New evaluation scheme for two-dimensional isotope analysis to decipher biodegradation processes: Application to groundwater contamination by MTBE

Article abstract of DOI:10.1021/es049650j

Compound-specific analysis of stable carbon and hydrogen isotopes was used to assess the fate of the gasoline additive methyl tert-butyl ether (MTBE) and its major degradation product tert-butyl alcohol (TBA) in a groundwater plume at an industrial disposal site. We present a novel approach to evaluate two-dimensional compound-specific isotope data with the potential to identify reaction mechanisms and to quantify the extent of biodegradation at complex field sites. Due to the widespread contaminant plume, multiple MTBE sources, the presence of numerous other organic pollutants, and the complex biogeochemical and hydrological regime at the site, a traditional mass balance approach was not applicable. The isotopic composition of MTBE steadily changed from the source regions along the major contaminant plume (-26.4‰ to +40.0‰ (carbon); -73.1‰ to +60.3‰ (hydrogen)) indicating substantial biodegradation. Constant carbon isotopic signatures of TBA suggest the absence of TBA degradation at the site. Published carbon and hydrogen isotope fractionation data for biodegradation of MTBE under oxic and anoxic conditions, respectively, were examined and used to determine both the nature and the extent of in-situ biodegradation along the plume(s). The coupled evaluation of two-dimensional compound-specific isotope data explained both carbon and hydrogen fractionation data in a consistent way and indicate anaerobic biodegradation of MTBE along the entire plume. A novel scheme to reevaluate empiric isotopic enrichment factors (ε) in terms of theoretically based intrinsic carbon (¹²k/¹³k) and hydrogen ( ¹k/²k) kinetic isotope effects (KIE) is presented. Carbon and hydrogen KIE values, calculated for different potential reaction mechanisms, imply that anaerobic biodegradation of MTBE follows a S_N2-type reaction mechanism. Furthermore, our data suggest that additional removal process(es) such as evaporation contributed to the overall MTBE removal along the plume, a phenomenon that might be significant also for other field sites at tropic or subtropic climates with elevated groundwater temperatures (25°C).

Full text of DOI:10.1021/es049650j

Environ. Sci. Technol. 2005, 39, 1018-1029
potential reaction mechanisms, imply that anaerobic
New Evaluation Scheme for
Two-Dimensional Isotope Analysis
to Decipher Biodegradation
Processes: Application to
Groundwater Contamination by
MTBE
biodegradation of MTBE follows a SN2-type reaction
mechanism. Furthermore, our data suggest that additional
removal process(es) such as evaporation contributed to
the overall MTBE removal along the plume, a phenomenon
that might be significant also for other field sites at
tropic or subtropic climates with elevated groundwater
temperatures (25°C).
Introduction
L U C Z W A N K , † M I C H A E L B E R G ,
With an annual production of 21 million t in 1999 (1), MTBE
is among the organic chemicals with the highest production
volume worldwide. Despite its relatively recent introduction
as gasoline additive (2), MTBE has become one of the most
frequently detected groundwater contaminants (3). MTBE
tends to form widespread contaminant plumes in ground-
water due to its negligible sorption (4, 5) and relatively slow
biodegradation (6-8). As chemical transformation of MTBE
is insignificant in soil and groundwater biodegradation
commonly is thought to be the dominant removal mechanism
in the subsurface.
M A R T I N E L S N E R , ‡
T O R S T E N C . S C H M I D T , §
R E N EÄ P . S C H W A R Z E N B A C H , A N D
S T E F A N B . H A D E R L E I N * , §
Swiss Federal Institute for Environmental Science and
Technology (EAWAG), and Swiss Federal Institute of
Technology Zurich (ETHZ), Ueberlandstrasse 133,
CH-8600 Duebendorf, Switzerland
Under oxic conditions, biodegradation of MTBE has been
demonstrated in pure cultures (9-13). Although MTBE
biodegradation may proceed co-metabolically in the presence
of short-chain alkanes (14-16), the presence of other gasoline
components such as BTEX may slow or even completely
inhibit the degradation of MTBE (17). In the absence of
oxygen, MTBE degradation was described for microcosms
under denitrifying conditions (18, 19), sulfate-reducing
conditions (18, 20), iron-reducing conditions (18, 21, 22) and
methanogenic conditions (19, 23, 24). Degradation rates
under methanogenic conditions are slow (t1/2 ) 2.2-5.0 yr-1
in the field, 3.5 yr-1 in microcosms with added alkylbenzenes,
and 3 yr-1 in unamended microcosms) (25). The reaction
mechanism(s) for anaerobic biodegradation of MTBE are
unknown, but enzyme-catalyzed hydrolysis has been sug-
gested as a potential pathway (26).
Compound-specific analysis of stable carbon and hydrogen
isotopes was used to assess the fate of the gasoline
additive methyl tert-butyl ether (MTBE) and its major
degradation product tert-butyl alcohol (TBA) in a groundwater
plume at an industrial disposal site. We present a novel
approach to evaluate two-dimensional compound-specific
isotope data with the potential to identify reaction
mechanisms and to quantify the extent of biodegradation
at complex field sites. Due to the widespread contaminant
plume, multiple MTBE sources, the presence of numerous
other organic pollutants, and the complex biogeochemical
and hydrological regime at the site, a traditional mass balance
approach was not applicable. The isotopic composition
of MTBE steadily changed from the source regions along
the major contaminant plume (-26.4‰ to +40.0‰
tert-Butyl alcohol (TBA) is considered as the major product
of MTBE biodegradation under both oxic and anoxic condi-
tions. Compared to MTBE, little is known about the bio-
degradation of TBA (for an overview, see ref 27). TBA often
accumulates in the presence of MTBE and therefore has been
used in field studies as an indicator of MTBE biodegradation.
Biodegradation of MTBE without significant accumulation
of TBA (28-30) or preferred biodegradation of TBA (30) has
also been described. The potential lack of TBA accumulation
and the fact that TBA may be present as a constituent of
spilled gasoline (21, 31) complicate the assessment of in-situ
degradation of MTBE by a mass balance approach even under
well-defined boundary conditions (32, 33). At our and at many
other contaminated field sites, a “traditional” mass balance
approach to assess in-situ biodegradation is not applicable
due to the widespread contamination plume, the existence
of multiple contamination sources, and complex biogeochem-
ical and hydrological conditions. Hence, alternative methods
are necessary to assess quantitatively sources and the fate
of MTBE and TBA in the field.
(carbon); -73.1‰ to +60.3‰ (hydrogen)) indicating
substantial biodegradation. Constant carbon isotopic
signatures of TBA suggest the absence of TBA degradation
at the site. Published carbon and hydrogen isotope
fractionation data for biodegradation of MTBE under oxic
and anoxic conditions, respectively, were examined and
used to determine both the nature and the extent of in-
situ biodegradation along the plume(s). The coupled
evaluation of two-dimensional compound-specific isotope
data explained both carbon and hydrogen fractionation
data in a consistent way and indicate anaerobic biodegrada-
tion of MTBE along the entire plume. A novel scheme to
reevaluate empiric isotopic enrichment factors (ꢀ) in terms
of theoretically based intrinsic carbon (12k/13k) and
hydrogen (1k/2k) kinetic isotope effects (KIE) is presented.
Carbon and hydrogen KIE values, calculated for different
* Corresponding author: phone: +49 (0)7071 297 3148; fax: +49
(0)7071 295 139; e-mail: haderlein@uni-tuebingen.de.
† Present address: s&e consult, B.P. 2453, L-1024 Luxembourg.
‡ Present address: Stable Isotope Laboratory, Department of
Geology, University of Toronto, 22 Russell St., Toronto, ON, M5S 3B1
Canada.
Here, we describe a two-dimensional compound-specific
carbon and hydrogen stable isotope approach to evaluate
biodegradation of MTBE and TBA in a contaminated
groundwater plume at an industrial disposal site. Compound-
specific isotope analysis (CSIA) has been applied to different
organic groundwater contaminants including chlorinated
solvents (34-39) or BTEX (40, 41) and is recognized as a
§ Present address: Center for Applied Geosciences (ZAG), Eber-
hard-Karls University Tuebingen, Wilhelmstr. 56, D-72074 Tuebingen,
Germany.
9
1018 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 4, 2005
10.1021/es049650j CCC: $30.25
2005 American Chemical Society
Published on Web 12/22/2004

