Traditional or adaptive design of experiments? A pilot-scale comparison on wood delignification.

Traditional design of experiments and response surface methodology are widely used in engineering and process development. Bayesian optimization is an alternative machine learning approach that adaptively selects successive experimental conditions based on a predefined performance measure. Here we compared the two approaches using simulations and empirical experiments on alkaline wood delignification to identify important benefits and drawbacks of Bayesian optimization in the context of design of experiments. The simulations showed that the selection of initial experiments and measurement noise influenced the convergence of the Bayesian optimization algorithm to known optimal conditions. Both methods, however, showed comparable pilot-scale results on optimal digestion conditions, where high cellulose yields were combined with acceptable kappa numbers and pulp viscosities. Bayesian optimization did not enable a decrease in the number of experiments required for reaching these conditions but provided a more accurate model in the vicinity of the optimum based on additional modelling and cross-validation. These results shed light on the practical differences between the two methodologies for process development and are an important contribution to the chemometrics and machine learning communities.

Mapping Nanocellulose- and Alginate-Based Photosynthetic Cell Factory Scaffolds: Interlinking Porosity, Wet Strength, and Gas Exchange.

To develop efficient solid-state photosynthetic cell factories for sustainable chemical production, we present an interdisciplinary experimental toolbox to investigate and interlink the structure, operative stability, and gas transfer properties of alginate- and nanocellulose-based hydrogel matrices with entrapped wild-type Synechocystis PCC 6803 cyanobacteria. We created a rheological map based on the mechanical performance of the hydrogel matrices. The results highlighted the importance of Ca2+-cross-linking and showed that nanocellulose matrices possess higher yield properties, and alginate matrices possess higher rest properties. We observed higher porosity for nanocellulose-based matrices in a water-swollen state via calorimetric thermoporosimetry and scanning electron microscopy imaging. Finally, by pioneering a gas flux analysis via membrane-inlet mass spectrometry for entrapped cells, we observed that the porosity and rigidity of the matrices are connected to their gas exchange rates over time. Overall, these findings link the dynamic properties of the life-sustaining matrix to the performance of the immobilized cells in tailored solid-state photosynthetic cell factories.

Wood-Water Relations Affected by Anhydride and Formaldehyde Modification of Wood.

The moisture uptake of wood is influenced by accessible hydroxyl groups acting as sorption sites and the water-available cell wall space. To what extent do these mechanisms control the moisture uptake in wood needs to be addressed. For this purpose, we modified sorption site density and cell wall space by wood treatments with acetic anhydride or formaldehyde and investigated their effects on moisture uptake. Chemical changes at the cell wall level caused by the treatments were first determined by confocal Raman imaging. Following this, the deuterium exchange method was used to gravimetrically measure the hydroxyl accessibility, while the moisture uptake and the consequent swelling of the wood were determined by dynamic measurements of mass and dimensions within the hygroscopic range. The results showed that the effectiveness in reducing the moisture content of untreated wood across the hygroscopic range differed between the anhydride- and formaldehyde-modified wood. We also observed a poor correlation of accessible hydroxyl concentration in formaldehyde-modified wood with weight percentage gain and water uptake. Moreover, the dynamic mass and dimension analysis indicated that the reduction in swelling in formalized wood was affected by an unidentified mechanism in addition to reduced moisture content.

Chemical Characterization and Visualization of Progressive Brown Rot Decay of Wood by Near Infrared Imaging and Multivariate Analysis.

Brown rot fungi cause a type of wood decay characterized by carbohydrate degradation and lignin modification. The chemical and physical changes caused by brown rot are usually studied using bulk analytical methods, but these methods fail to consider local variations within the wood material. In this study we applied hyperspectral near infrared imaging to Scots pine sapwood samples exposed to the brown rot fungi Coniophora puteana and Rhodonia placenta to obtain position-resolved chemical information on the fungal degradative process. A stacked-sample decay test was used to create a succession of decay stages within the samples. The results showed that the key chemical changes associated with decay were the degradation of amorphous and crystalline carbohydrates and an increase in aromatic and carbonyl functionality in lignin. The position-resolved spectral data revealed that the fungi initiated degradation in earlywood, and that earlywood remained more extensively degraded than latewood even in advanced decay stages. Apart from differences in mass losses, the two fungi produced similar spectral changes in a similar spatial pattern. The results show that near infrared imaging is a useful tool for analyzing brown rot decayed wood and may be used to advance our understanding of fungal degradative processes.

