract. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Chemistry) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Beckman, Arnold O. (advisor) Dickinson, Roscoe Gilkey (advisor) Thesis Committee: Unknown, Unknown Defense Date: 1 January 1931 Funders: Funding Agency Grant Number Carnegie Institution of Washington UNSPECIFIED Record Number: CaltechTHESIS:09062024-171104247 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:09062024-171104247 DOI: 10.7907/hjtw-a149 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 16697 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 06 Sep 2024 21:57 Last Modified: 06 Sep 2024 22:00 Thesis Files PDF - Final Version See Usage Policy. 10MB Repository Staff Only: item control page

THE PHOTOCHEMICAL DECOMPOSITION OF HYDRAZINE Thesis by Ralph hichter Wenner In partial fulfillment of the requirements for the degree of Doctor of Philosophy California Institute of Technology Pasadena, California 1930 TABLE OF CONTEN TS Page Objectives of the research l Acknowl edgement 1 Introduction 2 Apparatus for investigating the products of the dec oraposition 3 The products of photoche~ical decompo sition 10 Apparatus for determining quantum yields 13 ~ua.ntum yie lds 21 Mecha..~ism of the decompositi on 33 Biblio.sraphy 37 THE PHOTOCHE~'-1: ICAL DECOMPOSI-rI ON OF' HYDRAZINE OBJ 2CTIVES OF THE RESEA."ftCH The research to be described in this thesis was undertaken in an effort to improve the technique of photochemical investigations in the far ultra.violet a.s already carried out by Beckman and Dickinson. The chief improvement in the method consists of the use of a. system for the continuous recording of the energy incident upon and absorbed by the reaction mixture. From the chemical standpoint, the problem is of special interest since hydrazine is the only hydride of nitrogen that has not been subjected to a quantitative quantum yield investigation. ACKNOWLEDGEMENT This research was carried out under the supervision of Professor A.O.Beckman. The author wishes to express his sincere ap- preciation for the advice and personal assistance so lavishly bestowed by the latter during the course of the investigation. Thanks are due Professor R.G.Dickinson for his valuable advice. The expense of the investigation was defrayed by a grant from the Carnegie Institution of Washington to Professor A.A.Noyes. 2. INTRODUCTION The photochemical decomposition of hydrazine has been 1 studied by Elgin and Taylor , who did not however attempt to carry out an absolute determination of the quantum yield. Indeed, in spite of the importance of the nitrogenhydrogen compou~ds, perhaps the only unquestionable determination of a quantum yield has been ma.de in the decomposition of gaseous hydrogen azide by Beclonan and Dickinson 2 • quantum was found. A yield of 3.0 ~ .5 molecules per The quantum efficiency in the decomposition of am- monia has been studied independently by Warburg 3 who found 0.25 molecules per quantum absorbed and by Kuhn4 , who found 0.4 - 0.5 molecules per quantu.~. Owing to the uncertainties in both determinations the solution of the problem is still far from satisfactory. The experiments of Elgin and Taylor were perfonned in a quartz cell using the full radiation from a quartz mercury vapor la~p. They concluded that ammonia, nitrogen and hydrogen were formed and that hydrazine decomposes approximately ten times as fast as ammonia under the srune conditions. An actinometric determination of the quantum yield in the mercury sensitized reaction gave thirteen molecules of hydrazine decomposed per quantum of~ 2537 absorbed. The method of investigation to be described in the following pages depends upon the actual measurement of the quanta absorbed during the course of the reaction a.nd the calculation of the molecules decomposed from the analysis of the decomposition products. 3. APPARATUS FOR INVESTIGATING THE PRODUCTS OF DECOMPOSITION The apparatus originally constructed for the investigation and measurement of the decomposition products was designed to measure both the moles of nitrogen plus hydrogen and the moles of ammonia fanned. The microanalytical sy stem so constructed operated in the following manner: Immediately after irradiation of the hydrazine, liquid air was applied to the cell trap in order to condense out everything but the nitrogen and hydrogen. These gases were then passed in- to a modified McLeod gage by means of a Toepler pump and their PV product obtained. The liquid air was then replaced by ether-carbon dioxide snow mixture, which according to Elgin and Taylor permits all the ammoni a to pass into the gas phase and retains the hydrazine. The ammonia was then Toeplerized as before and the number of moles measured in the McLeod gage. From these two readings it is possible by simple stoichiometry to calculate the moles of hydrazine decomposed. An attempt was made to check the analysis by analyzing the fixed gases as well as the alleged ammonia. To do this, an electrically heated platinum filament was sealed into the McLeod gage so that after the moles of nitrogen plus hydrogen had been measured, pure oxygen was introduced into the gage and the hydrogen burnt to water. After freez- ing out the water on a side trap, the contraction in volume was measured and the mole fraction of hydrogen calculated. It was a lso thought pos- sible to prove that the gas non-condensable on ether-carbon dioxide was ammonia by decomposing it thermally on the hot filament, measuring the increase in volume and burning the hydrogen so produced. 