promising tool to elucidate transformation processes of such
pollutants in the subsurface. Our two-dimensional stable
isotope approach is based on a comparison of measured
field data with carbon and hydrogen isotope fractionation
data reported for biodegradation of MTBE under oxic and
anoxic conditions (42, 43). Specifically, a novel evaluation
procedure allowed for the first time the interpretation of
observed empirical fractionation of MTBE in terms of
theoretically based intrinsic kinetic isotope effects (KIE). It
was therefore not only possible to assess the nature and extent
of in-situ biodegradation but also to decipher the pathway
of anaerobic MTBE degradation.
entire molecule and not only of the atom(s) involved in the
reaction. It is, however, the isotopic fractionation of the
atom(s) at the reactive position that has to be known in order
to use this information for mechanistic interpretations. Such
intrinsic fractionation is commonly expressed as kinetic
isotope effect (KIE), which is the ratio between the reaction
rate constant of the light isotope and the heavy isotope (e.g.,
1k/2k and 12k/13k for hydrogen and carbon KIEs, respectively).
Hence, if molecules contain more than one atom of the
element for which isotope fractionation is studied, two
different effects must be taken into account. One effect is the
“dilution” of the measured isotopic shift due to the presence
of atoms of the studied element at nonreacting positions in
a molecule. In the case of MTBE oxidation for example, 13
C
Theoretical Background
and 2H isotope fractionation will occur solely at the methoxy
group, whereas the isotopic composition of the tert-butyl
group will remain unchanged. This invariant isotope ratio in
the nonreacting group will, however, contribute to a great
extent to the average isotopic signature of the whole molecule;
hence, the observed effect will be much smaller than the
intrinsic isotope effect occurring at the reactive bond.
As derived in the Supporting Information, this dilution of
the observed isotopic shift ∆δ13C ) δ13Cx - δ13C0 (and ∆δ2H
) δ2Hx - δ2H0) can be corrected by an appropriate factor
n/x, where n is the total number of atoms of one element
present in the molecule and x is the number of atoms of this
element that are located at reactive positions. In the case of
MTBE oxidation mentioned above, the correction factor
would then equal 5/1 and 12/3 for 13C and 2H isotope
fractionation, respectively (see Table 3 for correction factors
n/x for all three possible reaction mechanisms). It should be
mentioned that the underlying assumption for the application
of this correction is a statistical distribution of the heavy
isotopes in the parent compound, here MTBE. The validity
of this assumption has been discussed to some detail in ref
44. The term n/x may be replaced by the experimentally
determined isotopic distribution if such data are available.
For example, as discussed below, an analysis of commercially
available MTBE in this study showed that thetert-butyl group
was enriched in δ13C by about 2.6‰ and the methoxy group
was depleted by about 8‰ as compared to the average δ13C
of MTBE (-28.13‰). Consequently, because the relative
abundance of 13C was 0.8% lower in the methoxy group, a
correction factor of n/x ) 5/(1-0.008) ) 5.04 would be
appropriate instead of (n/x ) 5/1) ) 5.00 if a sample of this
particular production batch reacted by microbial oxidation.
The fact that the deviation between the two values is very
low illustrates that with carbon isotopes the effect of a
nonstatistical isotope distribution is generally negligible;
hence, the factor n/x is a good approximation.
As bond cleavage may be accompanied by a kinetic isotope
effect, the isotopic signatures of the reactant and the pro-
duct(s) may change during a given transformation reaction.
The remaining fraction of the parent compound may thus
become progressively enriched in heavier isotopes since
bonds containing light isotopes are slightly less stable and
thus more reactive than those with heavy isotopes. Isotopic
enrichment factors (ꢀ, in per mil) are commonly used to
quantify the apparent isotopic shift during a reaction. ꢀvalues
(in per mil) can be derived from a linearized form of the
Rayleigh eq 1. A detailed derivation of eq 1 can be found in
ref 44:
δ13Cx + 1000
ꢀ
ln
)
‚ln f
(1)
13
(
)
1000
δ C0 + 1000
Here, δ13Cx and δ13C0 refer to the carbon isotopic signatures
of MTBE at well x and in the contamination source,
respectively. The average carbon isotopic signature of MTBE
within the source zone was used as δ13C0. f is the remaining
fraction of the reacting compound and is defined as the ratio
of the MTBE concentration at well x as compared to a
reference concentration of MTBE (e.g., at the source of
contamination):
[MTBE]x
f )
(2)
[MTBE]0
Solving eq 1 for f yields
1000/ꢀ
δ13Cx + 1000
f )
(3)
13
(
)
δ C0 + 1000
which allows the calculation of the expected concentration
ratio using the determined isotopic signatures and assuming
that the difference between C0 and Cx is only due to a
transformation reaction with a given ꢀ. On the basis of this
ratio, the extent of (bio)degradation (B) can be estimated
(45, 46):
If eq 1 is rewritten as:
1000 + δ13C0 + ∆δ13C
ꢀ
ln
)
ln f
(6)
1000 + δ13C0
(
)
1000
Applying the correction factor n/x leads to:
B ) (1 - f ) × 100 (%)
(4)
Assuming (i) similar transport processes for TBA and MTBE,
(ii) MTBE degradation as the only source of TBA, and (iii) the
absence of TBA (bio)transformation, the expected TBA
concentration in well x may be estimated using B and the
MTBE concentration found at this well using eq 5:
n
x
1000 + δ13C0 + ∆δ13C
ꢀreactive position
ln
)
ln f (7)
(
)
1000 + δ13C0
1000
The whole correction procedure will be discussed in
greater detail in a critical review of Elsner et al. (currently in
preparation), and a mathematical derivation of eq 7 is
included as Supporting Information.
B
1 - B
[TBA]x )
[MTBE]x (µM)
(5)
Determination of 12k/13k and 1k/2k Values from Enrich-
ment Factors. Isotopic signatures determined in degradation
studies reflect the average isotopic composition over the
Note that for reevaluation, original data ofδ13C0 and ∆δ13C
(or δ2H0 and ∆δ2H) rather than published enrichment factors
ꢀ must be used. Whereas this is a straightforward process if
9
VOL. 39, NO. 4, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 1019

FIGURE 1. Map of the investigated field site depicting the major MTBE disposal pond, monitoring wells, and groundwater heads. The
main MTBE source is located at well PM 04. In addition to the major disposal pond PM04, there are other infiltration ponds of minor
significance for MTBE input where phenol and other organic compounds were disposed.
newly acquired experimental data are available, in past
studies these values are generally not reported. In such cases
it is convenient to use an approximate equation:
After applying both corrections, the intrinsic kinetic
isotope effect (KIE) is then obtained by:
13k
12k
ꢀreactive position
1
KIE
)
) z
+ 1
(9)
ꢀreactive position ≈ n/x × ꢀ
(8)
1000
where ꢀreactive position is an approximate estimate of the position-
specific enrichment factor, and ꢀ is the “traditional” value
reported in most literature studies. Equation 8 gives good
approximations as long as changes in isotopic signatures
∆δ2H or ∆δ13C do not become too large (see Supporting
Information).
The second effect besides dilution that has to be taken
into account is intramolecular competition. Depending on
the symmetry of the molecule, it is possible that the
investigated element is present at z chemically equivalent,
reactive positions in the molecule. In the case of the oxidation
of MTBE, this is the case for the three hydrogen atoms of the
methoxy group. At low natural isotopic abundance only one
of the three positions at most is occupied by deuterium.
Hence, if a bond to this heavy isotope is broken, the reaction
is in competition with breakage of bonds to the two light
hydrogen isotopes in equivalent positions. To correct for
this effect, the ꢀreactive position value resulting from eq 7 has to
be multiplied by z (see Supporting Information and review
by Elsner et al. (in preparation)).
For the reevaluation of literature data, the corresponding
data sets were obtained from graphs found in the cited
publications using an Excel-based digitizing software (GrabIt!
XP, Datatrend Software, Raleigh, NC).
Experimental Section
Site Description. The investigated field site (Figure 1) is a
former industrial landfill in South America where phenol
and MTBE have been disposed in open ponds. Phenol had
been disposed for over 20 yr until 1989. Over a span of several
years, MTBE was used as solvent in chemical synthesis and
was also disposed. The ponds are located on a shallow hill,
leading to a radial groundwater flow pattern (Figure 1).
The water table varies between 10 and 20 m below the
topsoil; the aquifer is shallow (10-15 m) and has a low
hydraulic conductivity (10-5 cm/s). It is confined by an
underlying zone of basaltic rock from volcanic origin. This
impermeable zone is fractured, and preferential flow along
these fissures is likely to occur. The average annual ground-
9
1020 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 4, 2005

