Identifying the fingerprint of permanganate oxidizable carbon as a measure of labile soil organic carbon using Fourier transform mid‐infrared photoacoustic spectroscopy

Abstract

AbstractPermanganate oxidizable carbon (POXC) is a sub‐pool of labile soil organic carbon (SOC) and is defined as the carbon (C) that can be oxidized by potassium permanganate (KMnO4). Determination of the content of POXC has been used as a way to measure labile C and may be related to soil microbial biomass and particulate organic C. Fourier transform mid‐infrared photoacoustic spectroscopy (FTIR‐PAS) has been used for characterizing the chemical composition of soil organic matter. The objective of the current study was to identify the chemical information in the FTIR‐PAS spectra that is unique to POXC and therefore associated with labile C. FTIR‐PAS was used for characterizing SOC using a set of soil samples collected in four different countries (Laos, Malaysia, Peru and Thailand, n = 575). Partial least squares (PLS) regression was used to correlate the collected FTIR‐PAS spectra with both POXC and total SOC. The FTIR‐PAS spectra combined with the PLS method was useful for predicting both SOC and POXC. However, SOC and POXC were also highly correlated (R2 = 0.84), and because POXC can be seen as a fraction of total SOC, a method for extracting the FTIR‐PAS fingerprint of the POXC model that is independent of SOC was applied for the first time. The results showed that the POXC prediction was mainly based on its strong correlation with SOC and there was little information unique to POXC. However, the results do indicate four dominant peaks at 1,574, 1,495, 1,138 and 1,041 cm−1 to be specific for POXC. These peaks can be ascribed to amide II, aromatics, polysaccharides and carbohydrates, respectively.Highlights Permanganate oxidizable carbon (POXC) was predicted using FTIR‐photoacoustic spectroscopy (FTIR‐PAS). The POXC prediction using FTIR‐PAS is mainly based on its high correlation with SOC. A method extracting the unique spectral features of POXC independently of SOC was applied. Around 11% of the information used to predict POXC was identified as unique. Four dominant peaks were found to represent the POXC fingerprint.