Environ. Sci. Technol. 2003, 37, 5645-5650
Photochemical Alkylation of Inorganic Selenium in the Presence of Low Molecular Weight Organic Acids XUMING GUO,† RALPH E. STURGEON,* Z O L T AÄ N M E S T E R , A N D GRAEME J. GARDNER Institute for National Measurement Standards, National Research Council of Canada, Ottawa, Ontario, Canada K1A 0R9
Using a flow-through photochemical reactor and a low pressure mercury lamp as a UV source, alkyl selenium species are formed from inorganic selenium(IV) in the presence of low molecular weight organic acids (LMW acids). The volatile alkyl Se species were cryogenically trapped and identified by GC-MS and GC-ICP-MS. In the presence of formic, acetic, propionic and malonic acids, inorganic selenium(IV) is converted by UV irradiation to volatile selenium hydride and carbonyl, dimethylselenide and diethylselenide, respectively. Se(IV) was successfully removed from contaminated agricultural drainage waters (California, U.S.A.) using a batch photoreactor system Se. Photochemical alkylation may thus offer a promising means of converting toxic selenium salts, present in contaminated water, to less toxic dimethylselenide. The LMW acids and photochemical alkylation process may also be key to understanding the source of atmospheric selenium and are likely involved in its mobility in the natural anaerobic environment.
Experimental Section
Introduction Formation of volatile metal adducts in nature is frequently considered to be primarily a result of bacterial activity, although other major sources include abiotic processes (usually methylation). The latter involve natural methyldonor compounds, such as iodomethane, methylcobalamine or humic substances. Chemical methylation of tin in the marine environment has been reported, and biomethylation of Sn, Sb, Se, As Bi, Tl and Hg has additionally been documented (1). Selenite is much more toxic than selenate, both in vivo and in vitro (2, 3); however, dimethylselenide (DMSe) is 500- to 700-fold less toxic (to rats) than aqueous selenite and selenate (4, 5). As one of several detoxification processes, it has been known for over a century that bacteria are capable of reducing selenium salts to elemental selenium (6). Many suboxic sediments and soils contain Fe(II, III) oxides which reduce Se(VI) to Se0 (7). Bacteria (8-13), plants (14, 15), decaying plant detritis (16), and marine algae and plankton (17-19) as well as animals (20) have all been shown capable of methylating selenium from its major oxidation states. Biomethylation of selenium in soils, sediments, plants, * Corresponding author phone: (613)993-6395; fax: (613)993-2451; e-mail:
[email protected]. † On leave from the Department of Chemistry, Xiamen University, Xiamen, China. 10.1021/es034418j CCC: $25.00 Published on Web 11/11/2003
freshwater systems, and the marine environment into DMSe is considered to be a major source of atmospheric selenium (21, 22). Losses of selenium from soils by biomethylation may, in some cases, give rise to an insufficient supply of selenium to ruminant animals (23). The reduction of Se oxyanions to Se0, followed by further reduction and methylation to a volatile methylated form, is commonly regarded as the most important detoxification route in biological systems. Such processes have also been widely applied for reducing Se contamination in wastewater arising from agricultural drainage systems, power plants, oil refineries and semiconductor industries (24-28). Although emphasis has been placed on biomethylation, parallel photochemical processes abound in nature. A methylchromium bond was shown to be formed during the photolysis of tertiary-butoxy radicals and chromium(II) in aqueous solution (29). In the presence of acetate ion or acetic acid in aqueous solution, Hg2+ gives rise to methylmercury following photolysis by sunlight (30). In some phototrophic bacterial cultures amended with tellurate or elemental Te (powered metal), dimethyltelluride was detected after growth in the light (31). In the case of selenium, use of a TiO2 photocatalyst and UV irradiation permitted removal of selenate ions from an aqueous solution, presumably by formation of volatile SeH2 (32). Synthetic seawater, spiked with organo-selenium compounds and exposed to sunlight, produced methylated selenium, which was not the case with spikes of inorganic selenium (33). These studies clearly indicate that low molecular weight organic acids and photolysis may play a significant role in the transformation of heavy metals in the environment. To date, however, there have been no reports of the possible role of direct photochemical alkylation of inorganic selenium. Experiments performed in this laboratory demonstrate that photochemical alkylation by UV light may serve as a new pathway for the transformation of inorganic selenium to its volatile hydride, carbonyl, methylated, or ethylated analogues in the environment.
