122s
JOE
CO(NH~)~C~ as+ + compared to Co(NH3)40H2.C1++, but the former undergoes chloride release in aqueous solution a t a rate only a few times greater than the latter,33as against a factor of 60 or more in the corresponding carbonato compounds. The explanation must be that, while each substitution event (whether by HC03- or H20) results in carbonate exchange in N5, HzO substitution for H 2 0 in the bicarbonato-aquo species, which results in no carbonate exchange, may well be the favored path.34 Furthermore, the dynamic (33) D. R . Stranks in "Modern Co-ordination Chemistry" (ref. 2 9 ) , p. 129. (3.1) Since only t h e immediately adjacent water molecules of t h e solvent sheath can take part in t h e substitutions, t h e geometry of t h e transition states for the H*O/HzO and HzO/HCOC reactions may be almost identical, differing only by small shifts in t h e 0 - H , @C and 0-Co bond distances. T h e more easily attainable orientation, assumed here t o be t h a t for HzOI HzO exchange, will thus be favored a t t h e expense of the other, especially as labilization of the OHz ligand by t h e adjacent hasic O C H ligand" will assist in formation of this "more favorable" transition state. ( : l a ) I t is significant t h a t water exchange between water solvent and t h e ris-Co(en)nOH.H?O-' ion occurs a t 23' a t a rate over .50 times greater than
[ COSTRIBUTIOS FROM
THE
1101. s5
V. MICHAELAND W. ALBERT NOYES,JR.
equilibrium of reaction 1 provides an additional noncarbonate-exchange reaction path which is not available to the X5 complex. The tn2 complex undergoes carbonate exchange an order of magnitude more slowly than do the other bident a t e carbonato complexes. This is logically explained in terms of the reduced value of Kl for tn2, since the rate constants of both types of carbonate substitution reactions include the factor K1(H20)/K1(H20) 1. Equilibrium isotope effects, hydrogen-bonding and spectral evidence for a relatively small K1value in tn2 have been discussed above. Acknowledgment.-Support of this research by the U.S. Atomic Energy Commission through Contract AT(30-1)-1578 is gratefully acknowledged.
+
the rate of the corresponding reaction of the cis-Co(en)z(Hz0)2+3ion (W. Kruse and H. Taube, J . Ain. Chent. Soc , 83, 1280 (1901)), and over 400 times greater than for cis-Co(en)2P;H3.HzO-~ ( D . F . Martin and hf. I,. Tobe, J . Chem. Soc., 1388 (1962)). This commonly observed hydroxideinduced labilization of t h e adjacent ligand (see ref. 3 3 ) can be expected t o be paralleled by a similar effect by the strongly hasic bicarbonate group.
DEPARTMEXT O F CHEMISTRY, VSIVERSITY O F ROCHESTER, ROCHESTER 20,
s.Y.]
The Photochemistry of Methylaminel BY JOE V. MICHAEL^
AND
W. ALBERTNOYES,J R .
RECEIVEDSEPTEMBER 29, 1962 Hydrogen, methane, nitrogen, ethane, ammonia, ethylenimine, dimethylamine, azomethane and a polymer have been identified as products in the photochemical decomposition of methylamine. Quantum yields of most of these products have been determined under a variety of experimental conditions at room temperature. By use of CH3NDl and of C D 3 S H l as well as by use of scavengers i t has been shown t h a t the main primary process is the elimination of a hydrogen atom. This is followed by abstraction from the substrate to form hydrogen gas. Other steps in the mechanism are suggested and evidence for some of them presented.
