Cryogenic Photolysis Studies Part 2.-Inf'rared Spectrum of Nitrosomethane Monomer J. BARNES,~ STEPHENBYAUSTIN HARRYE. HALLAM,* WARING~ AND J. RONALDARMSTRONG Department of Chemistry, University College of Swansea, Singleton Park, Swansea SA2 8PP Received 8th July, 1974 The photolysis of trans-t-butyl nitrite in an argon matrix at 20 K gave rise to absorptions due to acetone, nitrosomethane and cis-t-butyl nitrite. Isomerisation was dominant at higher concentra- tions ;decomposition was dominant at lower concentrations. Vaporisation of nitrosomethane dimer and trapping of the products at 20 K yielded the spectrum of nitrosomethane monomer and the trans dimer. Ultra-violet irradiation dissociated the dimer, leaving nitrosomethane monomer. On warming to room temperature and recooling to 20 K the cis-nitrosomethane dimer was obtained, which could be dissociated by photolysis to regenerate the monomer. A vibrational assignment of the monomer is presented.The structure and properties of C-nitroso compounds have been reviewed by Gowenlock and Luttke.l Nitrosomethane is produced in the pyrolysis or photo- lysis of t-butyl nitrite and by the reaction of methyl radicals (from e.g., the photolysis of iodomethane) with nitric oxide. Early results were confused since the final product obtained is nitrosomethane dimer or, in some cases, the tautomer formaldoxime. Gowenlock and Trotman carried out a detailed investigation of the pyrolysis and photolysis products of t-butyl nitrite and established that the nitrosomethane dimer exists in two isomeric forms, cis and trans : CH3 CH3 CH3 0 \/ \/IN=N N-N / 4 d\ 0 0 0 CH3 The cis dimer was produced on warming the trapped pyrolysis products to ca.230 K or on ultra-violet irradiation of the trans dimer. The tram dimer was produced on photolysis of t-butyl nitrite or on heating the cis dimer or on dissolving the cis dimer in a solvent of low dielectric constant. Monomeric nitrosomethane could be generated by heating the dimer in the gas phase (the activation energy for dissociation was found to be ca. 90 kJ mol-I). Tautomerisation to formaldoxime occurred most readily in aqueous solution. Gowenlock et aL2* and Luttke reported infi-a-red and ultra-violet spectra of the two dimers. Luttke attempted to record the infrared spectrum of the monomer by dissociat- ing the dimer in a heated gas cell, but experienced difficulty due to rapid tautomerisa- j. present address :Department of Chemistry &Applied Chemistry, University of Salford, Salford M54WT.$present address :Atomic Energy Research Establishment, Harwell Oxfordshire OX11 ORA. 11-1 1 CRYOGENIC PHOTOLYSIS STUDIES tion. He reported the N-0 stretching mode at 1564 cm-I and the C-N stretching mode at ca. 842 cm-l. The electronic spectrum of the monomer has been reported by Dixon and Kroto following flash photolysis of t-butyl nitrite ; absorptions at 660 and 677 nm were observed. Coffey, Britt and Boggs obtained the microwaye spectrum of the monomer by vaporisation of the trans dimer at room temperature.The matrix isolation technique is a powerful method of stabilising transient species for spectroscopic stu~ly.~ Thus we embarked on an investigation of the photolysis products of t-butyl nitrite isolated in argon matrices in an attempt to obtain the complete vibrational spectrum of nitrosomethane monomer. EXPERIMENTAL The cryogenic apparatus and procedure have been described previously. O t-Butyl nitrite, obtained from Fluka Chemicals, and acetone were dried over anhydrous calcium chloride and fractionally distilled. Photolysis experiments were carried out using a Philips medium pressure mercury lamp. In the first series of experiments (photolysis of t-butyl nitrite in