Determining whether
aqueous diffusion and dispersion lead to significant
isotope fractionation is important for interpreting the isotope ratios
of organic contaminants in groundwater. We performed diffusion experiments
with modified Stokes diaphragm cells and transverse-dispersion experiments
in quasi-two-dimensional flow-through sediment tank systems to explore
isotope fractionation for benzene, toluene, ethylbenzene, 2,6-dichlorobenzamide,
and metolachlor at natural isotopic abundance. We observed very small
to negligible diffusion- and transverse-dispersion-induced isotope
enrichment factors (ε < −0.4 ‰), with changes
in carbon and nitrogen isotope values within ±0.5‰ and
±1‰, respectively. Isotope effects of diffusion did not
show a clear correlation with isotopologue mass with calculated power-law
exponents β close to zero (0.007 < β < 0.1). In
comparison to ions, noble gases, and labeled compounds, three aspects
stand out. (i) If a mass dependence is derived from collision theory,
then isotopologue masses of polyatomic molecules would be affected
by isotopes of multiple elements resulting in very small expected
effects. (ii) However, collisions do not necessarily lead to translational
movement but can excite molecular vibrations or rotations minimizing
the mass dependence. (iii) Solute–solvent interactions like
H-bonds can further minimize the effect of collisions. Modeling scenarios
showed that an inadequate model choice, or erroneous choice of β,
can greatly overestimate the isotope fractionation by diffusion and,
consequently, transverse dispersion. In contrast, available data for
chlorinated solvent and gasoline contaminants at natural isotopic
abundance suggest that in field scenarios, a potential additional
uncertainty from aqueous diffusion or dispersion would add to current
instrumental uncertainties on carbon or nitrogen isotope values (±1‰)
with an additional ±1‰ at most.