The photochemistry of methylamine has been extensively s t ~ d i e d . The ~ results were not always in good agreement. Methylamine is a photochemically interesting molecule. There is some direct production of molecular hydrogen in the primary process following absorption by methanol vapor.4 An analogous reaction in methylamine would be possible. Methylamine probably shows a predissociation type ~pectrum.~ By analogy with NH2 the reactions of CH3NH with oxygen and with nitric oxide might also prove to be The photochemistry of methylamine is complex, a fact to be expected from the difficulties in interpreting results on the photolysis of ammonia.* The main primary process is the formation of hydrogen atoms, but there may be small amounts of dissociation to (1) This work was supported in part by t h e Directorate of Chemical Sciences, Air Force Office of Scientific Research under Contract AF49(638)c70. (2) Sational Science Foundation Cooperative Fellow, 1959-1900; Eastm a n Kodak Company Fellow, 1000-19G1; National Science Foundation Predoctoral Fellow, 1901-1902. (3) H. J. Ernelkus and H . S.Taylor, J . A m . C h e m SOL.,63, 3370 (1931); H. J EmelCus and L J . Jolley, J . C h e m SOL.,1012 (1935); 0. C. Wetmore and H . A. Taylor, J . Chein. Pkys., 12, 01 (1944); C . I. Johnson and H. A. Taylor, ibid., 19, 013 (10.51); J. S . Watson and B. deB. Ilarwent, ibid., 20, 1011 (1K;Z). For reviews see W.A. Noyes, J r , and P. A. Leighton, "The Photochemistry of Gases," Reinhold Publishing Corp , S e w York, N . Y., 1931, p. 382, and E. W. R Steacie, "Atomic and Free Radical Reactions," Reinhald Publishing Corp., New York, S . Y . , 19.54, pp. 214, 030. (-1) R P . Porter and \V. A . S o y e s , Jr., J . A ~ I ICkein. , SOL.,81, 2307 ( 10.>!)), Lj) See H. J . Emeleus and I,. J . Jolley, ref. :3, A . B. F. Duncan, Phys. Rev , 47, 822 (1935), and R . S. hfulliken, J . C h e m Phys., 3 , 500 (1933), have discuised ammonia spectrum and bands. (0) H . Gesser, J . A m . Chem. Soc., 77, 202ij (1933). ( 7 ) A . Serewicz and VV'. A . S o y e s , J r . , J . Phys. Chenr., 63,813 (1959). ( 8 ) Cf.C. C . hIc1)onald and H . E. Gunning, J . Cheiii. Phys., 23, 532 ( I9):.
methyl and amine radicals as well as to form molecular hydrogen. Experimental An unfiltered Hanovia S-100was used for all experiments. Since polymer formed on t h e front window of the cell, it was necessary t o clean the cell after each run. The radiation was not monochromatic. -111 of the wave lengths between 1940 and 2440 il. are probably absorbed by methylamine. Actinonietry measurements were made by the hydrogen bromide-mercury vapor system in tandem with the photolysis cell. This actinometer is drscribed in detail e l s e ~ h e r e . ~ ~ ~ ~ One molecule of hydrogen is assumed t o be formed per photon when mercury is present t o react with the bromine. Quantum yields should be valid within 55%. Reagents.-Methylamine was prepared from Eastman Kodak Co. JVhite Label methylamine hydrochloride. I t was recrystallized three times from water, and the amine was liberated by anhydrous calcium oxide." The purity was 99.95;. Methylamine-C-& was prepared in t h e same manner as was niethylamine. Methylamine-C-& hydrochloride was supplied by D r . R . J. Cventanovih of the Sational Research Council, Ottawa, Can., to whom the author$ are indebted. T h e purity determined by vapor phase chromatography of t h e final sample was 99.97; and t h e isotopic purity determined by mass spectrometry was a t least 93:';. Methylamine-X-& was obtained from Merck, Sharp, and Dohme of Canada, L t d . , and had a purity of 99.5%. T h e principal impurity was S D 3 . The isotopic purity was a t least 99%. A4mmonia was obtained from the Matheson Co., Inc. A middle third was taken from a bulb-to-bulb distillation. Vapor phase chromatography and mass spectrometry showed it t o be 99.9% pure. Azomethane was prepared by the method of Jahn and was purified by vapor phase chromatography.12 Research grade methane and ethane (Phillips Petroleum Co.) were used without further purification. (9) W.A . Koyes, Jr., and P . A. Leighton, "The Photochemistry of Gases," Reinhold Publishing Corp., New York, N . Y., 1941, p. 83. (10) W. A. S o y e s , Jr., J . Chem. Phys., 6 , 807 (1937). (11) A . P. Gray a n d R . C . Lord, ibid.,2 6 , 090 (1957). (12) F. P . Jahn, J . A m . Chem. Soc., 5 9 , 1761 (1937).
