Environ. Sci. Technol. 1908, 22, 1447-1453
(35) Schuetzle, D.,unpublished results, 1986.
(33) West, W. R.; Lee, M. L. HRC CC, J. High Resolut. Chro-
matogr. Chromatogr. Commun. 1986, 9, 161-167. (34) Henderson, T.R.; Sun, J. D.; Royer, R. E.; Clark, Ch. R.; Li, A. P.; Harvey, T. M.; Hunt, D. H.; Fulford, J. E.; Lovette, A. M.; Davidson, W. R. Environ. Sci. Technol. 1983, 17,
443-449.
Received for review December 21,1987. Accepted May 16,1988. This work was supported by the Coordinating Research Council, Project No. CAPE-30.
Organic Photochemistry. 20. A Method for Estimating Gas-Phase Rate Constants for Reactions of Hydroxyl Radicals with Organic Compounds from Their Relative Rates of Reaction with Hydrogen Peroxide under Photolysis in 1,1,2-Trichlorotrifluoroethane Solutiont Wendell L. Dllllng," Stanley J. Gonslor, Glenn U. Boggs, and Celia G. Mendoza
Analytical and Environmental Chemical Research Laboratory, The Dow Chemical Company, Midland, Michigan 48674 The reaction with hydroxyl radicals appears to be the major transformation route for many organic compounds in the atmosphere. To avoid the difficulties of measuring the rate constants for these reactions in the gas phase for some compounds, e.g., those with very low vapor pressures, we have developed a solution-phase system for measuring relative rates. This system involves photolysis of continuously extracted 90% hydrogen peroxide into 1,1,2-trichlorotrifluoroethane solution, which contains two or more organic compounds, one of which serves as a reference standard whose gas-phase rate constant is known. Reasonable correlations (1.2 = 0.84,0.87) were obtained between the relative solution-phaserates and the absolute gas-phase rate constants, which varied over 4 orders of magnitude for n-hexane, 2,2,4-trimethylpentane, cyclohexane, l,l,ltrichloroethane, cyclohexene, trichloroethene, tetrachloroethene, ethyl acetate, toluene, n-propylbenzene, o-xylene, p-isopropyltoluene, 1,2,4-trimethylbenzene, 1,3,5-trimethylbenzene, biphenyl, naphthalene, hexafluorobenzene, o-dichlorobenzene, m-dichlorobenzene, 1,2,4-trichlorobenzene, anisole, and nitrobenzene. This method allows estimations of the gas-phase rate constants to within 1 order of magnitude or better. Introduction (2) Numerous organic compounds, such as DDT, PCBs, other chlorinated hydrocarbons, polychlorodibenzodioxins, polycyclic aromatic hydrocarbons, organic acids, etc., with vapor pressures in the 10-12-10-5 Torr range at ambient temperatures have been detected in the troposphere or have been suspected to be present there (3-6). Little information is available on the chemical reactivity of these compounds under tropospheric conditions (7) because of difficulties associated with laboratory gas-phase studies on such compounds (8). The major chemical reaction for many of these compounds in the troposphere appears to be with hydroxyl radicals (HO') (9-11). Knowing the rate constants of these reactions would allow calculation of the transformation rates of these compounds in the atmosphere. To avoid the difficulties of measuring these rate constants in the gas phase for many of these compounds, various workers have suggested calculations or solutionphase rate measurements as alternative methods for est For Part 19 of this series, see ref 1.
* Current address: Central Research-Organic Chemicals and Polymers Laboratory, The Dow Chemical Co., Midland, MI 48674. 0013-936X/88/0922-1447$01.50/0
timating these rate constants. Presently, calculations employing group or substituent constants ( 8 , I I ) are applicable to only a limited number of types of organic compounds. At least two groups (12,13) have reported correlations between gas- and aqueous solution-phase rate constants for the reaction of HO' with a series of organic compounds. Reasonable correlations were obtained that would allow prediction, within ca. 1 order of magnitude for most compounds, of the gas-phase rate constant from the measured solution-phase rate constant. However, water is not a satisfactory solvent for many compounds of interest, such as those noted above, because of their low solubility. We have developed a system for measuring relative rate constants for the reactions of organic compounds with H20z under photolysis in 1,1,2-trichlorotrifluoroethane (CF2C1CFC12,F-113) solution, a satisfactory medium for many compounds of