Faculté des sciences

Isotope fractionation of volatile organic compounds in porous media under variably saturated conditions

Jeannottat, Simon ; Hunkeler, Daniel (Dir.)

Thèse de doctorat : Université de Neuchâtel, 2012.

Compound-specific isotope analysis (CSIA) has proven to be an effective tool to assess in-situ biodegradation of volatile organic compounds (VOCs) in groundwater and to distinguish between different contaminant sources. There is an increasing interest to apply the CSIA method also in unsaturated zone studies. However, under variably saturated conditions, physical processes such as phase... Plus

Ajouter à la liste personnelle
    Summary
    Compound-specific isotope analysis (CSIA) has proven to be an effective tool to assess in-situ biodegradation of volatile organic compounds (VOCs) in groundwater and to distinguish between different contaminant sources. There is an increasing interest to apply the CSIA method also in unsaturated zone studies. However, under variably saturated conditions, physical processes such as phase partitioning and diffusion might influence isotope ratios as well. The main aim of this study was to quantify isotope fractionation associated with physical processes relevant for the unsaturated zone and to explore how isotope ratios evolve if different processes interact. The study focussed on chlorinated hydrocarbons, which are among the most common contaminants detected in the subsurface. In addition to carbon, the behaviour of chlorine isotopes was also investigated. Since the mass difference between the stable chlorine isotopes is two, larger isotope effects are expected to occur than for carbon, especially for diffusion. In order to make chlorine isotope analysis feasible, some analytical method development was required as well.
    A new analytical method for chlorine isotope ratio analyses based on gas chromatography quadrupole mass-spectrometry (GC-qMS) measurements was evaluated by comparing it with established IRMS based methods. The study highlighted the need to perform a two-point calibration and to bracket samples with standards of the same concentrations. Using this approach, a precision of 1σ ≈0.2-0.5% (n=10) was reached for TCE measurements with an Agilent GC-qMS. Additionally, the relationship between isotope and isotopologue fractionation during vaporization and diffusion was explored theoretically. While fractionation takes place among isotopologues it is usually more convenient to express results as isotope ratios. It was demonstrate that during vaporization and diffusion chlorine isotope and isotopologue fractionation is proportional in good approximation even when several isotopologues with multiple heavy isotopes are present. As a consequence, isotope ratios can be determined even if only some isotopologues are quantified and isotope fractionation factors can be derived from either isotope or isotopologue ratios.
    In a next step, isotope fractionation during processes relevant for the release of compounds from NAPL (vaporization) or groundwater (air-water partitioning) was quantified in isolation for the trichloroethylene (TCE) using various laboratory experiments. During NAPL-vapor equilibration, carbon and chlorine isotope ratios evolved in opposite directions although both elements are present in the same bond, with a normal isotope effect for chlorine (εCl = -0.39±0.03‰) and an inverse effect for carbon (εC = +0.75±0.04‰). During air-water partitioning, no significant chlorine isotope fractionation occurred, while an inverse carbon isotope effect (εC = +0.38±0.04‰) was observed, that was however weaker than for vaporization.
    Then, the combined effect of release of compounds from NAPL or groundwater and transport by diffusion through unsaturated porous media was investigated. Contaminant release to the unsaturated zone from an NAPL source was simulated using a 1D column. A TCE NAPL source was emplaced at one end of the column and the other end was open to the atmosphere to maintain a concentration close to zero. In the initial phase of the experiment, gaseous TCE became depleted in heavy isotopes with increasing distance from the source due to the more rapid diffusion of light isotopologues. Once steady state concentration profiles were established, isotope ratios of gaseous TCE were constant along the column. At the NAPL source, carbon isotope ratios remained constant (εC = +0.10±0.05‰) while chlorine isotope ratios became enriched (εCl = -1.39±0.06‰) as vaporization proceeded. Based on theoretically consideration, it is expected that the observed isotope effect is a combination of isotope fractionation associated with vaporization and diffusion. For carbon, the inverse isotope fractionation associated with NAPL-vapor equilibration cancelled out with the normal diffusion isotope fractionation while for chlorine, both processes followed normal isotope fractionation and hence they cumulated.
