File(s) under permanent embargo
Is demineralization with dilute hydrofluoric acid a viable method for isolating mineral stabilized soil organic matter?
journal contributionposted on 2017-10-15, 00:00 authored by J Sanderman, M Farrell, Peter MacreadiePeter Macreadie, M Hayes, J McGowan, J Baldock
Hydrofluoric acid (HF) is a powerful tool in the investigation of soil organic matter (SOM) due to its ability to dissolve minerals but not break the chemical bonds of organic matter. These properties make the use of HF a common pretreatment step for removing paramagnetic interferences and concentrating carbon prior to solid-state 13C NMR spectroscopy with the working assumption that any SOM lost during HF treatment will not bias the resulting NMR spectra. Hydrofluoric acid is also used to isolate a mineral-stabilized OM fraction with the working assumption that most mineral-stabilized OM is primarily low molecular weight compounds bound to mineral surfaces and when the minerals are dissolved in HF, the OM bound to these surfaces will be lost to solution. The working assumptions behind these two uses of HF dissolution appear to be contradictory. To address this apparent conundrum, we treated a number of simple organic compounds, soil and sediment samples with HF in 2 and 10% concentrations and tracked C and N loss as well as chemical shifts observed in solid-state 13C NMR spectra. For the soil and sediment samples there were inconsistent C and N losses but no difference in loss between the 2% and 10% HF concentrations. There were no obvious soil properties that could explain the differences in C or N loss. Overall, there were significant shifts in NMR-observable organic chemistry after treatment with both 2 and 10% HF with anoxic fine grained sediments under a seagrass meadow exhibiting strong preferential loss of O-alkyl C while terrestrial soils generally lost OM with more of a mixed chemical character. For many samples, the degree of selective loss was enough to significantly bias the interpretation of OM composition. Given the lack of ability to explain the large differences in C loss between samples with observed soil properties, this study suggests that caution should be used when interpreting HF-soluble C to indicate a mineral-stabilized fraction without considering the soil physicochemical environment and putative mechanisms for organo-mineral associations in that particular soil.