4. After considerable testing, the method had to be abandoned for the following reasons: (1) Ammonia can not be accurately measured in glass apparatus because of the "clean-up" action of' ammonia on the walls of the app aratus. (2) Oxygen combines rapidly with mercury in the presence of a red hot filament at lov1 pressures. (3) The thermal decomposition of ammonia on platinum, even if complete at these temperatures and pressures, is uncertain because of "clean-up" and outgassing effects of hydrogen on the hot filament. Another complication discovered during the course of these experiments was the strong adsorbing action of hydrazine on glass, which made it difficult to be removed by pumping and resulted in the liberation of non-condensable gases from the glass walls. The analytical system finally adopted was that used by Beckman and Dickinson in their studies on hydrogen azide and is shown in Figure 1. R TO CELL CUT-OFF TO PUMPS A C D E 6 -. A series of "seal-off" bulbs w:en a.lso a.tta.ched to the line during the preliminary experiments in order to ma.ke the qualitative tests for ammonia.. The method of analysis depends essentially upon the mea.sure- ment of the f~ of the nitrogen plus hydrogen by means of the quartz fiber gage G a.nd their total pressure by means of the McLeod gage. The use of the quartz fiber ga.ge as a means of analyzing nitrogen-hydrogen mixtures has been fully tested by Beckrnan and Dickinson 5 and the construc6 tion of an efficient form of this device has been described by Beckman. Additional accuracy in the use of the ga.ge has been secured by a special mounting. The ga.ge was held ri gidly in a hole bored in a solid brass ulock, which in turn wa.s clamped rigidly to an upright 2x3 joist. The latter was atta ched to the vacuum bench at the top and bottom through several la.yers of ha.ir felt and a loose fitting lag bolt, which served very efficiently to damp out vibrations transmitted to the gage by means of the bench structure. The use of the quartz fiber gage depends on the fact that the reciprocal of the time for a vibration of the fiber to die down to a certain fraction of its original value is a linear function of the summation of the IfM 's for the gases present. Thus, if 'X. represents the time required for a vibration to die down between two -definite marks on the observing telescope sea.le, then it is a well known fact that the following relation exists: I t:~ ~ 1: V(~l \-- \:> The gage was calibrated with air at known pressures and from the calibration plot so obtained subsequent 11 ha.lf-lives" were transla.ted 7. into the corresponding L \\.~"'~ of the gas present. Since these gases could only consist of nitrogen plus hydrogen, the composition ~-~o - The source of radiant energy used was that set up by Beckman and Dickinson for their studies on hydrogen azide . It consisted of a condensed aluminum spark energized by 13,000 volts from the secondary of a transformer whose primary operated at 220 volts and 40 - 60 amperes. The spark was attached directly to the collimator of a Bausch and Lomb quartz monochromator, as shown in Figure 2. 9. To prevent melting of the electrodes and overheating of the monochromator parts, a stream of compressed air was blown thru the spark in order to direct it away from the monochromator. periments were ~arried out using the radiation of The quantum yield ex- l\ 1990 exclusively, but the qualitative experiments for the detection of ammonia required the full radiation i 'C)om an aluminum sp ark placed 20 cm. in front of the cell. In the preliminary experL~ents, hydrazine was introduced from the reservoir R (Figure 1) thru the doser D' into the cell, care being taken to have cut-offs A and D closed and the mercury in the McLeod gage as hi gh as possible. The hydrazine was then condensed in trap B by means of ether-carbon dioxide mixture, cut-off A lowered and am..monia and then other non-condensaole gases pwnped off. This procedure was repeated three times in order to insure complete removal of gases other than hydrazine. F'inally, the hydrazine was allo·111ed to expand into the cell, the cell cut-off raised and the sample illumina ted for a definite length of t ime. After irradiation liquid air was applied to the cell trap, all cut-offs were set on definite marks, and the '[_r~ and the total pressure of non-condens a ble gases measured. The volwne of the sy stem was obtained by trapping a small amount of air in the McLeod gage, measuring; its pressure, expanding it into the rest of the evacuated sy stem and again measuring its pr essure. 