Fast and quantitative compositional analysis of hybrid cellulose-based regenerated fibers using thermogravimetric analysis and chemometrics.

Cellulose can be dissolved with another biopolymer in a protic ionic liquid and spun into a bicomponent hybrid cellulose fiber using the Ioncell® technology. Inside the hybrid fibers, the biopolymers are mixed at the nanoscale, and the second biopolymer provides the produced hybrid fiber new functional properties that can be fine-tuned by controlling its share in the fiber. In the present work, we present a fast and quantitative thermoanalytical method for the compositional analysis of man-made hybrid cellulose fibers by using thermogravimetric analysis (TGA) in combination with chemometrics. First, we incorporated 0-46 wt.% of lignin or chitosan in the hybrid fibers. Then, we analyzed their thermal decomposition behavior in a TGA device following a simple, one-hour thermal treatment protocol. With an analogy to spectroscopy, we show that the derivative thermogram can be used as a predictor in a multivariate regression model for determining the share of lignin or chitosan in the cellulose hybrid fibers. The method generated cross validation errors in the range 1.5-2.1 wt.% for lignin and chitosan. In addition, we discuss how the multivariate regression outperforms more common modeling methods such as those based on thermogram deconvolution or on linear superposition of reference thermograms. Moreover, we highlight the versatility of this thermoanalytical method-which could be applied to a wide range of composite materials, provided that their components can be thermally resolved-and illustrate it with an additional example on the measurement of polyester content in cellulose and polyester fiber blends. The method could predict the polyester content in the cellulose-polyester fiber blends with a cross validation error of 1.94 wt.% in the range of 0-100 wt.%. Finally, we give a list of recommendations on good experimental and modeling practices for the readers who want to extend the application of this thermoanalytical method to other composite materials. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s10570-021-03923-6.

Identification of cellulose textile fibers.

Distinguishing different textile fibers is important for recycling waste textiles. Most studies on non-destructive optical textile identification have focused on classifying different synthetic and natural fibers but chemical recycling requires more detailed information on fiber composition and polymer properties. Here, we report the use of near infrared imaging spectroscopy and chemometrics for classifying natural and regenerated cellulose fibers. Our classifiers trained on images of consumer textiles showed 100% true positive rates based on model cross-validation and correctly identified on average 8-9 out of 10 test set pixels using images of specifically made cotton, viscose and lyocell samples of known compositions. These results are significant as they indicate the possibility to monitor and control fiber dosing and subsequent dope viscosity during chemical recycling of cellulose fibers. Our results also suggested the possibility to identify fibers purely based on polymer chain length. This finding opens the possibility to indirectly estimate dope viscosity and creates entirely new hypotheses for combining imaging spectroscopy with classification and regression methods within the broader field of cellulose modification.

Hyperspectral near infrared image calibration and regression.

Reference materials are used in diffuse reflectance imaging for transforming the digitized camera signal into reflectance and absorbance units for subsequent interpretation. Traditional white and dark reference signals are generally used for calculating reflectance or absorbance, but these can be supplemented with additional reflectance targets to improve the accuracy of reflectance transformations. In this work we provide an overview of hyperspectral image regression and assess the effects of reflectance calibration on image interpretation using partial least squares regression. Linear and quadratic reflectance transformations based on additional reflectance targets decrease average measurement errors and make it easier to estimate model pseudorank during image regression. The lowest measurement and prediction errors were obtained with the column and wavelength specific quadratic transformations which retained the spatial information provided by the line-scanning instrument and reduced errors in the predicted concentration maps.

Distribution and curing reactions of melamine formaldehyde resin in cells of impregnation-modified wood.

Wood modification improves the properties of wood as a building material by altering the wood structure on a cellular level. This study investigated how dimensional changes of wood on a macroscopic scale are related to the cellular level chemical changes on the micron level after impregnation modification with melamine formaldehyde (MF) resin under different heat curing conditions. Our results showed that the curing conditions affected the polycondensation reactions and the morphological structure of the MF resin within the cell lumen. The diffusion of the resin into the cell wall was estimated based on the triazine ring vibration of melamine in the Raman spectrum at 950-990 cm-1. Thereby, it was shown that macroscopic changes in wood dimensions do not provide a reliable estimate for the cell wall diffusion of the resin. The removal of cell wall constituents during the modification, which was facilitated by the alkaline pH of the impregnation solution, counterbalanced the cell wall bulking effect of the resin. This was particularly evident for wet cured samples, where diffusion of MF resin into the cell wall was observed by confocal Raman microscopy, despite a reduction in macroscopic wood dimensions.

