Most of the chemical pretreatment technologies that have been described herein are effective on one or more factors that contribute to lignocellulosic recalcitrance, as shown in Table 8.6. Table 8.7 summarizes the main advantages and disadvantages of these pretreatment technologies. Each method discussed shows the ability to take the complex carbohydrate and depolymerize the substrate to a lower fraction for enzymatic saccharification in the subsequent step. There are a number of feasible routes, each of which has their own merits and disadvantages, and consequences on the enzymatic hydrolysis.

Tars in the product gas can be tolerated in some systems where the gas is used as a fuel in applications, such as burners. However, in most of applications, tars in the raw product gases, even at low concentrations, can create major handling and disposal problems. Two basic approaches have been used to remove tars from product gas streams: (1) physical removal technologies similar to those used for particulate removal, such as wet scrubbers, ESP; and 2) catalytic and thermal tar-reduction methods where tars are converted to permanent gases. The catalytic approaches can potentially destroy tars in either vaporized or condensed state. The second approach is discussed below.

10.5.2.1 Thermal Cracking

By increasing the gasifier temperature all organic compounds will crack to smaller hydrocarbons (~ 1,200 °C). Oxygen or air can be added to the gasifier to allow partial combustion of the tar to raise its temperature. Using electrical arc plasma for tar cracking is another option. It is a simple technique but it produces gas with lower energy content.

10.5.2.2 Catalytic Cracking

This technique can be applied in the gasifier or in a secondary reactor. The gasifier is commercially used in many plants for the removal of undesired elements from product gas. This method is explained in more detail in Sect. 4. For most syngas applications, the optimization of operating conditions, catalytic gasifier material or additives combined with secondary hot gas cleaning (catalytic cracking) are the most preferred methods.

Independent components models incorporate the notion of ‘pseudo-component’. Fig­ure 11.2d gives a typical example of pyrolysis with multiple decomposition levels. The combination of ‘pure’ components decomposing in the same system at differ­ent temperatures would be equivalent to the reaction pattern in Fig. 11.2b. As the temperature of the biomass particles increases, the stability of the solid macromolec­ular matrices changes accordingly. This variation is mostly due to the release of low molecular weight compounds as previously explained. The reactivity of the particu­late material will thus vary during pyrolysis and this behaviour is similar to having different components decomposing with specific reaction kinetics.

The pyrolysis of wood, for example, can be described by the combination of hemicellulose, cellulose and lignin pyrolysis kinetics, which all show single or two steps decomposition [37]. In this case, the combination of the decomposition steps for each of the three individual components can accurately model the overall pyrolysis behaviour of wood. However, ideal cases like wood are rare and the concept of ‘pseudo-component’ may not be useful for other biomass such as manure and MSW, for example, which are highly heterogeneous. One major disadvantage of this second

Temperature/ C

type of models is its complexity and the large number of reaction parameters involved. With materials having a variable composition from one provider to another, the value of these parameters will change as well as the weighting of each ‘pseudo-component’ species.

AC plasma torches are more promising for industrial applications and technologies requiring relatively high powers (e. g., waste treatment). Figure 12.3a represents the typical schematic diagram of a power-supply circuit. The power-supply system in this case is essentially cheaper and more reliable, rather than power-supply systems of DC plasma torches. Their maintenance is easier. The thermal efficiency of AC

plasma torches is high. In AC power-supply systems, the losses do not exceed several percent as the reactive ballasts stabilizing an arc are used. Reactive power losses are minimized using standard capacitor compensators.

Existing AC plasma torches can be divided into three basic groups: the single­phase [72, 73], the three-phase single-chamber, and the three-phase multi-chamber plasma torches. In some cases, a DC plasma torch with linear circuit and cylindrical electrodes is taken as a basis for single-phase AC plasma torches, thus AC power source is used. Working gas is usually supplied tangentially into such system. Another option of a single-phase plasma torch is a construction with the central rod electrode and ring or toroidal electrode. Usually, the arc is stabilized by the magnetic field rotating the arc in the interelectrode gap. Such type plasma torches use axial supply of the working gas.

