The speciation of aluminum (Al) in soils is important for Al cycling and Al toxicity in terrestrial ecosystems. Current soil Al speciation methods are mostly wet-chemical, discriminating among operationally defined fractions rather than distinct Al species. We performed synchrotron-based X-ray absorption near edge structure (XANES) spectroscopy at the Al K-edge (1560 eV) for Al speciation in different horizons of three forest soil profiles and O layer density fractions. We deconvoluted the spectra by linear combination fitting (LCF; 1550–1690 eV), using reference spectra of 22 diluted inorganic and organic Al-bearing soil constituents. Our first time application of Al XANES spectroscopy + LCF for soil Al speciation revealed methodological challenges and difficulties. Thus, a minimum sample Al content of 5 mg g−1 was necessary to yield spectra with satisfactory signal-to-noise ratio in reasonable acquisition time, and Al contents > 15 mg g−1 resulted in spectrum distortion by self absorption, requiring adequate sample dilution prior to XANES analysis. Nevertheless, Al K-edge XANES spectroscopy allowed for estimating the relative contribution of different Al species to total Al in our soil and density fraction samples. The Al XANES spectra of different Al carboxylates were very similar. The same was true for different Al phenol species, different 2:1 clay minerals (illite, montmorillonite, chlorite), different 1:1 clay minerals (kaolinite, halloysite), and different feldspars (orthoclase, anorthite), jeopardizing a reliable differentiation and quantification of these specific compounds. We therefore combined these compounds into the groups SOM carboxyl(ate)-bound Al, SOM phenol-bound Al, 1:1 clay mineral-bound Al, 2:1 clay mineral-bound Al, and feldspar-bound Al. XANES results showed a decreasing contribution of organically bound Al to total soil Al with soil depth (O layers 20–43 %; Ah horizons 10–22 %; B horizons 0–12 %). Major inorganic Al-bearing constituents in the fine earth of all soils were 2:1 clay minerals and feldspars, followed by 1:1 clay minerals (kaolinite), and Fe oxyhydroxides with partial Al substitution. Gibbsite or allophane did contribute only marginally to fine-earth Al in our soils. Different Oa layer density fractions with different SOM content and SOM decomposition status also differed in Al content and Al speciation, indicating the existence of spatially separated forest floor constituents with different Al speciation. Aluminum in the mineral-dominated density fraction > 1.6 g cm−3 with advanced SOM decomposition, comprising about 50 % of total Oa mass, was entirely bound in clay minerals (65–70 % of total Al) and feldspars (30–35 % of total Al). In contrast, Al in the mineral-poor density fraction < 1.0 g cm−3 was entirely bound to SOM as Al-organo complex, with Al bound to SOM phenol groups (>75 % of total Al) being more relevant than Al bound to SOM carboxyl(ate) groups. However, the latter density fraction comprised only 4 % of total Oa mass in both soils. Aluminum K-edge XANES spectroscopy, particularly when combined with total element analysis and XRD, is a promising novel tool for speciation of Al in soils and in SOM-mineral associations, with a great potential to promote our understanding of Al biogeochemistry in terrestrial ecosystems.
Read full abstract