Methyl t-Butyl Ether Mineralization in Surface-Water Sediment Microcosms under Denitrifying Conditions

Abstract

Mineralization of [U-¹⁴C]methyl t-butyl ether (MTBE) to ¹⁴CO₂ without accumulation of t-butyl alcohol (TBA) was observed in surface-water sediment microcosms under denitrifying conditions. Methanogenic activity and limited transformation of MTBE to TBA were observed in the absence of denitrification. Results indicate that bed sediment microorganisms can effectively degrade MTBE to nontoxic products under denitrifying conditions.

Pervasive contamination of groundwater (2, 7, 22–24) and surface-water systems (1, 4, 13, 16, 25) by the fuel oxygenate methyl t-butyl ether (MTBE) makes crucial an accurate understanding of its environmental fate. A drinking water advisory exists for MTBE for taste and odor of 20 to 40 μg/liter (26), and MTBE is classified by the U.S. Environmental Protection Agency as a possible human carcinogen (7, 26). The U.S. Geological Survey national water use summary estimated that 60% of the drinking water consumed in the continental United States comes from surface water (20). Because these systems are not easily shielded from numerous point (5, 9) and non-point (1, 2, 4, 6, 10, 13, 14, 16) sources of MTBE contamination, identifying natural sinks for MTBE in surface waters is particularly important. Although microorganisms in surface-water bed sediments can mineralize MTBE to CO₂ under oxic conditions (5), geochemical constraints on oxygen transport (27) may limit the importance of aerobic MTBE mineralization in these systems (5). Evidence for anaerobic MTBE biodegradation in surface-water sediments currently is limited to a single methanogenic microcosm exhibiting incomplete transformation to the toxic product, t-butyl alcohol (TBA) (11). Such findings (5, 11) suggest that efficient anaerobic degradation of MTBE to nontoxic products may require relatively oxidizing terminal electron-accepting conditions.

Of the anaerobic terminal electron-accepting processes common to surface-water environments, denitrification is the most energetically favorable and is widely observed in both pristine and waste-impacted systems (8, 15). The ability of surface-water microorganisms to degrade MTBE under denitrifying conditions was examined in bed sediments from Charleston, S.C., and Pensacola, Fla. At the Charleston site (5), MTBE-contaminated groundwater discharges to a shallow freshwater stream containing poorly sorted sandy bed sediments. Charleston sediments contained significant concentrations of dissolved CH₄ (100 ± 15 μM), SO₄ (70 ± 10 μM), and sulfide (>200 μM) but no detectable O₂ (method detection limit [MDL] < 2 μM), NO₃ (MDL < 2 μM), Fe(II) (MDL < 1 μM), or MTBE (MDL = 2 μg/liter). Pensacola sediments were well-sorted medium sands collected from a shallow freshwater wetland with no detectable MTBE (MDL = 2 μg/liter) and no history of MTBE exposure. Pensacola sediments contained significant concentrations of dissolved NO₃ (60 ± 10 μM), SO₄ (2 ± 0 mM), and CH₄ (100 ± 15 μM) but no detectable O₂ (MDL < 2 μM) or Fe(II) (MDL < 1 μM).

MTBE mineralization was investigated using [U-¹⁴C]MTBE (5). The radiochemical composition (mean ± standard deviation [SD]) of the [U-¹⁴C]MTBE stock was evaluated in our lab by radiometric detection high-performance liquid chromatography and gas chromatography and found to be 97.4% ± 0.3% as [¹⁴C]MTBE and 2.6% ± 0.2% as [¹⁴C]TBA. The presence of TBA as a trace contaminant in commercially available MTBE is not uncommon (P. M. Bradley, unpublished results) and must be considered when evaluating the significance of TBA as an intermediate in MTBE biodegradation. Bed sediment microcosms were prepared as described previously (5) and were composed of 5 ml of saturated sediment and an atmosphere of air (oxic treatment) or helium (anoxic treatment) in 10-ml serum vials. An anoxic KNO₃ solution was added to NO₃-amended treatments to yield initial dissolved NO₃ concentrations of 4.6 ± 0.1 mM and 7.3 ± 0.2 mM in the Charleston and Pensacola microcosms, respectively. All microcosms were amended with 0.5 μCi of [U-¹⁴C]MTBE (specific activity, 10.1 mCi/mmol) to yield initial dissolved MTBE concentrations of 17.2 ± 0.3 μM and 19.9 ± 0.2 μM in the Charleston and Pensacola treatments, respectively. Headspace concentrations were monitored periodically using radiometric detection gas chromatography combined with thermal conductivity detection. The radioactivity associated with C₁ to C₄ organic acids, TBA, and MTBE was assessed using radiometric detection high-performance liquid chromatography. NO₃ and SO₄ concentrations were determined by ion chromatography. Radiometric detectors were calibrated by liquid scintillation counting using H¹⁴CO₃ and [U-¹⁴C]MTBE. The results presented below in Table 2 and Fig. 1 were corrected for losses due to sampling.

TABLE 2.

