Adsorption is another unit operation widely employed in the industry for etha­nol dehydration. In this operation, the ethanol-water mixture passes through an apparatus (usually cylindrical) that contains a bed with an adsorbent material. Due to the difference in the affinity of molecules of water and ethanol with respect to the adsorbent, the water remains entrapped in the bed while the ethyl alcohol passes though this same bed increasing its concentration in the stream leaving the apparatus.

In early work carried out in the 1980s, the utilization of different biomaterials as adsorbents for ethanol dehydration, such as corn, xylan, pure cellulose, corn sto­ver, corn flour, wheat straw, and sugarcane bagasse, was proposed. Such materials whose structure is based on polysaccharides demonstrated their ability to separate water and ethanol (Hong et al., 1982; Ladisch and Dyck, 1979; Ladisch et al., 1984; Westgate and Ladisch, 1993). In particular, the broken corn grains showed good properties to absorb water in aqueous solutions of ethanol. In pilot plant stud­ies, the possibility of concentrating ethanol from 91% to more than 99% using a bed of broken corn grains was demonstrated. The bed can be used during several adsorption-regeneration cycles (up to more than 30; Tanaka and Otten, 1987).

The bed can then be utilized as a starch source for ethanol production. In principle, the energy costs were lower than those of azeotropic distillation. More recently, some works have shown the great adsorption ability of different starchy materials that have adsorptive properties similar to those of the inorganic adsor­bents when the mixture to be separated contains about 10% water. In addition, the enzymatic modification using a-amylase contributes to enhance the adsorptive properties of starch (Beery and Ladisch, 2001a, b).

However, the adsorption of water employing the so-called molecular sieves to dehydrate ethanol has been the technology that has acquired more development in past years in the fuel ethanol industry. In fact, this technology has been replacing the azeotropic distillation. The molecular sieves are granular rigid materials with a spherical or cylindrical shape manufactured from potassium aluminosilicates. They are classified according to the nominal diameter of the large amount of internal pores that provide access to the interstitial free volume found in their microcrystalline structure (Madson and Monceaux, 1995). For ethanol dehydra­tion, sieves with an average diameter of the interstitial passageways of 3 ang­stroms (Type 3A sieves) are commonly used. The water molecule has a diameter lower than that of the interstitial passageways of this type of sieves, while the ethanol molecule does not. In addition, the water can be adsorbed onto the inter­nal surface of the passageways in the molecular sieve structure. Thus, ethanol molecules pass out of the apparatus without adsorbing onto the bed.

The adsorption operation requires that, once the adsorbent bed is saturated with the substance to be separated, the desorption of this substance should be accomplished to make possible the reutilization of the adsorbent material (regen­eration cycle). For regeneration of the sieves, hot gas is needed that rapidly dete­riorates them especially if the bed is fed with a liquid stream during the previous

water adsorption cycle. To counter this deterioration, the pressure-swing adsorp­tion (PSA) technology was developed. This technology involves the use of two adsorption beds. While one bed produces vapors of anhydrous ethanol super­heated under pressure, the other one is regenerated under vacuum conditions by recirculating a small portion of superheated ethanol vapors through the saturated sieves (Figure 8.7). The system feeding is carried out using the overhead vapors from the rectification column. The ethanolic vapors obtained in the regeneration cycle and that can contain 28% water are recirculated to the rectification column (Montoya et al., 2005; Wooley et al. 1999). In this way, the molecular sieves life can be prolonged for several years, which, in turn, leads to very low costs related to the replacement of adsorbent material and, therefore, reduced operating costs (Guan and Hu, 2003; Madson and Monceaux, 1995).

8.2.6 Pervaporation

Different applications of the membranes for both concentration of ethanol solu­tions and ethanol dehydration have been developed recently. Although the reverse osmosis was first proposed (Leeper and Tsao, 1987), the pervaporation definitively boosted the introduction of membranes into the fuel ethanol industry (Sanchez and Cardona, 2005). The pervaporation (evaporation through membranes) is an operation based on the separation of two components by a selective membrane under a pressure gradient in which the component passing across the membrane is removed as a gaseous stream (permeate), while the other component remains in the liquid phase and is removed as a more concentrated stream (retentate), as shown in Figure 8.8. This operation began to exhibit industrial feasibility when polyvinyl alcohol (PVA) composite membranes were developed at the end of 1980s. These membranes present a high selectivity by favoring the water passing across them

