Molecule-specific soil analysis using shifted excitation Raman difference spectroscopy: from laboratory investigations towards field deployment of portable sensors

Precision agriculture relies on detailed knowledge about the current soil composition to make informed treatment decisions for site-specific farming. Fast and non-destructive optical methods are a promising approach to complement existing complex and time-consuming standard laboratory analyses. Raman spectroscopy is well-suited for this task as it permits to obtain molecule-specific soil information, but it suffers from the weak Raman effect and dominant background interferences, e.g., caused by fluorescence or daylight. Here, shifted excitation Raman difference spectroscopy (SERDS) is applied to overcome this challenge using a physical approach with two slightly shifted laser wavelengths. An in-house developed dual-wavelength diode laser with two emission lines around 785 nm is used as excitation light source for SERDS. In preparation for on-site soil analysis, initial SERDS laboratory investigations are performed on 150 samples collected from an agricultural field in Germany. Results show the ability of our approach for qualitative (detection of soil constituents) and quantitative (prediction of organic matter and carbonate contents) soil investigations. The transfer of SERDS from the laboratory setting to the field environment is realized by an in-house developed portable SERDS system specifically designed for soil analysis. On-site investigations on the above-mentioned field are exemplarily carried out at selected points along a distance of 480m revealing the spatial distribution of carbonaceous matter and selected soil minerals. Moreover, a successful prediction of the contents of organic carbon (R2 = 0.89, root mean square error of cross-validation RMSECV = 0.3%) and carbonate (R2 = 0.86, RMSECV = 2.5%) as important soil parameters could be realized. Results demonstrate the capability of portable SERDS for qualitative and quantitative on-site soil analysis directly on agricultural fields thus paving the way for applications in precision agriculture.