Hyperspectral Near-Infrared Image Assessment of Surface-Acetylated Solid Wood.

Visualization of acetic anhydride flow and its heterogeneity within the wood block necessitates the development of a reliable and robust analytical method. Hyperspectral imaging has the potential to acquire a continuous spectrum of chemical analytes at different spectral channels in terms of pixels. The large set of chemical data (3-dimensional) can be expanded into relevant information in a multivariate fashion. We quantified gradients in acetylation degree over cross sections of Scots pine sapwood caused by a one-sided flow of acetic anhydride into wood blocks using near-infrared hyperspectral imaging. A principal component analysis (PCA) model was used to decompose the high-dimensional data into orthogonal components. Moreover, a partial least-squares (PLS) hyperspectral image regression model was developed to quantify heterogeneity in acetylation degree that was affected by the flow of acetic anhydride through wood blocks and into the tracheid cell walls. The model was validated and optimized with an external test data set and a prediction map using the root-mean-squared error of an individual predicted pixel. The model performance parameters are well suited, and prediction of the acetylation degree at the image level was complemented with confocal Raman imaging of selected areas on the microlevel. NIR image regression showed that the acetylation degree was determined not only by the time-dependent flow of the acetic anhydride through the wood macropores but also by the diffusion of the anhydride into the wood cell walls. Thereby, thin-walled earlywood sections were acetylated faster than the thick-walled latewood sections. Our results demonstrate the suitability of near-infrared imaging as a tool for quality control and process optimization at the industrial scale.

Cellular level chemical changes in Scots pine heartwood during incipient brown rot decay.

The heartwoods of many wood species have natural resistance to wood decay due to the accumulation of antifungal heartwood extractives. The natural durability of heartwoods has been extensively investigated, yet very little information is available on the initiation of heartwood decay. This experiment examined the onset of Rhodonia placenta brown rot decay in Scots pine heartwood in order to identify the key changes leading to heartwood decay. An imaging approach based on Raman imaging and multivariate image analysis revealed that the degradation of heartwood began in the innermost cell wall layers and then spread into the remaining cell walls and the middle lamella. Pinosylvins were extensively degraded in the cell walls, middle lamella and extractive deposits, while unidentified material most likely consisting of hemicelluloses and/or lipophilic extractives was removed from the inner cell wall layers. Changes similar to inner cell wall degradation were seen in the remaining cell walls in more advanced decay. The results indicate that the key change in incipient heartwood decay is the degradation of antifungal heartwood extractives. The inner cell wall degradation seen in this experiment may serve a nutritive purpose or facilitate the penetration of degradative agents into the cell walls and middle lamella.

Hyperspectral near infrared imaging quantifies the heterogeneity of carbon materials.

For many applications heterogeneity is a direct indicator of material quality. Reliable determination of chemical heterogeneity is however not a trivial task. Spectral imaging can be used for determining the spatial distribution of an analyte in a sample, thus transforming each pixel of an image into a sampling cell. With a large amount of image pixels, the results can be evaluated using large population statistics. This enables robust determination of heterogeneity in biological samples. We show that hyperspectral imaging in the near infrared (NIR) region can be used to reliably determine the heterogeneity of renewable carbon materials, which are promising replacements for current fossil alternatives in energy and environmental applications. This method allows quantifying the variation in renewable carbon and other biological materials that absorb in the NIR region. Reliable determination of heterogeneity is also a valuable tool for a wide range of other chemical imaging applications.

Process water properties from hydrothermal carbonization of chemical sludge from a pulp and board mill.

Hydrothermal carbonization (HTC) can be used to break down sludge structure and generate carbonaceous hydrochar suitable for solid fuel or value-added material applications. The separation of char and the reaction medium however generates a filtrate, which needs to treated before potential discharge. Thus, this work determined filtrate properties based on HTC temperature and sludge moisture content and estimated the discharge emissions and the potential increase in analyte loads to an industrial wastewater treatment plant based on derived regression models. Direct discharge of HTC filtrate would significantly increase effluent emissions at the mill, indicating the filtrate treatment is crucial for the future implementation of HTC at pulp and paper mills. Recycling the HTC filtrate to the wastewater plant would lead to only a nominal increase in effluent flow, but would increase the suspended solids, BOD, COD and total nitrogen loads by 0.1-0.8%, 3.8-5.3%, 2.7-3.1% and 42-67%, respectively, depending on HTC temperature.

