5162
J . Phys. Chem. 1989, 93, 5162-5165
Reaction of CH,OH with O,, NO, and NO, at Room Temperature Palle Pagsberg, Jette Munk, Department of Chemistry, Riso National Laboratory, DK-4000 Roskilde, Denmark
Christopher Anastasi,* and Victoria J. Simpson Department of Chemistry, University of York, Heslington. York. YO1 SDD, England (Received: September 20, 1988: In Final Form: February 8, 1989)
The reaction of CH,OH with 02,NO, and NO2 has been studied using pulse radiolysis to generate the radicals and ultraviolet absorption to observe the kinetics. Rate constant values of (0.88 f 0.02) X lo-", (2.5 f 0.02) X IO-", and (2.3 f 0.4) X lo-" cm3 molecule-l s-I have been measured at room temperature and 1 atm pressure for the 02, NO, and NO2 reactions, respectively. Absorptions due to long-lived or stable products were observed in the same spectral region. A simple analysis of these observations suggests that formation of an adduct may dominate in the reaction of CHzOH with NO and NO2 but that this process accounts for only a minor route in the 0, reaction.
Introduction Hydroxyalkyl radicals are important intermediates in the oxidation processes that occur in both combustion' and atmospheric c h e m i ~ t r y . ~ ,Despite ~ this, there have been very few studies involving this class of intermediates; this is in contrast to the many studies involving the isomeric species, the alkoxy The simplest member of the hydroxyalkyl class of intermediates is the hydroxymethyl radical, C H 2 0 H . Previous studies involving this species have concentrated on the mutual reaction and its Radford' used laser magnetic resonance reaction with 02.7-11 (LMR) to monitor the formation of H 0 , in the reaction CHzOH
+0 2
-+
CH2O
+ HO2
(la)
and obtained a rate constant of 2 X cm3 molecule-' for k I a a; similar value of 1.4 X cm3 molecule-' s-I was obtained by Wang et aL8 using photofragment emission (PFE) detection of the product H 0 2 . More recently, Grotheer et aL9 measured cm3 molecule-' s-I using a much higher value of k , , = 9.5 X a discharge-flow apparatus with mass spectroscopic detection of C H 2 0 H ; higher rate constants values of k , , = 10.5 X and cm3 molecule-' s-I were also measured by Dobe et a1.I0 8.6 X using LMR and Payne et al." using mass spectrometry, respectively. Clearly there is still considerable uncertainty in the value of this rate constant. Very recently, the ultraviolet absorption spectrum of C H 2 0 H has been reported by us,I2,I3and several bimolecular reactions involving this species have been measured.', The present work uses this spectrum to monitor the kinetic behavior of the radical in a study of its reaction with 0 2 NO, , and NOz at room temperature and I atm pressure.
described previ0us1y.l~ Briefly, argon/SF6/CH30H gas mixtures are made up in a 1-L stainless steel reaction vessel before irradiation with a 30-11s pulse of 2-MeV electrons from a field emission accelerator (Febetron 705B). The hydroxymethyl radicals are generated in the following way: Ar Ar* F
-- ++
+ 2-MeV e-
+ SF,
Ar
+ CH30H
-
+ eSF, + F
Ar*
HF
CH2OH
(2)
(3) (4a)
HF + CH30 (4b) where k4a/k4b= 1.27.13 Typical experimental conditions were 2.4 X IO1, molecule cm-) C H 3 0 H and 1.2 X lo1*molecule SF, in 2 X IOi9 molecule cm-3 argon; this mixture yields 8 X loi4 molecule F atoms per electron pulse. The absorption by the C H 2 0 H intermediates at a selected wavelength (285.5 nmlZ)is monitored by use of a fast photomultiplier (Hamanatsu) at the output of a monochromator (Hilger and Watts); the signal is stored on a transient digitizer (Biomation 8 100) and then transferred to computer for further analysis. When 02,NO, and N O z are added to the A r / S F 6 / C H 3 0 H mixture, C H 2 0 H decays by the reactions CH2OH