Final distribution of ¹⁴C radioactivity in anoxic bed sediment microcosms after 77 days^a

Sediment source	Treatment	Compound	14C distribution (%)
			Experimental	Control
Charleston	NO3 amended	MTBE	70 ± 1	97
		TBA	NDb	3
		CO2	26 ± 10	ND
		CH4	ND	ND
		Total recovery	96 ± 10	100
	Unamended	MTBE	85 ± 5	91
		TBA	10 ± 2	3
		CO2	3 ± 3	ND
		CH4	1 ± 1	ND
		Total recovery	99 ± 5	94
Pensacola	NO3 amended	MTBE	71 ± 4	94
		TBA	1 ± 1	4
		CO2	23 ± 5	ND
		CH4	ND	ND
		Total recovery	95 ± 5	98
	Unamended	MTBE	78 ± 4	101
		TBA	9 ± 2	4
		CO2	5 ± 2	ND
		CH4	5 ± 1	ND
		Total recovery	97 ± 4	105

^aNo radiolabeled C₁ to C₄ organic acids were detected in this study. For each treatment, experimental data are means ± SD for triplicate microcosms, and control data are from a single microcosm. 

^bND, not detected. The MDL for radiometric detection was equivalent to 1% final recovery. 

FIG. 1.

Percent mineralization of [U-¹⁴C]MTBE to ¹⁴CO₂ or ¹⁴CH₄ in microcosms containing bed sediments from the Charleston and Pensacola sites. The symbols indicate ¹⁴CO₂ in the oxic (●), anoxic (▾), NO₃-amended anoxic (○), and control (▿) treatments and ¹⁴CH₄ (■) in the anoxic Pensacola treatment. Experimental data are means ± SD for triplicate microcosms, and the control data are from a single autoclaved control microcosm. For each sediment, superscript letters adjacent to final data points indicate statistically significantly different final mean ¹⁴C recoveries according to the Kruskal-Wallis one-way analysis of variance on ranks (P < 0.05).

The observed production of N₂-N (Table 1) and the lack of detectable SO₄ reduction or methanogenesis indicated that, under NO₃-amended anoxic conditions, both sediments were dominated by denitrification. The fact that NO₃-N loss and N₂-N production were not statistically different confirmed that denitrification was the primary sink for NO₃ under these conditions (Table 1). Under unamended anoxic conditions, production of N₂-N also was observed in Pensacola sediments over the first 36 days but production was insignificant between 36 and 77 days (Table 1). SO₄ reduction was substantial (42% ± 8% decrease in dissolved SO₄ concentrations over 77 days), and methanogenesis (1.5 ± 0.3 nmol of CH₄ · g⁻¹ · day⁻¹) became significant between 36 and 77 days. These observations indicate that, in the absence of added NO₃, anoxic Pensacola sediments shifted from denitrifying conditions to predominantly SO₄-reducing and methanogenic conditions. In contrast, under unamended anoxic conditions, the Charleston sediments were characterized by insignificant denitrification (Table 1), trace SO₄-reducing activity (data not shown), and extensive methanogenesis (19 ± 6 nmol of CH₄ · g⁻¹ · day⁻¹, expressed per gram of sediment [dry weight]).

TABLE 1.

Final amount of NO₃-N consumed and N₂-N produced after 77 days in anoxic microcosms containing bed sediment from the Charleston and Pensacola sites^f

Sediment source	Anaerobic treatment	NO3-N consumeda (μmol)	N2-N producedb (μmol)
Charleston	NO3 amended	5.4 ± 0.5	5.6 ± 0.8
	Unamended	—c	—d
Pensacola	NO3 amended	2.3 ± 1.4	3.2 ± 0.7
	Unamended	0.1 ± 0.0	0.3 ± 0.2e

^aNo changes in dissolved NO₃ concentrations were observed in sterile control microcosms. 

^bN₂-N level produced in experimental treatments minus that in sterile control microcosms. 

^cNo NO₃ (MDL = 2 μM) was detected in unamended Charleston sediments. 

^dN₂-N production did not differ significantly from that observed in sterile control microcosms. 

^eProduced in the first 36 days of incubation. 

^fData are means ± SD for triplicate microcosms. Transient production of N₂O was observed for both sediments in NO₃-amended microcosms, but no N₂O was detected in any treatment at the end of the incubation period. 

The microorganisms indigenous to both sediments demonstrated efficient mineralization of [U-¹⁴C]MTBE under NO₃-amended, denitrifying conditions (Fig. 1). Approximately 25% of the [U-¹⁴C]MTBE radioactivity was recovered as ¹⁴CO₂ in 77 days. The mineralization of [U-¹⁴C]MTBE observed in this study under NO₃-amended, anoxic conditions was comparable to that observed under oxic conditions (Fig. 1). The fact that the final combined recovery of [U-¹⁴C]MTBE and ¹⁴CO₂ in the experimental microcosms did not differ significantly from the final recovery of radioactivity in autoclaved control microcosms (≥98%) indicates that ¹⁴CO₂ was the only significant product of MTBE biodegradation under denitrifying conditions (Table 2). [U-¹⁴C]MTBE mineralization was attributable to biological activity, because the final recovery of ¹⁴CO₂ in killed control microcosms was less than 1% (Table 2). In earlier studies, the potential for anaerobic MTBE biodegradation under NO₃-enriched conditions was investigated in soils (28) and aquifer sediments (3, 11) and was reported to be insignificant. In contrast, the results of this study demonstrate that microorganisms indigenous to surface-water bed sediments can mineralize MTBE under denitrifying conditions.