interest. A related method using the same solvent, but not based on relative rate measurements, was reported recently (13). Experimental Section Equipment. Photolyses were performed in a two-compartment reactor shown in Figure 1. The lamp was a standard Hanovia 450-W medium-pressure mercury arc lamp. The photoreactor was connected to the extractor by two 6-mm-i.d. glass tubes, each of which had two 90" bends (as viewed from the top). The extraction of H202 into the F-113 solution and the pumping action were achieved by magnetically stirring the solution in the extractor with a Star Head Teflon-coated magnetic stirring bar. The outlet and inlet tubes on the extractor were offset so that the stirring action forced the solution out the lower tube and back into the extractor from the upper tube. Stirring was maintained at the highest allowable rate that did not form droplets of 90% H202that would be swept out of the extractor. This resulted in a flow rate of - 5 cm38. The extractor and connecting tubes were covered with black tape to minimize light penetration into the 90% aqueous HzOzlayer. The entire apparatus was immersed up to the reaction solution level in a refrigerated constant-temperature bath maintained at 25.0 f 0.1 "C. Materials. HzOzsolutions (go%,FMC Corp., and 30%, VWR Scientific Chemical Reagent) were used as received. Assays by the permanganate method (14) gave values of 90.47 f 0.19% and 29.91 f 0.11%, respectively. All organic test compounds were commercial samples and were used as received. Gas chromatographic (GC)
0 1988 American Chemlcal Society
Environ. Sci. Technol., Vol. 22, No. 12, 1988
1447
450 Watt Medium Pressure Hg Arc Lamp
portion (-5 mL) was retained as a GC standard, and the remainder (-545 mL) was placed in the photoreactor. Hz02(go%, 5 mL) was added to the extractor portion of the photoreactor, and 0.5 mL of 90% H202was added to the dark reaction flask. After the mixtures were stirred for 2 h to saturate the F-113solutions with H202,the lamp was turned on. Zero time was taken as the time when the lamp intensity increased to its maximum. Aliquots (-2 mL) were periodically removed from both reaction mixtures for GC analyses. Another run in the photoreactor was performed in the same manner as described above except that no H202was added. Analyses. GC analyses for the kinetic runs were performed with a Hewlett-Packard 5730A gas chromatograph and 3390 integrator. Samples (1.0 pL) of the F-113 reaction solution were injected directly (port temperature 200 "C) onto a 30 m X 0.32 mm fused silica, bonded SPB-5, capillary column held at 0 "C for 4 rnin and then heated at 16 "C m i d to 150 "C. The He carrier gas flow rate was 1cm3 min-l, and compounds were detected with a Hz flame ionization detector (FID, block temperature 300 "C). Retention times ranged from 4.4 rnin for (Figure 2) to 14.9 min for Ph,. Standard solutions in F-113 containing all the substrates (for a given experiment) and C14F24 were run several times each day to calibrate the detector response. The detector response was linear over the concentration ranges used and was independent of dissolved H202. GC analyses for reaction products from C6H12 in CC14 solution were performed on a 10 f t X 0.25 in. glass column packed with 100-120 mesh ULTRA-BOND PEGS operated at 70 "C for 2 min and then heated at 32 "C min-' to 180 "C (He carrier gas 25 cm3 mi&, FID). Analyses for traces of CHCl, in these reactions were performed on a 9 ft X 2 mm glass column packed with 3% Carbowax 20M and 0.5% H3P04on 60-80 mesh Carbopack B operated at 100 "C (He carrier gas -30 cms min-', FID). GC analyses for reaction products from C6H12in F-113 solution were performed on the following columns: (1)15 m X 0.32 mm fused silica, bonded Durawax, operated at 20 "C for 4 min and then heated at 16 "C m i d to 180 OC (N2carrier gas inlet pressure 6.5 psi, FID); (2) 10 m X 0.53 mm fused silica, 1.2 pm bonded Carbowax, operated at 20 "C for 4 min and then heated at 16 "C m i d to 180 "C (He carrier gas 4 cm3 min-', FID). GC-mass spectrometric (electron impact ionization) analyses on these reaction solutions were performed by use of a 4610 Finnigan GCMS, 15 m X 0.32 mm fused silica capillary column with 0.5 pm DB-WAX operated at 45 "C for 5 rnin and then heated at 15 "C min-l to 185 "C.
-
EXTRACTOR
PHOTOREACTOR
Flgure 1. Two-compartment photoreactor.