CH20H
CH,OH
Experimental Section The pulse radiolysis/kinetic absorption apparatus has been
+0 2
+ NO + NO2
+
-
CH20
+ H02
(la)
CHzO
+ HNO
(5a)
C H 2 0 + HNOz
-
(6a)
(M)
( I ) Westbrwk, C . K.; Dryer, F. L. Combust. Sci. Technol. 1979, 20, 125. (2) Niki, H.; Maker, P. D.;Savage, C. M.; Breitenbach, L. P. J . Phys. Chem. 1978, 82, 135. (3) Atkinson, R. Chem. Reu. 1985, 85, 69. (4) Wendt, H . R.; Hunziker, H. E. J . Chem. Phys. 1979, 71(12), 5202. ( 5 ) Gutman, D.; Sanders, N.; Butler, J. E. J . Phys. Chem. 1982, 86, 66. (6) Morabito, P.; Heiklen, J. J . Phys. Chem. 1985, 89, 2914 and references
therein. ( 7 ) Radford, H. E. Chem. Phys. Lett. 1980, 71(2), 195. (8) Wang, W. C.; Suto, M.; Lee, L. C. J . Phys. Chem. 1984, 81, 3122. (9) Grotheer, H. H.; Riekert, G.; Meier, U.; Just. T. Ber. Bunsen-Ces. Phys. Chem. 1985, 89, 187. (IO) Dobe, S.;Temps, F.; Bohland, T.;Wagner, H . Gg., Z . Nuturforsch. 1985. 40a, 1289. ( I I ) Payne, W. A.: Brunning, J.: Mitchell, M .B.; Stief. L. J. fnt. J . Chem. Kine?. 1988, 20, 63. (12) Pagsberg. P.; Munk, J.; Sillesen, A,; Anastasi, C. Chem. Phys. Lett. 1988, 146(5), 375. ( I 3 ) Rettrup. S.; Pagsberg, P.; Anastasi, C. Chem. Phys. 1988. 122, 45
0022-3654/89/2093-5l62$0l.50/0
HOCH20N0
(6~)
The concentrations of the reactants, C H 3 0 H , and SF6 were small enough to ensure direct irradiation of these species did not occur. Also, the use of excess methanol over the reactants Oz, NO, and NO, and a fast rate constant for reaction 4 ensured the reaction between F atom with reactants was negligibly small. SFS does not affect the kinetics of C H 2 0 H over the time lengths used in these experiments (20 b s ) .
Materials High-purity, commercially available Ar, SF,, NO, NO2, and 0, (Matheson) gases were used directly from cylinders without (14) Pagsberg, P. B.; Eriksen, J.; Christensen, H. C. J . Phys. Chem. 1979,
83. 582.
C 1989 American Chemical Society
The Journal of Physical Chemistry, Vol. 93, No. 13, 1989 5163
Reaction of C H 2 0 H with 02,NO, and NO2 A may,
1
TABLE I: Chemical Mechanisms Used To Model the Reaction of CHIOH with 0,.NO. and NO, rate const, cm3
reaction molecule-I s-l b CH20H + 0 2 CH2O + H02 1.6 X CH30 + O2 CH20 + H 0 2 CH20H + CHzOH CH3OH + CH2O 8.6 X lo-” 4.2 X 10‘” CH30 + CH30 CH30H + CH20 CH2OH + CHjO CH30H + CH2O 1.1 x 2.5 X HO2 + HO2 H202 + 0 2 HO2 + CH2OH CH3OH + 02 2.9 X IO-” 1.8 X lo-” H 0 2 + CH30 CH30H + O2 CHzOH + NO HOCH2NO b 3.0 X IO-” CH30 + NO CH30N0 CH2OH + CH2OH CH3OH + CH2O 8.6 X lo-” 4.2 X IO-” CH30 + CH30 CH30H + CH20 1.1 x IO-” CH2OH + CHjO CH3OH + CH20 4.0 X HNO + HNO products HNO + CH20H products 1.2 x 7.2 X HNO + CH30 products CHIOH + NO2 HOCH2N02 b CH3O + NO2 CH3ON02 1.5 X lo-” CH20H + CH2OH CHjOH + CH2O 8.6 X IO-” 4.2 X lo-” CHjO + CH3O CH3OH + CH20 1.1 x CHzOH + CH30 CH30H + CH20
- -- -----
~
-+
to
Tlme
+lQ
Figure 1. Experimental decay trace following radiolysis of a mixture containing [02]= 4.9 X 10I6 molecules cm-), [CH30H] = 2.5 X 10’’ molecules c ~ n - [SF6] ~ , = 1.2 X IO’* molecules and [Ar] 2.5 X 1019 molecules ~ m - T~ = ; 298 K; time span = 2.0 X 1O-j s. 1
I
-+
14
ref this work 16, 17 13 18
a 19
a a
this work 20 13 18
a 21
a a
this work 20 13 18 a
In the absence of any reported measurements, these rate constants were calculated by using the cross-combination rule.22 *Rate constant varied until the predicted half-life matches the experimentally derived value.
