The Journal of Physical Chemistry, Vol. 82, No. 1, 1978 7
BrONO, and Its Stratospheric Significance
(24) H. B. Singh, L. J. Salas, J. Shiegeshi, and L. Cavanagh, "Atmospheric Fates of Halogenated Compounds". First year summary report to Environmental ProtectionAgency, Stanford Research Institute, Menlo Park, Calif., 1976.
Agency, Washington State University, 1976; R. Rasmussen, data presented to EPA meeting, Research Triangle Park, N.C., Feb, 1977. (22) V. Ramanathan, Science, 190, 50 (1975). (23) C. E. Junge, Z . Naturforsch. A , 31, 482 (1976).
Bromine Nitrate and I t s Stratospheric Significance John E. Spencer and F. S. Rowland" Department of Chemistry, University of California, Irvine, California 927 17 (Received May 19, 1977) Publication costs assisted by the U.S. Department of Energy
Bromine nitrate, BrON02, has been synthesized and purified, and its ultraviolet and infrared spectra have been measured. The stratospheric photolysis of BrONOz is about twenty times more rapid than that of the analogous ClONO2. However, since HBr is not as important a sink in the Br0,-catalytic chain for removal of stratospheric ozone as HC1 is for the C10, chain, as much an 10-20% of the Br may be present as BrON02. Measurements of the rate of its formation from BrO + NOz + M are necessary for accurate estimate of the stratospheric importance of BrON02.
Introduction The discovery that chlorine atoms released in the stratosphere by solar UV photolysis of chlorofluorocarbon compounds can remove ozone by the C10, catalytic chain's2 (reactions 1 and 2) has focussed attention on the strato-
c1+ 0 , +0
c10
-+
-+
c10 + 0 ,
c1+ 0 ,
spheric reactions of other halogenated species as weL3s4 The reactions of stratospheric bromine atoms have been discussed by Crutzen? Watson: and by Wofsy et al.7 and the BrO, catalytic chain of (3) and (4) is also very effective Br
+ 0,
-+
BrO
+ 0,
BrO t 0 -+ Br t 0,
(3) (4)
in removing ozone. Indeed, Br atoms are less easily removed from the BrO, chain than are C1 atoms from the C10, chain because C1 can be diverted from the chain through formation of HC1 by reaction with CH4, H2, or HOz while Br does not react with CH4 or H2. Since the abstraction reactions from these major sources are endothermic, Br is converted to HBr only by reaction with less abundant stratospheric species such as H02,and HBr plays a less important role for the BrO, chain than HC1 in the C10, chain. Consequently, the injection of large quantities of bromine atoms into the stratosphere would also be expected to result in significant depletion of the natural levels of stratospheric ozone. While methyl bromide is used as a soil fumigant, its C-H bonds permit rapid reaction with tropospheric OH radicals, and only a small fraction of the Br atoms from its decomposition is released in the ~tratosphere.~ On the other hand, some bromofluorocarbon compounds (e.g., CBrF,; CBrF2CBrF2)are extensively used as flame retardants, and will frequently be ultimately released to the atmosphere. Moreover, these compounds are apparently not subject to any tropospheric removal processes, and will closely parallel the atmospheric chemistry of chlorine molecules such as CC13F and CC12F2.s Consequently, all (or a major fraction) of the bromine atoms from these compounds can be expected to be released eventually in the stratosphere. Bromine has recently been detected in the stratosphere 0022-3654/78/2082-0007$0 1 .OO/O
(as Br-) by aircraft- and balloon-borne experiments analogous to those used for HC1 detection for several year~.~JO Chlorine nitrate, C10N02, can be formed in the stratosphere by the reaction of C10 with NO2, and its photolytic lifetime in the lower stratosphere (e30km) is long enough that detectable amounts are expected a t these a1tit~des.ll-l~Since bromine nitrate, BrON02, can be expected from the analogous reaction 5 , we have invesBrO