10. THE PRODUCTS OF PHOTOCHEMICAL DECOMPOSITION Preparation of Anhydrous Hydrazine.- Anhydrous hydrazine was prepared from the best obtainable Eastman's hydr a zine hydrate, by a modification of the method of Hale and Shetterly 7 • The original procedure called for the use of barium monoxide as a dehydrating agent but unfortunately this substance yielded a product lower in hydrazine content than the original hydrate. A test with standard potassium permanganate solu- tion showed a negligible amount of peroxide present in the oxide. After a considerable period of experimentation with barium oxide, it was decided to abandon t his substance and distill the hydrazine from fused potassium hydroxide. Three successive distillations from the latter substance at reduced pressure in a current of hydrogen, yielded a product which when analyzed according to the method of Bray and Cuy 8 , showed 98.6% hydrazine. This sa..rnple was distilled as Defore directly into a "break-off-easy" bulb containing some pieces of potassium hydroxide, without exposure to the air. This in turn was broken and distilled in vacuo into several smaller "break-off" bulbs by ap plication of liquid air to the proper bulb. After five distillations, the product was assumed to be sufficiently pure without any further analytical tests oeing made on it. Preliminary Decompositions.- Previous experiments had shown that hydrazine at room temperat ures undergoes a reaction on glass or quartz resulting in the liberation of gases not condensable on liquid air. Since hydrazine has long been knovm to attack glass when hot, it was assumed that this action was partly due to a purely thermal reaction and partly to adsorption resulting in the displacement of gases not removed from the walls by previous baking of the sy stem. The plausibility of these ll. asswnptions was strengthened by the observation that the outgassing from the cell after frequent exposures to hydrazine for a week, was reduced to the small, practically constant amount of 5 x 10- 5 mm. in 27 minutes. Some attempts to determine the ?JM and hence the composi- tion of this gas s howed it to consist largely of hydrogen. Since .this determination involved readings of '.ton the fiber gage of the order of Ci., magnitude of 500-600 seconds, it was not considered~very important determination. The method of correcting the data. for outga.ssing effects will be resumed under the discussion of the quantum y ield experiments. Preliminary decompositions of hydrazine by l\ 1990 at pressures from 10 - 13 mm. gave non-condensaole gases with mol fractions of hydrogen around 0.56, which were not considered entirely correct because of the l arge uncertain outgassing correction. Tv10 experiments were made in which the hydrazine at lOP mm. was exposed to the full aluminum sp ark for three minutes, in order to produce sufficient ammonia to give a qualitative test. This test was ma.de by applying ether-carbon dioxide to the cell trap afte r exposure and condensing the ammonia in a seal-off bulb by means of liquid air. This bulb was then broken under distilled water and the cont ents emp tied into Nessler's reagent. A reddish-brovm colora"l\\E tion indic a ted the pres ence of ammonia. No coloration was obtained when /\ bulbs were sealed off as before with liquid air on the cell trap to retain the ammonia. Hydrazine produce& a grey deposit of mercury with Nessler 1 s reagent. The presence of runmonia and a mol fraction of hydrogen around 0.56 seemed t o i ndicate that the decompo sition proceeds a ccording to the reactions, 12. Reaction (a) above would give XH = 0.50 and reaction (b) would give 2 XH 2 = 0.66, whereas the simultaneous reactions would give mol fractions between o.50 and 0.66. 13 . APP ARATUS FOR DETERiHNI NG QUANTUM YI ELDS Fi gure 3 i s a schematic r epre s entati on of t he appa ratus f or determining quantum y i e lds. M ~ ~- <;}L2 L1 ~ C TO SYSTEM I TO RJMPS E I 14. The analytical sy stem wa. s the same as that already described. It was joined thru a cut-off and a quartz pyrex graded seal to the cell C, the whole system being designed to keep the total volume as small a.s possible. The monochromator, cell a.nd thermocouples were mounted on a table attached to the vacuum bench behind the analytical apparatus. The mono- chromator and spark set-up have already been described. ,, \\. The cell was built-up of pieces of polished 1fused quartz to the shape shown in the diagram. It had a volume of approximately 25 cc. and a The front window was built at an angle of 45° mean length of one inch. to the longitudinal axis of the cell in order to reflect part of the incident radiation on to the thermocouple T,. Light :iimerging from the monochromator was brought to a focus on the thermocouples by means of the two 20-diopter, cylindrical, quartz lenses L, a nd 1 2 • L, with axis verti- cal converges the light in a vertical plane and 1 2 , with axis horizontal, converges it horizontally, the result being a small rectangular image focussed on T2 • The light reflected onto T, was less sh arply def ined be- cause t he images reflected from the fr ont and rear surfaces of t he cell window did not quite coincide. Considerabl e coincidence was gai ne d by properly grinding the front surface of the window. The two thermocouples were connected to two Leeds and Northrup "High Sensitivity" galvanometers which were set up in conjunction with the a utomatic recorder s hown in Figure 4. 