Multi-phase multi-chamber AC plasma torches comprise various combinations of several single-phase plasma torches using multi-phase AC electrical grid. There are constructions consisting of three separate single-phase plasma torches. It is possible to connect three single-phase plasma torches sharing one mixing chamber, while the connection configuration can be different. There are systems designed by this principle, for example, a multi-phase electric arc heating system (see Fig. 12.3b) described in [73].

Multi-phase single-chamber AC plasma torches are described in [70]. Their fea­ture is installation of an electrode system of the plasma torch in a single-electric arc chamber. The electrode systems of multiphase single-chamber plasma torches can have the form of rings, toruses, or rods. In case of using toroidal or ring electrodes (Fig. 12.3c), generally the first and the last electrode are connected to the same phase. Electrodes are usually separated from each other by heat-resistant insulating pads. Stabilization and arc twirl are supported either by a magnetic field by means of the solenoid mounted on the plasma torch case, or by the creation of the tangential gas vortex setting the arc column on an axis of the electric arc chamber and moving the attachment points of the arc along the electrode surface.

Another widely used group of single-chamber multiphase plasma torches are plasma torches with rod electrodes. Several types of plasma torches, including the ones with rod electrodes have been developed, produced, and tested in IEE RAS. Research and development, and design experience are described in [67, 80, 90]. Figure 12.4a shows a single-chamber plasma torch with rod electrodes.

Ignition of several simultaneously burning AC arcs in a single chamber allowed creation of simple and reliable plasma torches transforming the electric current energy into plasma energy with high efficiency of 80-90 %.

Several different designs were developed. Tungsten or tungsten-containing rod electrodes were used for operation on inert gases, nitrogen, and hydrogen. Water — cooled copper tubular electrodes were used for operation on oxidizing media. Plasma torches with rod electrodes can be divided into three groups: with power up to 200 kW and 2 MW, and also working at short-time modes with power 4-50 MW [90]. Both types have the similar design comprising three main parts: case, arc chamber (nozzle), and electrode unit. Multi-phase mode of arc burning in the discharge chamber allows using low voltage of re-ignition due to the preliminary ionization of the discharge gap. Tungsten with additives of rear earth metals and compounds having low work function were used as an electrode material. The advantages of single-phase plasma torches with rod electrodes are: simple design, high efficiency providing by optimal relation of volume and surface area of the arc chamber, and also the possibility of electrode operation in the thermo-emission mode. In these systems, it is easier to stabilize burning of AC arcs.

A series of plasma torches with rail electrodes have been developed (Fig. 12.4b, 12.4c) [81]. A plasma torch with rail electrodes can provide stable operation with oxidizing (air) and neutral media (nitrogen, inert gases). The range of air flow rates varies from 15 to 70 g/s. Power input into the arcs varies from 100 to 700 kW. Thermal efficiency almost does not differ from the system (plasma torch and power supply) efficiency and is 70-95 %.

The basic principle of plasma torch operation is the rail-gun effect (arcs move along the electrodes in the field of their own current). The movement of arc attachment point along the electrode allows uniform distribution of thermal load, which gives the opportunity to use the water-cooling electrodes made of a fusible material with high thermal conductivity (copper tubes). The multi-phase single-chamber AC plasma

torch with rail electrodes uses an integrated single-phase high-voltage plasma torch of low power as an injector. It creates a plasma stream providing sufficient electron concentration in a zone of the minimal interelectrode gap for ignition of the basic arcs.

It allows stable ignition of arcs between the electrodes mounted with a gap up to 20 mm powering from the industrial grid with voltage about 380-500 V. Arcs fill the major part of the discharge chamber, moving in the longitudinal and transverse directions. The insulating layer is formed near the wall where cold gas moves, where concentration of charged particles dramatically decreases, and arcs extinguish. The above-described process repeats continually forming a low-temperature plasma jet with average mass temperature of about 1,500-6,500 K at the plasma torch nozzle.

High-voltage plasma torches with high thermal efficiency 80-95 % have been developed for operation with power up to 100 kW. These plasma torches have rod electrodes in cylindrical channels. Figure 12.5a represents their general view and design. High supply voltage 4-10 kV provides stable ignition and burning of the long arc.