+
water temperature is between 25 and 30 °C. The assessment
of the biodegradation of MTBE is difficult at this particular
site, because 13 other disposal ponds (not shown in Figure
1) are located in the area between wells PM 10 and PM 04.
Production wastes containing organic contaminants such
as BTEX, phenol, and isopropylbenzene have been disposed
in these additional ponds. MTBE was also disposed near
well PM01 although apparently of minor significance to the
overall MTBE contamination resulting primarily from dis-
posal in the pond near PM04. Two wetlands are situated
400-600 m downstream of the disposal site. To protect these
wetlands from contamination, three pumping wells and two
deep horizontal pumping wells have been installed as
hydraulic barriers. The installation of these wells has changed
the hydrological regime at the site, leading to potentially
higher transport velocities of the contaminants. Forty-seven
permanent monitoring wells have been installed at the site
and are sampled semiannually. The monitoring wells do not
all extend to the bottom of the aquifer, hence varying dilution
effects during sampling cannot be excluded.
Sample Collection and Storage. The samples used for
this study were taken during two sampling campaigns. Before
sampling, the different wells had been pre-pumped for several
hours. Water temperature and pH data were determined on
site. Measurements with an oxygen-sensitive electrode on
selected samples indicated an oxygen depletion (<1 mg/L)
of the groundwater. The samples for the determination of
inorganic parameters were collected in polyethylene flasks,
acidified, and kept at 4 °C until analysis. Samples for the
determination of concentrations and isotopic signatures of
MTBE and TBA were filled without headspace into 40 mL
glass vials. The vials were closed with PTFE-sealed screw
caps and stored at 4 °C until analysis.
spectrometer. H3 is formed in the ion source of the mass
spectrometer, and hence the precise detection at m/z ) 3 of
HD+ is hampered. The production of H3+ is correlated to the
amount of H2 entering the source and can be corrected with
the H3+ factor. To allow a good DH/H2 determination, a low
+
and stable H3 factor is a prerequisite. This factor was
determined daily by measuring a set of nine reference gas
peaks of increasing amplitudes. This factor was constant (3.21
( 0.14 ppm/nA). To verify the reproducibility of the isotopic
measurements during one sequence of samples as well as
throughout the measurement campaign, external standards
of known carbon isotopic signature were measured after 3-5
samples. If not stated otherwise, the isotopic signatures
correspond to the average of at least three replicate mea-
surements.
Complete MTBE Hydrolysis under Acidic Conditions.
MTBE of known isotopic composition was hydrolyzed to TBA
under acidic conditions. The hydrolysis reaction was carried
out in 20-mL headspace vials sealed with PTFE-lined crimp
caps. The vials were filled with 17 mL of deionized water,
and subsequently 1.7 mL of concentrated HCl was added.
The vials were placed in the agitator of a CombiPAL
Autosampler (CTC, Zwingen, Switzerland) heated at 60 °C.
After an equilibration time of 1 h, 0.5 µL of a pure MTBE
standard was spiked into the reaction vials. To determine
the reaction kinetics of this reaction, 75 µL of sample was
taken at appropriate time intervals from the reaction vials
and neutralized with NaOH. A control reaction at neutral pH
was sampled in the same way. Instead of adding NaOH, the
control samples were diluted accordingly with distilled water.
The concentrations of MTBE and TBA were measured using
DAI-GC/MS. The mass spectrometer was acquiring in the
scan mode (m/z 40-250) in order to verify that no additional
products formed during hydrolysis. The high methanol
concentrations produced during the reaction led to an
overestimation of the TBA concentrations, due to peak
broadening, but since no additional products were formed,
complete MTBE hydrolysis can be expected. For the isotopic
determination of the MTBE tert-butyl group, one vial was
spiked with 1 mL of pure MTBE of known isotopic signature,
and one vial was spiked with 200 mL of a methanolic solution
(1:10 v/v) of a field sample from the MTBE source that
contained no TBA. To ensure complete hydrolysis, the vials
were kept at 60 °C for 5 h. After hydrolysis the samples were
neutralized using NaOH (1 M) and kept at 4°C until analysis.
Carbon isotopic compositions were determined using purge-
and-trap extraction coupled on-line to a GC-IRMS system
following a method described elsewhere (47).
Analytical Methods. Concentrations of MTBE and of its
major degradation products TBA as well as BTEX were
determined using direct aqueous injection gas chromatog-
raphy/mass spectrometry (DAI-GC/MS) using the procedure
described by Zwank et al. (31), on a gas chromatograph (GC
8000, Fisons, Manchester, U.K.) coupled to a mass spectro-
metric detector (single quadrupole MD 800, Fisons, Manches-
ter, U.K.).
To achieve a sufficient sensitivity for carbon and hydrogen
isotopic analysis, the samples were extracted using a solid-
phase microextraction (SPME) procedure (47). The analytes
were extracted for 30 min from the samples (to which 4 M
NaCl was previously added) by directly immersing a Car-
boxen-PDMS fiber (75 µm, Supelco, Bellefonte, PA) into the
sample. The fractionation of the carbon isotopic signature
caused by SPME extraction was very small and highly
reproducible (47). After SPME extraction, the analytes were
thermally desorbed for 1 min in a split-splitless injector
(270°C) equipped with a deactivated SPME liner. To reduce
the peak tailing, a splitless time of only 0.75 min was used.
The chromatographic separation of the analytes for IRMS
was achieved on a gas chromatograph (Trace GC, Thermo
Finnigan) equipped with a 60 m × 0.32 mm Stabilwax fused
silica column (1 µm film cross-bonded Carbowax poly-
(ethylene glycol)) purchased from Restek (Bellefonte, PA)
with the following temperature program: 2 min at 45°C, then
to 90°C at 7.5 °C/min, 2 min at 90 °C, then to 200 °C at 30
°C/min, and 6 min at 200 °C. For carbon isotopic measure-
ments the analytes were combusted after separation in a
combustion interface (GC Combustion III, Thermo Finnigan
MAT, Bremen, Germany) maintained at 940 °C, and the
resulting CO2 was analyzed in an isotope ratio mass
spectrometer (DeltaPLUSXL, Thermo Finnigan MAT, Bre-
men, Germany). The catalyst in the combustion interface
was oxidized regularly after ∼40 samples. For hydrogen
isotopic analysis, the analytes were pyrolyzed at 1400 °C,
and the resulting H2 was analyzed in the isotope ratio mass
Results and Discussion
Position-Specific Isotope Signatures in MTBE. To evaluate
whether the isotopic signature of thetert-butyl group of MTBE
is significantly different from the isotopic composition of
the methoxy group, MTBE of a known isotopic composition
was hydrolyzed to TBA under acidic conditions. As MTBE
was completely transformed to TBA, which was stable under
the given conditions, the carbon isotopic signature of TBA
corresponds to the one of thetert-butyl group of MTBE. MTBE
was hydrolyzed with an observed pseudo-first-order constant
of -0.0233 min-1 (R2 ) 0.98) corresponding to a half-life of
30 min (at pH 0 and 60 °C). The reaction kinetics were in
rather good agreement with rates and activation energies
determined by O’Reilly et al. (26) (t1/2 ) 41 min).
To determine the position-specific carbon isotopic com-
position of MTBE, the isotopic composition of the MTBE
standard used in the hydrolysis experiment was measured
(δ13C ) -28.13 ( 0.15‰). After complete hydrolysis of MTBE,
the δ13C of the produced TBA was -25.49 ( 0.10‰, reflecting
the isotopic composition of the tert-butyl group of the parent
MTBE. With the help of an “isotopic mass balance” the δ13C
9
VOL. 39, NO. 4, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 1021

FIGURE 2. Extension of the MTBE (upper part) and TBA (lower part) plume as well as carbon and hydrogen isotopic signatures of these
two compounds. The upper isotopic value corresponds to δ13C vs VPDB and the lower value to δ2H vs VSMOW.
of the methoxy group was calculated to be δ13C ) -36.35 (
1.11‰. Thus, the isotopic signature of the methoxy group
and the tert-butyl group of MTBE differed by more than 10
‰. To our knowledge, this is the first indication of position-
specific isotope signatures in MTBE.
show almost identical carbon (-26.44 ( 0.15‰) and similar
hydrogen signatures (-71.4. ( 7.6‰) indicates that these
wells share MTBE of the same isotopic composition and that
biodegradation is not significant next to the source zone.
Aerobic degradation of MTBE is known to be accompanied
by a slight and robust carbon isotopic enrichment (-1.5 (
0.1‰ < ꢀc < -2.0 ( 0.1‰) (28) regardless of whether MTBE
was the sole carbon source or degraded co-metabolically
with 3-methylpentane as substrate. Similar carbon enrich-
ment factors were obtained in microcosms enriched from
contaminated sites (-1.5 ( 0.1‰ < ꢀc < -1.8 ( 0.1‰) and
in pure cultures (strain PM1; -2.0 ( 0.1‰ < ꢀc < -2.4 (
0.3‰) (48). Although these ꢀc values are relatively small, they
differ significantly from those caused by physical processes
such as organic phase/gas-phase partitioning (ꢀc ) 0.50 (
0.15‰), aqueous phase/gas-phase partitioning (ꢀc ) 0.17 (
0.05‰), and organic phase/aqueous phase partitioning (ꢀc
) 0.18 ( 0.24‰) (28). Hydrogen enrichment factors (ꢀH) for
aerobic MTBE biodegradation ranged from -29 ( 4‰ to
-66 ( 3‰ in microcosm experiments and from -33 ( 5‰
Qualitative Assessment of in-situ Biodegradation of
MTBE. As can be seen from Figure 2, the MTBE plume at the
contaminated site was very widespread (over a length of>500
m) with a maximum concentration of 1.7 g/L (aqueous
solubility of MTBE ) 48 g/L) (5) at the major source zone
(PM 04). At this well the TBA concentration was below the
detection limit (e.g., 2 µg/L) but increased with distance from
the MTBE source, suggesting that TBA was not present as
co-contaminant but rather was formed by biodegradation of
MTBE. Highest TBA concentrations (>22.6 mg/L) were
measured downstream of the highest MTBE concentrations.
The MTBE concentration contours depicted in Figure 2 show
a major hotspot (PM 04) from where the plume stretches out
radially. The fact that several wells near the suspected major
MTBE source near PM 04 (PM 10, PM01, PZ01, and PZ03)
9
1022 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 4, 2005