15. FIGURE 4 16. The recorder consisted of a carriage which was made to descend at a uniform rate of speed on a.n accurately cut screw d't':tven by a. Warren Telechron motor. The images from the two galvanometers were brought to a. point focus on photographic paperlmounted on the moving carria.ge\by means of a horizontally placed cylindrical lense. A 200-wa.tt projection la.mp mounted in an air-cooled ·box fitted with a. small slit served as light source for the galvanometers. The latter were placed in a. carefully shielded wooden oox a.nd arranged so as to be free from external vibrations. Complete freedom from electrical disturbances was attained by shielding all electrical parts and connections. The ga.lvs.nometers ·were placed at a distance of one meter from the recording paper and the scale tested for linea.ri ty. 'fhe sea.le was impressed directly on the photographic paper before developing by exposing the paper for 30 seconds in a. RIZEAU to an overhead carbon lamp. This device consists of a.n ordinary photographic printing frame, the window being a silvered glass accurately ruled in 1/20-inch squares. The whole recording system was contained in a. blackened) light-tight compartment underneath one of the laboratory benches. This type of reaction cell and method of energy measurement permitted the continuous recording of the radiation incident upon and absorbed during the course of the photochemical decomposition. Sensitive vacuum thermocouples of the Coblentz-Petit type were used for measuring the radiant energy. is shown in Figure 5. The construction of such a. couple 17. FIGUR E 5 ' ' ! I I I ' \, I I 18. The vanes used in these thermocouples were ma.de of thin aluminum foil, approximately 3.8 x 7 mm. in size. They were blackened with a coating of ordinary commercial dead-black lacquer. The junctions were ma.de of fi"fle bismuth-bismuth (10% tin) wires and were arranged six in series. Three junctions were attached to each vane by means of Du Pont "Household Cement". A piece of fluorescent uranium glass placed behind the vanes was a necessary aid in focussing the light. The region around the cell and thermocouples was protected from stray light and air currents by means of a cardboard shield and several layers of cotton batting. This was very helpful in preventing shifting of the thermocouple zero. Calibration of thermocouples~~-- The calibration of thermocouple T2 was performed according to Bureau of standards specifications, using two different standard lamps. A brass plate having a hole 0.100 inches in diameter permitted a known amount of the light flux to reach the thermocouple. Since the window of the couple was made of polished fused quartz, it was necessary to determine the transmission of this window or a similar one to the radiation from the standard lamp. This measurement was made by recording the deflections of the couple with and without the plate in the berun of light from the lamp. ferent pieces of quartz gave the following transmissions; Piece Transmission I 0.909 0.914 II 0.919 _0.918 Mean 0.914 Two dif- 19. In view of the fa.ct tha.t the maximum deviation of any determination from the mean was less than one percent, it was assumed that the transmission of any thin piece of polished fused quarts may be ta.ken to be 0.914. This was to be expected since the loss of inten sity is chiefly a reflection loss for these wave lengths. In order to translate thermocouple deflections obtained during the course of a run into a bso lute units it was necessary to measure the 1\ 1990 ( at normal incidence) of the rear window of the transmissions to cell and the window of T2 • This was done by measuring the deflection of T2 with and without the window in the beam of~ 1990 emerging from the monochromator. A piece of quartz exactly like the window of T2 was measured instead of the one on the thermocouple. In order to allow for changes in light intensity during t hese determinations, it was necessary to set up the two couples in the posi.tions shown in Figure 3 and an auxiliary piece of quartz placed at an angle in front of L2 for the reflection of light onto T,. The window under investigation was mounted on a metal bracket, a.s close to T2 as possible and could readily be swung in and out of the beam of light as desired. In making a transmission determina tion, the zero reading of the couple was obtained by running the spark and the