argon matrices) spectra were recorded on a Perkin-Elmer 225 spectrometer ; wavenumbers quoted are accurate to _+ 1 cm-l.In the second series of experiments (deposi- tion in an argon matrix of the product of the vaporisation of nitrosomethane dimer) spectra were recorded on a Perkin-Elmer 457spectrometer ;wavenumbers quoted should be regarded as accurate to +3 cm-l. RESULTS Idrared spectra of t-butyl nitrite in argon matrices at 20 K were recorded at matrix to absorber (M/A) ratios of 2000, 500 and 100. No significant changes were observed between the spectra. Spectra of t-butyl nitrite in an argon matrix at M/A 500 were recorded after different photolysis times with the medium pressure mercury lamp. A number of new absorptions appeared (fig. 1) which increased in intensity with increasing photolysis time. Identical behaviour was observed in several experi- ments under different conditions, including photolysis during deposition.Photo-Iysis of more dilute matrix samples (M/A 2000) gave rise to a different distribution of intensities of the product absorptions (table 1). Analogous results were obtained when the experiments were repeated in nitrogen matrices. Also recorded in table 1 are the more intense bands observed in the spectra of acetone trapped in argon matrices at M/A 500. There is an excellent correspondence between a number of the absorptions produced on photolysis of the t-butyl nitrite and the spectrum of acetone. The only medium or strong band in the acetone spectrum not observed in the photolysis product spectrum (2972 cm-') would be obscured by a t-butyl nitrite absorption.The small shifts observed for a number of bands, particularly the C=O stretch, may be accounted for by the presence of another photolysis product in the same or an adjacent trapping site. Coe and Doumani and Tarte l2 proposed that the photolysis of t-butyl nitrite proceeds via intramolecular rearrangement to give acetone and nitrosomethane as primary products : (CH3)3CONO+hv + (CH3)JCO+CHSNO. However McMillan, Calvert and Thomas l3 found that there is an induction period in the appearance of nitrosomethane, whereas acetone is produced immediately. Dixon and Kroto 'also observed an induction period in nitrosomethane production in their flash photolysis experiments.Thus the mechanism is apparently : (CH3)3CONO+]ZV + (CH3)3CO* +NO (CH3)3CO* + (CH3)2CO+CH3 CH3+NO + CHSNO. A. J. BARNES, H. E. HALLAM, S. WARING AND J. R. ARMSTRONG I I I-lI 1 I 1800 I600 lL00 1230 1000 wavenumber/cm-l II I000 8 00 600 400 FIG.1 .-Infrared spectra of t-bury1 nitrite in argon matrices at M/A = 500 before photolysis and after 150 min photolysis with a medium pressure mercury lamp. Absorptions which appear on photo- lysis are arrowed. (a) 1800 to 1000cm-' region. (6) 1000 to 400 cm-' region. In a matrix at 20 K it is unlikely that the fragments could escape from the cage and thus it would not be expected that appreciable quantities of methyl radicals and nitric oxide would be stabilised. No evidence was found for infra-red absorption due to either methyl or nitric oxide; consequentIy it would seem reasonable to assign the CRYOGENIC PHOTOLYSIS STUDIES absorptions, not due to acetone, observed after the photolysis of t-butyl nitrite to nitrosomethane.However, there are two sets of bands with different concentration dependences. The set which is more intense in the M/A 2000 spectrum parallels the intensity variation of the acetone absorptions and may thus be assigned to nitroso- methane. The complexity of the spectra undoubtedly causes several nitrosomethane bands to be obscured, thus an attempt was made to obtain nitrosomethane monomer trapped in an argon matrix without other interfering species. Nitrosomeihane dimer was prepared by vapour phase photolysis of t-butyl nitrite in a silica bulb with the bottom part blackened to prevent photolysis of the liquid (cf.Gowenlock and Trot- man '). The following series of experiments was performed, infrared spectra being recorded after each stage. TABLE1.