Physical pre-treatments are the key to control feedstock properties, which sig­nificantly influence gas/solid hydrodynamics as well as heat/mass transfer in the pyrolysis reactor. Recommended feedstock pre-treatments depend on the initial biomass characteristics, the pyrolysis conditions as well as the reactor type. Pre­treatments also allow the homogenization of the feedstock characteristics with time.

Intrinsic feedstock properties such as specific heat, thermal conductivity and density (dry and true) cannot be easily modified and constitute limitations for ther­mal processes. On the other hand, feedstock moisture and particle size are the main physical parameters that can be adjusted to optimize the pyrolysis process performance.

Particle fluidizability has been correlated to its average size and density. Geldart classified particles into four groups (Geldart classes A, B, C and D) based on their fluidization behaviour at ambient conditions [27]. Figure 11.1 illustrates the Geldart classification of powders and indicates the properties of common feedstocks for biorefineries.

In the Geldart classification, class C (cohesiveparticles) and D (large particles) can be detrimental to gas/solid mixing as well as heat/mass transfer. For example, these particles cannot be easily and uniformly fluidized in a reactor: class C particles lead to channelling [28]. Furthermore, large particles (class D) are more subject to internal temperature gradients and species diffusion effects, which affect the final pyrolysis product distribution and composition. Species intra-particle diffusion increase the species exposure time to the pyrolysis conditions (additional time for reactions) while temperature gradients lead to uncertainties related to the characterization of the pyrolysis conditions.

Based on typical wood densities (various species) and Fig. 11.1 sawdust particles (50-500 ^m) are classified as Geldart A. On the other hand, coarser bark or wood residue particles are classified as Geldart B or Geldart D.

Determination of the most promising types of biomass for the plasma gasification is a big challenge. For its solution it is necessary to carry out researches in abso­lutely different areas: gasification technologies (taking into account special features of plasma usage), wood industry, agriculture, and also other industries producing biomass as end — or by-product. Therefore, in this work the problem of prospectiv­ity is considered not among the vast diversity of biomass types, but among several representatives chosen for the consideration.

Let us discuss plasma gasification features versus autothermal one. Currently gasification utilizes only thermal plasma. Correspondingly, the plasma influence on the process is defined by its contribution to energy balance (the higher share of plasma energy is the grater difference it makes for the process). In the plasma injection zone of a reactor the temperatures grows significantly in comparison with the similar autothermal gasifiers that results in increase of chemical reaction rates. Content and yield of valuable products of gasification (hydrogen and carbon monoxide) increase due to additional energy injected with plasma. Moreover, high efficiency of energy introduction makes it possible to gasify feedstock by pure steam and carbon dioxide.

The gasifier type and the duration of main stages of gasification determine the syngas composition. The gasifier type defines the direction of mass stream of the process. Plasma is used in two general types of gasifiers: counter-current fixed bed (updraft), co-current fixed bed (downdraft). In the former, the gas stream (oxidant and gasification products) is directed upward, whereas in the latter it is downward, but in both cases the feedstock stream is directed downward. Solid gasification products mainly consist of inorganic components and are discharged from the bottom of the reactor.

In the updraft process, the organic matter during gasification is first devolatilized and then oxidized. Pyrolysis products (volatiles) are formed and leave the reactor

volume at rather cold zones where tars, water, and some permanent gases are not con­verted to syngas. Sometimes in such cases the separate unit for tar plasma conversion is used. Plasma usage in the updraft gasifier allows liquid slag discharge.

In the downdraft process, the organic matter during gasification is also first de­volatilized and then oxidized. However, in this case the pyrolysis products pass through the high-temperature oxidizing zone, where they are converted by plasma. It results in considerable reduction of tar concentration of syngas and allows using plasma energy to increase H2 and CO yield and content.

There are parameters permitting preliminary estimation of the prospects and suitability of raw materials for plasma gasification (see Table 12.1).