recorder for two minutes with the vita glass screen in the monochromator closed. The screen was then opened for two minutes and the deflection with the window in the beam recorded. The zero was again ta.ken by closing the shutter for two minutes, during which interval the window was removed from in front of T2 • Again the shutter was opened for two minutes a.nd the full defJ.ection recorded, followi ng which another zero reading was taken. The ratio of the two deflections with the proper correction for 20. changes in light intensity during the exposures gave the desired transmissions. Window Transmission Thermocouple window 0.650 0.645 0.660 0.659 0.650 Mean --- 0.653 o. 724 0.704 o. 714 o. 729 o. 727 Cell window Mean 0.720 After sealing on the rear window of the cell and mounting it rigidly in place, several determina tions of the ratio of the deflection of the two thermocouples were made with the cell evacuated. This ratio had to be redetermined whenever the set up was disturbed for any reason. The ratio was obtained for intervals of time corresponding to the actual exposures during a quantum yield determination. 21. QUANTUM YIELDS By means of the apparatus just described it was possible to decompose hydrazine photochemically and to measure the quantum efficiency of the reaction. The method of introducing hydrazine into the cell has alr eady been described. Usually. the sample of hydr a zine which was to be irradiated was frozen out on the cell trap by means of ether-c a rbon TO dioxide mixture and left exposed to the pumps for t hree qua rters,~ one hour. During t his time the room was completely darkened, all ventilation shut off, and t he sy stem allowed to come to temperature equilibrium in order to eliminate shifting thermocouple zero. After standing, the cell cut -off was raised, the ether- carbon dioxide removed and the hydrazine allowed to expand into the cell. Suf fi cient time wa s allowed for the hydrazine to vaporize before starting the spark. With the vita glass shutter closed, the spark was run for one and a half mi nutes i n order to record the the rmo couple zero. The shutter was then opened to expose the hydrazine for a defini t e time and the zero again t aken at the conclusion of the exposure. The rec ording paper was im.~ediately exposed in the Rizeau and developed in the still darkened room. Vvhen dry• the deflections were cut out and the mean deflection for the run obtained by weighing. The pressure of the hydrazine wa s then read on the cell cut-off, li qui d air ap plied to the cell trap , and the total time fr om the removal of t he ethe r-c a r bon dioxide until the application of liquid air noted. After expansion of the non-condensables in- to the sy stem, the cell cut-off wa s raised until communica t ion between the cell a nd the rest of the s;y stem was just broken. 1'his last pr ecau tion was taken to avoid errors in the quartz fiber readings due t o the out gassing a ction of hydrazine retained in the cell. 22 •. Ovring to the liberation of non-condensable gases during the time the hydrazine was in contact with the cell walls, a correction had to be applied before the number of molecules decomposed could be calculated from the analysis. To do this, non-condensa.ble gases were pumped off the sample after reading the gages and the hydrazine was allowed to stand in the cell for the same length of time as the total duration of the experiment, without irradiation. Liquid air was then restored to the cell trap and the pressure of the non-condensable gases libera.ted)measured on the McLeod. This amount was subtracted from the previous pressure read- ing in order to obtain the pressure of nitrogen plus hydrogen produced by purely photochemical means. The mol fraction of hydrogen, however, was calculated on the ass~~ption that the composition of these nonphotochemical gases would not differ very much from the composition of hydrazine and since they were in small amount would not have a.n appreciable effect in altering the composition of the photochemically produced nitrogen plus hydrogen. This assumption was necessary inasmuch as quartz fiber gage readings a.round ten minutes were of doubtful significance, hence the composition of the liberated gases could not be determined accurately. From these data. the number of molecules of hydrazine decomposed may be calculated by simple stoichiometry with no a.ssUl:lptions as to the mechanism of the reaction. Let a= moles of ammonia. produced b = moles of nitrogen plus hydrogen produced x = mole fraction of hydrogen Since the nitrogen and hydrogen in all the products proceeds only from the decomposition of hydrazine, the total number of hydrogen atoms must 23. equal tv:ice the nitrogen atoms and the moles of hydrazine decomposed equal one half the total number of nitrogen atoms. 