TABLE 1. Concentrations and Isotopic Signatures (δ13C and δ2H) of MTBE and TBA
MTBE
TBA
sampling
well
concn
(µg/L)
δ13C (‰ vs
VPDB)
δ2H (‰ vs
VSMOW)
concn
(µg/L)
δ13C (‰ vs
VPDB)
δ2H (‰ vs
VSMOW)
PM 04a
PZ 01
PZ 04
PM 01
PZ 06
PZ 03
PZ 07
PM 10
PZ 02
PM 19
PM 24
PM 23
PB 01
PZ 11
PM 03
PB 02
PM 14
PB 03
PZ 15
PZ 14
PM 02
PZ 05
PM 16
PB 06
PB 05
PM 09
PZ 13
1 700 000
23 700
10 800
6 060
5 740
4 690
4 590
3 900
3 870
2 690
1 420
726
-26.35 ( 0.73
-26.47 ( 0.25
-25.92 ( 0.19
-26.49 ( 0.02
-21.11 ( 0.06
-26.64 ( 0.17
-18.12 ( 0.02
-26.26 ( 0.44
-21.07 ( 0.07
-25.18 ( 0.54
-13.63 ( 0.07
-23.35 ( 0.08
-22.57 ( 0.38
-22.25 ( 0.17
-6.54 ( 0.07
-23.03 ( 0.33
-25.98 ( 0.44
-25.12 ( 0.09
-19.81 ( 0.07
+40.04 ( 1.15
-16.21 ( 0.15
-25.83 ( 0.06
-20.97 ( 0.18
-22.77 ( 0.05
-22.53 ( 0.06
+3.52 ( 0.05
-13.20 ( 0.83
-71.3 ( 1.9
-61.6 ( 3.3
-42.5 ( 5.4
-80.9 ( 1.4
-56.5 ( 1.2
-76.2 ( 3.8
-56.7d
-67.0 ( 3.1
-48.6 ( 1.6
-66.8 ( 9.5
-61.8 ( 2.8
-66.0 ( 3.0
-58.2 ( 4.3
ndc
-37.6 ( 10
-74.0 ( 6.3
-61.3 ( 2.2
-63.8 ( 1.2
-51.1 ( 1.8
+60.3 ( 11
-54.2 ( 1.9
ndc
-57.9 ( 4.5
-59.3 ( 2.5
-61.3 ( 10
-16.6 ( 1.3
-109.6 ( 11
<2b
6 610
8 790
10 200
22 600
5 190
8 740
6 140
2 120
10 300
4 440
459
1 270
2 030
7 130
1 730
309
296
746
2 270
1 540
1 780
363
183
43
1 820
1 520
-25.37 ( 0.63
-24.42 ( 0.34
-25.44 ( 0.13
-24.61 ( 0.20
-27.63d
-24.76 ( 0.02
-24.75 ( 1.00
-24.57 ( 0.16
-25.26d
-25.10 ( 0.08
-25.29 ( 0.11
-24.64 ( 0.17
-26.15d
-24.69 ( 0.08
-24.91 ( 0.13
-25.90 ( 0.34
-24.69 ( 0.12
-23.37 ( 0.13
-24.62 ( 0.07
-24.61 ( 0.11
-25.31 ( 1.33
-24.99 ( 0.52
-24.95 ( 0.13
-24.92 ( 1.39
-21.98 ( 0.09
-24.56 ( 0.17
ndc
ndc
-116.2 ( 11
-61.3 ( 7.0
ndc
-68.8 ( 0.6
ndc
-60.0 ( 18
ndc
-102.2 ( 3.8
-78.0 ( 1.1
-98.2 ( 4.3
ndc
-76.7 ( 10
-100.9 ( 6.3
-104.5 ( 14
-95.8 ( 7.6
-48.3 ( 4.4
-82.0 ( 3.1
-63.3 ( 3.6
ndc
-63.4 ( 21
-92.0 ( 11
ndc
-75.5 ( 9.7
-95.4 ( 13
626
575
506
493
401
348
318
278
270
265
204
198
88
58
46
a Source well. b Below method detection limit. c Not determined. d n ) 1.
to -37 ( 4‰ for the pure culture (strain PM1) (48). It was
proposed that the larger but less reproducible hydrogen
isotopic fractionation was a sensitive qualitative indicator of
in-situ degradation, whereas the more reproducible carbon
isotopic fractionation was best suited for quantification of
biodegradation in the field. Aerobic co-metabolic degradation
of TBA in the laboratory resulted in a carbon isotopic
enrichment (ꢀc) of -4.21 ( 0.07‰ (28). In a “mainly” anoxic
plume, TBA degradation was accompanied by a change of
δ13C values from -28.6‰ in the source zone to -22‰ with
decreasing TBA concentrations (49).
Anaerobic biodegradation of MTBE appears to also cause
a consistent but substantially higher carbon isotopic enrich-
ment (ꢀc ) -8.1 ( 0.85‰ in the field and ꢀc ) -9.2 ( 5.0‰
in microcosms) (50) as compared to aerobic conditions.
Provided that the few studies available in the peer-reviewed
literature are representative, this difference of carbon isotope
fractionation may indicate that different reaction mecha-
nisms are involved in aerobic and anaerobic biodegradation
of MTBE and thus may be used to characterize the nature
of biodegradation in the field. Furthermore, the hydrogen
isotopic fractionation under anaerobic field conditions
appears to be less pronounced than under aerobic conditions
(ꢀH ) -11.4‰) (51).
FIGURE 3. Rayleigh plot of measured δ13C-field data for MTBE.
Open circles represent data from the major plume near well PM
04. The concentration decline (f ) C/C0) in the field does not match
with carbon isotopic shifts expected for anaerobic (E ) -8.62‰)
or aerobic (E ) -1.82‰) biodegradation.
enrichment in 13C as compared to the wells downstream
that had lower concentrations, suggesting substantial bio-
degradation despite high concentrations. The apparent
mismatch of MTBE concentration and δ13C value in this
particular well might be due to changes in flow patterns and
mixing imposed by the intense pumping in wells PB 04, PB
05, and PB 06 (see Figure 1). With the exception of well PM
09 (-20.98 ( 0.09‰), the carbon isotopic signature of TBA
was more or less invariable within the plume (-25.02 (
0.75‰) but slightly enriched in 13C as compared to MTBE
at the source. This effect can be explained by a position-
specific difference in the carbon isotopic signature of the
methoxy and tert-butyl group in MTBE as found for a MTBE
standard (see above). The hydrogen isotopic signature of
TBA did not show a consistent trend possibly due to a variable
hydrogen isotopic signature of the tert-butyl group of MTBE
In both studies δ13C values of TBA did not show a
pronounced fractionation. Under oxic conditions, the initial
step of MTBE degradation takes place at the methoxy group
which is eliminated during the course of the reaction. Thus,
a substantial carbon isotopic change of the product TBA is
not expected.
Table 1 shows measured concentrations of MTBE and
TBA together with their carbon and hydrogen isotopic
signatures. A significant isotopic enrichment of MTBE with
increasing distance from the source is evident withδ13C values
of up to +40.04‰ versus VPDB and δ2H of up to +60.3‰
versus VSMOW. MTBE in well PM24 was present at high
concentrations but at the same time showed a significant
9
VOL. 39, NO. 4, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 1023

