HCA e PCA de espectros FT-IR e UV-VIS de ácidos húmicos e DRX de solos de TPI

Abstract

The formation of Terra Preta (TPI) intrigued many researchers, besides it has too many hypotheses, the most likely involves human activities, whose origin is related to the deposition of material remained of pre-Columbian populations. It resulted in soils rich in humic acid (HA) that are regarded as natural binders most widely distributed in nature, which greatly contributes to soil fertility. This study applied statistical methods of multivariate analyzes such as hierarchical cluster analysis (HCA), Principal Component Analysis (PCA), infrared spectra Fourier Transform (FT-IR) and ultraviolet-visible spectra (UV-Vis) AH ICC for eight sites in the Amazon region, and applied to spectra of X-ray diffraction (XRD) of soil same eight sites of the ICC. Statistical methods were applied using the R software that makes use of normalization of the data obtained from each spectrum to generate the results of PCA and HCA. Using r software tools, it was possible observe which functional groups and stretches contributed positively OH, C=O and C=C, to generate graphics of PCA and HCA extracted from the FT-IR spectra. As for the results of UV-Vis spectra it was not possible to observe that groups could contribute to generation of HCA and PCA. For the XRD spectra it was observed that the PC1 that kaolinite and illite were the most positive contribution to the graphics, the PC2 that in most of the contribution was quartz. Thus, the simple application of HCA and PCA in the spectra allowed to compare the eight sites and classifies them according to similarity of the spectra.