Although some desired properties can be obtained by means of adjusting the particle size, recycling of nanoparticle materials can be seen as an application barrier due to adsorption and viscous effects of reaction mixture. Carbonaceous solid acids ground to nanosize (10-100 nm) had high catalytic activity for the cellulose that was similarly grounded with product selectivity higher than 90 % [120]. It is interesting to note that in some cases, the recycled catalysts exhibited no deactivation in the catalytic hydrolysis of fresh non-treated cellulose. In our research, the mixed metal oxide, Zn — Ca-Fe oxide, exhibited moderately good catalytic activity for hydrolyzing crystalline cellulose [121]. Cellulose conversion rate and glucose selectivity was 42.6 % and

69.2 %, respectively.

By introducing magnetic nano-components into nano-catalysts, it may be possible to separate the catalysts with an external magnetic field [122]. Magnetic supported nano-catalysts are highly efficient and environment-friendly. They have double func­tions of magnetism and acidity. They are obtained according to the following steps: (1) preparation of magnetic nanoparticles (magnetic core); (2) coupling anchor points on the surface of the nanoparticles; and (3) coordination of active sites to the anchor points [123]. By additional magnetic field, magnetic nano-catalysts disperse into liquid phase equably, avoiding the aggregation of nanoparticles and increasing the contact area between reactants and catalysts. In our work, hydrotalcite nanoparticles were synthesized and activated, and used for catalytic cellulose hydrolysis. Cellu­lose conversion rate and glucose selectivity of 46.6 % and 85.3 %, respectively, were obtained and remained stable for four cycles [27].

However, hetero-junctions between different active oxide and magnetic core lead to the decrease of activity as compared with single phase active oxide. Beydoun et al. [124] synthesized magnetic photocatalysts by coating TiO2 particles on Fe3O4 magnetic nanoparticles. It was found that the photo-activity of titania-coated magnetite decreased. Moreover, nanometer-scale apertures are not large enough to allow the transport of sub-micron scale cellulose. Increasing the aperture in nanoparticles would facilitate mass-transfer of oligosaccharides to catalytic sites. Techniques of increasing aperture include change of synthesis conditions, addition of pore-formers and change of templates. The external surface hydrolysis reactions result in a non-shape-selective reaction as well as deposition of un-hydrolyzed cellulose, leading to short lifetime for the catalysts. These issues can be avoided by adopting other treatment methods such as ultrasound. To use the internal acid sites effectively, cellulose or its constituents can be dissolved in organic solvents or ILs for further reactions. It is expected that acid-functionalized magnetic nano-catalysts are promising materials for the hydrolysis of biomass.

15.4 Summary

Solid acid catalysts, which have favorable characteristics such as efficient activity, high selectivity, long catalyst life, and ease in recovery and reuse have great potential for efficiently transforming lignocellulosic biomass into biofuels and chemicals, and can replace many conventional liquid acids for hydrolysis and pretreatment. Besides specific surface area, pore size, and pore volume, the active site concentration and acidic type are important factors for solid acid performance. Solid acid catalysts being considered for biomass hydrolysis should have a large number of B acid sites, a good affinity for the reactant substrates, and good thermal stability. Catalyst composition, porosity, and stability in the presence of water are other important properties for solid acids in biomass hydrolysis. A good solid catalyst with sufficient catalytic activity combined with appropriate reactor design should make it possible to realize biomass hydrolysis on a practical scale. The development of highly acidic solid catalysts with nanometer size that have special characteristics (e. g., paramagnetic properties) is an interesting area of research for developing practical systems for biomass hydrolysis. In the near future, through the combination of green solvents, nanoparticle techniques, and functional solid acid catalysts, it can be expected that chemical processes based on the catalysis of biomass will begin to replace petroleum — based processes to reach a sustainable economy.

It is favored that the feedstock is low FFAs virgin oils which can be blended with lower quality oils without seriously compromising production of high-quality biodiesel. Although this technique is widely practiced by biodiesel producers such pretreatment method doesn’tsolve the long-term problem of low-quality oil and also would not offer any potential solution for the conversion of FFAs to biodiesel.