Anhydrous

ethanol Vacuum Pump

f^nden*:

Water

FIGURE 8.8 Schematic diagram of pervaporation for ethanol dehydration. Tin > Tout.

and a high retention of several organic solvents. Just the PVA membranes allow the change of the phase equilibrium of ethanol-water solutions in such a way that the mass transfer between the liquid and vapor phases is determined by the semi­permeable membrane and not by the free interphase (Sander and Soukup, 1988). Thus, it is possible to obtain a concentrated ethanol stream with a composition above the azeotropic. The driving force of pervaporation is maintained thanks to the application of a vacuum in the permeate side of the membrane. This driving force is evidenced by the difference of partial pressures or activities of the com­ponent passing across the membrane. This difference of partial pressures can be increased making the feed temperature as high as possible. The permeate is later condensed to generate a liquid stream that can be recycled back to the rectification column. Diverse chemical modifications of PVA membranes have been proposed to improve the hydrophilic/hydrophobic ratio in order to enhance the selectivity and flux of these membranes (Chiang and Chen, 1998).

The pervaporation offers a series of advantages compared to azeotropic or extractive distillations because the product contains no entrainer or solvent traces. In addition, this technology is easily adjustable and very flexible with respect to the changes of the feed concentration. The start-up and stop of the process require minimum labor and supervision. Finally, the pervaporation units are compact and need no large areas compared to the big towers of azeotropic distillation schemes (Sanchez and Cardona, 2005; Sander and Soukup, 1988). As this operation con­sumes less energy than the conventional operations based on distillation, the use
of pervaporation has become very promising considering the permanent develop­ment of membrane manufacturing technologies, which have allowed the increase of membrane selectivity and permeate flux.

Besides the flat membranes, hollow-fiber membranes have been manufactured to dehydrate ethyl alcohol. Tsuyomoto et al. (1991) produced this type of mem­brane by the partial hydrolysis of polyacrylonitrile in order to introduce carboxy­lic groups and obtain a polyionic complex. These membranes were successfully tested to concentrate 95% ethanol solutions at 60°C showing separation factors higher than 5,000. Currently, there exist about 100 plants worldwide employing ethanol and isopropanol dehydration (Baker, 2004), some of them located in facil­ities for ethanol production from corn. In practice, the chemical process industries use only plate-and-frame modules for pervaporation. These units are sealed with graphite compression gaskets that universally resist organic liquids. Spiral-wound and hollow-fiber modules require adhesives that are not resistant to all solvents (Wynn, 2001).

Other separation technology based on membranes has been proposed (the vapor permeation), which has been much less studied than pervaporation (Sanchez and Cardona, 2005). This process has the same principle of pervapo­ration, but the feed stream is also gaseous. Vapor permeation offers certain advantages over pervaporation (Jansen et al., 1992). First of all, the construc­tion of vapor permeation modules is simpler because evaporation heat is not required, i. e., there is no need to heat the fractions of the retentate feeding the subsequent modules. In practice, the flux of vapor permeation is higher than the pervaporation flux because the driving force is less affected by the polarization of concentration and by the reduction of partial pressures. Second, increasing the pressure or decreasing the superheating can significantly increase the flux. Moreover, the membrane can be impregnated with different substances (espe­cially salts) to enhance or regulate its selectivity and flux, which is favored by the fact that the membrane is not in contact with the liquid phase. For instance, the impregnation of composite PVA membranes with cesium fluoride (CsF) pro­voked almost a two-fold increase of the flux during the vapor permeation of 95% ethanol mixtures.

