Detection of calcium phosphate species in soil by confocal mu-Raman spectroscopy

Background: Raman spectroscopy is a promising but largely underexplored tool for the detection of phosphates (P) in soil. Although it requires minimal sample preparation, it has been demonstrated mainly in test matrices or substrates to circumvent the typical signal interference caused by fluorescence of organic matter in actual agricultural soils. Aims: The aim of this study was to highlight the Raman spectroscopic detection and identification of distinct calcium phosphate species amended in contrasting soil matrices-including a real arable soil. Methods: Pure calcium dihydrogen phosphate [Ca(H2PO4)(2).H2O], calcium hydrogen phosphate (CaHPO4), and beta-tricalcium phosphate [beta-Ca-3(PO4)(2)] were each amended in Luvos(R) healing earth, loess from a C-horizon, and a loam arable soil from an Ap-horizon at a dose of 1 mmol (10 g soil)(-1). The unique Raman signature of each pure calcium phosphate species was determined by confocal mu-Raman spectroscopy and used subsequently as a reference spectrum to identify the compound in each soil matrix. Controls without added P were also analyzed. Results: Ca(H2PO4)(2).H2O, CaHPO4, and beta-Ca-3(PO4)(2) were each unambiguously detected in the treated soils. Native hydroxyapatite [Ca-5(PO4)(3)(OH)], quartz (SiO2), feldspar (NaAlSi3O8), calcite (CaCO3), and dolomite [CaMg(CO3)(2)] were also identified, for example, in the Raman microscopic image of the control Luvos (R) healing earth sample. Intrinsic beta-Ca-3(PO4)(2) and Ca-5(PO4)(3)(OH) present in Ap-horizon loam were detected and distinguished from each other by a Lorentzian fitting, which deconvoluted the individual Raman signals from an unresolved peak. Conclusions: The usefulness of confocal mu-Raman spectroscopy to detect distinct P species present in agricultural soil could be shown as a proof of concept.