Published 2003 by the Am. Chem. Soc.
Instrumentation. A flow through photoreactor, consisting of a 5 m length of 1.1 mm i.d. × 1.7 mm o.d. poly(tetrafluoroethylene) (PTFE) tubing (Cole Parmer Instrum. Co., Vernon Hills, IL) wrapped around a low-pressure Hg vapor UV lamp (254 nm, 15W, Cole Parmer, U.S.A.) was constructed. This provided for an incident radiation of ∼48 mW/cm2. A schematic of this system is illustrated in Figure 1. A simple scanning monochromator was used to record the spectral output and relative energy distribution of the UV light source before and after passage through the thickness of the PTFE tube wall used in the photoreactor. The flow through photoreactor system was enclosed in an aluminum metal box for safety of operation. Samples were propelled through the system with the aid of a Minipuls 2 peristaltic pump (Gilson, Middleton, WI) operating at a flow rate of 2 mL min-1. UV vapor generation was accomplished in a continuous-flow mode as the sample was pumped through the PTFE tube, resulting in a 2 min irradiation. A continuous flow system limited the rise in temperature of the reaction medium to less than 38 °C over the duration of all experiments. The effluent from the photoreactor was directly merged with a 55 mL min-1 flow of He in a gas-liquid separator, and the resultant volatile species were transported either directly to a heated quartz tube atomizer, via a 10 cm length of PTFE transfer tubing, or through an intermediate cryogenically cooled U-shaped glass condensation tube. The VOL. 37, NO. 24, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
5645
FIGURE 1. Schematic of the photoreactor. quartz tube atomizer temperature was maintained at 900 °C. A flow rate of 15 mL min-1 H2 was introduced via a second line into the quartz tube to aid in the atomization of selenium. The quartz tube atomizer was mounted in the burner (optical) compartment of a Perkin-Elmer Model 4100 atomic absorption spectrometer fitted with a Perkin-Elmer Se electrodeless discharge lamp. Simultaneous deuterium background correction was applied for all measurements at 196.0 nm. Both peak-height and integrated absorbance measurements were recorded. Measurements were made in a continuous mode to characterize the yield of the reaction products and in an off-line mode to cryogenically trap them for subsequent characterization using GC-MS and GC-ICP-MS techniques. For this latter purpose, a series of cryogenically cooled U-tube traps (0.8 cm o.d. × 0.6 cm i.d. × 13 cm deep × 4 cm across) was inserted between the generator and the detector to condense the analyte species. The first trap removed water using a dry ice methanol bath, the second, packed with glass wool and immersed in liquid nitrogen (-196 °C), was used to sequester the volatile species swept from the dry ice bath. The normal temperature of the dry ice and methanol trap is about -78 °C; the actual temperature experienced by the gas flow was likely higher because the volatile products were rapidly removed from the trap by the He carrier gas. Continuous on-line monitoring of the AAS signal revealed that no decrease in intensity of the Se signal occurred when the water trap alone was used, and the signal was completely lost when the liquid nitrogen trap was introduced, indicating that the volatile selenium species were completely trapped in this second U-tube. Helium was chosen as the carrier gas because it is not condensed in the U-tube at liquid nitrogen temperatures and does not introduce any mass spectral interferences in the subsequent identification of trapped species by GC-ICP-MS and GC-MS. A Hewlett-Packard (HP) model 6890 gas chromatograph was interfaced to an HP5973 mass selective detector (mass range 60-270 Daltons). Selenium compounds were separated on a 30 m × 0.25 mm i.d. × 0.25 mm film (J&W Scientific) DB1 capillary column (1% phenyl, 99% poly(dimethylsiloxane)) using UHP helium carrier gas. A Perkin-Elmer SCIEX ELAN 6000 (Concord, Ontario, Canada) ICP-MS instrument was interfaced to a Varian 3400 GC (Varian Canada, Georgetown, Canada) equipped with a 15 m length of DB-1 column (0.32 mm) through a heated transfer line. The sampling process and GC conditions were similar to those used for GC-MS measurements. Reagents and Samples. All solutions were prepared using 18 MΩ-cm deionized, reverse osmosis water (DIW) obtained from a mixed bed ion-exchange system (NanoPure, model D4744, Barnstead/Thermoline, Dubuque, IA). Calibration solutions were prepared daily by diluting 1000 mg L-1 stock solutions prepared from selenite (Na2SeO3 99%, Aldrich, U.S.A.) and selenate (Na2SeO4 99%, Aldrich). Solutions of low molecular weight acids (LMW) were prepared from analytical reagent grade materials: formic acid (23 M, 5646