Currently, AC plasma torches with cylindrical electrodes operating with high arc voltage drop (up to 5 kV) are the most promising ones. Figure 12.5b shows a plasma torch of this type.

The plasma torch except high efficiency has the following advantages: long life­time of electrodes (more than 1,000 h) and possibility to change the plasma heat content over a wide range, providing at this a range of average mass temperature for air plasma from 1,500 to 7,500 K. Of special note is the ability to provide plasma temperature less than 2,000 K, which is claimed for some technological processes. Moreover, we have developed 6 MW AC plasma torches, but their electrode lifetime did not exceed 100 h [90].

Solid acids are widely used for the study of biomass hydrolysis that will be introduced in Sect. 15.3 in detail. The hydrolysis mechanistic route of a water-soluble polysac­charide consists of the following steps: (1) The soluble polysaccharide diffuses onto the surface acidic sites of a solid acid; (2) в-1,4 glycosidic bonds access to the acidic sites; and (3) These bonds cleave randomly, and the polysaccharide is hydrolyzed to mono-sugars. Therefore, ideally, cellulosic materials need to be dissolved in a solvent and converted into short-sugar chains to make full use of the acid sites.

Liquid acid has some important limitations including the corrosion of reactor, recycle, and neutralization of the acid for microbial fermentation. Formation of degraded products and release of fermentation inhibitors are other characteristics of liquid acid pretreatment.

Unlike hydrolysis, in the pretreatment process, few mono-sugars are formed. Therefore, it performs under milder conditions, such as lower temperature, short processing time, and weaker acidity that can be selectively controlled. After pretreat­ment, biomass can be further hydrolyzed by either enzymatic or chemical method. Solid acid catalysts mediated pretreatment is superior in the destruction of biomass structure as compared with traditional liquid acid pretreatment because it achieves low reaction rate, and fewer side-reactions occur. After pre-hydrolysis of cellulose, the formed oligosaccharides can extend into the vicinity of Brpnsted or Lewis acid sites where catalytic hydrolysis occurs. Pre-converting cellulose into oligomers is important to the subsequent hydrolysis with solid acid catalysts.

Solid acid pretreatment is a novel method for biomass pretreatment. Although little work is reported, it can be considered as the initial stage or lesser degree of hydrolysis, and is easily controlled by changing the parameters of solid acid hydrolysis conditions. With the development and application of solid acids in biomass hydrolysis, they will find applications in pretreatment. So, hydrolysis is focused on the following sections.

Inherent properties of SB/SL cell wall make them resistant to cellulase mediated act for releasing monomeric sugars from hemicelluloses or cellulose. Pretreatment of SB/SL increase the accessible surface area and crystallinity of holocellulose which ameliorates the enzymatic action, in turn yielding optimum sugars recovery [14]. Table 16.4 summarizes the enzymatic hydrolysis profile of SB/SL after various kinds of pretreatment methods employed.

The hydrolysis rate of SB/SL directly depends upon the efficiency of pretreatment method used. Removal of lignin from the substrate determines the enzyme accessi­bility to the carbohydrate fraction of cell wall. A direct correlation exists between the removal of lignin and hemicelluloses on cellulose saccharification [11, 17]. Dilute H2SO4 pretreated SB followed by NAOH pretreatment (1 % NaOH, 60 min, 120 °C) produced 35g/L sugars (86.2 % hydrolytic efficiency) after 96 h of enzymatic hy­drolysis (10 FPU/g; 15 beta-glucosidase IU/g) [70]. Scanning electron microscopic (SEM) analysis also reveals the effect of dilute-acid pretreatment on native SB, alka­line pretreatment on cellulignin followed by enzymatic hydrolysis of cellulose. The un-homogeneity in structure and the disruption of first hemicelluloses followed by lignin and cellulose is clearly evident in SEM analysis (Fig. 16.3).