    Contaminant release from a groundwater plume to the unsaturated zone was simulated in a 2D laboratory system. A 2D system with horizontal groundwater flow in the lower part was chosen because the rate of contaminant release is strongly influenced by transverse vertical dispersion. PCE was selected as a model compound since it does not degrade under aerobic conditions, which were maintained in the experiment. Unlike in the NAPL experiment, no significant depletion of the heavy isotopes with distance from the groundwater plume was observed. This can be explained by the slower release of contaminants from groundwater compared to the NAPL source. Hence, constant quasi-steady state isotope profiles are maintained throughout the experiment. However, numerical modeling suggested that, depending on the thickness of the unsaturated zone and the lithology, depletion in heavy isotopes could occur with distance during the transient migration of contaminant through the unsaturated zone. Additionally, a small offset of about 0.5 ‰ was observed in the isotope ratios between and aqueous and gaseous values across the capillary fringe for both carbon and chlorine isotopes. Numerical modeling indicated that this offset is due to isotope fractionation during air-water partitioning, as well as to the preferential loss of light isotopologues from the unsaturated zone by diffusion. Variations of chlorine isotope ratios of 1.5‰ were also observed in the unsaturated zone during water table fluctuations.
    Finally, the combined effect of biodegradation and transport by diffusion on the isotope ratios of VOCs migrating through the unsaturated zone was studied. Vinyl chloride (VC) was selected as model compound. Carbon isotope fractionation associated with biodegradation only was investigated in microcosm experiments using material from a contaminated site. An average carbon isotope enrichment factor of -7.2‰ was obtained. The coupling of degradation and diffusive transport in the unsaturated zone was also evaluated with numerical modeling. It was shown that the effective enrichment factor of biodegradation coupled to diffusion in the unsaturated conditions is reduced compare to saturated conditions (εC = -2.1‰). Numerical modeling suggested that biodegradation could be identified in a diffusion-controlled system using CSIA if degradation rate is fast enough. Shifts in isotope ratios due to biodegradation were amplified if layers of fine sediments were present or if the unsaturated zone was thick.
    In conclusion, this work confirms that significant isotope fractionation can occur during physical processes affecting VOCs in the unsaturated zone even for molecules with a relatively large mass such as TCE and PCE. Moreover, during the combination of several processes, isotope effects associated with different processes can reinforce or cancel each other. In the case of diffusion-controlled vaporization from an unsaturated zone NAPL source, diffusion will always be associated with a normal isotope effect, while during vaporization either a normal (chlorine isotopes) or inverse (carbon isotopes) effects can occur. Hence, for diffusion-controlled vaporization, a fairly strong isotope effect can generally be expected for chlorine because the two processes reinforce each other. During the transfer of compounds though the capillary fringe from a groundwater plume to the unsaturated zone, air-water partitioning (normal and inverse isotope effect for chlorine and carbon, respectively) and the preferential lost of light isotopologues from the unsaturated zone due to diffusion cancel for carbon isotopes, but reinforce each other for chlorine. In unsaturated conditions where diffusion is the major transport process, the net isotope fractionation for biodegradation is reduced compare to saturated conditions, as the preferential removal of light molecules due to biodegradation is partly counterbalanced by the larger mass flux of light molecules brought along by diffusion from the source. Finally, the extent of the observed isotope shifts strongly depends whether a system is under steady state or in a transient state conditions. A significant isotope fraction is expected at transient state in the case of NAPL source, as well as if the unsaturated zone is thick and/or layer of fine sediments are present in the case of a groundwater source. In contrary, if the unsaturated zone is thin and/or if coarser sediments are present above groundwater level, isotope profiles generally do not vary with distance from the water table at a given time, even at transient state. Constant isotopic ratios are however always observed at steady state across the unsaturated zone.
    This study has several implications for the application of CSIA methods to characterize sources of contamination and/or to assess biodegradation of chlorinated ethenes in unsaturated zone contamination studies. It can be expected that the isotope signature of the vapor around an unsaturated zone NAPL source does not vary with distance as long as steady state conditions are reached. A small offset equivalent to the vaporization enrichment factor is expected between the vapor contamination and the NAPL source. This offset will nevertheless not affect the source identification methods as it will be the same for each NAPL source. For groundwater source, a small isotopic shift is expected between the groundwater and the gas phase (in the unsaturated zone) depending on subsurface lithology. Constant isotopic values are then expected through the unsaturated zone. Since constant isotopic values were expected across the unsaturated zone at steady state in absence of biodegradation, an enrichment of heavy isotopes indicates the presence of significant biodegradation. It seems thus possible to access biodegradation with CSIA in unsaturated zone studies, as long as the net isotope fractionation factor is sufficiently high.