Proximate analysis data show that organic matter of chicken manure consists of more than 90 % (moisture + volatile matter) from substances turning into gaseous phase during the devolatilization stage, for other types of biomass this value is more than 80 %. In general, more the content of volatile matter, easier and more effective is its use in the downdraft plasma gasification [31]. In the updraft process, it becomes a disadvantage because volatile matter is not exposed to high-energy plasma flow and the produced syngas is heavily contaminated by tars [32]. Conversion of these tars is an important line of plasma usage development. For example, Europlasma is developing a reactor module comprising an autothermal gasifier (similar to updraft one on organization of mass flows) and plasma system for tar conversion, thus the most complete conversion occurs at energy consumption of ~ 1.8 MW, while syngas yield possesses ~ 10.2 MW of chemical energy [33]. Non-equilibrium plasma is used for conversion of synthesis gas with very low tar content (0.7-1.9g/Nm3). In order to decrease tar concentration by ~20 % it is required that 27-39 % of electricity is produced by a gasifier [34]. The content of tars in the downdraft plasma pro­cess is already significantly reduced because they pass through the high-temperature oxidizing zone and almost does not affect the energy balance [35].

Feedstock’s fixed carbon/ash mass ratio is important for downdraft process since feedstock and oxidizer flows are co-current; the gasification rate decreases dramati­cally downstream. Concentration of an oxidizing component in gas phase and carbon in char-ash residue decrease, and multiplication of these concentrations determines the mass exchange rate. Thus, fuels with high carbon content in a char-ash residue are preferred for downdraft plasma and autothermal gasification processes. Wood waste (~93 % carbon in the char-ash residue) and switchgrass (~64 %) are the best raw materials according to this characteristic.

In the updraft process feedstock and oxidizer flow in counter-current configu­ration. The plasma flow at the inlet contacts with the extremely carbon depleted char-ash residue. This method versus the downdraft gasification allows achievement of higher carbon conversion level of char-ash residue. Moreover, position of the high temperature zone at the bottom of the reactor simplifies liquid slag removal and slag vitrification [36].

Data on proximate and ultimate analysis allow determination of oxygen con­sumption required for complete gasification of carbon and for complete combustion of feedstock. These values should be considered simultaneously with the heating value. Jointly they determine adiabatic combustion temperature having higher im­pact on practical value of a feedstock for energy industry than heating value. The fuel can have a significant heating value, but the higher amount of oxygen is required for its combustion, the larger amount of energy is spent for heating of neutral nitro­gen at operation on air. The decrease of adiabatic combustion temperature is mainly caused by three factors: excessive moisture, high content of oxygen in a feedstock, and in a less degree, high ash content. Only moisture can be easily altered, its re­moval leads to a decrease of oxygen and hydrogen content in the feedstock. Updraft plasma gasification process compared to downdraft one allows decreasing of energy consumption for fuels with high ash and moisture if their removal is impossible or leads to a decline in economic viability of the whole process. However, usage of such types of feedstocks for energy needs usually is not rational.

One of the key parameters for all plasma processes is a relationship of energy and oxygen consumption for stoichiometric carbon gasification conditions. However, it is impossible to determine energy consumption without calculation of the gasification process. At the analysis initial stages we could examine ratios of chemical energy yield to a heating value and oxygen consumptions for combustion and gasification. The higher these values are the harder optimum parameters of plasma gasification are achieved. The most challenging feedstock in the given context is cattle manure (ratios: ~1.31 and to, respectively), and the simplest switchgrass (~ 1.09 and ~9.07).

It should be noted that estimations of chemical energy yield limit are correct if the downdraft plasma gasification or any other method with complete tar conversion is used. It is impossible to determine clearly the most promising feedstock by this value, because it is necessary to spend energy and funds for feedstock acquisition and to take into account transportation expenses. Both these parameters depend on used methods and technologies of feedstock collection and transportation. Assuming that thermal energy of gasification products in all cases amounts to 3 MJ in energy balance on 1 kg (this rough approximation is admissible for stoichiometric gasification of many types of feedstocks with LHV less than ~ 15 MJ/kg), synthesis gas will be used in the combined cycle with efficiency of ~60%[ 37] and electricity cost amounts to 5c/kW h, we define that treatment of chicken manure will be the most profitable. According to the energy balance (per 1 kg), energy consumption will be ~1.38 kW h (~6.9 c), electricity yield ~2.3 kW h (~11.7 c), and income from treatment of manure ~1.3 c. In this case, the income from treatment of 1 kg of feedstock will be ~6.1 c.