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 24, 2003
FIGURE 2. Relative energy distribution of the source lamp in the UV wavelength range 190 to 390 nm. A: UVC lamp; B: UVC lamp after transmission through the PTFE tube wall of the photoreactor. Anachemica, Canada), acetic acid (6.3 M, BDH, Canada), propionic acid (13 M, BDH) and malonic acid (BDH). High purity HNO3, NaNO3 and NaNO2 (Fisher Scientific) were used, except where indicated otherwise. Ottawa River water was sampled from the shore of the river using precleaned polypropylene bottles and directly transported to the laboratory for immediate study. Selenium contaminated agricultural drainage waters (California, U.S.A.) were treated using a 200 mL quartz immersion batch photochemical reactor fitted with a UVC pen lamp (Analamp, Claremont, CA, No. 80105701, 79 µW cm-2, λmax: 253.7 nm). A flow rate of 5 mL min-1 Ar or N2 was used to sparge any gaseous species produced during the irradiation. Hydride generation atomic absorption spectrophotometry was employed to determine the initial concentration of selenium in the wastewater. Procedure. Volatile selenium compounds were generated by UV photolysis of selenium standard solutions containing various LMW acids at different concentrations. The gaseous products were flushed into the heated quartz tube atomizer for AAS measurements so as to determine optimal generation conditions or passed through the successive U-tubes using a stream of He carrier gas. Prior to commencing the trapping experiment, the U-tubes were flushed with He for 10 min at room temperature in order to remove any air entrained in the system and eliminate memory effects. Following cryocondensation, the second U-tube was closed at both ends by stoppers consisting of rubber septa and removed from the liquid nitrogen bath to equilibrate to room temperature for about 15 min. A 250 µL volume of the volatile Se species was sampled through the septum of the U-tube, using a gastight syringe, and injected into the GCMS for species identification.
Results and Discussion Figure 2 shows the spectral intensity of the excitation source in the 190 to 390 nm region used in the flow through photochemical reactor. As expected, the most intense output occurs at the 253.7 nm excitation line for mercury. Following passage through the wall thickness of the PTFE tubing used in the reactor, the deep UV portion of the spectrum is severely attenuated, with less that 5% of the energy being transmitted. Despite this attenuation, sufficient irradiation of the samples could be achieved within a 2-minute time frame that encouraging and efficacious results were obtained. It should be noted that when a UVB lamp (no output below 280 nm) was used under identical conditions, no volatilization of the selenium occurred. As a consequence, the reaction is deemed to be quite efficient under UVC irradiation and the rate of the reaction should be enhanced if quartz tubing had been substituted for the PTFE tubing in these experiments.
FIGURE 3. Effect of low molecular weight organic acid concentration on AAS signals arising from the continuous photochemical treatment of solutions containing 0.1 mg L-1 Se(IV).
FIGURE 4. Effects of NaNO3, NaNO2 and HNO3 on the AAS signals arising from the continuous photochemical treatment of solutions containing 0.1 mg L-1 Se(IV). Preliminary experiments revealed that the presence of TiO2 in the photolysis cell was unnecessary for the generation of a volatile selenium product (32). Formic acid was initially investigated as it has the simplest structure among the organic acids studied. Generation efficiency was dependent on the acidity at which the reaction is performed, as shown by the data in Figure 3. A plateau in the range of 0.4-1.0 M HCOOH is evident. By maintaining the pH of the solutions to be irradiated at