-cOMPARISON OF THE INFRARED SPECTRUM OF ACETONE WITH ABSORPTIONS (Cm-') PRODUCED BY THE PHOTOLYSIS OF t-BUTYL NITRITE IN AN ARGON MATRIX (CH 3)3CONO photolysis products 3019 (W -3018 (m) CH3 asym. stretch 301 5 (w) -I --2972 (m) CH3 asym. stretch -2962 (vw)1718 (m) (s) 1722 (vs) C=O stretch 1599 (w)1597 (s) 1555 1551 1460 1435 1443 (m) CH3 asym. defs.1429 (s) 1406 (m) CH3 asym. def. 1361 (s) CH3 sym. def. 1354 (s) CH3sym. def. 1223 (s) C-C stretch1216 1091 (m) CH3rock 865 -860 736 (ms) (m) 729 (m) (m679 (w)--570 (w)531 (w) (mw) 529 (ms) C==O i.p.bend * T. Shimanouchi, Tables of Molecular Vibrational Frequencies (NSRDS-NBS 39, 1972). (I) The vapour from the heated nitrosomethane dimer was deposited on the cold window of the cryostat simultaneously with a stream of argon matrix gas (no attempt was made to estimate the matrix ratio achieved). (2) The deposited products were irradiated with a medium pressure mercury lamp for 45 min using a filter to cut out wavelengths below ca. 250 nm. A, J. BARNES, H. E. HALLAM, S. WARING AND J. R. ARMSTRONG (3) The window was allowed to warm up to 77 K (thereby evaporating off the argon matrix gas) and then recooled to 20 K.(4) The window was allowed to warm up to room temperature and then recooled to 20 K. (5) The material on the window was irradiated with the medium pressure mercury lamp for 40 min using a filter to cut out wavelengths below CIZ. 250 nm. (6) The irradiation was repeated without the filter. The absorptioiis observed (fig. 2) after stages 1, 2, 4 and 6 are recorded in table 2. The nitrosomethane dimer vaporisation products originally deposited apparently contained at least two species, since a number of the absorptions disappeared on photolysis. The species remainirrg is designated A, that removed by the photolysis is called B in table 2. Heating the window to 77 K and then recooling to 20 K had little effect on the spectrum, despite the loss of the argon matrix material : the 1297 cm-l absorption of species B reappeared only weakly and the 1410 cm-l absorption of species A split into a doublet. However, on warming to room temperature and recooling to 20 K the absorptions due to species A were replaced by an entirely new spectrum (designated species C in table 2).The effect of photolysis with the filtered mercury lamp was to reduce the intensities of the bands due to species C, while the spectrum of A reappeared. Unfiltered photolysis led to complete conversion of species C to species A. __ I, I I 1 I 3200 28001600 1400 1200 1000 800 600 400 wavenumber 1cm-I FIG.2.-Infrared spectra of (a) products of vaporisation of nitrosomethane dimer deposited in an argon matrix at 20 K; (6) deposited products after 45 min filtered mercury lamp photolysis ;(c) Droduct after warming to room temperature and recooling to 20 K :(d)product as (c) after 60 min photolysis, using an untiltered lamp.Similar results were obtained when the experiment was repeated with a different flow rate of argon gas (the rates used were in the range 10-20mmol h-l) and the experiments should perhaps be regarded as solid state, rather than matrix, photolysis in view of the slight changes resulting from the loss of the matrix in stage 3. CRYOGENIC PHOTOLYSIS STUDIES TABLE2.