+ NO, + M
+
BrONO,
+M
(5)
tigated its ultraviolet and visible absorption characteristics, and have used these for estimates of its stratospheric lifetimes at various altitudes. In general bromine nitrate has about a 20-fold greater weighted absorption coefficient than chlorine nitrate in the lower stratosphere, and therefore should have a 20-fold shorter lifetime than its chlorine counterpart. However, the fraction of bromine as BrO should usually be larger than that of chlorine as C10 so that BrON02may play a fractional role in the BrO, chain not much less than that of CIONOz in the C10, chain. The literature on BrON02as a pure compound is sparse. Its synthesis and some of its reactions were first reported by Schmeisser and Taglinger in 1961.15J6 More recently Schack and Christie have studied some of its reactions in the condensed phase,17and Schack also has reported an incomplete infrared spectrum.ls
Experimental Section Bromine nitrate was synthesized by the reaction of BrCl with C10N02.15 Our preparation of chlorine nitrate has been discussed earlier.11J4 Typically, several milliliters of CION02 were mixed with an equal amount of BrCl and the mixture was allowed to remain at -50 to -70 "C for as long as 1week. The unreacted starting materials were then distilled off a t -78 "C, leaving behind a small amount of BrON02. The unreacted mixture was then allowed to continue to react and the process repeated. Sufficient quantities of BrONOz for subsequent experiments were produced by collecting the yields from several batches. Yields were always low and appeared to be little affected by varying reaction conditions. @ 1978 American Chemical Society
The Journal of Physical Chemistry, Vol. 82,No. 1, 1978
8
J. E. Spencer and F. S. Rowland
TABLE I: Vapor Pressures of Bromine Nitrate for Various Temperaturesn Vapor Vapor pressure, pressure, T,K Torr T,K Torr 226 233 238 242 250 a
0.4 1.2 2.3 3.4 6.3
257 261 264 266
TABLE 11: Infrared Absorption Bands of Bromine Nitrate tion,c cm-'
10.4 13.8 16.9 19.3
The observed impurities in BrONOz were Br2, HN03, and NzO5. The reaction of Nz05with BrONOz forms an adduct,16a white solid which remained as a residue during distillation of BrON02 at -20 "C. Nitric acid could not be separated by distillation, but was found to be efficiently removed by condensing impure BrONOz onto dried, thoroughly degassed NaF powder and allowing the two to remain in contact for about 2 h at -25 "C. Under these conditions, H N 0 3 reacted with NaF to form H F and NaN03, which were readily removed from BrONOZ. Molecular Brz was removed by distillation at -78 "C. Handling of BrONOz on the vacuum line usually resulted in some decomposition, as evidenced by the presence of about 5% Brz impurity in the optical spectrum. Since no absorption lines characteristic of NOz were ever observed in the bromine nitrate samples, the decomposition presumably occurred in accord with eq 6. In a UV --f
Br, t N,O, t 1 / , 0 2
Approx descriptiond
3400 w 2978 m 2558 w 1711 vs 1440 w 1285 vs 1110 w 950 w 802 s 722 w 720.5 w (690 w)" 560 s (390)b
The visually observed melting point is 240 K.
2BrON0,
Band locations in CIONO,,e cm-'
Band loca-
Asymmetric NO stretch Symmetric NO stretch NO, scissoring mode NO, out-of-plane bend BrO stretch NO, in-plane bend N-OBr stretch
3447 3009 2573 1735 1424 1292 1119 988 780 714 711 (809, C10) 560 (434, N-OCl)
Described in footnote to ref 18; not observed in this work. Expected from 560- and 95O-cm-' combination band, by analo y with ClONO,; not observed in this work. This work. fReferences 18-20. e Reference 19. I.