1229
PHOTOCHEMISTRY OF METHYLAMINE
May 5 , 1963
TABLE I QUANTUM YIELDSAS A FUNCTION OF TIME is assumed to be unity Cell volume, 191.4 ml.; room temperature; P C H ~ = NH 140~ i 5 m m . ; Hanovia S-100 lamp; Time, sec. '%CH3)2KH Q(CH2hNH -*CHaNHz *(CHdzNz ON2 *C2H6 *CHp QKHJ 0.49 0.003 0,077 0.033 0.014 450 1.08 0 020 ,028 ,036 .57 ,013 ,046 1.11 ,019 900 ,041 .55 ,081 ,013 ,026 1.04 ,027 1800 .42 .12 ,027 ,010 ,023 3636 ,036 1.07 .30 2.2 ,025 .16 ,013 ,019 7200 ,049 1.04 .24 1.7 .15 ,020 ,028 ,015 ,051 10800 0.93 .17 1.9 ,019 ,024 .15 ,013 ,076 21762 1.03 1.9 i0.2 0.013 f 0.001 1.04 f 0.04 t
Dimethylamine was prepared and purified in the same manner as was methylamine from Eastman Kodak Co. White Label dimethylamine hydrochloride. T h e purity was 99.9%. Propylene (Phillips Petroleum Co.) was purified by taking a middle third from a bulb-to-bulb distillation. Cylinder hydrogen and deuterium were used without further purification. Mass spectrometric analysis showed only trace impurities. Hydrogen deuteride was prepared by the reaction of lithium aluminum hydride with heavy water. T h e sample was 987' pure as shown by mass spectrometric analysis. Ethylenimine was obtained from Chemical Intermediates and Research Laboratories, Inc. A middle third was collected from a bulb-to-bulb distillation, \.apor phase chromatography and mass spectrometry showed t h e sample t o be a t least 99% pure. Hydrogen bromide was obtained from the Matheson Co., Inc., and was used without further purification. T h e stated purity was 99.87,. Procedure .-Methylamine was measured in a calibrated volume and transferred t o the evacuated cell b y distillation. The sample was allowed t o stand a t least 15 minutes before irradiation. X conventional high vacuum distillation technique was employed t o obtain all products except dimethylamine and ethylenimine. These two products were obtained from mass spectrometric analysis of the photolyzed mixture. After t h e sample had been photolyzed, t h e break-seal was broken and the contents of the cell distilled into a t r a p a t - 196". Fraction 1 contained t h e gases not condensed by solid nitrogen (-215') and was oxidized over copper oxide a t 220". Mass spectrometric analysis showed t h a t hydrogen, methane and nitrogen were present. T h e hydrogen was determined as the difference between this fraction and the second fraction. Fraction 2: -4fter oxidation the residue was measured and placed in a sample tube for mass spectrometric analysis. This fraction contained only methane and nitrogen. T h e mole fraction of each could be determined and, thus, the absolute amount of each component calculated. Fraction 3 : T h e unreacted methylamine a n d products condensable by solid nitrogen were transferred t o a LeRoy still maintained at -16E1".'~ T h e gas was removed by a Toepler pump and measured. Analysis by the mass spectrometer showed this fraction t o contain only ethane. Fraction 4 was obtained by distillation a t -78" and contained almost all of the unreacted methylaniine. The sample was analyzed on a Perkin-Elmer model 154 vapor phase chromatograph with Column \T. Polyethylene glycol was the solute phase and the stationary phase was powdered Teflon. T h e column was 2 meters in length. At relatively slow flow rates, the products were separated from the unreacted methylamine. Azomethane was eluted after 2 minutes and was determined by the mass spectrometer. At 3.5 minutes, ammonia, identified in the same way, was eluted. Methylamine was eluted a t 7 minutes. The azomethane was determined by collection and chromatographic analysis with a Perkin-Elmer Column D. This is a 3-meter column of tetraisobutylene upon 60-80 mesh firebrick. Fraction 5: The final fraction was taken a t room temperature. Mass spectrometric analysis showed t h a t this sample contained dimethylamine and ethylenimine as well as unreacted methylamine. The quantitative results obtainablc for dimethylamine and for ethylenimine were never reproducible because Ethylendimethylamine has a slight vapor pressure at -78'. inline is adsorbed on glass so strongly- t h a t it was impossible t o col!ect all of it. The mass spectra always showed peaks in the m / e = 35-45 region but reiative peak heights changed with conditions. T h e ratio of peak 45 t o peak 44 was constant a t 0.56 and compared favorably with the ratio for dimethylamine. T h e height of peak 45 was taken t o measure dimethylamine. The spectrum after peaks due t o dimethylamine were subtracted was identical, within experimental error, with the spectrum of ethylenimine. T h e ratio of peak heights in methylamine, peak l7/peak 31, is 0.0139. Thus, by using this ratio, t h e relative peak height for -~ (1:j)
I ) J . I,eR(ly. Cnu. J . Rus
B28, 1 Y 2 (19.50).