3 a + 2 b x a+ 2b(l-x) ( 1) 2 a= b(4 - 6x) 2b(l-x) ( 2) b was = 2 a 4b(3/4 - x) calculated from the perfect gas laws knowing the pressure, volume and temperature of the nitrogen plus hydrogen. The number of quunta absorbed were calculated in the following manner: Let DR= mean deflection of thermocouple 1 DT = mean deflection of thermocouple 2 R = ratio of transmitted to reflected deflections with the cell evacuated T2 .. transmission of window of couple 2 TC = transmission of rear window of cell k = thermocouple sensitivity (ergs per unit deflection) Ee= energy entering cell per second (ergs) E~ = energy leaving cell per second (ergs) t = time of exposure NQ. = total number of quanta of 1\ 1990 a bsorbed during run Ee = E = 24. Total energy abs orbed = (E e. Substituting T2 = 0.653 T0 = 0.720 k = 1.28 (3) NQ _ E )t = k(R DR __"."__D_Tl:_ \.. T2 ~c = 2.76 x loll (R DR_ - DT)t 25. In Table I are given the data from which the number of molecules of hydrazine decomposed in each run were calculated. The data for the calculation of the number of quanta absorbed are given in Table II. The quantum yields shown in Table II were obtained by dividing the molecules of hydrazine decomposed by the quanta absorbed. An inspection of Figure 6 illustrating the variation of the quantum yield with the pressure of hydrazine shows a falling off in the yield as the pressure of the hydrazine decreases. r'igure 7 shows an in- crease in the mole fraction of hydrogen of the non-condensable gases as the pressure of hydrazine decreases. With the exception of Runs 15, 18 and 23 the results show a quite consistent dependence on the pressure of the hydrazine. The results of Run 18 are more liable to error than the others, because of the low light absorption and consequent small quantity of products available for analysis. The utmost. acc uracy :i.s necessary in the determination of X, since the number of molecules of hydrazine as calculated by Equation 1 is extremely sensitive to errors in this quantity. 26. TABLE Run p N~,)-14 mm . Temp. Ao p N2 +H2 I p mmxl03 co r r ection mm X 10 5 m XH Molecules N2H4 3 13 300 4.15 6.0 0.0126 o.582 12.22 X 1015 4 13 300.5 3 .83 5.0 .01 64 .582 ll.27 5 8.0 299.5 4.06 5.0 . Oll75 .618 9.39 6 8.0 301 3.91 7.0 .0113 .621 8.78 7 11.0 29 8.5 3.99 5.0 .Oll9 • 598 10.72 8 10.0 298 3. 64 5.0 .01072 .608 9.13 9 2.0 299 2.85 6.0 .00808 . 634 5.77 10 3.0 299.7 3.47 6.0 .01002 .622 7.76 11 6.0 299 3. 63 6.0 .0106 .613 8 . 73 14 10.5 301.6 3.11 8.0 .00890 .628 6.75 15 6.0 301.5 3.04 8.0 .00852 .644 5.75 16 3.0 302.4 2.59 8.0 .0072 6 .644 4.85 17 9.8 299.8 2.99 7.0 .00870 .616 7.19 18 2.0 298.3 2.33 8.0 .00667 .629 5.03 19 7.0 298.8 3.22 6.0 .00922 .629 7.05 20 7.0 298.4 3.07 8.0 .00890 .619 7.21 21 12.0 298.3 2.97 5.0 .00867 .613 7.41 22 12.0 298.0 3 .00 5.0 .00879 .611 7.57 23 5.0 298.4 2.81 6.0 .00792 .639 5.62 24 5.0 298.3 2.91 6.0 .00827 .634 6.10 25 3.5 298.8 2.68 5.0 .00760 .634 5.61 26 14 298.2 3.06 5.0 .00924 .587 9.05 NOTE: The volume of the system for huns 141. 9 cc. for the remaining runs. '2: V - 11 was 137.2 cc. and 27. TAB LE Run Time exp. ~ DT II R Quanta entering X lo- 15 Quanta leavin§ x 10- 1 Quanta absorbed N x 10-lS Q Quantum yield y 3 900 9.58 0.555 3.15 7.50 0.14 7.35 1.66 4 900 9.48 0.928 3.15 7.43 0.23 7.20 1.57 5 900 ll.40 1.68 3.15 8.92 o.42 8.50 1.10 6 900 9.87 1.13 3.15 7. 72 0.28 7.44 1.18 7 870 11.49 o.89 3.15 8.70 0.21 8.49 1.26 8 900 9.37 1.12 3.15 7.33 0.28 7.05 1.29 9 900 10.92 8.99 3.15 8.54 2.23 6.31 o.915 10 900 10.83 4.26 3.15 8.47 1.06 7.41 1.05 11 900 11.07 1.67 3.15 8.66 0.41 8.25 1.06 14 780 8.54 0.39 2.86 5.25 0.08 5.17 1.30 15 780 10.82 1.74 2.86 6.65 0.37 6.28 0.915* 16 780 9.17 5.57 2.86 5.63 1.20 4.43 1.09 17 787 9.58 o. 74 2.86 5.95 0.16 5.79 1.24 18 780 10.03 9.95 2.86 6.16 2.14 4.02 1.25* 19 780 11.13 1.55 2.86 6.84 0.33 6. 51 1.08 20 780 10.42 1.46 2.86 6.40 0.31 6.09 1.18 21 780 9.31 o.oo 2.86 5.73 o.oo 5. 73 1.29 22 780 9.64 o.oo 2.86 5.93 o.oo 5.93 1.28 23 780 10.22 1.90 2.86 6.28 0.41 5.87 0.958"' 24 780 9.81 2.24 2.86 6.03 o.48 5.55 1.10 25 780 9.93 4.55 2.86 6.11 0.98 5.13 1.09 26 780 9.38 o.oo 2.86 5.77 o.oo 5.77 1.57 . ' l~ +: l . t h-1:t' :t i 4r + -- ·r ' t :r: ' + ' r ' ' 1 "T - - ~ -+ H· - J ' :t '- -l- H- 1 ' : - - ' ~ ' ~1 ' 1-- L· i ~ l I+ ' - ±: 0 0) rl - - i.:r + '' - - ' ' ' f-'.· I .L ' I+ ' -- -+ 1-1- ,:J- I+ !- ..i :, tt' # ~ •+ i-h· it ITT: Li- - I I -i -I ' ' -:- - :-[trrtt -rm ,.R , u..:~ j L ..w. , f tll ra l r h -g ' -f-r"f~flffl t,+i::1tt~ r-'IT 1:r: I r ' J l ,, ' I I ,_ 1 L l= f:- - Lf- 1-1 i H i + T -t I ' j ' '1" tt ''- t 1~ .41 ,- I- ; '' - H- -1:j::i.. ,+ ' -H -- ~. -, ' ' ' !ft t- - ' ' ' ' ' ' ' .-H- -H-H ' ' ' ; - ' ' I- L ' ' 1 + - l r-l r-l L +;--l - I:<) ,:: ,, l -- + I N rl r--1 0 ,-I m co t-- co LO ti) . ;...i g HJ-I- ' rd - .L -++ -~ fi ! '-\ ,, ➔+-+ - :r J - + ' 1 - -R ' t t-+· l- 0 r-i I ,+ - -1-t ' t -- ' ' . •-' ' IT - _, !-- H- '' - • r-i l-'-1-1+ ' - • 0 r-+1 1+' ·H- r:t a I I,() rl N :+ -- ' 0 i- fflr ' ' ' ' ' r- :-- . r- $.,·--t ltf i:.i+l ' ' T - .1 '' ' i '++- ' • rl f- ' ' 0 'Sf! rtrWltUH-'a+lr-t..jlJ lr 11_ ~ + ' ' tt L J. ~.