I
I
I
I
I
I
(6)
or IR cell BrONOz at pressures of several Torr was stable for several hours, as determined spectrophotometrically. Typically, less than 10% decomposition occurred in 2 h in a cell at room temperature in the dark. The compound was characterized as BrONOz by its of its infrared ~ p e c t r u mand ,~~~~~ ~ y n t h e s i s , 'the ~ , ~features ~ chemically by its reaction with HCI. In the latter experiment, 6 Torr of BrONOz was placed in an infrared cell and its spectrum recorded. Then 1 2 Torr of HC1 was added, and the spectrum recorded again. Quantitative conversion to HON02 was observed, as expected from the known reactions of BrON02.16 Pressure measurements were made with a Model FA-141 Wallace-Tiernan gauge. Ultraviolet spectra were recorded with a Cary 14, and infrared measurements were made with a Perkin-Elmer IR 283.
Results Physical Properties. The vapor pressure of BrONOz was measured over the temperature range 226-266 K, and melting was observed at 240 K. This melting point is in disagreement with the previously reported value of 231 K.15 The visually observed melting point of 240 K is corroborated by a change of slope, indicating a phase change, at this temperature on a semilogarithmic plot of vapor pressure vs. 1/T. The measured temperatures and the respective vapor pressures are listed in Table I. The measured vapor pressures of the liquid phase correspond to the expression log PTorr = 8.898 - 2.025 X 103/T. Application of the Clausius-Clapeyron equation yields a heat of vaporation of -9 kcal/mol and a heat of sublimation of -16 kcal/mol. Since decomposition was often observed in handling, the vapor pressure measurements were made on freshly purified batches. The material entering the gauge was discarded after the measurement, and the process repeated several times, thus ensuring that higher vapor pressure decomposition products were distilled off. This method
IW
180
I
I
220
260
I
300 nm
I
340
I
1
380 420
Figure 1. Ultraviolet absorption spectra for BrON02 (A) and C10N02 (0)in the 186-390-nm range.
yielded vapor pressures of excellent reproducibility; scatter was generally less than 2 or 3%. The source of the discrepancy between these two observations of the melting point is probably attributable to impurities of Brz and/or HON02 in the earlier sample paralleling our observations that these are the common impurities generated in synthesis and handling. The reported elemental a n a l y ~ i s 'of ~ Br:N03 = k1.06 is reasonably satisfactory, but would not be very sensitive if the chief impurities were Br2 and HON02. Infrared Spectrum. The infrared spectrum of BrONOZ is very similar to that of C10NOz,19and the band positions match those expected from the analysis of halogen nitrates by Christie, Schack, and Wilson.20 The major band locations and their descriptions following Christie et al. are listed in Table 11. The BrO stretch reported by Schackls at 690 cm-I was not observed, presumably because of its weak intensity. An attempt to observe the low frequency bands of BrON02 (e.g., the band expected at 390 cm-')
BrONOp and
The Journal of Physical Chemistry, Vol. 82, No. 1, 1978
Its Stratospheric Significance
TABLE 111: Ultraviolet Absorption Cross Sections (186-390nm) of Bromine Nitrate h , nm
390 380 370 360 350 345 340 335 330 325 320 315 310 305 300 295 290 285 280 275 270 265 260 255
a,(I
cm’
h,
250 245 240 235 230 225 220 21 5 210 205 200 195 190 186
2.8(-20) 4.0 4.9 6.2 7.7 8.5 8.7 9.5(-20) 1.0(-19) 1.1 1.2 1.4 1.5 1.8 1.9 2.2 2.4 2.7 2.9 3.1 3.4 3.9 4.8 6.1(-19)
2.8(-20) signifies a = 2.8
nm
X
a,
cmz
7.8(-19) 1.0(-18) 1.3 1.7 1.9 2.1 2.4 2.7 3.2 4.3 7.2(-18) l.O(-17) 1.3 1.5(-17)
cm*.
