Environ. Sci. Technol, 1990,2 4 , 1326-1332
(29) Shafer, W. E.; Schonherr,J. Ecotoxicol. Environ. Saf. 1985, 10, 239-252. (30) Schonherr,J. Biochem. Physiol. Pflanz. 1976,170,309-319. (31) Schonherr,J.; Kerler, F.; Riederer, M. Deu. Plant Biol. 1984, 9,491-498. (32) Shafer, W. E.; Bukovac, M. J. Plant Physiol. 1987, 83, 652-656. (33) Shafer, W. E.; Morse, R. D.; Bukovac, M. J. HortScience 1988,23, 204-206. (34) The large majority of the simple and valence zero- to sixorder molecular connectivity indexes correlate far worse than the 3x" index. Their explained variances are between 50 and 70%. More specifically,the 'x and lxVindexes have explained variance of 54 and 66%,respectively. The only exception was the 4xvindex with explained variance of 80%. (35) Irvin, J. A.; Quickenden, T. I. J . Chem. Educ. 1983, 60, 711-712.
(36) Maes, F. W. J. Theor. Biol. 1984, 111, 817-819. (37) York, D. Can. J.Phys. 1966,44, 1079-1086. (38) Tsantili-Kakoulidou,A.; El Tayar, N.; van de Waterbeemd, H.; Testa, B. J. Chromatogr. 1987, 389, 33-45. (39) Pearlman, R. S.; Yalkowsky, S. H.; Banerjee, S. J. Phys. Chem. Ref. Data 1984,13, 555-562. (40) Mackay, D. Enuiron. Sci. Technol. 1982, 168, 274-278. (41) Davies, R. P.; Dobbs, A. J. Water Res. 1984,18,1253-1262. (42) Harnisch, M.; Mockel, H. J.; Schulze, G. J. Chromatogr. 1983,282, 315-332. (43) Isnard, P.; Lambert, S. Chemosphere 1989,18,1837-1853. (44) Krell, H. W.; Sandermann,H., Jr. Plant Sci. 1985,40,87-93.
Received for review November 20,1989. Accepted March 30,1990. A S . gratefully acknowledges the financial support by the Znternationales Buro, Forschungsanlage Julich (KFA).
Kinetics of the Atmospherically Important Reactions of Dimethyl Selenide Roger Atklnson," Sara M. Aschmann, and Davld Hasegawa
Statewide Air Pollution Research Center, University of California, Riverside, California 92521 Ellsabeth 1.Thompson-Eagle and Wllllam 1. Frankenberger, Jr.
Department of Soil and Environmental Sciences, University of California, Riverside, California 92521 The biomethylation of selenium into dimethyl selenide from soils, sediments, and plants is a major source of atmospheric selenium. The photolysis and kinetics of the gas-phase reactions of dimethyl selenide with OH and NO3 radicals and O3 were investigated at 296 f 2 K and atmospheric pressure of air. No evidence for photolysis of dimethyl selenide was observed. Rate constants for reaction with the OH radical and O3 of (6.78 f 1.70) X and (6.80 f 0.72) X lo-'' cm3 molecule-' P', respectively, were obtained. The rate constant for reaction with the NO3 radical was observed to increase with increasing NOz concentration, with an extrapolated value of 1.4 X lo-'' cm3 molecule-' s-l at zero NOz concentration. Combining these rate constants with estimated ambient tropospheric concentrations of OH radicals (1.5 X lo6 molecule cm-7, NO, radicals (2.4 X los molecule cm-,) and O3 (7 X 10" molecule cm-,) leads to the following calculated lifetimes of dimethyl selenide due to reaction with these species: OH radicals, 2.7 h; NO, radicals, 5 min; and O,, 5.8 h. The products of these reactions of dimethyl selenide are presently unknown but are likely to be involved in gasto-particle conversions and aerosol scavenging during the global cycling of selenium.