Ozone treatment is one way of reducing the lignin content of lignocellulosic wastes. This results in an increase of the in vitro digestibility of the treated material, and unlike other chemical treatments, it does not produce toxic residues. Ozone can be used to degrade lignin and hemicellulose in many lignocellulosic materials such as wheat straw [149], bagasse, green hay, peanut, pine [150], cotton straw [151], and poplar sawdust [152]. Research indicated [153] ozone is highly reactive toward compounds incorporating conjugated double bonds and functional groups with high electron densities. Therefore, the moiety, most likely to be oxidized in ozonization of lignocellulosic materials, is lignin due to its high content of C=C bonds. Thus, during the ozonolysis, the degradation is mainly limited to lignin. Ozone attacks lignin releasing soluble compounds of less molecular weight, mainly organic acids such as formic and acetic acid [153]. The main advantages linked to this process are the lack of any degradation products that might interfere with subsequent hydrolysis or fermentation and the reactions occurring at ambient temperature and normal pres­sure. Furthermore, the fact that ozone can be easily decomposed by using a catalytic bed or increasing the temperature means that processes can be designed to minimize environmental pollution. A drawback of ozonolysis is that a large amount of ozone is required, which can make the process expensive and less applicable [154]. However, recently Hu et al. [155] demonstrated that a lower charge of ozone could be used to enhance the enzymatic digestibility of cellulose, if the ozone-treated biomass was not washed and the in-situ generated acids were employed in a subsequent DAP.

8.2 Summary

The effects of different chemical pretreatment technologies on the structure of lig- nocellulose are summarized in this section. In addition, the environment impacts of these pretreatments are also briefly discussed. Some directions and perspectives are also proposed for the future chemical pretreatment technologies.

Compared to fossil fuels, biomass has a high concentration of alkali salts, and their removal from the product gas is a very important step in biomass gasification. Alkali salts will condense below 600 °C, which causes serious corrosion problems. If the temperature of the gas decreases below 600 °C, the alkali salts condense and can be separated in a cyclone or Alters. However, in some applications, gas cannot be cooled so alumosilicates, such as bauxite, kaolinite, bentonite, and naturally occurring zeolite, can be used for alkali removal at temperatures up to 700 °C [59].

10.5.1.1 Electrostatic Precipitators

Electrostatic precipitators (ESP) are used to remove fine solids and liquid droplets from gas stream; however, they are not very efficient in terms of tar removal at high temperature. In order to have an efficient tar removal, gas stream should be quenched before ESP. In wet ESP, gas is ionised upon passing between high voltage electrodes and a grounded electrode. The produced ions attach themselves to dust particles or tar and water droplets. The charged particles and droplets are attracted to the grounded electrode, flowing to the bottom of the ESP where they are collected.

Many simplified pyrolysis global reaction models were proposed in the scientific lit­erature. Babu [36] proposed to regroup these conceptualizations of pyrolysis kinetics into three categories: (1) single-step models, (2) independent components models and (3) parallel and series reactions models. Figure 11.2 illustrates these models.

11.3.2.1 Single Decomposition Step Models (1-Step Models)

The simplest pyrolysis models consider a single decomposition step (Fig. 11.2a). Biomass decomposition directly yields a stream of bio-char, bio-oil and non­condensable gas. These models have the advantage of simplicity and possess a limited number of parameters. These models can be accurate for a limited range of pyrolysis conditions where the temperature is constant (isothermal system) or relatively low (<450 °C, thus conventional pyrolysis) and the product composition does not vary significantly.

In reality, biomass decomposition is more complex. In the case of non-isothermal systems at high heating rates, it may appear as if different components are reacting at different temperatures with their specific reaction kinetics. The notion of ‘pseudo­component’ emerges from that behaviour. Even if a system is operated isothermally at high temperature, pyrolysis will happen during the heating step where different products composition will be obtained. Single decomposition step models are thus only suitable in specific pyrolysis situations and are generally inadequate to reproduce industrial fast pyrolysis behaviour.

Until now, most gasification and pyrolysis systems use DC plasma torches. Fig­ure 12.2a shows the typical schematic diagram of power-supply circuit. The ballast resistor is used to stabilize burning of DC arc that causes considerable real power losses.

The most widespread DC plasma torches are Westinghouse Plasma Corp plasma torches with cylindrical electrodes. Sketch of this type of plasma torch is presented in Fig. 12.2b.

Power ranges for plasma torches MARC-3A and MARC-11L are 130-300 kW and 300-800 kW [87, 88], respectively, which are most attractive for industry. These are the most advanced designs. Their efficiency of energy transfer from the arc to plasma (thermal efficiency) is 70-85 %. The total efficiency of the system including

ohmic resistance losses in power-supply system is significantly lower. The lifetimes of electrodes for these models are 600 and 1,000 h, respectively.

EUROPLASMA designed a series of plasma torches of the same type which could be fed by CO2, CO, CH4, H2, N2, and their mixtures. Figure 12.2c shows a 300 kW plasma torch [89].