a bimolecular rate constant k l = 8.9 (f0.3)
X
cm3 molecule-I
S-‘.
2
4 6 Concentration/ x 10‘6molec-’ c m 3
“0
Figure 2. Plot of the reciprocal CH20Hdecay half-life as a function of concentration of added reactant; (A)02,( 0 )NO, and (0)N02. T = 298 K.
further purification. Methanol (Merck, 99.8% pure; inert impurities) was thoroughly degassed before use.
Results I . Reaction Kinetics. Figure 1 shows the change in absorption due to the fast formation and subsequent decay of C H 2 0 H by reaction with 02,measured at 285.5 nm and 298 K over a 2 0 - ~ s time scale. Although the concentration of O2 is high (4.9 X I O i 6 molecules ~ m - ~there ) , is also a bimolecular contribution to the decay due to the fast mutual reaction (7):12
(MI
2CH2OH (CHIOH), C H 3 0 H + CHzO -+
(7a) (7b)
A series of experiments were carried out where the O2 concentration in the gas mixture was varied ((1.2-7.4) X 10l6 molecules clearly as this concentration is reduced the bimolecular contribution due to reaction 7 is increased. Two methods of analysis were used to extract the required rate constant. In the first, a simple analytical expression is used that describes the half-life ( t ) of the experimental decay:15 I/t = l/t,
+ kI[O2]/In 2
(1)
where to refers to the half-life of reaction 7, which controls the decay of C H 2 0 H in the absence of 02,k , = kla + klb,and square brackets denote concentration. A residual absorption is observed at long times which decreases the apparent rate of decay of CH,OH. The true half-life is taken as half the time required for [CH,OH],,, to fall to this residual absorption (see Figure I ) . Figure 2 shows a plot of i / t versus [02] to be a straight line with gradient k , / l n 2; a least-squares analysis of these results yields (15) Bjarnov, E.; Munk, J.; Nielsen, 0. J.; Pagsberg, P.; Sillesen, A. Riso-M-2366, April 1983.
A computer-based analysis has also been used to extract the kinetic information. Here, a chemical model in which all the reactions that are thought to play a role in this system Table I is used to generate decay curves for C H 2 0 H . k l is altered until the predicted half-life is the same as that measured experimentally at the O2concentration; a mean of these values is 8.6 (f0.8) X cm3 molecule-I SKI,in excellent agreement with that derived in the simple analysis. When the simple expression I is used to analyze the decay curves for the reaction of C H 2 0 H with NO and NO2,rate constant values of 2.5 (f0.2) X IO-” and 2.2 (f0.4) X lo-’’ cm3 molecule-’s-I are derived respectively; Figure 2 again shows the NO and NO2 dependence on 1/ t . The full chemical models outlined in Table I yield 2.5 ( f 0 . 2 ) X and 2.4 ( f 0 . 2 ) X IO-” cm3 molecule-’ s-l for the NO and NO2 reactions respectively; once again the two methods of analysis are in excellent agreement. A sensitivity analysis was used to quantify possible errors due to assumptions made in the models. All rate constants for radical-radical reactions not available in the literature were calculated according to the cross-combination rule;22the cross-combination number of 2 in this equation was varied by *IO%. In addition, the estimated rate constants for the reaction of methoxy radicals with NO and NO2 were varied by f30%. In both these cases the effect of the variation was negligible at high concentrations of NO and NO2,increasing to a maximum of 8% for the NO system at the lowest concentration used in our experiments. An area of uncertainty