1 Center for BioProcess Engineering, Department of Chemical and Biochemical Engineering, Technical University of Denmark2 Department of Chemical and Biochemical Engineering, Technical University of Denmark3 unknown
Enzyme technology provides key strategies to green chemistry as many processes have undergone re-design to serve increasing demands towards being sustainable. While the population is rapidly increasing on our planet it is leading to accumulative problems in terms of production of waste, depletion of natural fossil resources and increasing demands for food and energy. Biorefinery, in particular, deals with related challenges, as it is defined to deal with the conversion of biomass using enzyme technology to produce renewable energy, in terms of heat, power and fuel. Furthermore, biorefinery intends to extract value-added compounds from biomass to avoid downcycling effects prior to e.g. biofuel production. Those value-added compounds are highly attractive to be utilized as food ingredients, bio-chemicals or precursors for pharmaceutical products and represent high market potentials for related industries. However, as biorefinery concepts are implemented in many industrial processes an increasing demand for Process Analytical Technology (PAT) evolves to monitor, understand and steer processes optimally. Biomasses can be very diverse and are usually of complex chemical nature. Conventional univariate analytical methods therefore require time-consuming sample preparation which is mostly cumbersome to analyze biomass conversion related processes. Throughout this project alternative approaches will be presented to deal with the individual challenges. As outlined it seems obvious that more advanced techniques are necessary to monitor such difficult reactions as enzymatic biomass degradations. Such techniques should be of multivariate nature to capture and understand complex patterns in comparison to univariate techniques which can only capture information in a highly specific sense which does not allow interference of information. Vibrational spectroscopy (e.g. infrared) represents such multivariate techniques and is mostly used throughout the project. Data is analyzed by chemometric methods to extract the underlying patterns from the complex datasets. Hence, this project focuses on chemometric approaches utilizing mostly Fourier Transform Infrared (FTIR) spectroscopic data to provide descriptive and predictive insights into biomass conversion related processes. Two main study fields are introduced to the reader. First, two-way chemometric methods are used to establish Process Analytical Technology (PAT) solutions for prediction of monosaccharide release efficiency of pretreated destarched corn bran using Near Infrared (NIR) spectroscopy (PAPER 1). Throughout this study predictive and descriptive models were established to evaluate the pretreatment effect without the need to perform the subsequent enzymatic hydrolysis itself. Furthermore, the efficiency (and quality) of differently extracted pectin from lime peel could be predicted from measuring FTIR spectra (PAPER 2). The prediction models were compared to results retrieved from carbohydrate microarray analysis which additionally enhanced the understanding of the structural properties of the extracted pectin. Secondly, enzyme kinetics of biomass converting enzymes was examined in terms of measuring enzyme activity by spectral evolution profiling utilizing FTIR. Chemometric multiway methods were used to analyze the tensor datasets enabling the second-order calibration advantage (reference Theory of Analytical chemistry). As PAPER 3 illustrates the method is universally applicable without the need of any external standards and was exemplified by performing quantitative enzyme activity determinations for glucose oxidase, pectin lyase and a cellolytic enzyme blend (Celluclast 1.5L). In PAPER 4, the concept is extended to quantify enzyme activity of two simultaneously acting enzymes, namely pectin lyase and pectin methyl esterase. By doing so the multiway methods PARAFAC, TUCKER3 and NPLS were compared and evaluated towards accuracy and precision.
Main Research Area:
Hansen, Per Waaben, Egebo, Max, Mikkelsen, Jørn Dalgaard