Series of “Linde” plasma torches are described in [83-85]. These devices are designed for short-time operation at high pressures and power up to 20 MW. These plasma torches use cylindrical copper electrodes. Some models have the working gas enthalpy up to 10 MJ/kg. Stabilized rectifiers power these devices. Voltage of power sources is up to 10 kV. Typically, the flow rates of cooling agents for ballast resistor and plasma torch are virtually equal. Pressure in the cooling system is about 4 MPa. Plasma torches MDC-200 and MDC-300 [86] use air as a working gas and operate in relatively short-time mode, they are intended for special experiments. Maximum working pressure of gas in the plasma torch chamber is 25 MPa, a range of operating currents 100-1,200 A, voltage 1.2-14 kV, thus power changes from 0.7 to 10.5 MW.

Chemical pretreatment is a widely used method. It effectively removes and recovers most of hemicellulose portion as soluble sugars, and disrupts lignin to be partially dissolved in aqueous acidic/basic solution (e. g., H2O2 and ammonia) [1]. Chemical pretreatment with acids has proven to be effective for breaking down hydrogen bonds leading to intra-crystalline cellulose swelling. Acid pretreatment is a process in which hydronium ions break down or attack inter- and intra-molecular bonds among cellu­lose, hemicellulose, and lignin. It increases the porosity of substrate and accessibility of cellulose to enzymes for subsequent hydrolysis. During acid pretreatment process, very little cellulose is hydrolyzed. However, the process is usually accompanied by further degradation of monomers, and it has drawbacks such as equipment corrosion and issues in the recovery and recycle of the acid.

Alkali (e. g., NaOH, KOH) solution is a swelling agent for both crystalline and amorphous celluloses that can destroy the linkages between lignin and carbohydrates by saponification of intermolecular ester bonds [20]. Kumar et al. [21] reported that NaOH pretreatment increased hardwood digestibility from 14 % to 55 % by reducing lignin content from 24-55 % to 20 %. Some chemical agents, such as peroxides and acidic alcohol solutions, have advantages for dissolving lignin and loosening hemicellulose from insoluble crystalline cellulose. Pan et al. [22] pretreated woody biomass with an organic solvent (1.25 % H2SO4 and 50 % ethanol) under conditions of 180 °C for 60 min, about 75 % lignin was removed from the substrate.

Ionic liquids (ILs) are efficient for the pretreatment and hydrolysis of lignocel- lulosic materials, and can dissolve biomass and overcome many of the physical and biochemical barriers for hydrolysis at ambient conditions. Many ILs have been shown to be effective solvents for cellulose. After IL pretreatment, glucose yield from subsequent cellulose hydrolysis is greatly increased with 97 % enzymatic glucose conversion rate being reported [23].

Organosolv process involves the use of an organic or aqueous organic solvent mixture (methanol, ethanol, acetone, ethylene glycol, triethylene glycol, tetrahydrofurfuryl alcohol, glycerol, aqueous phenol, aqueous я-butanol) and water, with or without ad­dition of an acid (oxalic, salicylic, and acetyl salicylic acid) or base as catalyst agents at high temperature (150-200 °C) for the lignin degradation and partial solubilization of hemicellulose sugars [10]. Solvents used during the pretreatment can be elimi­nated by evaporation and condensation and recycled to reduce the operational costs of the process. This step allows avoiding the presence of inhibitory compounds, which mayreduce the rate of enzymatic hydrolysis [14,27]. During organosolv, low boiling point alcohols such as methanol, ethanol seems more effective due to low cost and easy recovery of solvents. Lignocellulosic substrates after organosolv pretreatment have high crystallinity and increased surface area for better cellulolytic enzymes ac­tion. However, organosolv pretreatment methods are expensive at large scale. Zhao et al. [67] comprehensively reviewed the various organosolv pretreatment methods (alcohol mediated, ethanol, acetone, per acetic acid, per formic acid, etc). They further recommended developing the pretreatment strategies based on continuous processes in reactors with fewer loads of organic solvents with high solid biomass and exploration of the increased applications of by-products generated during the organosolv pretreatment.

Mesa et al. [23] described a combination of a dilute-acid pretreatment followed by the organosolv pretreatment with NaOH (60 min, 195 °C, ethanol 30 % v/v) to fractionate the SB which resulted in 67.3 % (w/w) glucose released.

16.3.1 Biological Pretreatment

The selective delignifying microorganisms usually applied to SB/SL for the degra­dation of lignin. This process can also be referred as in-situ microbial delignification (ISMD). Generally white-rot fungi are used for the degradation of lignin and hemi- celluloses fraction of SB/SL. The myco-SB/SL (SB/SL after fungal growth) can be used subsequently for enzymatic hydrolysis which in turn yields high amount of fermentable sugars [13, 68]. Figure 16.2 shows the mechanistic demonstration of biological pretreatment applied to SB/SL.