-ABSORPTIONS OBSERVED (Cm-') AFTER DEPOSITION OF THE PRODUCTS OF THE VAPORISATION OF NITROSOMETHANE DIMER IN AN ARGON MATRIX species species species stage stage stage stageA B C (1) (2) (4) (6)* * * 3059 (w) * 3055 (mw) *3019 (w) * * * 2991 (mw) * * 2955 (w) *2951 (mw) *2917 (w) * * * 2901 (mw) * * * 1549 (s) * 1511 (w) *1484 (w) *1443 (w) *1422 (m) 1411 (m) * * * * 1410 (s) *1401 (m) * 1388 (m)1383 (vs) * (w)*1371 (mw) * * *1348 (s) 1343 (m) * * 1311 (m) * 1304 (w) *1297 (s) * * 1162 (w) * 1142 (m) * * 1128 (w) 1104 (m)1069 (m) 1057 (m) 1028 (s) * * 967 (w) * 949 (m) * * * 916 (mw) * * * 870 (m) *749 (m) *622 (m) *619 (s) * * 574 (m) 536 (s) * *480 (m) *470 (m) * = observed.DISCUSSION From the detailed study by Gowenlock and Trotman of the isomers of nitroso-methane dimer, the most probable identification of species A, B and C is monomer, trans dimer and cis dimer respectively. Species A is produced by the photolysis of either B or C, as would be expected if A is the monomer and B and C are the dimers.A. J. BARNES, H. E. HALLAM, S. WARING AND J. R. ARMSTRONG Warming species A to room temperature leads to species C, corresponding to Gowen-lock and Trotman's observation that the cis dimer is produced from the monomer under similar conditions. The possibility of species B being formaldoxime may be rejected on two grounds : (i) the absorption maximum of formaldoxime is below 210 nm, but species B is photolysed by wavelengths above 250 nm (the absorption maxi- mum of the trans dimer is at CQ. 275 nm); (ii) the absorptions observed fur species B do not include a strong band above 1600 cm-I expected for formaldoxime.'" The (relatively) slow dissociation of species C on filtered-mercury-lamp photolysis may be explained since the absorption maximum of the cis dimer is to lower wavelength than that of the trans ciimer,2 and thus nearer to the cut-off of the filter used.The possibility of secondary photolysis may be rejected, since the different species may be readily interconverted by using the appropriate conditions. These identifications of species €3 and C may be confirmed by comparison of the absorptions observed in these experiments (table 2) with the spectra reported by Gowenlock et d The trans dimer in a KBr disc displays strong absorptions at 2p 3049, 1397, 1286, 1134 and 936 cm-'. Species €3 has bands at 1297, 1142 and 949 cm-1 corresponding to the last three, while the first two would be obscured by absorp- tions due to species A.The strong band observed for species B at 536 cm-I was out- side the range of the original study. Similarly the cis dimer in a KBr disc was reported 2, to exhibit strong absorptions at 3049,2967,1667,1471,1387, 1341,1107, 1061, 1017 and 740 cm-l. It can be seen from table 2 that species C exhibits strong absorptions corresponding to all these bands with the exception of that at 1667 cm-'. The low tenperature spectra obtained here do, as expected, show more detail than the previous (KBr disc) spectra. The weak absorption observed at 1303ern-' coincides with a characteristic band of cis nitrosoalkane dimers, attributed to the N--N stretch, which had not previously been observed for the cis dimer of nitroso- methane.Species A may thus be assigned with confidence as nitrosomethane monomer. Comparison of the absorptions produced by the photolysis of t-butyl nitrite in argon matrices with those of species A shows that the bands which are more intense at high M/A ii; ihe photolysis of t-butyl nitrite (1555, 1411, 1351, 865 and 570 cm-l) corre- spond to the most intense absorptions of nitrosomethane monomer. The absorptions which are more intense at low M/A (1598, 938, 736 crn-I and other, weaker, bands) do not correspond with the spectrum of any of species A, B or C. The possibility that this set of absorptions is due to formaldoxime may be rejected, since 1598 cm-' is too low a