analyses (see below) indicated some samples contained minor amounts of impurities. These were inconsequential because the rates at which the compounds disappeared were based on the individual GC peak areas. The inert internal standard, perfluoroperhydrophenanthrene (C14F24, SCM Specialty Chemicals) was a mixture of at least three major compounds (probably isomers, area ratio 0.72:1.00:0.19 in order of elution, 93% of total area) and three minor compounds. All of the components were completely stable under the photolysis conditions in the presence or absence of H202. The combiged GC peak areas of the three major components were used as the GC internal standard. F-113 (Aldrich spectrophotometric grade Gold Label) was used without further treatment. GC analysis showed only traces of impurities having retention times shorter than that of F-113. Determinations of Maximum Concentrations of H20zin Organic Solvents Equilibrated with Aqueous H202Solutions. These concentrations in F-113, CFC13, CC14,and perfluoro-2-butyltetrahydrofuran(C4F70C4F~) were determined by stirring 5-10 mL of either 30% or 90% H202with 400 mL of the organic solvents at 22 f 2 "C and periodically removing N 25-mL aliquots of the organic layers for titration. The 30% HzOZmixtures were stirred for -20 h. The 90% HzOzmixtures required -1-4 h to saturate the organic layer with H202. To the weighed aliquots were added 25 mL of water, purified via a Milli-Q system, and 15 mL of 1:4 HzS04. These solutions were titrated with standardized aqueous 0.01 N KMn04 (14). Kinetic Runs. In a typical procedure 600 mL of F-113 was aerated by stirring it for 2 h in a flask open to the air. Test compounds and Cl4FUwere dissolved in this aerated F-113 to an extent of 1.0 mM each. A portion (50 mL) of this solution was placed in a 125-mL flask that had a ground-glass stopper and stirred magnetically in a dark constant-temperature bath at 25.0 f 0.1 "C. Another 1448
Environ. Sci. Technol., Vol. 22, No. 12, 1988
Results Because photolysis of H202appeared to be the most simple method for generating HO' in solution (15,16),we determined the maximum concentrations of H202 achievable in several perhalogenated organic solvents, which were not expected to react with HO' (111, by equilibrating the organic solvent with an excess of either 30% or 90% aqueous H202(Table I). The reason for the discrepancy in the concentration of H202in F-113 when equilibrated with 30% Hz02between our work and that reported in ref 13 is not known. Different analytical methods for H202were used. Suspension of tiny 30% H202droplets in the F-113 could cause significant errors. Initial competition HO' rate studies were performed with CBHl2and Ph2 as test compounds in CC14solution because, of the solvents examined, the highest initial HzO2 concentrations were achieved in that solvent. Analyses of
hexane C6H14
iso0 ct ane C8H18
cjq
c1
c1
tetrachloroethene c2c14
o-xylene
cyclohexeae ‘gH1O
Joethyl acetate EtOAc
toluene PhMe
propylbenzene PhPr
pseudocumene 124T
p-cymene IPMB
DMB
methylchlorof orm MCF
c1
trichloroethene ‘zHC13
cyclohexane C6H12
mesitylene 135T
biphenyl Ph2
c1
-d; cldcl UO‘ \ /
\ /
m-dichlorobenzene MDCB
U
1,2,4-trichlorobenzene TCB
anisole Ph OMe
N
O
2
nitrobenzene PhN02
Flgure 2. Organic test compounds used for correlation of solution- and gas-phase HO’ kinetics.
Table I. Maximum Concentrations of HzOzin Halogenated Organic Solvents Equilibrated with Aqueous HzOz Solutionsa
solvent
H,Oz concn (mM) in solvent on equilibn with 30% H202 90% HzOz
CC14 0.43b 4.86b CFC1, 0.36b 3.31b F-113 0.22bve 2.12bsd 0.338 C4F70C4FB 22 & 2 “C. b A 10-mL aliquot of HzOz solution and 400 mL of organic solvent. value of 1.5 mM was reported previously (13). dA value of 2.10 mM was obtained when 4 mL of HzOz solution and 320 mL of F-113 were used. e A 3-mL aliquot of HzOz solution and 200 mL of C4F70C4Fg. (I
reaction mixtures showed the presence of significant quantities of CHC in addition to CHL, CHN, and numerous unidentified minor products (eq l ; Table 11). The
CHL
CHN
CHC
products from Ph2 were not identified. The formation of CHC1, could not be confirmed because of the presence of
traces (0.30 mM) of it in the CC14used as the solvent. The concentration of CHC1, was less after 3-h irradiation than it was initially, C2C&was not found at the detection limit of -0.04 mM. Because trichloromethyl radicals and/ or chlorine atoms probably were involved in the CC14solution reactions (see below), we turned to F-113 as the most convenient solvent for the competition kinetic experiments. No CHC was detected on photolysis of C6HI2,Ph2, and H202in this solvent (Table 11). CHL, CHN, and numerous unidentified minor productqwere formed. Because the concentration of H202in F-113 was low, the conversions of C6HI2and Ph2 were too low for us to determine reliable relative rate constants when a turntable apparatus (17)was used for the irradiation. To increase the amount of H202in the F-113 solution over the course of the reaction, we developed a two-compartment reactor (Figure 1). In one compartment, H202was continuously extracted from an excess of 90% aqueous H202in the dark. The F-113 solution, saturated with H202,was pumped into the photoreactor and then back into the extraction compartment. With this system, satisfactory conversions of the test compounds were achieved in reasonable lengths of time. Several runs were performed in which two or more test compounds were irradiated in the presence and absence of H202. Dark-control reactions in the presence of Hz02 showed that all compounds were completely stable for time periods that at least equaled the duration of the kinetic runs. Typical results, for PhOMe, are shown in Figure 3. Environ. Scl. Technol., Vol. 22, No. 12, 1988 1449
Table 11. Products from Photolysis of CBHLZ, Phz,and HzOzin CCll and F-113Solutionsn
concn, reactnt init mM PhZ HzOz
solvent
reactn time, h
C6H12
cc1,c
2.0 2.1 0.71 0.79 0.98 0.98 1.00
CCI~'
F-113' F-113' F-113e F-113' F-113e
2.0 2.1 0.67 0.77 0.98 0.98 1.00
4.6d 0 1.7 0 -2.lf 2.If 0
3.0 3.0 12.0 12.0 7.0
-
reactnt conv, % C6H12 PhZ
43 24 28 10 68 92 44
25 12
f f
15.5
13.5
prodct yields,* % CHN CHC
CHL
19 9
18 13 30 33 10
35
22 14
46 74 9
4
3
6
15
16 12