failed when the compound reacted with CsI windows in the cell. Ultraviolet Spectrum. The ultraviolet spectrum was recorded between 186 and 390 nm, as shown in Figure 1 and Table 111. All values for wavelengths greater than or equal to 340 nm were corrected for the presence of Br2 a t levels of 3 4 % in various Bamples. The magnitude of the Br2 correction was identified by a nonreproducible relative maximum at 420 nm and calculated with the assumption that all absorption between 450 and 500 nm was caused by Br2. Literature values of the cross sections of Br2 were used.21 At the pressures used, absorption by BrONOz in this region should have been negligible based on an extrapolation of shorter wavelength cross sections. The estimated uncertainty in these measurements is &30% for X 1340 nm and &20% for 186 nm < X < 335 nm. The maximum amount of BrON02 available for any measurements of spectra corresponded to about 20 Torr pressure in a 10 cm long cell. Stratospheric Significance. The measured absorption cross sections, extrapolated toward 500 nm, were combined with attenuated solar flux at 15 km22to estimate the rate of photolytic decomposition of BrON02. A value of J = 1.5 X s-l was determined for this altitude from direct absorption only. Since most of the absorption takes place at wavelengths greater than 300 nm, which are not strongly attenuated by the atmosphere, this photolytic rate is almost constant throughout the lower stratosphere. However, molecules with appreciable absorption for h >300 nm are also strongly affected by multiple scattering of this radiation with the result that the effective J value is approximately double the J calculated for direct radiat i ~ n The . ~ ~summed absorption cross sections of BrON02, weighted by the direct UV visible fluxes, are 20 times larger than those of C1ONO2. This 20-fold relative absorptivity of BrON02/C10N02 should be rather insensitive to the corrections for multiple scattering which do not vary sharply with wavelength in the 300-420-nm region. Detailed calculations of the stratospheric importance of BrON02 cannot be made without knowledge of its rate of formation from BrO and NO2 by eq 5. However, an estimate of k5 can be made by analogy with the corre-
9
sponding reaction of C10 with NO2. If k5 = 4 X cm6 molecule-l s-l is estimated for 220 K, corresponding to the extrapolated C10N02rate,%then the partitioning between BrO and BrONOz may be estimated from the steady state expression given in eq 7. Solar photolysis is so rapid that B r O N 0 2 / B r 0 = k[NOzl [MI lJ
(7)
other processes such as reaction with 0 or OH would contribute negligibly to BrON02 removal. For this estimate of k , and an assumption that J with multiple scattering is approximately double the direct J, the BrONOe/BrO ratio is about 0.2 in the 20-30-km altitude range. Since BrO is expected to be the principal form of stratospheric bromine at these altitudes, these ratios indicate that BrONOz could be a significant reservoir (10-20%) for bromine in the lower stratosphere. More accurate estimates will require actual measurement of kb at stratospheric temperatures. We have no direct information concerning the mechanism of the photodecomposition of BrON02. Measurements of the quantum yields from laboratory photolysis of C10N02 are best fitted by the assumption that it decomposes to ClONO plus 0.25The corresponding photodecomposition to BrONO plus 0 is the most reasonable assumption for BrON02 with current knowledge. No information is available about the atmospheric lifetime of BrONO, but there is no reason to expect it to be much longer than that of ClONO which photolyzes in a few minutes in overhead sun at all altitudes, and it may well have even a shorter lifetime. Acknowledgment. we acknowledge a valuable discussion on bromine nitrate with Dr. Carl J. Schack, and thank him for furnishing an unpublished infrared spectrum of BrONOZ. This research was supported by ERDA Contract No. E(04-3)-34, Project Agreement 126.