Introduction Volatile selenium compounds are released into the environment from both anthropogenic and natural sources ( I , 2). Gaseous selenium species detected in the atmosphere include dimethyl selenide, dimethyl diselenide, dimethyl selenone, and methaneselenol ( 3 , 4 ) . Hydrogen selenide may also be released into the atmosphere under reduced conditions, but it is quickly oxidized into elemental selenium in the presence of air (5). One of the most important vapor-phase selenium species found in the atmosphere appears to be dimethyl selenide (1). Biomethylation is a major source for this metalloid-carbon bond formation in the natural environment (6). Although the biological methylation of Se has been poorly quantified in the past and was once considered to be less important than the generation of volatile selenium compounds from 1326 Environ. Sci. Technol., Vol. 24, No. 9, 1990
the combustion of fossil fuels and refuse, biological transformations are now believed to constitute an important link in the global cycling of this element. Within the accuracy of current estimates, anthropogenic and natural source emissions are of the same order of magnitude, being approximately (1-10) X lo9 g of Se/year (1). Higher levels of selenium are detected in atmospheric and rainfall samples collected in Denmark, Scotland, and the United States during the warmer spring and summer months, when biological activity is at its peak, than in the winter months when combustion of fossil fuels is a t its highest (7, 8). Selenium is an element that is required in trace amounts by some microorganisms. It is considered to be an essential element for plants and animals, but is toxic a t higher concentrations. Relatively high concentrations of selenium (>2 ppm) can occur in certain arid, saline, alkaline, and poorly drained soils including the west side of the San Joaquin Valley in California. Agricultural irrigation water percolates through these soils and displaces the shallow saline- and selenium-enriched groundwater. In this manner, soluble salts including toxic selenium ions are removed from the cropland via tile drains into drainwater evaporation ponds, and the evaporation of agricultural irrigation water in these ponds may lead to levels of selenium that are toxic to wildlife (9). Laboratory and field research has shown that microorganisms naturally present in the contaminated soil and water transform selenium compounds into dimethyl selenide, which is liberated into the atmosphere (10-20). Dimethyl selenide has been found to be nonhazardous to rats (21-23). Characterization of this naturally occurring microbial selenium detoxification and removal process has led to the discovery that biomethylation can be accelerated to the point where there is a significant drop in the initial selenium inventory within a relatively short period of time (10-12, 16-20). Biomethylation could play a significant role in the bioremediation of selenium-contaminated sites in the California Central Valley and be used to prevent the harmful accumulation of selenium in agricultural areas
0013-936X/90/0924-1326$02.50/0
0 1990 American Chemical Society
(13). If this biotechnology is to be effectively utilized, it is important to know the fate of dimethyl selenide in the atmosphere. The majority of organic compounds present in the troposphere undergo photolysis and chemical reactions with hydroxyl and nitrate radicals and ozone (24-26). Dimethyl selenide may be transformed by these processes into more oxidized and less volatile species, which can be either particle-associated or distributed between the gas and particle phases. To date, however, no information is available concerning the atmospherically important reactions or the tropospheric lifetime of dimethyl selenide. Because of lack of data, residence times of volatile selenium compounds in the atmosphere are usually estimated by using mass balance equations in combination with sulfur data. In this work, we have experimentally investigated the potentially atmospherically important gas-phase loss processes of dimethyl selenide and determined the rate constants a t room temperature for its reactions with OH and NO, radicals and 0,. Experimental Section The experimental techniques used were generally similar to those described previously (27-31), and only the relevant details are given here. Photolysis and Reaction with the OH Radical. The OH radical reaction rate constant for dimethyl selenide was determined by using a relative rate technique in which the decay rates of dimethyl selenide and a reference organic were measured in the presence of OH radicals. Hydroxyl radicals were generated by the photolysis in air of methyl nitrite (CH,ONO)
CH30N0 + hu
-
+ Oz HOz + NO
CH,O
-
CH,O
HCHO
-+
OH
+ NO
+ H02
+ NO2