frequency for the C=N stretch of formaldoxime, nor do the other unassigned absorptions correspond with the spectrum of formaldoxime. The absorp- tions cannot be attributed to secondary photolysis products, since they are observed after.a short photolysis time and increase steadily in intensity with increasing photoly- sis time.Unless photolysis of t-butyl nitrite aggregates in an argon matrix can lead to hither- to unsuspected decomposition products, the only remaining possibility is that iso- merisation of the t-butyl nitrite is occurring. The nitro compound, 2-methyl-2- nitroprspane, may be ruled out as the major product since no strong absorptions corresponding to the NO2stretchin2 modes of the compound (1543 and 1346 cm-I) l6 were observed.Brown and Pirnentel l7 observed cis-trans isomerisation of methyl nitrite on photolysis in an argon matrix. Tarte has investigated the geometrical isomerism of the alkyl nitrites, and found that t-butyl nitrite is predominantly in the trans form, He gives the vapour phase N-0 stretching frequency of the trans and cis isomers as 1655 and 1610 cm-I respectively. The argon matrix frequency for the trans isomer (the only conformer present in the low temperature matrix) is 1638 cm-l, CRYOGENIC PHOTOLYSIS STUDIES thus 1598 cm-l is very close to the value expected for the cis isomer. A further verification is provided by the characteristic ON0 bending frequency, normally found l8 at ca.600 cm-l for the trans isomer and ca. 680 cm-l for the cis isomer. This mode may be assigned as 580cm-l in the spectrum of unphotolysed trans-t- butyl nitrite, and the photolysis product shows a corresponding absorption at 679 cm-l. Comparison of the principal bands of the trans-t-butyl nitrite with the bands produced by photolysis, which are not assignable to either acetone or nitrosomethane, TABLE3.-cOMPARISON OF THE PRINCIPAL BANDS (Cm-l) OF tranS-t-BUTYL MTRITE IN AN ARGON MATRIX WITH THE ABSORPTIONS ASSIGNED TO Cis-f-BUTYL NITRITE ~~UIIS(CH 3) JONO cis(CH3))30NO assignment 3002 (m) 3015 (w)2995 (s) 2962 (vw) CH3 asym. stretches 2989 (s) 2945 (m) 2914 (m) CH3 sym. stretches 2883 (m) 1638 (vs) 1598 (s) N-0 stretch 1476 (s) 1464 (m) 1460 (w) CH3 asym.defs 1395 (s) 1371 (s) 1363 (w) CM3 sym. defs1369 (s) 1268 (s) *1252 (m) 1246 (m) CH3rocks 1197 (vs) * 1180 (m) 1037 (s) 957 (m) 938 (s) 808 (vs) 916 (m) skeletal stretches 736 (ms)769 (vs) 729 (m)761 (vs) 580 (w) 679 (w) ON0 bend * these bands are more intense in the photolysed spectrum. TABLE4.-ASSIGNMENT OF THE FUNDAMENTAL VIBRATIONS OF NITROSOMETHANE MONOMER IN AN ARGON MATRIX mode CH3NO/Ar CH,NO(g) a CHjCHO b A' AN CH, asym. stretch CH3 sym. stretch N-0 stretch CH3 asym. deformation CH, sym. deformation CH3rock C-N stretch CNO bend CH3 asym. stretch CH3 asp. deformation CH3 rock torsion 2991 (mw) 2901 (mw) 1549 (s) 1410 (s) 1348 (s) 967 (w) 870 (m) 574 (mw) 2955 (w) 1410 (s) 916 (mw) 1564 842 3005 (m) 1441 (s) 919 (m) 2917 (-) -1352 (s) -I 2967 (m) 1420 (s) 867 (m) 150 (w) a ref.(6). T. Shimanouchi, Tables of Molecular Vibrational Frequencies (NSRDS-NBS 39, 1972). A. J. BARNES, PI. E. HALLAM, S. WARING AND J. R. ARMSTRONG shows a clear parallel in each spectral region (table 3). Thus, photolysis of trans-t- butyl nitrite aggregates leads to a mixture of cis and trans isomers, presumably via a process involving adjacent t-butyl nitrite molecules. The bands observed for nitrosomethane monomer may be assigned to the funda- mental modes by comparison with the isoelectronic molecule acetaldehyde (table 4). Only one band was observed in the region of the CH3 asymmetric deformations, thus these are both assigned at 1410cm-l.The C-N stretching frequency is nearly 30 cm-I higher than the