Assessing poorly crystalline and mineral-organic species by extracting Al, Fe, Mn, and Si using (citrate-) ascorbate and oxalate

Abstract Poorly crystalline and mineral-organic species are commonly assessed by oxalate (Ox) extraction of aluminium (Al), iron (Fe), manganese (Mn), and silicon (Si). Drawbacks regarding selectivity towards analytical targets led to test alternatively, but mostly with topsoils, citrate-ascorbate (CA), combined with a separate citrate extraction. Here, we tested the extraction by Ox, CA, and citrate with 61 carbonate-free subsoils characterized by pedogenic element accumulation, and with 17 minerals, including primary silicates, clay minerals, and (Al, Fe) oxides. Iron extraction from the minerals, apart from pyrite, by citrate was negligible (≤2% of total Fe), like from Fe oxides (≤1.5% of total Fe), while oxalate-extractable Al, Fe, and Mn contents were larger than those extracted by CA, the relative fraction of which was below 9%, except for allophane. In subsoils, extractable element contents spanned several orders of magnitude whatever the extractant. With several samples, Mn was entirely solubilized by any extractant. The AlOx, FeOx, and MnOx amounts were very similar among soils lacking allophanic minerals. However, the ratio FeCA:FeOx increased with Bl and Bg horizons, which received Fe2+ by ground- or stagnic water. Citrate-extraction data revealed, depending on the soil type, a potentially large contribution of organic forms to CA- and Ox-extractable Al and Fe. Thus, we recommend combining Ox and CA extraction with citrate extraction to approximate the contents of poorly crystalline pedogenic Al and Fe species, rather than using CA- or oxalate-extraction data alone. Our results challenge the appropriateness of oxalate-extraction data for species identification and mineral quantification.