-t 1i :~H tH-- ~ +-- -- It J - -f:rdfl~l -4ft:T Ii\,-+ 1·Ptt[it·H, t+t, :!1 l:: I +' ~; '' '/- t=rt-n.: 1ill 11. ,_ti 'l+ ! r i ~ -'-Ft J::,t f ' 7 LO rl t .±; 1 r ;-·!-!-I ,-~1{:111· 11 'ff ' -lt-t- ~ ri·4 l-td- ~mmltf-1 _ft !l ' ' t±U. _, ,±- H : - - I' -r:t }1+- l±L 1 :,• ;: n r:i-TI ,_r I~ h :-- i · +· 0 :f 1:r11. ~- _-;-tr+- ' J. rt ; I -- +,- ' ·"I t rl -H+ H--lr ' (_(_ ~ ,+1~r-1 'ij'r r..U_,,- ~t~ t: t' + -~ t '_,J ~ H- H+~ ,· ~!- - lf - ' r - .:/= f-1 1-l- ' ' ' • rl +-:i.-" • ri q-cr H - J ~J H--~ ' l: L,. tWtl ~~:tr- .L' Hf =i,-= It ' r ' ' tJ ,:' - ' ;--;- '' ' 0 (.() + r-f · - t -- ' T 17" ~ H T +-+- --, . I ' T 4 - +1- +1 -'-- J1 - ' 1-7,R= ' H- -! : =r-... ' t J }i I - ' - ~ + f' u ..i l ' :rt t 1-i. 1. :f:_ l -:- ~~ - ~ + ,_ it= !-+- ' ' I- : : + - : I ~ ,+· >-+ J - + I- H-' 'j 1-t - ' ' ' IJt H 11-j-- ' t~ j -+ ii{ ' ' '' ' f .,, + '' r T Pi rt ..Ll- +-h I t· ' ,, - rt Lt '"'"" ·~: i ti:tl I ,_ ' ' +±!" H: -! '' L U, I+ fPI + .' .:t I- - tL ' ' :J ''P: u:-l- 0 -!:'+ ti}1 t- : ' - s ' - . 1-FtI- ..I' - '_j_ ' + :± 1 - - =l - - - .. f : ~ -+ l - ' , + 1:t=f --'- ::i: i1 -,r ' i..-l rl ' ' 1--'- - ' ' ' - - - r--4 f - ;..! t-i -l- - -t:~ I + I.Q <.O f 0 +· - • t 1-'-+--H--H--<--t-+I+-tt+;- •-+ r J L -f CQ L <.O I-, 0 t ~- ~ uH---H-+-H--H--H-1-+H-++++++I ; f~~ 1-t}I d I o-++-H--H++O++++++;-++,-,--H-+--:'- H ij{rilft it-r~Jt·mt r •t l-+- 1-l 1-+ - . 0 I - f+ H H-l-1--l-l-l~ 1 -rt~ I I t.t r I .J._J 11 t J If + -U' fr ,_ _ • I I I -t I +· '1--: •-+ 1 ---i 1 +-++-t-+-,--,H--+-1--<--t-+-M-t-H-,-, µ ''. 0 T :0 ' 4 t- t- ff ~-!+:s+i:~ - _I 0) I I.O Ht 0 P=!:i ]_ Fffl H-H++,_,_.__,<++->+'-1++++ ::t - • ~ f . a) I.O +Hl-l+l+l-10 ► ' ', i ttt- 1!ctttH tt·tttrt-J-+H H-1~rttu±1+ H-+++!-H+lt H--+ ~ + d M tr.) C\l r-l r-l r-l r-l 0) a) .:- • TJ±t-t!±-,-±lli±il:J±i±lP c-- co tr.) N 0 • 30. DISCUS SION OF RESULTS The results of this investigation are in agreement with the results of Elgin and Taylor insofar as the products of the decomposition have been identified as ammonia, nitrogen and hydrogen. These investi- gators presented data which showed that at .the init~al stage of the decomposition, the mole fractions of hydrogen and nitrogen in the gases which are non-condensable at liquid air temper atures are equal, i.e., each equal to 0.5, but as the decomposition proceeds the mole fraction of hydrogen increases toward the ultimate va lue of 0.66 at complete decomposition. These results were taken as evidence that the decomposition of hydrazine proceeded according to the stoichiometric equation followed by the subsequent decomposition of the ammonia The results of the present investigation are in disagreement - above, for the observed mole fractions of hydrogen never fell with the below o.58, even though less than a tenth of one percent of the hydrazine was decomposed during the run. In the data reported by Elgin and Taylor the minimum decompositiofi of hydrazine was about thirty percent, for which a mole frnction of hydrogen of 0.52 was obtained •. An analysis of the data shows. however, that they are rather inaccurate and are not necessarily in disagreement with the present results. As was stated pre- viously (Equation 1) a stoichiometric relation exists between the a.mount of 8llh~onia. the amount of non-condensable gas and the mole fraction of hydrogen in the latter gas. If the gases are measured in a system of the sarae volume e.nd at the same temperature, Equation 1 may be written 31. The data presented by Elgin and Tayler are given in Table III. TABLE III Exp't p total P1rn 3 P~+N2 X 4 - 6X PNH3 (calc) mm. (calc) 37 6.1 2.2 3.9 0.52 0.88 3.43 35 12.2 3.8 8.4 .50 1.00 8.4 38 13.4 4.2 9.2 • 56 0.64 5.8 .. 36 16.2 5.1 11.1 .55 • 70 7.77 39 24.1 1.8 22.3 .60 .40 8.92 28 26.5 ~0.5 26.0-26.5 .65 .10 72.6 The pressures of non-condensable gases were obtained by subtracting the pressure of ammonia from the total pressure of the products. ·From Equation 4 were obtained the calculated pressures of ammonia for comparison with the observed pressures. A rather large disagreement is observable. For the present discussion, however, it is more suitable to c&.lculate from Equation 4 the mole fractions of hydrogen which correspond to the observed pressures. These calculated mole fractions are given in Table IV. TABLE IV Exp't X(obs) Xcalo 37 o.52 o.57 35 .50 .59 36 .56 .59 · 36 • 55 .59 39 .60 .65 28 .65 .66 32. It will be seen that the calculated values are much more nearly in agreement with the present results. The authors state that their method of analysis was somewhat unsatisfactory as it involved long periods of time (five to six hours) for the diffusion of the gases into the copper oxide combustion region. The fact may also be mentioned that in the previous work the runs lasted as long as three hours• It would appear that the trouble caused in the present work by wall effects would be even more pronounced in the case of such long runs. 33. MECHANISM OF THE DECOMPOSITION In the work of Elgin e.nd Taylor both the purely photochemical and the mercury photosensitized decompositions of hydrazine were investigated but the quantum yield determination was attempted in the latter case only. A quantum yield of thirteen was obtained, which was independent of the presence of added nitrogen, hydrogen or ammonia. The inability of nitrogen and ammonia to lower the quantum yield is easily explained since both substances are known to be quite inefficient in quenching the fluorescence of excited mercury atoms. 