and NO was included in the reactant mixtures to avoid the formation of 0, and hence of NO, radicals. Experiments were carried out in a 6400-L all-Teflon chamber equipped with two parallel banks of blacklamps. Irradiations were also carried out in the absence of CH3ON0 to investigate the importance of photolysis under the experimental conditions employed. Isoprene (2-methyl1,3-butadiene) was used as the reference organic, and as in previous studies (27,32),2,3-dimethyl-2-butenewas also included in the reactant mixtures to check on the absence of 0, and NO, radical reactions. Thus, while the reactions of isoprene and 2,3-dimethyl-2-butenewith the OH radical have similar rate constants (33),the room-temperature rate constants for the reactions of 2,3-dimethyl-2-butene with 0, and with the NO, radical are higher than those for the corresponding reactions of isoprene by factors of -80 (34) and -95 (30),respectively. The initial reactant mixtures were as follows (in molecule cm-, units): CH,ONO (when present), 2.5 X 1014;NO, (0-2.4) X 1014;dimethyl selenide, and isoprene and 2,3-dimethyl-2-butene, 2.4 X 5X lo1, each. Irradiations were carried out at 20% of the maximum light intensity, and the irradiation times varied from 1 to 15 min. Providing that the dimethyl selenide and the reference organics reacted only with the OH radical, then In
[(CH3)2SeIto k, [isoprene],, = - In [(CHJZS~I, k2 [isoprene],
(1)
where [ (CH3)2Se]t,and [isoprene],, are the concentrations
of dimethyl selenide and isoprene, respectively, a t time to, [ (CH3&3e] and [isoprene] are the corresponding concentrations a t time t , and k, and k2 are the rate constants for reactions 1 and 2, respectively. Hence, a plot of In
,
,
--
+ (CH3)2Se OH + isoprene
OH
products
(1)
products
(2)
([(CH,)2Se]t,/ [(CH,),Se],) against In ([isoprene],,/ [isoprene],) should be a straight line of slope k,/kz and zero intercept. Dimethyl selenide, isoprene, and 2,3-dimethyl-2-butene were monitored by gas chromatography with flame ionization detection (GC-FID), using a 20 f t X 0.125 in. stainless steel column with 5% DC703/C20M on 100/120 AW, DMCS Chromosorb G, operated at 333 K. Reaction with the NO, Radical. The rate constant for the reaction of the NO3 radical with dimethyl selenide was also determined by a relative rate method in which the disappearance rates of dimethyl selenide and a reference organic were monitored in the presence of NO, radicals. NO, radicals were generated from the thermal decomposition of N2O5 M
N205ZNO3 + NOz and since NO, radicals, NOz, and N205are in reasonably rapid equilibrium at room temperature and atmospheric pressure (35),NO2 was added to the reactant mixtures to vary the NO, radical/NZO5 concentration ratios. The reactant mixtures had the following initial concentrations (in molecule cmW3 units): dimethyl selenide, (3.1-5.2) X lo1,; reference organic, 2.4 X lo1,; NO2, (1.2-12) X 1014. Four to six additions of N20s, at initial concentrations in the chamber of (2.4-15) X 10l2molecule cm-,, were introduced into the chamber during the experiments and rapid mixing (> [031initial, then -d In [03]/dt = k6 + k5[(CH3),Se] (111) where k, is the O3 decay rate in the absence of added dimethyl selenide and k, is the rate constant for reaction 5. Hence, plots of the ozone decay rate, -d In [03]/dt, against the dimethyl selenide concentration should have a slope of k5 and an intercept of k6. Experiments were carried out in a 160-L all-Teflon reaction chamber, which was initially divided into two approximately equal subchambers by metal barriers. Ozone was introduced into one subchamber, and the dimethyl selenide into the other, each in synthetic air diluent at concentrations designed to achieve the desired initial reactant concentrations in the entire reaction chamber. After the ozone and the dimethyl selenide concentrations had been measured in the appropriate subchamber, the barriers were removed and the contents of the entire reaction chamber mixed for 1 min. Ozone was monitored throughout the experiments by a Monitor Labs Model 8410 chemiluminescence analyzer, while dimethyl selenide was monitored by GC-FID as described above. In one experiment, excess n-hexane was added to the reactant mixture to scavange any OH radicals produced in the O3 reaction with dimethyl selenide. The initial reactant concentrations were (in units of ioi3 molecule cm-9 as follows: dimethyl selenide, 0-13.4; n-hexane (when present), -240; and 03,-2.4. Chemicals. The sources of the chemicals used, and their stated purity levels, were as follows: dimethyl selenide, Strem Chemicals, Inc.; n-hexane (99+%), isoprene (99+%) and 2-methyl-2-butene (99+%), Aldrich Chemical Coo;2,3-dimethyl-2-butene (99%), Chem Samples, Inc.; trans-2-butene (295%) and NO (99.0%), Matheson Gas Co. Methyl nitrite was prepared as described by Taylor et al. (37) and stored under vacuum at 77 K. NO, was prepared by reacting NO with an excess of O2prior to use. Nz05was prepared by collecting the products of the reaction of NO, (Matheson Gas Co., 199.5%) with O3 at 196 K and was stored at 77 K under vacuum. O3 in 0, diluent was generated as needed from a Welsbach T-408 ozone generator.
r
-
-
0
I
I
I
1
0.4
0.8
1.2
I .6
In(CISOPRENElto/ C ISOPRENE] t 1
Figure 1. Plot of eq I for the reaction of the OH radical with dimethyl selenide, with isoprene as the reference organic.
appearance rates in the CH30NO/NO/air irradiations were similar, with the measured rate constant ratio of k2(2,3-dimethyl-2-butene)/kz(isoprene) = 1.06 f 0.02 being in excellent agreement with the literature value of 1.09 (33). This good agreement of the measured and literature rate constant ratios for isoprene and 2,3-dimethyl-2-butene shows that any contributions of NO3 radical and O3 reactions to the observed disappearance rates of dimethyl selenide, isoprene, and 2,3-dimethyl-2-butene were negligible (
