Contents lists available at ScienceDirect Industrial Crops & Products journal homepage: www.elsevier.com/locate/indcrop Application of microscopy techniques for a better understanding of biomass pretreatment Victoria Rigual ⁎ , Juan C. Domínguez, Sandra Rivas, Antonio Ovejero-Pérez, M. Virginia Alonso, Mercedes Oliet, Francisco Rodriguez Department of Chemical Engineering and Materials Science, Faculty of Chemistry, Complutense University of Madrid, Avda. Complutense s/n, 28040, Madrid, Spain ARTICLE INFO Keywords: Image processing Scanning electron microscopy Digestibility Fractal dimension Lacunarity Colocalisation ABSTRACT In this work, the application of digital image analysis to micrographs obtained by two microscopy techniques, confocal fuorescence microscopy and scanning electron microscopy, is proposed. Two pretreated biomasses (Pinus radiata and Eucalyptus globulus) are characterized. Image processing methodologies are employed to de- termine equivalent diameters distribution (using x100 SEM micrographs), fractal dimension and lacunarity (using x1000 SEM micrographs), and surface composition and colocalisation (using confocal fuorescence mi- croscopy). Particle size distributions result in median equivalent diameters in the range of 7.9–15.4 μm that follow an exponential tendency, and are strongly afected by the severity of the autohydrolysis pretreatment and the biomass type. Biomass surface composition (in terms of holocellulose/lignin ratios) is not the same than the obtained with standard protocols (NREL), and let distinguish the lignin deposition in the surface, especially in severe autohydrolysis conditions with pine. Fractal dimension (in the range of 2.65–2.74) and lacunarity (in the range of 0.05-0.16) have been found to have a strong dependence of the solid digestibility (linear correlations of 0.93 and 0.70) for the two biomasses, independently of the pretreatment employed. 1. Introduction Lignocellulosic biomass is an abundant and cost-efective-renewable resource with a production of 15–17 × 10 17 Mt annually (Chen et al., 2017). In order to convert native biomass into an efective substrate for enzymatic hydrolysis, pretreatments are necessary (Agbor et al., 2011). The goal of any pretreatment technology is to alter or remove structural and compositional impediments to hydrolysis, so enzyme hydrolysis rate is improved and fermentable sugars yields of cellulose or hemi- cellulose are increased (Miyamoto et al., 2018). In this way, some of the key factors that increase digestibility are the cellulose crystallinity, hemicellulose disruption, accessible surface area (porosity), lignin protection, and association of hemicellulose to lignin (Putro et al., 2016). Particle size reduction is also afected by the kind of pretreat- ment employed, as lower particles sizes enhances glucose yields during enzymatic hydrolysis (Deb et al., 2016). Another factor that strongly afects cellulose digestibility is the porous structure, as pores defne pathways that enhance cellulases reach cellulose (Arantes and Saddler, 2010). Pore size can be defned as the diameter of the largest tracer molecule that can pass through walls in a given time. Porosity can be measured through ultrastructural methods (electron microscopy), bulk exclusion techniques through solute exclusion, and functional assays that detect the transport of molecules across an intact wall (Read and Bacic, 1996). In order to measure all these key aspects of biomass di- gestibility, chemical and structural characterization using analytical tools such as lignin characterization, NMR, infrared and Raman spec- troscopy have been employed (Foston et al., 2016). Recently, ToF-SIMS has also been employed to obtain chemical information directly from the surface (Jung et al., 2018). Complementarily, microscopy techni- ques such as scanning electron microscopy or confocal fuorescence microscopy enrich the characterization through the visualization of structures and cellulose and lignin distribution (Li et al., 2011; Rigual et al., 2018a; Singh et al., 2015). Two of the most employed techniques are confocal fuorescence microscopy and scanning electron microscopy (SEM). Confocal fuor- escence microscopy enhances the understanding of biomass decon- struction and lignin changes, through the visualization of cell wall modifcation. This technique lets visualize the spatial and temporal interactions between enzymes and substrate (Ding et al., 2012; Donaldson and Vaidya, 2017). Furthermore, the employment of mul- tichannel images enhances the acquisition of images where diferent cell wall components can be observed, showing the positions where https://doi.org/10.1016/j.indcrop.2019.111466 Received 27 November 2018; Received in revised form 15 April 2019; Accepted 5 June 2019 ⁎ Corresponding author. E-mail address: vicrigua@ucm.es (V. Rigual). Industrial Crops & Products 138 (2019) 111466 0926-6690/ © 2019 Elsevier B.V. All rights reserved. T