RELIABILITY OF COMMON CARBONATE ISOLATION TECHNIQUES AND SUBSEQUENT FIDELITY OF STABLE CARBON ISOTOPE SIGNATURES

In near-shore, carbonate-rich marine sediments of mixed composition, organic detritus (e.g., from seagrasses, mangroves) must often be separated from carbonate material—and vice versa—prior to stable isotopic analysis of each carbon fraction. In the process, the weight percent of each fraction may be determined, potentially informing interpretations of changing depositional environments through time and space.

The removal of carbonate material from mixed sediment samples by digestion in hydrochloric acid is a well-established approach. However, there is less agreement among geochemists regarding methods for the isolation of carbonate material by destruction or removal of organic matter ("carbonate isolation"). Despite an array of rigorously tested methodologies for carbonate isolation, disagreement remains as to the strengths and weaknesses of methods with respect to one another, as well as the optimal parameters (e.g., times, temperatures, reagent concentrations) for individual methods. Furthermore, these techniques vary in cost, can be time-consuming, and do not always allow for accurate weight percent determinations.

Here, we compare the reliability of three commonly used methods for carbonate isolation: (1) loss-on-ignition (LOI) at 550°C; (2) digestion in hydrogen peroxide; and (3) digestion in sodium hypochloride. We produced and tested sample mixtures containing known weight percentages of carbonate and organic matter. We find the weight percent LOI to be consistent with expected values, indicating near or total destruction of organic matter. However, in samples containing greater than ~80% organic matter by weight, the reliability of LOI breaks down. We also find that hydrogen peroxide and bleach digestions at low concentrations yield visibly and measurably incomplete destruction of organic matter.

Samples were dissolved in 10% phosphoric acid to produce carbon dioxide for analysis by cavity ring-down spectroscopy. Using carbonate and organic matter of known carbon isotopic values in our samples, we observe that LOI does not bias the measured isotopic composition of inorganic carbon for samples with up to 1.0 wt.% organic matter. Future research will be necessary to determine if similar effects are observed for the hydrogen peroxide or bleach digestions or in untreated samples.