References and Notes (1) M. J. Moiina and F. S.Rowland, Nature(London), 249, 810 (1974). (2) F. S.Rowland and M. J. Molina, Rev. Geophys. Space Phys., 13, l(1975). (3) “Fluorocarbons and the Environment”, Report of Federal Task Force (4) (5)
(6) (7) (8) (9) (IO)
(11) (12) (13) (14) (15) (16) (17) (18) (19) (20)
on Inadvertent Modification of the Stratosphere (IMOS), US. Council on Environmental Quality, Washington, D.C., June, 1975. “Halocarbons: Effects on Stratospheric Ozone”, Panel on Atmospheric Chemistry, H. S. Gutowsky, Chairman, U S . National Academy of Sciences, Washington, D.C., Sept., 1976. P. Crutzen, Can. J . Chem., 52, 1569 (1974). R. T. Watson in “The Natural Stratosphere of 1974”, CIAP Monograph 1, DOT-TST-75-51, Sept., 1975, pp 5-147-5-152. S.C. Wofsy, M. B.McElroy, and Y. L. Yung, Geophys. Res. Lett., 2, 215 (1975). F. S.Rowland, J. E. Spencer, and C. C. Chou, presented at the 174th National Meeting of the American Chemical Society, Chicago, IiI., Aug., 1977. W. Sedlacek, private communication. A. L. Lazrus, 8.W. Gandrud, R. N. Woodard, and W. A. Sedlacek, J . Geophys. Res., 81, 1067 (1976). F. S.Rowland, J. E. Spencer, and M. J. Molina, J . fhys. Chem., 80, 2711 (1976). F. S.Rowland, J. E. Spencer, and M. J. Molina, J . fhys. Chem., 80, 2713 (1976). D. G. Murcray, A. Goidman, W. J. Williams, F. H. Murcray, F. S. Bonomo, C. M. Bradford, G. R. Cook, P. L. Hanst, and M. J. Molina, Geophys. Res. Lett., 4, 227 (1977). L. T. Molina, J. E. Spencer, and M. J. Molina, Chem. fhys. Lett., 45, 158 (1977). M. Schmeisser and L. Taglinger, Berichte, 94, 1533 (1961). M. Schmeisser and E. Schuster in “Bromine and Its Compounds”, 2 . E. Jolles, Ed., Academic Press, New York, N.Y., 1976, pp 209-213. C. J. Schack and K. 0. Christie, Inorg. Chem., 13, 2378 (1974). C. J. Schack, quoted in C. J. Schack, K. 0. Christie, D. Pilipovich, and R. D. Wislon, Inorg. Chem., IO, 1078 (1971). R. H. Miller, D. L. Bernitt, and I. C. Hisatsune, Spectrochim. Acta, Part A , 23, 223 (1967). K. 0. Christie, C. J. Schack, and R. D. Wilson, Inorg. Chem., 13,
2811, (1974). (21) D. J. Seery and D. Britton, J . fhys. Chem., 68, 2263 (1964).
10
The Journal of Physical Chemistry, Vol. 82, No. 1, 1978
2. Diaz and R. D. Doepker
(22) M. Ackerman in “Mesospheric Models and Related Experiments”, G. Fiocco, Ed., D. Reidel Publishing Co., Dordrecht, Holland, 1971, pp 149-159. (23) D. J. Wuebbles and F. Luther, unpublished calculations. (24) M. T. Leu, C. L. Lin, and W. B. Demore, J . Phys. Chem., 81, 190
(1977). (25) W. S. Smith, C. C. Chou, and F. S. Rowland, Geophys. Res. Lett., submitted for publication; presented at the 173rd National Meeting of the American Chemical Society, New Orleans, La., March, 1977. (26) L. T. Molina and M. J. Moiina, Geophys. Res. Lett., 4, 83 (1977).
Gas-Phase Photolysis of Tetrahydrofuran at 147.0 and 123.6 nm Zalda Dlaz and Richard D. Doepker” Department of Chemistty, University of Miami, Coral Gables, Fiorida 33 124 (Received May 3 1, 1977) Publication costs assisted by the University of Miami
The vacuum-ultraviolet photolysis of tetrahydrofuran, 1:1 mixtures of tetrahydr0furan:tetrahydrofuran-d8, and tetrahydrofuran-3,3,4,4-d4 was investigated using xenon (147.0 nm) and krypton (123.6 nm) resonance radiation. Nitrogen oxide and hydrogen iodide were extensively employed as a means of determining radical yields. Mass spectral analysis of major components contributed to the identification of radical species, intermolecularprocesses, and intramolecular rearrangements. Major product of the photolysis include ethylene, propylene, cyclopropane, carbon monoxide, and hydrogen. Evidence is presented for the occurrence of seven primary reaction channels to which quantum yields have been assigned.