FIGURE 4. Plot of MTBE concentration versus chloride concentration
(2) and conductivity (O).
FIGURE 5. Hydrogen isotopic enrichment vs relative concentrations
of MTBE calculated from carbon isotopic enrichment factors for
anoxic conditions (2) and oxic conditions (O). y-axis: measured
hydrogen enrichment; x-axis: fraction of remaining MTBE calculated
from measured carbon isotopic fractionation according to eq 3. The
lines correspond to a linear regression model; their slopes represent
the corresponding hydrogen enrichment factors. Note the different
scales of the x-axis for aerobic and anaerobic biodegradation.
disposed over the years, and/or hydrogen exchange at the
hydroxyl group of TBA.
Quantification of Biodegradation. In the evaluation of
field isotope data usually “field enrichment factors” are
calculated by constructing a Rayleigh plot from measured
concentration and isotope ratio data according to eq 1. The
ꢀvalues thus determined are subsequently compared to those
obtained from laboratory experiments conducted under
controlled conditions. Clearly, because along a contaminant
plume in the field concentrations do not only decrease due
to transformation processes but also owing to other processes
such as dilution and dispersion, the true isotope fractionation
might be underestimated if concentration data were not
corrected appropriately.
In a tentative approach to determine “true" field enrich-
ment factors, we tried to minimize effects that can be caused
by the above-mentioned variable hydrological conditions as
well as pumping and focused on the wells that are located
around the major contamination source (PM 04). As can be
seen from Figure 3, there is no apparent simple correlation
between the MTBE concentration and the carbon isotopic
enrichment patterns (R2 ) 0.25). The field data did not match
with the range expected for carbon isotopic fractionation of
MTBE under aerobic or anaerobic conditions. In most cases,
the observed relative concentrations were too low to explain
the observed isotopic values. Hence other processes such as
dilution and elimination processes (e.g., volatilization) that
do not cause significant isotopic fractionations also have to
be taken into account.
To correct for dilution processes, MTBE concentrations
can be normalized to a suitable conservative tracer. At this
site, however, due to the multiple contamination sources
and the long contamination history of the disposal site, no
suitable conservative tracer could be identified. As an
example, concentrations of MTBE and chloride together with
conductivity data are shown in Figure 4. Hence, the traditional
Rayleigh-type approach was not appropriate to delineate
biodegradation of MTBE at this complex field site.
An alternative approach to determine the extent of (bio)-
degradation has been suggested by refs 40, 41, 52, and 53.
Assuming a constant 13C signature of the source, the fraction
of MTBE is calculated that results from the enrichment in
13C of MTBE downgradient of the source using eqs 3 and 4.
This calculated f represents the concentration decrease that
would be expected along a streamline assuming plug flow
(no mixing) and one degradation process with a constant
enrichment factor. However, because the nature of the
degradation process (i.e., aerobic or anaerobic biodegrada-
tion) cannot be determined in our case, and because both
processes are associated with significantly different enrich-
ment factors (ꢀ), an unambiguous determination of B by eq
4 was not possible. For example, assuming aerobic biodeg-
radation associated with an average isotopic enrichment
factor of ꢀ ) -1.82‰ (28, 48), the extent of expected
biodegradation in wells PZ 06 and PM 23 is 95% and 82%,
respectively. However, if the average isotopic enrichment
factor for anaerobic biodegradation is used (ꢀ ) -8.63‰)
(51), biodegradation only accounted for 46% and 30% in the
same wells. Thus, an approach has to be developed that allows
differentiating between anaerobic and aerobic biodegrada-
tion of MTBE at hydrogeologically and biogeochemically
complex field sites.
A first useful approach is to subject the values of f that
may be calculated according to eq 3 to a plausibility test. For
example, in well PZ 14 a highly enriched carbon isotope ratio
of δ13C ) +40‰ was measured (see Table 1). If one assumes
that this value has been reached by fractionation during
anerobic degradation of MTBE, a calculation according to
eq 3 with (ꢀC(anaerobic) ) -8.63‰) would give a value of ln f
≈ -7 (see value of x-axis belonging to the black triangle
farthest to the left in Figure 5). Conversely, under the
assumption that fractionation leading to δ13C ) +40‰ had
taken place during aerobic fractionation, the calculation
according to eq 3 with (ꢀC(aerobic) ) -1.82‰) would give a
value of ln f ≈ -37 (see value of x-axis belonging to the open
circle farthest to the left in Figure 5). The latter concentration
would be clearly too small to be analyzed by CSIA so that for
this particular point the aerobic pathway can be excluded.
An analysis, however, is less conclusive for wells where more
moderate enrichment has taken place so both aerobic and
anaerobic transformation are possible scenarios. (Calculated
ln f then corresponds to the x-axis values associated with
triangles and circles farther to the right in Figure 5.) Clearly,
in such cases an alternative “plausibility test” is needed.
Fortunately, the microbial transformation of MTBE under
oxic and anoxic conditions is not only associated with
different carbon isotopic enrichments but also simulta-
neously with significantly different hydrogen isotopic en-
richments. Our key to the solution, therefore, was to correlate
on one hand the ln f values that were calculated from
measured carbon isotope ratios according to eq 3 (x-axis of
Figure 5) and on the other the hydrogen isotope ratios that
were measured in the same samples (y-axis of Figure 5). (Note
that Figure 5 has two scales for the x-axis, one for anaerobic
degradation (associated with 2) and one for aerobic deg-
radation (associated with O).)
9
1024 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 4, 2005

TABLE 2. Calculated Extent of Biodegradation and Measured
vs Predicted MTBE and TBA Concentrations
measd pred
extent of TBA
TBA
sampling f (MTBE)
well measd
f (MTBE)a
calcd
biodeg concn concn [TBA]meas
/
(%) (µM) (µM) [TBA]pred
PM 04b 1.00E+00 0.99E+00
0
38
-2
-3
PZ 03c
PZ 01c
2.75E-03 1.03E+00
1.39E-02 1.01E+00
-3
-1
0
1
4
5
6
70
89
140
83
4
120
24
140
4
6
23
2
17
1
27
300
29
5
-35
-29
PM 01c 3.56E-03 1.01E+00
-1 -140
PM 10c 2.29E-03 9.87E-01
1
0
7
0
5
1
4
3
1
4
1
4
56
38
2
4
86
7
83
17
PM 14
PZ 04
PZ 05
PM 19
PB 03
PM 23
PB 02
PB 06
PB 01
PB 05
PZ 11
PZ 06
PZ 02
PM 16
PZ 15
PZ 07
PM 02
PM 24
PZ 13
PM 03
PM 09
PZ 14
2.36E-04 9.55E-01
6.26E-03 9.48E-01
1.55E-04 9.38E-01
1.58E-03 8.68E-01
2.05E-04 8.62E-01
4.25E-04 6.99E-01
2.90E-04 6.73E-01
1.16E-04 6.52E-01
3.68E-04 6.37E-01
5.18E-05 6.34E-01
3.38E-04 6.13E-01
3.37E-03 5.36E-01
2.27E-03 5.33E-01
1.20E-04 5.27E-01
1.87E-04 4.59E-01
2.69E-03 3.76E-01
1.59E-04 3.00E-01
8.35E-04 2.22E-01
2.71E-05 2.11E-01
2.98E-04 9.67E-02
3.40E-05 3.01E-02
1.63E-04 4.78E-04
13
14
30
33
35
36
37
39
46
47
47
54
62
70
78
79
90
97
100
28
4
1.5
7.6
2
4.2
1
6.7
5.4
0.8
2.5
2.5
1.4
3
FIGURE 6. Plot of hydrogen versus carbon isotopic shifts for MTBE.
(O) aerobic biodegradation in a batch experiment (48); (4) anaerobic
biodegradation at different field sites (51); (b) field data from the
major MTBE plume of our study.
in-situ biodegradation would have been impossible. Due to
the fact that we applied the isotopic data of one element
(carbon in this case) for the correction of the concentration,
we were able to use isotopic information of the second
element to get hints on the process responsible for the
observed isotopic fractionation. This new approach relies
on the combined evaluation of carbon and hydrogen isotopic
data collected at the field site and allows to identify (i) whether
the observed isotopic enrichment of carbon and hydrogen
isotopes at the site is associated with one single process as
well as (ii) the nature of this process.
10
120
21
60
21
96
25
56
2
54
21
1.1
10
1.8
1.2
0
31 6600
a Calculated fraction of remaining MTBE under the assumption that
biodegradation is the only major elimination process. b Concentration
set as initial value (f (MTBE) ) 1) for the downgradient wells. c Well
within the estimated source zone next to the major MTBE disposal
pond at PM04.
Hence, based on a plausible hypothesis for the type of
process causing the monitored fractionation, it was possible
to calculate the extent of biodegradation using eq 4.
Furthermore, expected TBA concentrations were calculated
for each sampling well, using eq 5. The results of this
calculation are also shown in Table 2 and are based on the
following assumptions: (i) no differences in the transport
behavior of MTBE and TBA, (ii) MTBE biodegradation is the
only source of TBA formation, (iii) constant isotopic signa-
tures of TBA, and (iv) the absence of TBA biodegradation.
The TBA concentrations were generally underestimated
with this approach. Only in well PZ 14, where the highest
TBA concentration would be expected based on the large
isotopic enrichment of MTBE, much less TBA was found
than would be expected. However, the general underestima-
tion can be explained by the fact that the measured MTBE
concentrations used for this calculation ([MTBE]x) are not
only lowered by biodegradation (quantified using isotopic
fractionation) or dilution (which should be roughly the same
for TBA) but by an additional process that does not yield TBA
and does not change the isotopic composition of the parent
compound. Owing to the relatively high temperature of the
groundwater (∼25 °C) and the higher water-air partition
constant of MTBE as compared to TBA (5), volatilization might
be significant for the attenuation of the MTBE concentrations
at this particular site. Volatilization is known to be associated
with a small inverse isotopic effect, which means that
molecules with the heavy isotope preferentially partition into
the gas phase (54, 55). For MTBE, this effect is, however,
small (+0.17 ( 0.05‰) (28) as compared to the overall
fractionation observed at the field site. The calculated TBA
concentration in the source wells is far off the measured
value since even small deviations in isotopic signatures from
the average carbon isotopic signature of the source wells
lead to high deviations in the calculated TBA concentrations
considering the high initial MTBE concentrations. Generally,
however, the differences between measured and calculated
TBA concentrations corroborate the hypothesis that alterna-
As is evident from Figure 5, the calculation of f based on
carbon isotopic fractionation resulted in a strong correlation
(R2 ) 0.94) suggesting that this approach allows to distinguish
MTBE concentration decrease associated with in-situ bio-
degradation from dilution or other nonfractionating pro-
cesses. From the slope of the correlations shown in Figure
5, hydrogen isotopic enrichments factors (ꢀH) could be
determined for both aerobic (-3.3 ( 0.4‰) and anaerobic
(-15.6 ( 2.0‰) conditions. The obtained ꢀH (-3.3‰) for
aerobic biodegradation, however, is much lower than those
previously reported (-29‰ to -66‰) (48). Also, the ex-
tremely low ln f values calculated for aerobic biodegradation
(ln f , -10) are not realistic because, as discussed above, the
resulting concentrations are far below those that can be
utilized or detected by both microorganisms and available
analytical instrumentation.
Assuming anaerobic biodegradation (filled symbols in
Figure 5), however, the obtained hydrogen enrichment factor
(ꢀH ) -15.6 ( 2.0‰) corresponds fairly well with the
enrichment factors published by Kuder et al. (51) for anoxic
conditions (-11.5‰), and the calculated fvalues appear more
appropriate.
Hence, the analysis of Figure 5 did not only exclude aerobic
biotransformation of MTBE as major degradation pathway,
it also allowed us to calculate the extent of biodegradation
using eqs 3 and eq 4 assuming an isotopic enrichment factor
(ꢀC) of -8.63‰ that is proposed for anaerobic biodegradation
of MTBE (51). The results of this calculation are shown in
Table 2.
Importance of a Two-Element CSIA Approach for the
Evaluation Scheme. This example illustrates the potential
of using a multi element approach in CSIA especially in the
case of field studies. If the data of only one element (e.g.,
carbon isotopic signatures) were present, a quantification of
9
VOL. 39, NO. 4, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 1025

