Predicting biochar cation exchange capacity using Fourier transform infrared spectroscopy combined with partial least square regression

• Published in: The Science of the total environment 2021-11, Vol.794, p.148762-148762, Article 148762
• Main Authors: Lago, Bruno Cocco, Silva, Carlos Alberto, Melo, Leônidas Carrijo Azevedo, Morais, Everton Geraldo de
• Format: Article
• Language: English
• Subjects: Chemometrics; Colloids with variable charges; Modeling; Multivariate analysis
• ISSN: 0048-9697; 1879-1026
• DOI: 10.1016/j.scitotenv.2021.148762

Determination of cation exchange capacity (CEC) in biochar by applying traditional wet methods is laborious, time-consuming, and generates chemical wastes. In this study, models were developed based on partial least square regression (PLSR) to predict CECs of biochars produced from a wide variety of feedstocks using Fourier transform infrared spectroscopy (FTIR). PLSR models used to predict CEC of biochars on weight (CEC-W) and carbon (CEC-C) basis were obtained from twenty-four biochars derived from several origins of feedstock, as well as compositions and mixtures, including four reference biochar samples. Biochars were grouped according to their CEC-W values (range of 4.0 to 150 cmolc kg−1) or CEC-C values (range of 6.0 to 312 cmolc kg−1). FTIR spectra highlighted features of the main functional groups responsible for biochar's CEC, which allowed a high prediction capacity for the PLSR models (R2pred ~ 0.9). Regression coefficients were associated to spectral variables of the organic matrix polar functional groups that contributed positively and negatively for biochar CEC. Phenolic and carboxylic were the main functional groups contributing to a higher biochar CEC, while CH and CC groups decreased the density of negative charges on the charred matrices. Chemometric models were highly robust to estimate biochar CEC, mainly on a weight basis, in a fast, reliable and economic way, compared to CEC conventional laboratory methods. [Display omitted] •A novel method for biochar CEC determination was developed.•The relationship between weight- and carbon-based biochar CEC was demonstrated.•PLS regression models were developed using an ample dataset FTIR spectra.•The models were effective in predicting the CEC of biochars in a wide range of values.•Functional groups polarity and amplitude were unraveled and related with biochar CEC.