This is the low energy and capital intensive process where mild environmen­tal conditions are required [68]. Selective lignin degraders such as Pycnoporous cinnabarinus, and Ceriporiopsis subvermispora are more useful in ISMD, having affinity toward lignin breakdown than cellulose and hemicelluloses [13]. SB was pre­treated with the C. subvermispora for 30 days of incubation. The results of chemical analysis and mass component loss showed that C. subvermispora was selective to lignin degradation. Pretreated SB after soda/anthraquinone pulping showed improved pulp yields, kappa number, and viscosity pretreated SB [69].

The results of PE and DE after each pretreatment are summarized in Table 19.3.

For samples after organosolv extraction (Group 1), the PE and DE were 51.4 % and 76.5 %, respectively. The results indicated that organosolv extraction had a significant effect on delignification. The delignification efficiencies are comparable or slightly higher than those reported in previous studies [42, 50]. The higher PE and higher DE might be accounted for by the slightly higher temperature (by 20 °C) employed in this work. Pasquini et al. [51] found that both temperature and pressure can affect PE and DE: A higher temperature and pressure could lead to an increase in the PE and DE. In a study by Pasquini et al., the DE was on the order of 93.1 % for P. tarda wood chips when 16.0 MPa and 190 °C were employed. This process also leads to the formation of acetic acid or other carboxylic acids that can act as catalysts accelerating the rupture of lignin-carbohydrate polymers in the presence of water.

For Group 2 samples (with organosolv extraction followed by ultrasonic treat­ment), slightly improved PE (53.3 %) and PE (77.2 %) were obtained. The results suggest that the ultrasonic treatment did affect PE and DE, although the effect was not significant. This could be caused by the decreased particle size of pine sawdust after ultrasonic treatment. A decrease in particle size would lead to an increase in specific surface area and thus release of hemicellulose and lignin. Ultrasonic treat­ment has been widely used to enhance the extraction of hemicellulose in alkaline solutions by introducing violent cavitation. In this study, the ultrasonic treatment of pine sawdust accounted for only 3-5 wt% weight loss, due to the fact that distilled water rather than alkaline solution was used as liquid phase in the ultrasonic treat­ment. Extractives from hemicellulose and cellulose could not be dissolved easily in water. The comparable DE values with and without ultrasonic treatment (76.5 % vs. 77.2 %) suggest that ultrasonic treatment was not very effective for removing lignin.

For Group 3 samples (after organosolv extraction followed by NaOH treatment), the PE (57.7 %) and DE (81.5 %) were significantly higher, implying the high effi­ciency of NaOH treatment in the extraction of hemicellulose and lignin. When the organosolv extracted solid was added into NaOH solution, its color turned to dark brown, and the mixture was much finer and more viscous than it was before the treat­ment. This could be explained by the fact that the NaOH treatment of lignocellulosic materials can cause swelling which leads to an increase in internal surface area, a decrease in the degree of polymerization, a decrease in crystallinity, and disruption of the lignin structure [13, 32].

It is not surprising to observe that organosolv extraction followed by ultrasonic and NaOH treatment in sequence (Group 4) led to the highest PE (61.1 %) and DE (86.4 %), approximately 10 %, respectively higher than that for the Group 1 pretreatment, as shown in Table 19.3.

Due to the poor performance of the original cellulase preparation, two industrial en­zymes specifically designed for cellulosic ethanol production were evaluated, namely CTec2 (cellulase and xylanase) and HTec2 (xylanase). Results using these enzymes on the pretreated material approached the maximum theoretical yield (Fig. 9.3). After 2 days of enzymatic hydrolysis, the mixed enzyme (CTec2) can extract almost 60 % total sugars in wood at 6 % enzyme loading. When using higher enzyme loadings of 30 %, the mixed enzyme can reach maximum sugar yield in 2 days. The xylanase (HTec2) preparation could also yield 60 % total sugars within 2 days, which indicated this enzyme had cellulase activity.

The total sugars yield after 3 days enzymatic hydrolysis reached 78 % sugar yield using the mixed enzyme at 6 % loading, while the xylanase achieved 58 % sugars. Therefore, the mixed enzyme can be used for further saccharification on hybrid poplar at a loading between 3 and 6 %. To note, the enzyme loading appears high since it is in a stabilizing buffer solution but it is actually a dilute protein solution (actual protein concentration is proprietary information).