lies in the products of the reaction of C H 2 0 H with 02,NO, and NO2 and their contribution to the (16) Lorenz, K.; Rhaza, D.; Zellner, R.; Fritz, B. Ber. Bunsen-Ges. Phys. Chem. 1985,89, 341. (17) Sanders, N.; Butler, J. E.; Pasternack, L. R.; McDonald, J. R. Chem. Phys. 1980, 48, 203. (18) Batt, L. Int. Rev. Phys. Chem. 1987, 6(1), 5 3 . (19) Kircher, C. C.; Sander, S. P. J . Phys. Chem. 1984, 88, 2082. (20) Finlayson-Pitts, B. J.; Pitts, J. N . Atmospheric Chemistry; Wiley: New York, 1986. (21) Baulch, D. L.; Drysdale, D. D.; Horne, D. G. Eualuated Kinetic Data for High Temperature Reactions, Vol. 2: Homogeneous Gas Phase Reactions of the H2-N2-02 System; Butterworths: London, 1973. (22) Anastasi, C.; Arthur, N. L. J . Chem. SOC.,Faraday Trans. 2 1987, 83, 277.
5164
The Journal of Physical Chemistry, Vol. 93, No. 13, 1989
TABLE 11: Comoarison of Rate Constant Data’ system rate comparison (ref) 2.0 x 10-12 (7) CH2OH + 0 2 1.4 X (8)
9.5 x 10.5 X
+ NO C H 2 0 H + NO2 CH20H
8.6 X 8.8 X 2.2 X
10-12
(9) (10) (1 I )
(this work) (23)
2.5 X 10-I‘ (this work) 8.3 X (23) 2.3 X IO-” this work)
Pagsberg et al.
technique DF/LMR DF/PE DF/MS DF/LMR DF/MS PR/KA DF/MS PR/KA DF/MS PR/KA
;1\ 30
‘Units of cm3 molecule-’ s-l. DF, discharge flow; L M R , laser magnetic resonance; PFE, photofragment emission; MS, mass spectrometry; PR, pulse radiolysis, KA, kinetic absorption. I
I
0
2 20
240
260 Wavelength/ nm
2 80
300
Figure 4. UV spectrum obtained 4 X lo4 s after radiolysis of a mixture containing [ N O ] = 2.0 X 10l6molecules ~ m - [~C,H 3 0 H ] = 2.5 X 10’’ molecules c m 3 , [SF,] = 1.2 X 10l8 molecules and [Ar] to 2.5 X lOI9 molecules T = 298 K. (A correction has been made to remove absorption due to CH,ONO.)
Wavelength / nm
Experimental “prompt” absorption signals measured following radiolysis of [ 0 2=] 1.2 X lo’’ molecules [CH,OH] = 1.2 X lo1’ molecules [SF,] = 1.2 X 10” mokcules cm-3, and [Ar] to 2.5 X I O i 9 molecules The dashed line represents the expected HO, absorption assuming reaction la is the only process removing C H 2 0 H in the oxygen system (see text). Figure 3.
C H 2 0 H decays. In the O2 reaction either H 0 2 or HOCH202 may be formed by abstraction or addition, and assuming these species are of similar reactivity, no correction needs to be made. The same simplification is possible for the N O 2 system where one of the two stable molecules may be formed, H N 0 2 or HOCH2NO2. However, the N O system may yield either H O C H 2 N 0 or H N O and the latter species may continue to react with C H 2 0 H or CH,O. The rate constant for the reaction of C H 2 0 H with NO quoted in the summary Table I1 assumes 100% adduct formation; the rate constant value falls by only 12% if 100% abstraction is assumed. Clearly, the models outlined in Table I are a good representation of the chemistry involved in the three systems. We suggest that, for each reaction, a mean of the two values derived is used and the larger of the error limits is applied; this leads to rate constant values of 0.88 (f0.08) X lo-”, 2.5 (f0.2) X lo-”, and 2.3 (f0.4) X lo-” cm3 molecule-’ s-I for the reaction of CH,OH with 02, NO, and NO,, respectively. 2. Reaction Pathways. The long-lived or stable products of the reactions absorb in the ultraviolet region, and the nature of the spectra may be used to try to elucidate the reaction pathways involved. The background absorption was measured at wavelengths between 220 and 300 nm over time scales up to 4 X S.