In the other study, oxalic acid pretreated SB (160 °C, 16 min, 3.5 % w/v Oxalic acid) produced total reducing sugars 56.3 g/g bagasse (92.30 % hydrolytic efficiency) after 120h of enzymatic hydrolysis (20 FPU/g; 25 beta glycosidase (IU/g)) [71]. Among alkaline pretreatments, hydrated ammonia based pretreatment methods have found great interest recently [8]. Hydrated ammonia precisely act on lignin removal from the SB leaving hemicelluloses and cellulose together but in fragile form which yields appreciable sugar recovery upon enzymatic hydrolysis [8]. In association, our laboratory reveals the maximum sugars recovery (28 g/L) after 96 h of enzymatic hydrolysis of ammonia pretreated SB (20 % ammonia, 24 h, 70 °C). The optimum enzyme loadings (15 FPU/g and 17.5 beta glucosidase IU/g) were used (Chandel et al. Unpublished work).

Other important factors such as substrate concentration, cellulase loading, and end-product inhibition also plays an important role for the hydrolytic efficiency of SB/SL [9, 50]. In order to enhance the surface area of holocellulosic fraction in the cell wall, surfactants like Tween-20 have been used [44, 23]. To overcome the problems of end-product inhibition and process complexities, integrated process configurations such as SSF and consolidated bioprocessing (CBP) have been found successful [14, 50].

Effect of pretreatment (organosolv extraction, ultrasonic treatment, and NaOH treat­ment) on the microstructure of pine sawdust was analyzed using SEM. Comparison of the changes in microstructure of pine sawdust before and after the organosolv extraction pretreatment is shown in Fig. 19.1a-d. The SEM pictures clearly show that the microstructure of pine sawdust changed after pretreatments. Compared with the nicely presented cell wall structure in the raw pine sawdust, the pretreatments led to a significant swelling of the raw pine sawdust and caused disorder in the cell structure. After pretreatment, the cell wall seems to be destroyed. The original cell wall was twisted and disordered, suggesting that significant morphological changes occurred in the process. Similarly, the raw pine sawdust sample exhibited rigid and highly ordered fibrils. The fibers of pine sawdust samples after pretreatment were distorted and twisted. The microfibrils were also isolated from the initial connected structure and made fully exposable, thus increasing the external surface area and the porosity of the biomass structure.

From the XRD measurement of pine sawdust before and after the various types of pretreatments, there are three peaks occurring at 26 of 14.6 °, 16.5 °, and 22.4 ° with varying intensities. These peaks correspond to cellulose XRD, corresponding to cellulose’s crystalline planes of (1Ї0), (110) and (220), respectively [52]. Compared with the un-treated raw material, the cellulose peaks all increase considerably in the pretreated pine sawdust samples. This demonstrates that the concentration of cellulose in the residues after pretreatment increases owing to the removal of hemi — cellulose and lignin in the pretreatment processes. The intensity of the cellulose peaks in sawdust samples increases with an increase in the PE and DE, which is expected as the cellulose content in the pretreated sawdust samples increased with an increase in the PE and DE. The CrI calculated from the Eq. 19.5 is 43 %, 63 %, 55 %, 64 %, and 65 %, respectively, for the untreated sawdust sample and the pretreated samples with the organosolv extraction, organosolv + NaOH, organosolv + ultrasound and organosolv + NaOH + ultrasound. As a common observation from chemical or

organosolv pretreatment of biomass [32, 49], crystallinity of the pretreated samples increases due to the removal of hemicelluloses and lignin. However, it should be noted that an increase in crystallinity of pretreated material was not expected to negatively affect the product yield during enzymatic hydrolysis of biomass [32,49].

The functional groups of the raw and pretreated sawdust samples, as well as pure cellulose and pure lignin were characterized using the FTIR technique (while the IR spectra are not included in this chapter). The FTIR results show that the pretreated sawdust samples are very similar to the peak patterns of the pure cellulose in the whole wavenumber range, simply because the cellulose content is high in pretreated sawdust samples when compared to the raw pine sawdust. The absorption bands at 1,275 cm-1 and 1,516-1,700 cm-1 correspond to the functional groups of lignin. The absorption band at 1,275 cm-1 is related to vibrations of guaiacyl rings and the absorption bands at 1,516-1,700 cm-1 ascribe to aromatic ring vibrations. The pure lignin and raw pine sawdust materials possess stronger signal at these absorption band, due to the higher content of lignin. The pretreated pine sawdust had a weaker signal at these absorption bands, because the majority of lignin was removed by the pretreatment. The absorption band at 1,711 cm-1 is related to carbonyl absorption in hemicelluloses. The raw and pretreated pine sawdust samples all showed absorption at 1,711 cm-1, implying that the pretreatment were unable to remove all hemicellu — lose in the raw materials, as also reported by Sreenivasan et al. [53]. In summary, the FTIR measurement indicated that the pretreatment was more efficient for removing of lignin than hemicellulose and cellulose from the pine sawdust.