gas phase value of 842 cm-I reported by LuttkeY6 but this band was observed in the gas phase only as a shoulder on the intense 888 cm-1 band of formaldoxime. The 3059 cm-l band of nitrosomethane monomer may be assigned as 2 x 1549 = 3098 cm-I, and the 1128 cm-l band as 2 x 574 = 1148 crn-'. Only the weak band at 1162 cm-I remains unassigned. Using the structural data given by Coffey et aL8 and reasonable values for the force constants associated with the methyl group, a local syminetry force field treat- ment gave the following values : KN+ 1030 N m-1 Kc-+ 380Nm-l KCNo 160 N m-l. The N-O stretching force constant may be compared with values of 1010 N m-1 reported.by Shurvell et aZ.19 for CF3N0 and 1104 N m-1 reported by Jacox and Milligan2* for HNO (the latter value corresponds to the authors' assignment By which Is supported by recent data from electronic spectra 21). CONCLUSiONS The high yield of cis-t-butyl nitrite compared with that of acetone and nitroso- methane in the photolysis of trans-t-butyl nitrite in the more concentrated matrix samples supports the mechanism of dissociation into an excited (CH3)3C0 radical and nitric oxide, rather than the intramolecular rearrangement mechanism. Either the (CH3)3C0breaks down to give acetone and methyl, which combines with the nitric oxide trapped in the same cage to give nitrosomethane, or the (CH3)3C0 attacks a neighbouring t-butyl nitrite molecule, picking off NO to give cis-or trans- t-butyl nitrite according to the relative orientation of the two molecules.The latter process, favoured by the cage effect at high concentrations of t-butyl nitrite, results in the conversion of the initially predominantly trans-t-butyl nitrite to a mixture of cis and trans isomers. This is a useful method of generating a non-equilibrium mixture of conformational isomers for spectroscopic examination. Nitrosomethane monomer may be conveniently obtained in the low temperature solid phase by trapping a mixture of monomer and dimer from vaporisation of the dimer, followed by photolytic dissociation of the dimer component. We thank I.C.I. Ltd for the award of a Research Fellowship (to A.J. B.) and S.R.C. for financial support. We are grateful to Dr. S. Suzuki for performing the force constant calculations. B. G. Gowenlock and W. Luttke, Quart. Rev., 1958,12, 321. B. G. Gowenlock and J. Trotman, J. Chenz. Soc., 1955,4190. L. Batt, B. G. Gowenlock and J. Trotman, J. Chem. Soc., 1960, 2222. L. Batt, J. K. Brown, B. G. Gowenlock and K. E. Thomas, J. Chem. Soc., 1962, 37. W. Luttke, 2.Elektrocliem., 1957, 61, 976. W. Luttke, 2.Elekfrocliem., 1957, 61, 302. CRYOGENIC PHOTOLYSIS STUDIES R. N. Dixon and H. W. Kroto, Proc. Roy. Soc. A, 1964,283,423. D. Coffey, C. 0.Britt and J. E. Boggs, J. Chem. Phys., 1968, 49, 591. Vibrational Spectroscopy of Trapped Species, ed. H. E. Hallam (Wiley, London, 1973). lo A. J. Barnes and J.D. R. Howells, J.C.S. Faraday ZI, 1973, 69, 532. l1 C. S. Coe and T. F. Doumani, J. Amer. Chem. Soc., 1948, 70, 1516. l2 P. Tarte, Bull. SOC.roy. Sci. Lizge, 1953, 22, 226. l3 G. R. McMillan, J. G. Calvert and S. S. Thomas, J. Plys. Chem., 1964, 68, 116. l4 S. Califano and W. Luttke, 2.phys. Chem., 1956, 6, 83. l5 B. G. Gowenlock, H. Spedding, J. Trotman and D. H. Whiffen, J. Chem. SOC.,1957, 3927. l6 J. F. Brown, J. Amer. Chem. Soc., 1955, 77, 6341. H. W. Brown and G. C. Pimentel, J. Chem. Phys., 1958, 29, 883. P. Tarte, J. Chem. Phys., 1952, 20, 1570. l9 H. F. Shurvell, S. C. Dass and R. D. Gordon, Canad. J. Chem., 1974,52, 3149. 2o M. E. Jacox and D. E. Milligan, J. Mol. Spectr., 1973, 48, 536. 21 P. N. Clough, B. A. Thrush, D. A. Ramsay and J. G. Stamper, Chent.Phys. Letters, 1973,23, 155. (PAPER 4/1366)