One feature of all three systems is the formation of C H 3 0 by initial attack of F atoms on methanol; the possible products formed by this species must be considered in the spectral analysis. The reaction of SF, with 02,NO, and N O 2 to form stabilized S F 5 0 2 , SF,NO, and S F , N 0 2 species, respectively, may also complicate the spectral analysis of the reaction products. However, Pagsberg et al.24have recently shown that SFSO, does not absorb in the UV region considered in this study; it is also unlikely, then, that S F 5 N 0 and S F s N 0 2contribute to the absorption spectra observed. (23) Nesbitt, F. L.; Payne. W. A,; Stief, L. J. Private communication, 1988. (24) Pagsberg. P.: Ratajczak. E.: Sillesen, A,; Jodkowski. J. T. Chem. Phys. Lett. 1987, !4/(1,2), 88.
( a ) Reaction of CH20H with 4. At high oxygen concentration (1.2 X lo” molecules the lifetime of CH20Hdue to reaction 1 is very short (1 11s). The “prompt” absorption signal observed under these conditions at different wavelengths is shown in Figure 3. The profile is very similar to that expected for H0225and is consistent with abstraction of a hydrogen atom from the hydroxymethyl radical. The methoxy radical is known to react with oxygen only by abstraction to form H02:16J7 CH30
+0 2
---*
CH2O
+ HO2
(8)
but the rate constant for reaction 8 is low and H 0 2 absorption due to this process does not contribute to the “prompt” signal. The H 0 2 absorption profile may be calculated by using the known concentration of C H 2 0 H ,the path length, and the absorption cross section of H02;,, if this profile is then compared to the experimental spectrum (normalized to the measured value a t 220 nm), differences are evident at long wavelengths (Figure 3). It may be that this difference is due to a stabilized peroxy radical: CH2OH
+0 2
(MI
HOCH202
(Ib)
The possibility of a stabilized H O C H 2 0 2radical was first suggested by Niki,2 who proposed this species as a route to HCOOH product observed in his studies on the photooxidation of ethene. More recently, Veyret et a1.26have reported an ultraviolet spectrum for H O C H 2 0 2in their studies on the photooxidation of C H 2 0 ; their spectrum is broad and featureless and has an absorption cross cm2 molecule-’ at the peak wavelength of section of 3.7 X 235 nm. ( b ) Reaction of C H 2 0 H with NO. In both the N O and NO2 systems the residual absorption a t long times (4 X IO4 s) and high reactant concentrations was used to try and identify possible reaction pathways. Once again the methoxy radical must also be considered; a t 300 K and atmospheric pressure, the reaction of the methoxy radical with N O proceeds essentially by addition: CH30 + N O
(MI
CH30NO
(9)
The contribution of CH,ONO to the absorption can be calculated by using the known absorption cross section^;^' when this contribution is removed, the residual spectrum shown in Figure 4 is obtained and is attributed to the products from the reaction of ( 2 5 ) Chemical Kinetic and Photochemical Data for use in Stratospheric Modelling. Eualuarion No. 8; JPL Publication 87/41; Jet Propulsion Laboratory: Pasadena, CA, September 1987. (26) Veyret, B:; Llesclaux, R.; Cox, R. A,; Moortgat, G. K . Presented at the 13th International Conference on Photochemistry, Budapest, Hungary, August 1987. (27) Calvert, J . G.: Pitts, J. N. Photochemistry; Wiley: New York, 1966.
"he Journal of Physical Chemistry, Vol. 93, No. 13, 1989 5165
Reaction of C H 2 0 H with 02,NO, and NO2
Nitrous acid absorbs in this region,29 peaking at 220 nm but whereas the absorption cross section values are similar to those that can be derived from Figure 5, the spectral shape is different. By comparison, the ultraviolet spectrum of CH3N0:7 has a profile nearly identical with Figure 5 although the absorption cross sections are lower by a factor of 10; it is possible the residual absorption is due to the adducts formed in reactions 6b and 6c and that the large differences in cross section may be due to replacement of a hydrogen atom by the hydroxy group.