SCHEME 1. Reaction Mechanisms for Enzymatic Degradation of MTBEa
a Dashed line indicates the chemical bond at which isotopic fractionation will occur during the given reaction.
TABLE 3. Correction Factors Applied and Expected KIEs for Different Reaction Mechanisms for the Transformation of MTBE
oxidation
SN2
SN1
Carbon
no. of carbon atoms (n)
no. of carbon atoms in reactive
positions (x)
5
1
5
1
5
1a
correction for nonreacting sitesb
correction for intramolecular
competition (z)c
n/x ) 5
none (z ) 1)
n/x ) 5
none (z ) 1)
n/x ) 5
none (z ) 1)
expected KIE 12k/13
k
1.01-1.02
1.03-1.08
1.00-1.02
Hydrogen
no. of hydrogen atoms (n)
no. of hydrogen atoms in reactive
positions (x)
12
12
12
3 (primary effect)
3 (secondary effects in
R-position)
9 (secondary effects in
â-position)
correction for nonreacting sitesb
correction for intramolecular
competition (z)c
n/x ) 4
n/x ) 4
n/x ) 4/3
simultaneous secondary
effects, no
competition (z ) 1)
1.1-1.2
z ) 3, three equal positions
of which only
one reacts
simultaneous secondary
effects, no
competition (z ) 1)
0.95-1.05
expected KIE 1k/2k
3-8
a The three methyl carbon atoms that could show a secondary isotope effect are neglected. b n/x value used in eqs 7 and 8. c Correction factor
z is used in eq 9.
tive MTBE sinks (such as evaporation) exist, consistent with
too low MTBE concentrations in the “traditional” Rayleigh
analysis (see Figure 3).
Mechanisms of MTBE Biodegradation. A meaningful
application of the two-element isotopic approach presented
above requires that correlations between carbon and hy-
drogen isotopic fractionations are not coincidental but reflect
the underlying reaction mechanism(s).
In other words, the approachsusing carbon isotope data
to calculate the extent of biodegradation (1 - f ) combined
with the use of calculated f to determine hydrogen isotopic
enrichmentssshould only work if both isotope signatures
are influenced by the same reaction. In Figure 6, carbon and
hydrogen isotopic signatures reported during aerobic and
anaerobic biodegradation (51) are plotted together with the
data collected from the investigated field site.
A good correlation of the hydrogen and carbon isotopic
shifts was obtained for all three data sets. Aerobic biodeg-
radation causes a small carbon isotopic fractionation but a
very strong shift in hydrogen isotopic signatures. Conversely,
anaerobic biodegradation results in strong isotopic enrich-
ment for both carbon and hydrogen. The slopes of the
regression lines of our field data and the data published for
anaerobic biodegradation are similar, corroborating our
hypothesis of anaerobic in-situ biodegradation. Note that
an alternative explanation for a similar slope due to com-
mitment to catalysis could be ruled out (see Supporting
Information).
Conversely, the significant difference between the frac-
tionation patterns under oxic versus anoxic conditions is a
strong indicator for the existence of different reaction
mechanisms. In the presence of oxygen, the most likely
degradation pathway from MTBE to TBA starts with an
oxidation of the methoxy group by mono-oxygenases (Scheme
1) (12, 14, 56). The so formed tert-butoxy methanol either
reacts directly to TBA and formaldehyde or is further oxidized
to tert-butyl formate, which then hydrolyses to form TBA.
The reaction mechanism for the formation of TBA from MTBE
under anoxic conditions has not been elucidated so far. One
possibility is a nucleophilic substitution reaction as depicted
in Scheme 1 either via an SN2 mechanism or via an SN1
mechanism. These three different cases are also associated
9
1026 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 4, 2005

TABLE 4. Reevaluationa of Published Carbon and Hydrogen Isotope Effects by Conversion of Published Enrichment Factors E to
KIE Values for the Biodegradation of MTBE
reported
carbon
E (‰)
calcd 12k/13k
after correction
of reported Eb
reported
hydrogen
E (‰)
calcd 1k/2k
after correction
of reported Ec
original data from
experimental conditions
Gray et al. (48)
aerobic, pure strain PM1
-2.0 to -2.4
1.010 to 1.012e -33 to -37 oxidation: 1.70f,g
SN2: 1.16f,g
SN1: 1.05f,g
Gray et al. (48)
aerobic, mixed microcosm
-1.5 to -1.8
1.008 to 1.009e -29 to -66 oxidation: 1.69/2.54f
SN2: 1.16/1.25f
SN1: 1.05/1.09f
Hunkeler et al. (28) aerobic, direct or cometabolic
degradation
Kolhatkar et al. (50) anaerobic, field and microcosm data -8.1 to -9.2
Kuder et al. (51)
-1.52 to -1.97 1.008 to 1.009e ndd
1.042 to 1.048e ndd
anaerobic
-8.9 to -10.2 1.047 to 1.054e -11.5
oxidation: 1.16e
SN2: 1.05e
SN1: 1.02e
this work
anaerobic
-15.6
a The parameters n, x, and z used for the reevaluation are listed in Table 3. b For 12k/13k, identical n, x, and z values are obtained for all three
reaction mechanisms. c For 1k/2k, different n, x, and z values must be applied leading to three different calculated1k/2k. d Not determined. e Determined
according to the approximate eq 8. f Determined based on the reevaluation of original isotope data provided by the authors according to eq 7.
g Reevaluation was done with data of both duplicate experiments;
1
with different kinetic isotope effects (KIE, i.e.,12k/13k or k/
2k). To compare the isotopic fractionation observed during
the biodegradation of MTBE with KIEs for the different
reaction mechanisms, the published data must first be
reevaluated using the corrections (eqs 7-9). Table 3 shows
the correction factors applied for the different reaction
mechanisms.
(expected KIE ) 1.03-1.08) but mismatch with KIE values
expected for oxidation (1.01-1.02) or for SN1 (1.00-1.02)
reactions. This first indication of an SN2 reaction based on
carbon isotope effects is corroborated by our reevaluation
of hydrogen isotope fractionation data. A comparison of
measured and predicted hydrogen KIEs for an oxidation (1.16
vs expected 3-8) and for a SN1 reaction (1.02 vs expected
1.1-1.2) ruled out a primary isotope effect as well as a SN1
reaction. The hydrogen isotope effect obtained for a SN2
mechanism (KIE ) 1.05), however, matches very well with
the theoretical KIE range (0.95-1.05) for this type of reaction.
This reevaluation of isotopic fractionation data illustrates
that the differing isotopic enrichments found for aerobic and
anaerobic biodegradation indeed reflect the underlying
reaction mechanisms and therefore can be used as a powerful
tool for the quantification of in-situ biodegradation of MTBE.
A major advantage of this approach is its applicability at
biogeochemically and hydrologically complex field sites as
it is independent of mass balances, as is illustrated by this
case study.
In case of an oxidation at the methoxy group (i.e., the
formation of tert-butoxy methanol), a carbon-hydrogen
bond is broken in the first step of the reaction; hence, one
would expect a very high primary hydrogen isotope effect
combined with a primary carbon isotope effect. The Stre-
itwieser Semiclassical Limits for the breaking of a C-H-bond
are 6.4 for k/2k and 1.02 for 12k/13k (57). In the case of an
1
SN2 reaction mechanism, the expected carbon isotope effects
are relatively large (58). The secondary hydrogen isotope
effects, in contrast, are very small and may even be inverse
due to a more constrained bending motion of the hydrogen
in the transition state (58, 59). The results for SN1 reaction
mechanisms show the opposite trend (i.e., relatively small
carbon isotopic effects and quite large secondary hydrogen
isotope effects). The extent of the hydrogen fractionation
depends on the nature of the leaving group (58).
Acknowledgments
We are particularly thankful to Michel Schurter for his
assistance, permanent support, and coordination efforts that
greatly contributed to the success of this study. We are also
indebted to Ernesto Moeri and his team from CSD Geoclock
for organizing the sampling campaign and providing data
from groundwater modeling and former sampling campaigns.
We thank Caroline Stengel, Jakov Bolotin, and BACHEMA
for the concentration determination of inorganic and organic
contaminants; Thomas Erhart for his contribution in the
validation of the H/D determination of MTBE using SPME;
and Thomas Hofstetter for helpful comments on the manu-
script. Finally, we are grateful to Jennifer McKelvie (Gray) for
providing the original data on aerobic MTBE degradation
that allowed a reevaluation according to eq 7.
The reevalution was applied to isotopic enrichment factors
reported in the literature and results are shown in Table 4.
The carbon kinetic isotope effects calculated for the aerobic
oxidation of MTBE are comparable to the expected 12k/13k
values shown in Table 3. The suggested reaction mechanism
is also supported by the hydrogen data. Even though the
1k/2k values based on our reevaluation of experimental
literature data are significantly smaller than the predicted
values in Table 3, they still show a primary isotope effect in
case of a methyl group oxidation. In contrast, the hydrogen
KIEs from aerobic experiments in Table 4 calculated for a
SN2 reaction are too large to be in agreement with an SN2
reaction mechanism. They do however agree with results
expected for a SN1 mechanism. While, hence, a SN1 mech-
anism cannot be ruled out based on kinetic isotope effects,
an oxidation mechanism seems to be most plausible in the
presence of molecular oxygen. A possible explanation for
the reduced hydrogen KIE observed might be an enzymatic
effect known as commitment to catalysis (see Supporting
Information).
Note Added After Print Publication
Due to a production error, some units of measure were
incorrect in the version published on the web December 22,
2004 (ASAP) and in the February 15, 2005 issue (Vol. 39,
No. 4, pp 1018-1029); the correct version of the paper was
published on August 12, 2005 and an Addition and Correction
appears in the September 15, 2005 issue (Vol. 39, No. 18).
The carbon isotope effects calculated from reported
fractionation for the anoxic reaction (KIE ) 1.04-1.05) are
very similar to those expected for a SN2 reaction mechanism
9
VOL. 39, NO. 4, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 1027