Work by Kazi et al. [22] had estimated the cost of ethanol production from ligncel — lulosic biomass at US $1.14/kg (US $0.90/L), and this was dependent on feedstock and enzyme costs (assumed enzyme price as US $0.23/kg of ethanol produced). Due to the poor returns for ethanol, alternate uses of sugars for bio-products offer higher value propositions and some are listed in Table 9.4. For example, succinic acid mar­ket price is at US $5.9-8.8/kg [23] and can be used in a variety of food products and as a building block in polymers. Another valued bioproduct, PHA, has a market price at US $4.4-6.1/kg [24]. PHA can be produced from low value sugars and organic acids [1, 4] derived from woody-biomass feedstocks, rather than current practices of using refined sugar as a carbon source, which offers significant financial advances in reducing PHA manufacturing costs by 50 % [1]. Future work will investigate the use of these hydrolysates for the manufacture of PHA using mixed microbial consortia [6].

9.2 Conclusions

A simple hot-water pretreatment on hybrid poplar was achieved and optimized by using a response surface methodology on a 23 central composite design. The opti­mized pretreatment condition (temperature 200 °C, time 18 min, solid loading 20 %) was used in further experiments. A subsequent cellulase/xylanase hydrolysis step resulted in yielding high-level of sugars. Thus, pretreatment was shown to be an important step for cost effective enzymatic hydrolysis of wood with low levels of inhibitory side-products. Future work will focus on enzyme loading, conversion to targeted bioproducts (PHA), energy balance, and cost.

Acknowledgments The project was supported by USDA—National Institute of Food and Agriculture—Wood Utilization Research program grant 2010-34158-20938. The enzymes CTec2 and HTec2 were kindly provided by Dr. Kurt Creamer at Novozymes North America Inc., Franklinton, NC, USA.

Gasifier performance could be expressed in terms of gasification efficiency and product quality, such as heating value and the amount of desired product gas. Math­ematical modeling and simulation of the gasification process provides qualitative information on the gasifier performance, such as the effect of feedstock, design pa­rameters, and operating conditions of the process even though it is not very accurate. A mathematical model or simulation is, however, useless if it cannot reproduce exper­imental data with acceptable deviation [14]. Gasifier models could be classified into the following groups: thermodynamic equilibrium, kinetic, artificial neural network and computational fluid dynamics (CFD).

Biomass particles are generally difficult to fluidize, such that a denser and more homogeneous inert particle media (generally sand) is employed as a fluidization media to improve transport phenomena. Bench and pilot-scale continuous operation of a pyrolysis BFB has been demonstrated by Dynamotive. Char removal from the reactor can be an issue: if char is very fragile, its particle size will decrease within the bed by attrition. When char particle size reaches a critical particle value, it is entrained out of the fluidized bed reactor and it must be separated from the gas and recovered via a cyclone. Therefore, the disengagement region must be carefully designed to allow char particles to exit the reactor once they are sufficiently small. The main advantages of BFBs for pyrolysis applications include a uniform reaction temperature (minimizes the formation of cold/hot spots in the bed) and capability to operate the reactor continuously (continuous biomass feeding). On the other hand,

Fig. 11.3 Pyrolysis units global schemes: a Bubbling fluidized bed [41], b circulating fluidized bed [41] and c rotary drum reactor [51]. (Reprinted from Ref. [41], Copyright 2012, with permission from Elsevier; Reprinted from Ref. [51], with permission from Prof. Dr-Ing. Roman Weber) the main disadvantage of BFBs is that the volatile will be mixed with the inert fluidizing gas. Therefore, the bio-oil can be recovered, but the non-condensable gas is diluted such that it can hardly be used as a primary energy source. Thus, energy must be obtained from the solid char, which is significantly detrimental to the process profitability.