14.3.1 Introduction

OES, short for organic electrolyte solution, is defined as non-aqueous or aprotic solvent solution of electrolytes, which usually contains fraction of polar organic solvents and free ions.

Organic solvents mixed with certain lignocellulose-dissolving solutions, such as room temperature ILs [83-86] and concentrated phosphoric acid [87], form new types of homogeneous solutions that can be used to (1) swell and (or) dis­solve the component of cellulose and (or) lignin, as a pure lignocellulose-dissolving solution does, and (2) precipitate the cellulose and lignin selectively. These combina­tions include dimethyl sulfoxide (DMSO) + paraformaldehyde [88, 89], DMSO + [BMIM]Cl [90], DMSO + tetra-n-butylammonium fluoride (TBAF) [91, 92], N, N — dimethylformamide (DMAC) + 1-allyl-3-methylimidazolium chloride ([AMIM]Cl) [93], (acetone, pyridine, or hexamethylphosphoramide) + ILs [94], 3-dimethyl-2- imidazolidinone + 1-ethyl-3-methylimidazolium acetate ([EMIM]AcO) or 1-butyl-

3- methylimidazolium chloride ([BMIM]Cl) [95], DMAC + LiCl [96], and H3PO4 + acetone [97], etc. The addition of common lignocellulose-dissolving solution in or­ganic solvents allows OES to have the equivalent abilities to destroy the crystallinity of natural cellulose (acting as a cellulose solvent) [42]. Meanwhile, with the existence of organic solvents, some OES can remove the lignin component from the lignocel — lulosic homogeneous solutions by precipitating cellulose and hemicellulose (acting as a precipitating agent) [97]. In terms of the selectivity of dissolving abilities, OES — based pretreatment processes can be divided into two categories: OES-dissolving pretreatment and OES-depositing pretreatment.

Steam explosion can be considered as one of the most promising techniques of fractionation of SB/SL. In the process, hydrolysis of the hemicellulose is accelerated by the contact of the biomass through steam at high temperature (160-240 °C) and high pressure (between 0.7 and 4.8 MPa) followed by quick decompression [16], which leads to breakdown of biomass [33]. Singh et al. [34] reported the steam explosion of SB which eventually showed the enzymatic hydrolysis efficiency of 100 % after 24 h of incubation by using the cellulases from Penicillium sp. Steam explosion disrupts the compactness of Abrils, thus increasing the surface area for better enzymatic action [16].

Sendelius [35] investigated steam explosion in SB at varying temperatures (180, 190, and 205 °C) for different time periods (5 and 10 min) using different impreg­nating agents (water, 2 % SO2, and 0.25 g sulfuric acid (H2SO4) per 100 g dry matter). These optimization studies showed 80 % theoretical hydrolysis yield at SO2-impregnation-180 °C-5min conditions. One of the major advantages of the steam explosion process is the rare or no use of chemicals, eventually reducing the operational costs and minimum production of potential inhibitors [17].

16.3.2.3 CO2 Explosion

Carbon dioxide (CO2) explosion is based on the utilization of CO2 as supercritical fluid leading to lignin removal. CO2 molecules at high pressure are able to penetrate small pores of SB disrupting the chemical nature of the substrate [36, 37]. Carbonic acid is formed in aqueous CO2 solution penetrating into the small pores of lignocel — lulosic material under high pressure. This method is considered environment friendly due to minimum by-products generation [14]. CO2 in presence of water forms car­bonic acid which allows the depolymerization of holocellulose eventually increasing the surface area of substrate. Zheng et al. [38] pretreated SB using CO2 as super­critical fluid which showed an increase in glucose yield by 50 % after enzymatic hydrolysis.

Chunbao (Charles) Xu, Liao Baoqiang and Wei Shi

Abstract This chapter presents some recent research results on pretreatment of pine sawdust for bio-ethanol production, using organosolv extraction-based methods, combined with other methods including ultrasonic treatment, and sodium hydroxide treatment, and enzymatic hydrolysis of the pretreated pine sawdust samples. The pre­treatment efficiency (PE) and delignification efficiency (DE) of various pretreatment methods were studied. All the pretreatment methods, in particular the organosolv extraction, resulted in significant removal of lignin and hemicellulose. The results indicated that the combination of three pretreatment methods (organosolv extraction + ultrasound + NaOH) achieved the best PE (61.6 % + 1 %) and DE (86.4 % + 3 %). Enzymatic hydrolysis of pine samples treated with different pretreatment methods was comparatively studied. Glucose yields, total sugar yields, and total weight loss were obtained under various enzyme loadings (0~15.6 FPU cellulase) and reaction times (up to 48 h). The maximum glucose yield and the maximum total sugar yield were 5.8 % and 7.1 %, respectively, for un-pretreated raw pine samples, compared with 19.3 % and 22.40 % for the (organosolv extracted + ultrasound + NaOH) treated samples.