peak 17, due t o ammonia, can also be calculated. Calibrations were made for ethylenimine, dimethylamine and ammonia in the presence of excess methylamine. X straight line relationship between t h e mole fraction and the parent peak fraction was obeyed, and it is believed t h a t analyses performed in this way are fairly accurate.
Results The photolysis of methylamine was studied as a function of time, intensity and pressure. The results are given in Tables I, I1 and 111. TABLE I1 QUASTUMYIELDSAS A FUKCTIOX OF INTENSITY Cell volume, 191.1 m l . ; room temperature; P C H ~ = SH 144 ~ is assumed t o be unity mm.; Hanovia S-100 lamp; Time, sec.
Intensitya
*Car
*Xz
*5
*C2H6
0.98 0.045 0.013 0.024 ,005 ,023 ,037 1.05 ,005 ,026 0.98 ,029 ,021 ,030 ,004 ,l5 1 . 1 0 ,008 ,026 ,028 .04 1 . 1 5 0 . 0 2 4 f 0.002 1.05 f 0.06 ilrbitrary units.
i200 (ai2 T200 12626 12600
- --
1.00 0.56 .38
In all experiments a polymer formed on the front window and absorbed the photolyzing wave lengths. Rates of production of products appear to decrease with time even if they were truly time independent. Quantum yields for four runs in the range 1800 to 8900 sec. were determined with the hydrogen bromidemercury vapor actinometer and indicate that is close t o unity in this time range. The values given in Table I are based on @H! = 1.0 for the complete time range. The hydrogen quantum yield was constant and independent of intensity. Since @H* is near unity the hydrogen yield is also assumed t o be unity over the complete intensity range and the other product quantum yields are calculated from this assumption. These data are shown in Table 11. The hydrogen bromide-mercury vapor actinometer was used to obtain the quantum yields as a function of pressure. These data are shown in Table 111. The rates of production and the isotopic distributionas of the products formed in the photolysis of CD3NH2 and CHsND2 are given in Table IV. The photolysis was also studied in the presence of the scavengers: propylene, oxygen and nitric oxide. The data are given in Table V. Azomethane was photolyze$ in the presence of CDBNH2 and of CH3ND2a t 3130 A. and room temperature. In both cases CHaDand CH4were observed. This indicates that CH2NH2 and CHaNH radicals are both formed by methyl abstraction. Attempts t o characterize the polymer were made. The formula calculated from material balance seemed to change from C2NH5to C2NHl as a function of time, but product determinations are not accurate enough to state this conclusively. The empirical formula determined from combustion analysis was (C2NzH3)z ,
1230
JOE
v. h k C H A E L AND I\'.
ALBERT
NOYES,JR.
TABLE I11 QUANTUM YIELDSAS A FUNCTION OF PRESSURE Cell volume. 191.4 ml. ; room temperature; time, 7200 sec. ; Hanovia $100 lamp ; hydrogen bromide-mercury actinometer P C H ~ S H ~ mm. ,
Hg
138.3 98.9 49.7 23,5 16.1 4.4
*Hz
*SE3
0.90 .89 .95 .83 79 .62
0.93 .89 .96 .78 .68 .27
*CHI
0.042 ,042 ,047 ,053 ,053 ,045
*h-2
0.011 ,011 ,020 ,038 ,041 ,049
*C2H6
*