00 2LO
260 280 Wavelength nm
300
Figure 5. U V spectrum obtained 4 X lo4 s after radiolysis of a mixture containing [NO2] = 2.0 X 10I6 molecules cm-), [CH30H] = 2.5 X IO1' molecules cm-), [SFJ = 1.2 X 10l6molecules ~ m - + ~ [Ar] , to 2.5 X lOI9 molecules cm-). T = 298 K. The absorption due to C H 3 0 N 0 2has been removed.
C H 2 0 H with NO. This reaction may proceed by addition or abstraction: HOCH2
+ NO
+ (M)
HOCHZNO
H N O CH2O H N O will decay by mutual reaction: HNO HNO products -+
+
(sa) (5b) (10)
The rate constant for reaction 10 is 4 X cm3 molecule-l s-I 21 and means that H N O has a lifetime of 0.43 s in this system; Le., it is essentially a stable species on the time scale of our experiment. Callear and Wood2*assigned a narrow band a t 207 nm to H N O which is a long way from the absorption at 220 nm observed in this study; it is possible that the residual spectrum shown in Figure 4 is due to a stabilized H O C H 2 N 0 adduct. (c) Reaction of CH20H with NO,. Assuming the C H 3 0 radical reacts with N O z to form an adduct: CH30
-
+ NG
(MI
CH30N02
(1 1)
the contribution due to this nitrate can be removed from the observed absorption, giving the residual spectrum shown in Figure 5. The C H 2 0 H radical can react with NO2 in the following ways: C H 2 0 H NO2 HONO + CH20 (6a)
+
(M)
HOCH2NO2
(6b)
HOCH2ONO
(6c)
(M)
(28) Callear, A. B.; Wood, P.M. Trans. Faraday SOC.1971, 67, 3399.
Discussion Table I1 compares the rate constant values for the reaction of hydroxymethyl radicals with 02,NO, and N O 2 obtained in this study with previously reported values. The reaction of C H 2 0 H with O2 has been studied before, and the rate constants summarized in Table I1 vary by nearly an order of magnitude. The first two measurements both monitored the formation of the reaction product H 0 2 , with Radford7 using the LMR technique and Wang et a1.* PFE. The two methods appeared to agree, giving and 1.4 X cm3 molecule-' rate constants of kla = 2.0 X s-l, respectively. More recent studies, however, have monitored the CHzOH radical directly. Grotheer et aL9 and Payne et al." both employed mass spectrometric detection, and these studies yielded higher cm3 molecule-' s-', and 8.6 X values for kla = 9.5 X respectively. In addition, Dobe et a1.I0 used LMR to follow the change in concentration of both C H 2 0 H and H 0 2 independently; both methods were in good agreement, giving k , , = 10.5 X cm3 molecule-' s-l. The large discrepancy between the rate constants summarized above has been commented on previo~sly.~In all cases the experiments were performed using flow-tube apparatus and the long total residence times involved increases the significance of secondary reactions; in addition, wall losses must be accounted for. Grotheer analyzed the effect of possible decay reactions of C H 2 0 H by means of computer simulations and was able to account for the majority of the difference between his data and that of previously reported data. All subsequent workers have shown that secondary reactions of C H 2 0 H were unimportant under their experimental conditions. Our results for reaction 4 are in very good agreement with the higher, more recent values reported. Very recently, Stief et al?3 have carried out low-pressure studies on the reaction of C H 2 0 H with NO and NO2. The rate constant values are shown in Table I1 and are considerably lower than those obtained in this work; the difference is probably due to the effect of diluent pressure on stabilizing the adducts. Work is currently under way to investigate the present effect on the rate constants for these reactions. Registry No. CH20H, 2597-43-5; CH,OH, 67-56-1; 02, 7782-44-7; NO, 10102-43-9; N02, 10102-44-0. (29) Okabe, H.Photochemistry of Small Molecules; Wiley: New York, 1978.