(21) Landmeyer, J. E.; Chapelle, F. H.; Bradley, P. M.; Pankow, J. F.;
Church, C. D.; Tratnyek, P. G. Fate of MTBE relative to benzene
in a gasoline-contaminated aquifer (1993-98). Ground Water
Monit. Rem. 1998, 18, 93-102.
(22) Finneran, K. T.; Lovley, D. R. Anaerobic degradation of methyl
tert-butyl ether (MTBE) and tert-butyl alcohol (TBA). Environ.
Sci. Technol. 2001, 35, 1785-1790.
(23) Mormile, M. R.; Liu, S.; Suflita, J. M. Anaerobic biodegradation
of gasoline oxygenates: extrapolation of information to multiple
sites and redox conditions. Environ. Sci. Technol. 1994, 28, 1727-
1732.
(24) Yeh, C. K.; Novak, J. T. Anaerobic biodegradation of gasoline
oxygenates in soils. Water Environ. Res. 1994, 66, 744-752.
(25) Wilson, J. T.; Cho, J. S.; Wilson, B. H.; Vardy, J. A. Natural
Attenuation of MTBE in the Subsurface under Methanogenic
Conditions; U.S. Environmental Protection Agency, National
Risk Assessment Research Laboratory: 2000.
(26) O’Reilly, K. T.; Moir, M. E.; Taylor, C. D.; Smith, C. A.; Hyman,
M. R. Hydrolysis of tert-butyl methyl ether (MTBE) in dilute
aqueous acid. Environ. Sci. Technol. 2001, 35, 3954-3961.
(27) Schmidt, T. C.; Schirmer, M.; Weiss, H.; Haderlein, S. B. Microbial
degradation of methyl tert-butyl ether (MTBE) and tert-butyl
alcohol (TBA) in the subsurface. J. Contam. Hydrol. 2004, 70,
173-203.
(28) Hunkeler, D.; Butler, B. J.; Aravena, R.; Barker, J. F. Monitoring
biodegradation of methyl tert-butyl ether (MTBE) using com-
pound-specific carbon isotope analysis. Environ. Sci. Technol.
2001, 35, 676-681.
(29) Landmeyer, J. E.; Chapelle, F. H.; Herlong, H. H.; Bradley, P. M.
Methyl tert-butyl ether biodegradation by indigenous aquifer
microorganisms under natural and artificial oxic conditions.
Environ. Sci. Technol. 2001, 35, 1118-1126.
(30) Bradley, P. M.; Landmeyer, J. E.; Chapelle, F. H. TBA Biodeg-
radation in surface-water sediments under aerobic and anaero-
bic conditions. Environ. Sci. Technol. 2002, 36, 4087-4090.
(31) Zwank, L.; Schmidt, T. C.; Haderlein, S. B.; Berg, M. Simultaneous
determination of fuel oxygenates and BTEX using direct aqueous
injection gas chromatography/mass spectrometry (DAI-GC/
MS). Environ. Sci. Technol. 2002, 36, 2054-2059.
(32) Schirmer, M.; Barker, J. F. A study of long-term MTBE attenuation
in the Borden Aquifer, Ontario, Canada. Ground Water Monit.
Rem. 1998, 18, 113-122.
(33) Schirmer, M.; Butler, B. J.; Barker, J. F.; Church, C. D.; Schirmer,
K. Evaluation of biodegradation and dispersion as natural
attenuation processes of MTBE and benzene at the Borden field
site. Phys. Chem. Earth B 1999, 24, 557-560.
(34) Sherwood Lollar, B.; Slater, G. F.; Sleep, B.; et al. Stable carbon
isotope evidence for intrinsic bioremediation of tetrachloroethen
and trichloroethen at Area 6, Dover Air Force Base. Environ.
Sci. Technol. 2001, 35, 261-269.
(35) Slater, G. F.; B., S. L.; Spivack, J.; Brennan, M.; Mackenzie, P.
Isotopic tracers of degradation of dissolved chlorinated solvents.
In International Conference on Remediation of Chlorinated and
Recalcitrant Compounds; Wickramanayake, G. B., Hinchee, R.
E., Eds.; Battelle Press: Columbus, OH, 1998; pp 133-138.
(36) Stout, S. A.; Uhler, R. M.; Philp, R. P.; Allen, J.; Uhler, A. D.
Source differentiation of individual chlorinated solvents dis-
solved in groundwater using compound specific carbon isotopic
analysis. 216th American Chemical Society National Meeting,
Environmental Chemistry Division; American Chemical Soci-
ety: Washington, DC, 1998,
(37) Song, D. L.; Conrad, M. E.; Sorenson, K. S.; Alvarez-Cohen, L.
Stable carbon isotope fractionation during enhanced in situ
bioremediation of trichloroethene. Environ. Sci. Technol. 2002,
36, 2262-2268.
(38) Hunkeler, D.; Aravena, R.; Butler, B. J. Monitoring microbial
dechlorination of tetrachloroethene (PCE) in groundwater using
compound-specific stable carbon isotope ratios: microcosm
and field studies. Environ. Sci. Technol. 1999, 33, 2733-2738.
(39) Sturchio, N. C.; Clausen, J. L.; Heraty, L. J.; Huang, L.; Holt, B.
D.; Teophilo, A.; Abrajano, J. Chlorine isotope investigation of
natural attenuation of trichloroethene in an aerobic aquifer.
Environ. Sci. Technol. 1998, 32, 3037-3042.
Supporting Information Available
Detailed mathematical derivation of eq 7 that is essential for
the described data reevaluation and a discussion of com-
mitment to catalysis as a rationale for similar slopes of
regression lines shown in Figure 6. This material is available
free of charge via the Internet at http://pubs.acs.org.
Literature Cited
(1) Krayer von Krauss, M., Harremoe¨s, P., Eds. MTBE in Petrol as
a Substitute for Lead; Office for Official Publications of the
European Communities: Copenhagen, 2001.
(2) EFOA. MTBE Resource Guide, Revised Version 2; European Fuel
Oxygenates Association: October 2002.
(3) Squillace, P. J.; Zogorski, J. S.; Wilber, W. G.; Price, C. V.
Preliminary assessment of the occurrence and possible sources
of MTBE in groundwater in the United States, 1993-1994.
Environ. Sci. Technol. 1996, 30, 1721-1730.
(4) Pankow, J. F.; Thomson, N. R.; Johnson, R. L.; Baehr, A. L.;
Zogorski, J. S. The urban atmosphere as a non-point source for
the transport of MTBE and other volatile organic compounds
(VOCs) to shallow groundwater. Environ. Sci. Technol. 1997,
31, 2821-2828.
(5) Schmidt, T. C.; Duong, H. A.; Berg, M.; Haderlein, S. B. Analysis
of fuel oxygenates in the environment. Analyst 2001, 126, 405-
413.
(6) Prince, R. C. Biodegradation of methyl tertiary-butyl ether
(MTBE) and other fuel oxygenates. Crit. Rev. Microbiol. 2000,
26, 163-178.
(7) Fayolle, F.; Vandecasteele, J.-P.; Monot, F. Microbial degradation
and fate in the environment of methyl tert-butyl ether and related
fuel oxygenates. Appl. Microbiol. Biotechnol. 2001, 56, 339-
349.
(8) Deeb, R. A.; Scow, K. M.; Alvarez-Cohen, L. Aerobic MTBE
biodegradation: an examination of past studies, current chal-
lenges and future research directions. Biodegradation 2000, 11,
171-186.
(9) Salanitro, J. P.; Diaz, L. A.; Williams, M. P.; Wisniewski, H. L.
Isolation of a bacterial culture that degrades methyl tert-butyl
ether. Appl. Environ. Microbiol. 1994, 60, 2593-2596.
(10) Mo, K.; Lora, C. O.; Wanken, A. E.; Javanmardian, K.; Yang, X.;
Kulpa, C. F. Biodegradation of methyl tert-butyl ether by pure
bacterial cultures. Appl. Microbiol. Biotechnol. 1997, 47, 69-72.
(11) Bradley, P. M.; Landmeyer, J. E.; Chapelle, F. H. Aerobic
mineralization of MTBE and tert-butyl alcohol by stream-bed
sediment microorganisms. Environ. Sci. Technol. 1999, 33,
1877-1879.
(12) Steffan, R. J.; McClay, K.; Vainberg, S.; Condee, C. W.; Zhang,
D. Biodegradation of the gasoline oxygenates methyl tert-butyl
ether, ethyl tert-butyl ether, and tert-amyl methyl ether by
propane-oxidizing bacteria. Appl. Environ. Microbiol. 1997, 63,
4216-4222.
(13) Hanson, J. R.; Ackerman, C. E.; Scow, K. Biodegradation of methyl
tert-butyl ether by a bacterial pure culture. Appl. Environ.
Microbiol. 1999, 65, 4788-4792.
(14) Hardison, L. K.; Curry, S. S.; Ciufetti, L. M.; Hyman, M. R.
Metabolism of diethyl ether and cometabolism of methyl tert-
butyl ether by a filamentous fungus, a Graphium sp. Appl.
Environ. Microbiol. 1997, 63, 3059-3067.
(15) Garnier, P. M.; Auria, R.; Augur, C.; Revah, S. Cometabolic
biodegradation of methyl tert-butyl ether by Pseudomonas
aeruginosa grown on pentane. Appl. Microbiol. Biotechnol. 1999,
51, 498-503.
(16) Garnier, P. M.; Auria, R.; Augur, C.; Revah, S. Cometabolic
biodegradation of methyl tert-butyl ether by a soil consortium:
effect of components present in gasoline. J. Gen. Appl. Microbiol.
2000, 79-84.
(17) Deeb, R. A.; Hu, H.-Y.; Hanson, J. R.; Scow, K. M.; Alvarez-
Cohen, L. Substrate interactions in BTEX and MTBE mixtures
by an MTBE-degrading isolate. Environ. Sci. Technol. 2001, 35,
312-317.
(18) Bradley, P. M.; Chapelle, F. H.; Landmeyer, J. E. Effect of redox
conditions on MTBE biodegradation in surface water sediments.
Environ. Sci. Technol. 2001, 35, 4643-4647.
(19) Bradley, P. M.; Chapelle, F. H.; Landmeyer, J. E. Methyl tert-
butyl ether mineralization in surface-water sediment micro-
cosms under denitrifying conditions. Appl. Environ. Microbiol.
2001, 67, 1975-1978.
(20) Somsamak, P.; Cowan, R. M.; Ha¨ggblom, M. M. Anaerobic
biotransformation of fuel oxygenates under sulfate-reducing
conditions. FEMS Microbiol. Ecol. 2001, 37, 259-264.
(40) Richnow, H. H.; Annweiler, E.; Michaelis, W.; Meckenstock, R.
U. Microbial in situ degradation of aromatic hydrocarbons in
a contaminated aquifer monitored by carbon isotope fraction-
ation. J. Contam. Hydrol. 2003, 1-20.
(41) Richnow, H. H.; Meckenstock, R. U.; Reitzel, L. A.; Baun, A.;
Ledin, A.; Christensen, T. H. In situ biodegradation determined
by carbon isotope fractionation of aromatic hydrocarbons in
an anaerobic landfill leachate plume (Vejen, Denmark). J.
Contam. Hydrol. 2003, 1-14.
9
1028 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 4, 2005