Keywords Softwood ■ Jack pine ■ Pretreatment ■ Organosolv extraction ■ Ultrasound Sodium hydroxide ■ Pretreatment efficiency ■ Delignification efficiency ■ Enzymatic hydrolysis ■ Glucose

19.1 Introduction

Fossil fuels, mainly coal, petroleum and natural gas, account to more than 80 % of the primary energy consumption in the world. The burning of fossil fuels emits around 21.3 billion tonnes of greenhouse gases (GHGs) annually. As such, it is

C. Xu (H)

Department of Chemical and Biochemical Engineering, Western University, London, ON,

N6A 5B9, Canada e-mail: cxu6@uwo. ca

B. Liao ■ W. Shi

Department of Chemical Engineering, Lakehead University, Thunder Bay,

Ontario P7B 5E1, Canada

Z. Fang (ed.), Pretreatment Techniques for Biofuels and Biorefineries, 435

Green Energy and Technology,

DOI 10.1007/978-3-642-32735-3_19, © Springer-Verlag Berlin Heidelberg 2013

strategically pivotal to pursue alternative and renewable energy sources due to the rapidly increasing demand for energy, the depleted fossil resources and the growing concerns over climate changes and energy security. The pursuit of bio-ethanol as an alternative energy source has attracted increasing interest recently.

As a typical bio-fuel, bio-ethanol is considered carbon neutral since the carbon dioxide released from combustion of ethanol produced from renewable lignocellu — losic materials is the CO2 sequestered by the plants during their growth. Bio-ethanol can be used in various ways for energy and chemicals, while more commonly as a blended fuel in gasoline. Nowadays, all gasoline engines can use up to 10 wt% ethanol blended fuel without any need of engine modification [1,2]. However, the main challenges for commercialization of the bio-ethanol technologies (particularly for the cellulosic ethanol) may be the combination of the low cost of conventional energy resources and the high biomass processing cost [1]. Compared with the con­ventional starch-based bio-ethanol manufacture, production of cellulosic ethanol using non-food lignocellulosic feedstock is advantageous as it does not compete with the food industry for feedstocks. Typical feedstock for cellulosic bio-ethanol production includes crop residues, grasses, forest biomass and waste, such as saw­dust and wood chips [1, 3]. Softwoods are the dominant wood species in North America. Softwoods, for example, pine and spruce, contain about 40-45 wt% cel­lulose, 20-25 wt% hemicelluloses, and 25-30 wt% lignin. Because of the structural characteristics of woody biomass (large polymeric molecules and high crystallinity) of the cell wall and the presence of hemicellulose and lignin, enzymatic hydrolysis of cellulose into glucose for bio-ethanol production is challenging due to its low ac­cessibility to enzymes. As a result, pretreatment of lignocellulosic biomass to loosen the cellulose crystalline structures and increase its accessibility is necessary, impor­tant, and the key in the bioconversion process. For instance, Zhu and his team in US Forest Service of Forest Products Laboratory reported a very high enzymatic hy­drolysis efficiency for softwood species (e. g., spruce, red pine) with combined dilute acid and sulfite pretreatment, or hot water, dilute acid or sulfite pretreatment, com­bined with disk-milling pretreatment [4-6]. With sulfite pretreatment using 8-10 % bisulfite and 1.8-3.7 % sulfuric acid on oven dry wood at 180 °C for 30 min, enzy­matic hydrolysis of spruce chips led to more than 90 % cellulose conversion at an enzyme loading of about 14.6 FPU cellulase plus 22.5 CBU в-glucosidase per gram of od substrate after 48-h hydrolysis [4]. Enzymatic hydrolysis of lignocellulosic materials into fermentable sugars can achieve a glucose yield from 10 % to 60 %, depending on the type of lignocellulosic materials [7]. There are a number of differ­ent pretreatment methods that can be generally classified into the following types: physical pretreatment, physical-chemical pretreatment, chemical pretreatment, and biological pretreatment, whose details will be overviewed as follows.

