Estimating the concentration of aluminum-substituted hematite and goethite using diffuse reflectance spectrometry and rock magnetism: Feasibility and limitations

Hematite and goethite in soils are often aluminum (Al) substituted, which can dramatically change their reflectance and magnetic properties and bias abundance estimates using diffuse reflectance spectroscopy (DRS) and magnetic techniques. In this study, synthetic Al-substituted hematites and goethites and two Chinese loess/paleosol sequences were investigated to test the feasibility and limitations of estimating Al-hematite and Al-goethite concentration. When Al substitution is limited (Al/(Al + Fe) molar ratio < ~8%), the reflectance spectrum provides a reliable estimate of the goethite/hematite concentration ratio. New empirical relationships between the DRS band intensity ratio and the true concentration goethite/hematite ratio are estimated as goethite/hematite = 1.56 × (I_{425 nm}/I_{535 nm}) or goethite/hematite = 6.32 × (I_{480 nm}/I_{535 nm}), where I_{425 nm}, I_{480 nm}, and I_{535 nm} are the amplitudes of DRS second-derivative curves for characteristic bands at ~425 nm, ~480 nm, and ~535 nm, respectively. High Al substitution (> ~8%) reduces DRS band intensity, which leads to biased estimates of mineral concentration. Al substitution and grain size exert a control on coercivity distributions of hematite and goethite and, thus, affect the hard isothermal remanent magnetization. By integrating DRS and magnetic methods, we suggest a way to constrain hematite and goethite Al substitution in natural loess. Results indicate that hematite and goethite in Chinese loess have Al contents lower than ~8% and, thus, that DRS can be used to trace hematite and goethite concentration variations.