(42) Kuder, T.; Philp, P.; Kolhatkar, R.; Wilson, J. T.; Landmeyer, J.
E. Compound-specific stable isotope analysis to demonstrate
in situ MTBE biotransformation. In In Situ and On-Site
Bioremediation-2003; Gagar, V. S., Kelley, M. E., Eds.; Batelle
Press: Columbus, OH, 2003; Paper I-08.
(51) Kuder, T.; Philp, R. P.; Kolhatkar, R.; Wilson, J. T.; Allen, J.
Application of stable carbon and hydrogen isotopic techniques
for monitoring biodegradation of MTBE in the field. In NGWA/
API Petroleum Hydrocarbons and Organic Chemicals in Ground
Water; American Petroleum Institute: 2002.
(43) Kuder, T.; Kolhatkar, R.; Wilson, J. T.; O’Reilley, P.; Philp, P.;
Allen, J. Compound-specific carbon and hydrogen analysis-
field evidence of MTBE bioremediation. In Proceedings of the
2003 NGWA Focus Conference on MTBE: Assessment, Remedia-
tion, and Public Policy; National Ground Water Association:
Westerville, OH, 2003; pp 112-121.
(44) Schmidt, T. C.; Zwank, L.; Elsner, M.; Berg, M.; Meckenstock,
R. U.; Haderlein, S. B. Compound-specific stable isotope analysis
of organic contaminants in natural environments: a critical
review of the state of the art, prospects, and future challenges.
Anal. Bioanal. Chem. 2004, 283-300.
(52) Richnow, H. H.; Meckenstock, R. U. Isotopen-geochemisches
konzept zur in-situ-erfassung des biologischen abbaus in
kontaminiertem grundwasser. TerraTech 1999, 38-41.
(53) Meckenstock, R. U.; Morasch, B.; Ka¨stner, M.; Vieth, A.; Richnow,
H. H. Assessment of bacterial degradation activities in the
environment by analysis of stable carbon isotope fractionation.
Soil, Water, Air Pollut. 2002, Focus 2, 141-152.
(54) Narten, A.; Kuhn, W. Genaue bestimmung kleiner dampfdruck-
unterschiede isotoper verbindungen II. der 13C/12C-isotopieef-
fekt in tetrachlorkohlenstoff und in benzol. Helv. Chim. Acta
1961, 44, 1474-1479.
(55) Jancso, G.; Van Hook, W. A. Condensed phase isotope effects
(especially vapor pressure isotope effects). Chem. Rev. 1974,
74, 689-750.
(45) Richnow, H. H.; Eschenbach, A.; Mahro, B.; et al. Formation of
nonextractable soil residues: a stable isotope approach.Environ.
Sci. Technol. 1999, 33, 3761-3767.
(46) Meckenstock, R. U.; Morasch, B.; Warthmann, R.; et al. 13C/12
C
(56) Hernandez-Perez, G.; Fayolle, F.; Vandecasteele, J.-P. Biodeg-
radation of ethyl tert-butyl ether (ETBE), methyl tert-butyl ether
(MTBE) and tert-amyl methyl ether (TAME) by Gordonia terrae.
Appl. Microbiol. Biotechnol. 2001, 55, 117-121.
isotope fractionation of aromatic hydrocarbons during microbial
degradation. Environ. Microbiol. 1999, 1, 409-414.
(47) Zwank, L.; Schmidt, T. C.; Berg, M.; Haderlein, S. B. Compound-
specific carbon isotope analysis of volatile organic compounds
in the low µg/L range. Anal. Chem. 2003, 75, 5575-5583.
(48) Gray, J. R.; Lacrampe-Couloume, G.; Deepa, G.; et al. Carbon
and hydrogen isotopic fractionation during biodegradation of
methyl tert-butyl ether. Environ. Sci. Technol. 2002, 36, 1931-
1938.
(49) Day, M.; Aravena, R.; Hunkeler, D.; Gulliver, T. Application of
carbon isotopes to document biodegradation of tert-butyl
alcohol under field conditions. Contam. Soil Sediment Water
2002, 88-92.
(57) Huskey, W. P. In Enzyme Mechanisms from Isotope Effects; Cook,
P. F., Ed.; CRC Press: Boca Raton, FL, 1991.
(58) Melander, L.; Saunders: W. H. J. Reaction Rates of Isotopic
Molecules; John Wiley & Sons: New York, 1980.
(59) Willi, A. V. Isotopeneffekte bei Chemischen Reaktionen; Georg
Thieme Verlag: Stuttgart, 1983.
(50) Kolhatkar, R.; Kuder, T.; Philp, R. P.; Allen, J.; Wilson, J. T. Use
of compound-specific stable carbon isotope analysis to dem-
onstrate anaerobic biodegradation of MTBE in groundwater at
a gasoline release site. Environ. Sci. Technol. 2002, 36, 5139-
5146.
Received for review March 5, 2004. Revised manuscript re-
ceived October 27, 2004. Accepted November 1, 2004.
ES049650J
9
VOL. 39, NO. 4, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 1029