9.2.1 Raw Materials

Hybrid poplar (Potlatch Corp., ID, USA) was milled to <40 mesh using a Wiley mill (Thomas Scientific, NJ, USA), vacuum dried, and stored in sealed plastic bags (moisture content of 4.6 %). Chemical composition analysis was determined using procedures described in [15].

9.2.2 Hot-Water Pretreatment

Pretreatment was conducted in a 76 mL pressure reactor, (Model 4740, Parr Instru­ment Co., IL, USA) connected to a temperature controlled block heater built in-house. Wood meal (5.00 g) was introduced to the reactor to which water was added giving a solid loading range from 20.0 to 46.8 % and sealed. The reaction temperature ranged from 170 to210 ° C. A23 full factorial design for temperature, time, and solid loading was conducted (Table 9.1). An additional temperature probe was used for controlling the outside temperature of the reactor vessel. The reaction vessel took 5 and 10 min, respectively, to reach 170 and 200 °C. After pretreatment, the vessel was placed in an ice-water bath to quench the reaction. The pretreated samples then were washed with hot-water (200 mL, 90 °C) [2, 3] to extract out the sugars and acids generated and centrifuged (10,000 rpm) to separate the solid and liquid fractions. The liquid frac­tion was named pre-liquid (PL) and pH was measured. The solid residue collected was used for enzymatic hydrolysis trials.

This type of reactor consists of a cylindrical reactor with a fixed bed of solid fuel. The gasifying agent is injected upward or downward through the reactor. These simple reactors can operate at high carbon conversion, for a long solid residence time, at

low gas velocity and they are suitable for small scale applications, that is, < 10 MW [65]. Fixed bed gasifiers are classified into three groups according to the way the biomass and gasifying agent enter the reactor.

One of the most evident synergic pyrolysis behaviours has been demonstrated by Brebu et al. [16]. By co-feeding pine cones with waste polyolefin, they showed that it was possible to significantly decrease the overall char yield and increase the amount of volatile produced. For binary blends of pine cones and polyethylene (PE), polypropylene (PP) or polystyrene (PS), the char yield decreased (by over 6 % units for PE) compared to the calculated value by linear combination, hence revealing synergic behaviour. When a blend of pine cones and the three polymers (PE/PP/PS) was pyrolyzed in the ratio 3:4:2:1, the synergy was even more significant. The reported char yield was lower than the average calculated value by 10 % units and the liquid product yield increased by over 11 % units. However, while these types of synergies are desired, they have been rarely observed. One typical example is the co­pyrolysis of biomass with coal. Weiland et al. [46] explored the possible synergetic interactions between coal and biomass in pyrolysis. Unfortunately, the interaction was almost nonexistent. Linear combination explained most of the variations with the various blends of coal/biomass.

The basic advantage of low-temperature plasma usage in biomass gasification is substantial growth of the hydrogen content and carbon monoxide content in syngas composition. At treatment of a chicken manure, the content of H2 + CO in syngas can be raised to ~97 % that allows an increase in the efficiency of its use in the Fisher-Tropsh process by 15-59 % and reach a specific yield of synthetic fuels ~240-260 g/kg, thus power inputs on the organization of the allothermal process will make only 4-5MJ/kg. These parameters can be reached in downdraft plasma gasification. Plasma energy is used for liquid slag removal in updraft gasification, and it has rather less influence on syngas composition. The most effective way of plasma generation is by systems on a basis of AC plasma torches thanks to a long lifetime of electrodes (more than 1,000 h), an effective energy transfer of the discharge to the plasma forming gas (to 80-95 %), high power at work in long-time modes (to 2 MW), and low losses in a power-supply system (no more than ~ 1-5 %). New AC plasma torches have arc voltage drops of about 1-10 kV and discharge currents of about 10-100 A. However, DC plasma torches are used more often. They typically have high currents (0.1-1 kA) and low voltage drops in the discharge (10-1,000 V) which are necessary for DC arc stabilization. Their main advantages are long-time operating experience and hence well-developed plasma torch models. Reliability of the executed estimations proves to be true by a good coordination with the experimental data which are obtained on the large-scale plasma gasifier. Total difference between mass balances was less than 1 %, and for energy balance less than 2 %. A series of long-time experiments on plasma gasification of wood proves that plasma gasification of biomass with use of AC plasma torches is ready for industrial implementation.

