Determining Composition of Bamboo using Thermogravimetric Analysis (TGA)-Chemometrics Technique

Abstract

The chemical composition of biomass plays a pivotal role in determining the processing routes, operational costs, and product quality of bioprocesses. Current wet chemistry analysis methods are time consuming and energy intensive making them unideal for industrial scale applications where productivity is important. In this study, suitability of thermogravimetric analysis coupled with chemometrics (herein referred to as TGA-Chemometrics technique) was evaluated in determining the composition of cellulose, lignin, and hemicellulose in bamboo. Bambusa balcooa and Bambusa bambos species were used and thermal analysis data was collected under oxidative and inert conditions. Then, the data was used to develop predictive models (regression models) to predict the composition of bamboo. The model-predicted results were compared to the results obtained via the wet chemistry analysis to assess the efficacy of the developed predictive models. The results indicated that the predictive models developed using the data from both oxidative and non-oxidative thermal decomposition were accurate at predicting the amount of cellulose, lignin, and hemicellulose with R2 and RMSEp values of 0.82–0.95 and 0.6–1.79, respectively. Moreover, the developed chemometric models developed from the data obtained under inert conditions displayed a better prediction than those developed from the data obtained under oxidative conditions, indicated by their lower RMSEp values. The difference in prediction was not large, showing an average absolute error of 1.49%. In conclusion, this study shows that TGA coupled with chemometric is a potential technique to obtain composition of biomass, making it an alternative method to the wet chemistry analysis method.