Monitoring complex monosaccharide mixtures derived from macroalgae biomass by combined optical and microelectromechanical techniques. (May 2018)

• Record Type: Journal Article
• Authors: Chudnovsky, Alexandra; Golberg, Alexander; Linzon, Yoav
• Is Part Of: Process biochemistry. Volume 68(2018)
• Page Start: 136
• Page End: 145
• Publication Date: 2018-05
• Subjects: Biorefinery -- Reflectance spectroscopy -- Micro- and nano-electromechanical systems (MEMS/NEMS) -- Monosaccharide mixtures; Biochemical engineering -- Periodicals; Biotechnology -- Periodicals; Biochemistry -- periodicals; Biotechnology -- periodicals; Chemical Engineering -- periodicals; Génie biochimique -- Périodiques; Biotechnologie -- Périodiques; Biochemical engineering; Biotechnology; Periodicals; 660.63
• DOI: 10.1016/j.procbio.2018.01.018
• Languages: English
• ISSNs: 1359-5113

Abstract

Graphical abstract: Highlights: Two orthogonal techniques (Reflectance spectroscopy and MEMS vibrometry) for the Ulva monosaccharides phenotyping were tested. Rhamnose, glucose, xylose and uronic acid contents released with acid hydrolysis were quantitated. MEMS provided better accuracy of estimation of all components (R 2 = 0.75–0.90). Abstract: To foster the development of macroalgal biomass for biorefinery applications, we tested two orthogonal techniques for rapid phenotyping of the green macroalga Ulva based on its glucose, rhamnose, xylose and glucuronic acid contents as derived for reference by acid hydrolysis. Partial Least Squares (PLS) regression analyses, calculation of slopes and correlations across different spectral ranges/frequencies were used to predict the monosaccharide contents using two complementary methods: near infrared reflection spectroscopy (NIRS) and microelectromechanical systems (MEMS) resonating membrane vibrometry. Both methods were found to perform sufficiently well in monosaccharide mixtures and to enable quantitative assessment of different monosaccharide contents with the relative Root Mean Square Error of Prediction (%RMSEP) ranging from 8 to 16% (with similar accuracy when using PLS analyses). The best estimation was found for rhamnose and glucose contents, whereas xylose and uronic acid content predictions were found to be less accurate using PLS analyses. For the two latter components, slopes across different spectral ranges and frequencies at certain signals provided better estimates for their concentrations (e.g. for NIRS slopes: R 2 values in the range 0.55–0.66 and with higher accuracy for MEMS: between 0.75 and 0.90). This result is pivotal for opening new perspective to the construction of simple, multi-functional sensors for biomass downstream processing control in biorefinery and biometric applications.