The results of this study also indicate that bed sediment microorganisms can mineralize TBA under denitrifying conditions (Table 2). TBA contamination of surface water is an environmental concern due to its presence in gasoline spills (7, 24), demonstrated carcinogenicity in laboratory animals, and significance as the presumptive initial intermediate in microbial degradation of MTBE (11, 17, 28). In the present study, no significant recovery of [¹⁴C]TBA was observed in NO₃-amended, experimental microcosms in spite of the fact that approximately 3% of the radioactivity in the original stock was [¹⁴C]TBA (Table 2). In contrast, the recovery of radiolabel as [U-¹⁴C]MTBE and [¹⁴C]TBA in autoclaved control microcosms was the same as that observed in the original added substrate (Table 2). The results are consistent with previous reports of enhanced biodegradation of TBA in soil samples under NO₃-amended conditions (28) and demonstrate that the microorganisms indigenous to surface-water bed sediments can oxidize [¹⁴C]TBA to ¹⁴CO₂ under denitrifying conditions.

A number of observations indicate that NO₃ availability was a primary determinant of the efficiency and the products of anaerobic [U-¹⁴C]MTBE biodegradation in this study. First, NO₃ amendment stimulated denitrification (Table 1) and MTBE biodegradation (Table 2) in both sediments. Second, mineralization of [U-¹⁴C]MTBE to ¹⁴CO₂ only occurred if dissolved NO₃ concentrations were significant (Table 2). In NO₃-amended, denitrification-dominated microcosms, ¹⁴CO₂ was the sole product of [U-¹⁴C]MTBE biodegradation (Table 2). For the unamended, anoxic Pensacola sediments, the percentage of radioactivity recovered as ¹⁴CO₂ increased during the initial period of denitrification (first 36 days) but decreased from 36 to 77 days as methanogenesis became significant (Fig. 1 and Table 1). The simultaneous decrease in ¹⁴CO₂ and increase in ¹⁴CH₄ indicate that ¹⁴CH₄ was formed autotrophically at the expense of ¹⁴CO₂ (Table 1 and Fig. 1). Finally, accumulation of [¹⁴C]TBA was not observed in this study under denitrifying conditions. [¹⁴C]TBA was oxidized to ¹⁴CO₂ under NO₃-amended conditions but increased in the absence of significant denitrifying activity (Table 2). Accumulation of TBA during biodegradation of MTBE under methanogenic conditions has been reported previously (11). The present results indicate that denitrification and anaerobic MTBE mineralization were limited by NO₃ availability and suggest that MTBE mineralization was coupled to denitrification.

The demonstrated ability of naturally occurring microorganisms to degrade MTBE to CO₂ under anoxic conditions without the accumulation of TBA has important implications for the fate of MTBE in groundwater and surface-water systems. Although the potential for aerobic biodegradation of MTBE to nontoxic products has been demonstrated for a number of groundwater (3, 9, 18) and surface-water sites (5), the actual contribution of this process to natural attenuation of MTBE is unclear because the onset and subsequent predominance of anoxic conditions are characteristics of hydrocarbon-contaminated waters. Consequently, engineered systems are being developed to support aerobic MTBE biodegradation under otherwise anoxic conditions (18). Unfortunately, recent groundwater and surface-water quality assessments indicate that environmental MTBE contamination is so widespread (1, 2, 4, 7, 13, 16, 22–25) that engineered solutions are realistic only for a small percentage of contaminated sites. For the remaining sites where natural attenuation would be expected to be the primary method for environmental restoration, identifying the conditions which support efficient anaerobic degradation of MTBE to nontoxic products is crucial. The demonstrated ability of bed sediment microorganisms to mineralize MTBE under denitrifying conditions indicates that anaerobic biodegradation of MTBE can be a significant contributor to the natural attenuation of MTBE in the environment. Because dissolved NO₃ concentrations are typically low (<10 μM) in uncontaminated groundwater (12), denitrification-associated MTBE biodegradation would be expected to be limited except in NO₃-contaminated aquifers or under engineered conditions. However, because NO₃ concentrations and denitrifying activity are often substantial in natural as well as waste-impacted surface waters (8, 15, 19, 21), these results hold considerable promise for the natural attenuation of MTBE in surface-water systems. Combined with previous demonstrations of a potential for aerobic MTBE mineralization in surface waters (5), these results indicate that bed sediment microbial processes represent a potentially important sink for MTBE in oxic and anoxic surface-water environments.