9 • 10 The fact that hydrogen did not lower the quantum yield is not so easy to explain, for hydrogen molecules are known to be practically 100 percent efficient in deactivating excited mercury a toms. 9 , 10 ro explain their results, Elgin and Taylor found it necessary to assume a chain mechanism for the sensitized reaction in which such intermediate products as H, N2 H3 and N~ were postulated. They stated further "there is no reason to believe that the purely photochemical decomposition proceeds by another meche.nism11 • There seems to be no reason for this latter statement and, in fact, the results of the present investigation indicate that this statement is definitely incorrect. It is difficult to see why the "same mec hanism" should give a quantum yield of somewhat greater than unity for the purely photochemical decomposition and a yield of thirteen for the photoscnsitized decomposition. The results of the present investigation can be plausibly explained on the basis of a rather simple mechanism without the intervention of such remarkable species as N?H3 , etc. From the work of Elgin and Taylor the absorption spectrum of hydrazine appears to be of the 34. ''predissocia.tion11 type. The primary act may therefore be represented by the following reaction: where N2 H4 1 represents a.n activated hydrazine molecule in its predissociated state. This activated molecule may then react in one of three ways, viz: 2b. N2 H4: 1 = N2 + 2H2 2c. N2 H4 • = N2 H41 + J t/ Reaction 2a obviously leads to a quantum yield of 2 and a value of X = 0.5. Reaction 2b leads to a quantum yield of unity and a value of X = 0.67. Reaction 2c leads to a quantum yield of zero. So f~r as is known, fluor- escence has never been observed in the case of predissociated molecules, as is postul ated by Reaction 2c, hOV'rever there is no appa rent theoretical reason why fluorescence should not occur. If it is assumed that the activated hydrazine molecule ca n react only by 2a and 2b, then the following relation exists between the quantum yield Y a.nd the mole fraction of hydrogen in the non-condensa.ble products ( 4) 3 - 4X y =--- 1 - X In Table V a.re given the calculated values of Y for compo. ri son with the experimental values. It will be seen that at the higher hydrazine pressure, the observed and calculated values of Y agree within the experimental error but at lower pressures, the observed values for the quantum yields a.re lower than those calculated by Equation 4. This is to be ex- pected if Reaction 2c occurs, since its effect would become increasingly important a.t the lov1er pressures. 35. TABLE Run p 26 V X yexp ytheor 14.0 0.587 1.57 1.58 3 13.0 .582 1.66 1.60 4 13.0 . 582 1.57 1.60 21 12.0 .613 1.29 1.42 22 12.0 .611 1.28 1.44 7 11.0 • 598 1.26 1. 52 14 10.5 .628 1.30 1.32 8 10.0 .608 1.29 1.45 17 9.8 .616 1.24 1.41 5 8 .618 1.10 1.39 6 8 .621 1.18 1.37 19 7.0 .629 1.08 1.29 20 7.0 .619 1.18 1.37 11 .:.6.0 .613 1.06 1.42 15 6.0 .644 0.915 1.18 23 5.0 .639 0.958 1.22 24 5.0 .634 1.10 • 1.26 25 3.5 .634 1.09 1.26 10 3.0 .622 1.05 1.35 16 3.0 .644 1.09 1.18 18 2.0 • 629 1.25 1.29 9 2.0 .634 o.915 1.26 N 2 H4 36. It would seem therefore, that the above me chanism although not necessarily the only one, offers at least a simple, reasonable explanation of the experimental facts. SUMMARY The decomposition of gaseous hydrazine by monochromatic radiation has been investigated over a range of pressures from ti.vo to fourteen millimeters of mercury. Hydrazine appears to decompose simultaneously into a..'!llnonia, nitrogen and hydrogen. The results clearly indicate that the composition of the decomposition products and the quantum yield depend on the hydrazine pressure. In this range of pressures, the quantum yield varies between the approximate limi t s of 1.00 and 1.60. A simple mechanism for the decomposition based on these experimental results has been offered. The photochemical studies of ~lgin and Taylor on hydrazine have been criticised in the light of the present . results. 3 7. B IB LI CGRAPHY 1. El gin and Taylor , J.Am.Chem.S oc., ~• 2059 (1 929) . 2. Beckman and Dickin son, J. Am . Chem . Soc., 52, 124 (1930). 3. Warburg, Sitz . Be r.Pr euss. Akad ., 746 (1 911); 216 (1912). 4. Kuhn, J.d.Chem.Phys ., 23, 521 (1926). 5. Beck~an and Dickinson, J . .F:...m . Chem . So c., 50, 1870 (1928) . 6. Beckman, J.Opt.Soc., 16., 276 (1928) . 7. Ha l e and Sh etterly, J . Am .C hem.Soc ., 33, 1071 (1911). 8. Bray and Cuy, J . P.m . 0 hem. Soc., ~• 860 (1924). 9. Stuart, Z.Physik, ~• 262 (1925). 10. Mitchell and Dickins on , J . Am.Chem.Soc., 49, 1478 (1927).