Introduction In contrast to the large volume of work that has been reported on the photochemistry of hydrocarbons, the photochemistry of simple heterocyclic compounds containing oxygen atoms in the ring has received relatively little attention. The only investigation of the vacuumultraviolet photolysis of a cyclic ether that has been reported in the literature is that of ethylene 0xide.l The thermal vapor phase decomposition of tetrahydrofuran (THF) was investigated by Klute and Waltem2 The major products found were ethylene, carbon monoxide, and methane, and acetaldehyde and formaldehyde were identified as intermediates. The photolysis of T H F with a full mercury arc was studied by Roquitte3v4 in the pressure range 10-165 Torr, with a variation of temperature from 30 to 120 “C. The major products were carbon monoxide, hydrogen, methane, ethane, ethylene, propylene, propane, cyclopropane, and formaldehyde. The author suggested the following decomposition scheme THFt
b-+[p]* +
c-C,H, t HCHO
-+
CH,CH,CH, t CO C,H, t CH, t H t CO
--f
(1) (2) (3)
The cyclopropane formed in reaction 1 is vibrationally excited and its yield increases with pressure. This increase is accompanied by a decrease of the yield of propylene with increasing total pressure, which agrees with the observation that thermally5 and chemically produced6,7 excited cyclopropane isomerizes to propylene. This study represents an attempt to examine the molecular decomposition processes of the T H F molecule at 147 and 123.6 nm. Deuterium labeling of the parent compound has been used to explore the molecular rearrangements leading to product formation, and the efficiency of each primary channel has been estimated at both wavelengths.
Experimental Section Materials. T H F (Aldrich), THF-d8 (Merck Sharp and Dohme), and THF-3,3,4,4-d4 (Merck Sharp and Dohme) 0022-3654/78/2082-0010$01 .OO/O
were purified by vapor chromatography using a 6-ft Chromosorb 101 column maintained at 60 “C. The purity of the resulting materials was better than 99.8%. Mass spectrometric analysis revealed the presence of 6.4% C4HD70 in the THF-dg, and 5.8% C4H5D30 in the THF-d4. Purification of other materials has been described elsewhereb8 Irradiation and Analysis. The vacuum-ultraviolet photolysis of T H F was investigated at room temperature in a standard static system using a 600-cm3reaction vessel and a “gettered” xenon (147 nm) or krypton (123.6 nm) resonance lamp.gJO Analysis was performed by vapor chromatography (25-ft Squalane column at 60 “C and 6-ft Chromosorb 101 column at 75 “C) and mass spectrometry (C.E.C. 21-103 C) as reported earlier.g The chemical = actinometers used a t 147 nm were cyclobutene (@czHz 0.61)’O and cyclopentene (@c& = 0.23),11while at 123.6 nm = O.19)I1 was employed. only cyclopentene (@c~H~
Results The quantum yields of the major products of the photolysis of T H F at 147 and 123.6 nm are reported in Tables I and 11, respectively. Hz and CO yields a t both wavelengths are collected in Table 111. In addition to the results presented in these tables the following observations should be reported. (1)A careful search was made for possible formation of HCHO, CH,CHO, furans, and dihydrofurans in this system, but with negative results. However, the presence of small amounts of formaldehyde would not be disclosed by the chromatographic analysis, which is about a factor of 7 more sensitive to methane than it is to formaldehyde. (2) In this study HI has been found to be a more effective radical interceptor than H2S, as suggested by Ausloos.l* Increasing the concentration of HI from 3 to 15% revealed that the optimum concentration of HI for radical interception in this system is 5%. This corresponds to the maximum in the radical yield vs. HI concentration curve. When T H F was photolyzed in the presence of 5% HI the quantum yields of CH4, CzH4, and C3Hsincreased, indicating the presence of CH3 (or CHz), CzH3, and C3H6 0 1978 American Chemical Society