The concentration and strength of acid sites in solid acid catalysts are usually measured by amine titration, infrared spectroscopy (IR), temperature-programmed desorption (TPD), solid-state nuclear magnetic resonance (NMR) spectroscopy and thermal analysis [46]. The traditional amine titration method can be used for quan­titative analysis of acid amount. However, it could not accurately analyze the types of acid and the distribution of acidic centers on the surface of solid acid catalysts. The analysis of total acid amount and strength of solid acid catalysts using TPD method is very accurate, except for distinguishing acid types. As a supplementary method, IR spectroscopy can be used well to distinguish B — and L acids but not suitable for exact quantitative analysis of acid amounts. Therefore, the combination of TPD method and IR spectroscopy is more accurate to get acidic characteristics of solid acid catalysts. Currently-developed solid-state NMR spectroscopy can realize qualitative and quantitative analysis of acidic center distribution, acid amount, and acid strength, but the analysis is complex and costly.

The acidic strength and acid site density are determined by the preparation process conditions, such as calcination temperature, crystal form of oxides, and sulfonation process. The main functions of calcination are as follows: (1) change amorphous oxides into crystals; and (2) promote solid phase reaction between sulfuric acid and metal oxides, forming acid structure of strong bonding of sulfuric acid with metal oxides. An appropriate calcining temperature is conducive to have good porous structure, many acid sites and high acid strength. Loss of sulfur species, decrease of specific surface and crystal transformation will occur if calcination temperature is too high. On the contrary, the ideal acid structure, crystal and pore structure will not be obtained if calcination temperature is too low. Pang et al. [29] found that the higher the sulfonation temperature, the higher the acid density of the sulfonated carbon. The active carbon sulfonated at 300 °C (AC-SO3H-300) showed an acid density of 2.19mmol/g, which was 15-fold that of the untreated active carbon. However, a higher temperature (>250 °C) caused the decomposition of sulfonate. The highest glucose yield of 74.5 % was achieved by AC-SO3H sulfonated at 250 °C, which possessed the highest density of -SO3H groups.

Mesoporous metal oxides with high catalytic efficiencies have the following spe­cial structure properties: (1) high specific surface area, (2) adjustable pore size, and (3) enhanced thermal stability [47]. However, it has been shown that highly crys­talline pore walls and high mesoporous order cannot always be achieved for the same material. High-temperature heat treatment helps to increase the crystallinity of pore walls, but leads to the collapse of mesoporous structure. Many reports on sulfated metal oxides showed that the amorphous metal oxides are needed for preparation of solid superacids except for y-Al2O3 [47].

During the preparation process of metal oxides, the pH should be controlled by adding precipitation agent to obtain the precipitates of metal hydroxides. Ammonia and urea are the most used precipitation agents. The precipitation process using am­monia is more slowly, thus improving crystal growth efficiency. B acid sites (proton donors) can be generated from highly polarized hydroxyl groups. They can also be formed on oxide-based catalysts via proton balance of a net negative charge intro­duced by substituting cations with a lower valence charge [48]. Sulfide type should be considered as the main factor in acid strength, acid type, and catalytic activ­ity. Commonly-used sulfides include H2S, SO2, SO3, CS2, H2SO4, (NH4)2SO4 and benzenesulfonic acid, but only high-valence sulfides perform acid action. In order to obtain enough strong superacid centers, the optimum calcination temperature should be slightly lower than the decomposition temperature of acid sites. Besides sulfated single metal oxides, manipulation of sulfate content and activation temperature also provides the means for controlling the strength of surface B acid sites in the sulfated mixed metal oxides [49].

Certainly, not only sulfides are used as active sites. Metal oxide supported Pt or Ru showed highly-active for converting cellulose to sugar alcohols with 31 % yield being reported [50]. They were active for catalyzing cellulose conversion and glucose isomerization, simultaneously with 88 % selectivity of glucose being obtained. It should be noted that they presented L acid catalysis.