Hyperspectral Imaging to Determine the Properties and Homogeneity of Renewable Carbon Materials.

Hyperspectral imaging within the near infrared (NIR) region offers a fast and reliable way for determining the properties of renewable carbon materials. The chemical information provided by a spectrum combined with the spatial information of an image allows mathematical operations that can be performed in both the spectral and spatial domains. Here, we show that hyperspectral NIR imaging can be successfully used to determine the properties of hydrothermally prepared carbon on the material and pixel levels. Materials produced from different feedstocks or prepared under different temperatures can also be distinguished, and their homogeneity can be evaluated. As hyperspectral imaging within the NIR region is non-destructive and requires very little sample preparation, it can be used for controlling the quality of renewable carbon materials destined for a wide range of different applications.

Predicting the drying properties of sludge based on hydrothermal treatment under subcritical conditions.

The effects of hydrothermal treatment on the drying properties of sludge were determined. Sludge was hydrothermally treated at 180-260 °C for 0.5-5 h using NaOH and HCl as additives to influence reaction conditions. Untreated sludge and attained hydrochar samples were then dried under identical conditions with a laboratory microdryer and an X-ray microtomograph was used to follow changes in sample dimensions. The effective moisture diffusivities of sludge and hydrochar samples were determined and the effect of process conditions on respective mean diffusivities evaluated using multiple linear regression. Based on the results the drying time of untreated sludge decreased from approximately 80 min to 37-59 min for sludge hydrochar. Drying of untreated sludge was governed by the falling rate period where drying flux decreased continuously as a function of sludge moisture content due to heat and mass transfer limitations and sample shrinkage. Hydrothermal treatment increased the drying flux of sludge hydrochar and decreased the effect of internal heat and mass transfer limitations and sample shrinkage especially at higher treatment temperatures. The determined effective moisture diffusivities of sludge and hydrochar increased as a function of decreasing moisture content and the mean diffusivity of untreated sludge (8.56·10(-9) m(2) s(-1)) and sludge hydrochar (12.7-27.5·10(-9) m(2) s(-1)) were found statistically different. The attained regression model indicated that treatment temperature governed the mean diffusivity of hydrochar, as the effects of NaOH and HCl were statistically insignificant. The attained results enabled prediction of sludge drying properties through mean moisture diffusivity based on hydrothermal treatment conditions.

Hydrothermal carbonization of industrial mixed sludge from a pulp and paper mill.

Mixed sludge from a pulp and paper mill was hydrothermally carbonized at 180-260°C for 0.5-5h with the use of HCl or NaOH for determining the effect of acid and base additions during sludge carbonization. Based on the results carbonization was mainly governed by dehydration, depolymerization and decarboxylation of sludge components. Additive type had a statistically significant effect on hydrochar carbon content and carbon and energy yield, of which especially energy yield increased through the use of HCl. The theoretical energy efficiencies of carbonization increased with decreasing reaction temperature, retention time and the use of HCl and suggested that the energy requirement could be covered by the energy content of attained hydrochar. The BOD5/COD-ratios of analyzed liquid samples indicated that the dissolved organic components could be treated by conventional biological methods.

Utilization of steel, pulp and paper industry solid residues in forest soil amendment: relevant physicochemical properties and heavy metal availability.

Industrial residue application to soil was investigated by integrating granulated blast furnace or converter steel slag with residues from the pulp and paper industry in various formulations. Specimen analysis included relevant physicochemical properties, total element concentrations (HCl+HNO3 digestion, USEPA 3051) and chemical speciation of chosen heavy metals (CH3COOH, NH2OH·HCl and H2O2+H2O2+CH3COONH4, the BCR method). Produced matrices showed liming effects comparable to commercial ground limestone and included significant quantities of soluble vital nutrients. The use of converter steel slag, however, led to significant increases in the total concentrations of Cr and V. Subsequently, total Cr was attested to occur as Cr(III) by Na2CO3+NaOH digestion followed by IC UV/VIS-PCR (USEPA 3060A). Additionally, 80.6% of the total concentration of Cr (370 mg kg(-1), d.w.) occurred in the residual fraction. However, 46.0% of the total concentration of V (2470 mg kg(-1), d.w.) occurred in the easily reduced fraction indicating potential bioavailability.