The stillage can be employed in composting processes as well. In particular, a mixture containing the organic fraction of the municipal solid waste (MSW) and stillage with a 3:1 ratio has been proposed to produce compost. In this case, the stillage supplies the nutrients needed for the process. The aerobic fermentation process inherent to the composting can be carried out during 30 to 50 days. Good results have been reported for this type of stillage treatment (Vaccari et al., 2005). In a similar way, the by-product from sugarcane cropping can be mixed with the stillage for their composting with a residence time of 35 days (Goyes and Bolanos, 2005). In Colombia, most of the sugar mills with co-located distilleries use the stillage, evaporated up to 30% solids, in mixtures with press mud for producing a solid fertilizer through composting for cane plantations (Gnecco, 2003). The use of concentrated stillage with 55% solids is also possible for its direct utilization as a fertilizer or for composting processes. The compost produced from the press mud, ash from boilers using cane bagasse, and concentrated stillage with 30% solids can supply a third of the nitrogen and much more phosphorous and potas­sium required by the sugarcane. In addition, the composting reduces by 50% the volume of the used materials since the oxidation reactions occurring during the aerobic solid-state fermentation are exothermal and, therefore, contribute to the vaporization of moisture contained in the stillage (Gnecco, 2003).

Nandy et al. (2002) describe the implementation of an effluent treatment pro­cess in an ethanol-producing plant integrated to a sugar mill in India. The plant produces 46,000 L/d of industrial rectified ethanol from the molasses generated in the same sugar mill, which processes 4,000 ton/d of sugarcane. The stillage undergoes anaerobic digestion in two fixed-film anaerobic reactors. The methane produced allows the reduction of the fossil fuel consumption for the generation of the steam required during the distillation step. Then, a 4-effect evaporation train is employed to concentrate the effluent from anaerobic digestion toward 50% solids. The evaporators employed are the down-flow film type. Finally, the efflu­ent of the evaporation step is used for composting the press mud generated in the sugar mill (140 ton/d). In this process of aerobic conversion through thermophilic microorganisms whose reactor has a residence time of 45 days, a biomanure that can be sold is obtained at a rate of 64 ton/d. In this way, the burdens are practi­cally eliminated. This plant represents an example of the integrated utilization of the bioresources generated and available in the joint production of cane sugar and ethanol.

10.1.2.2 Stillage as a Culture Medium

Considering the great amount of different organic compounds contained in the stillage, its use as a feedstock in other fermentation processes such as the produc­tion of single-cell protein (SCP) has been proposed (de la Cruz et al., 2004). SCP is employed as a component of animal feed due to its high content of proteins and vitamins. However, ethanol producers consider that both the anaerobic diges­tion and SCP production require high investments, have very prolonged residence times, and have difficult operation (Goyes and Bolanos, 2005). For Cuban situa­tions, de la Cruz et al. (2004) propose the combination of stillage irrigation and recycling to the fermentation step in ethanol production plants using sugarcane with a capacity of 55,000 L/d. For plants with a capacity of 70,000 L/d ethanol, the combination of stillage recycling and SCP production is suggested. For plants producing 90,000 L/d or more of ethanol, the installation of a plant producing SCP with a capacity of 15 ton/d other than the irrigation with stillage is proposed. These authors also point out the use of the entire stillage as fluidizing agents dur­ing the production of cement.

10.1.2.3 Stillage Oxidation

The oxidation of stillage from sugarcane in supercritical water (temperatures higher than 550°C and pressures above 25 MPa) is chemically equivalent to incin­eration. When the organic compounds make contact with the supercritical water in the presence of excess oxygen, the total oxidation reaction prevails allowing the degradation of the organic matter, which is converted into CO2, water, and nitrogen. When the oxygen is not in excess, the pyrolysis and hydrolysis reactions prevail, which can generate some value-added products. The main drawbacks of this technology are related to the corrosion and presence of insoluble inorganic salts that generate incrustations in the equipment. Goyes and Bolanos (2005) have carried out tests for total oxidation of cane stillage using hydrogen peroxide in a batch reactor achieving a 97% conversion of the organic matter with residence times lower than 3.5 min. Moreover, a carbonaceous material with an elevated surface area under partial oxidation conditions of stillage was obtained.

It is estimated that in the fuel ethanol production process from lignocellu — losic biomass, 15 liters of wastewater from each liter of ethanol produced are generated. Besides the conventional options for treating this liquid effluent by concentration-incineration and anaerobic digestion, catalytic oxidation has been proposed. The feasibility of oxidizing the generated stillage through heteroge­neous catalysts based on mixed metallic oxides (MnO2/CeO2) within a process for producing ethanol from timothy grass pretreated by steam explosion has been demonstrated (Belkacemi et al., 2000).

The International Food Policy Research Institute (IFPRI) estimates that rising bioenergy demand accounts for 30% of the increase in weighted average grain

TABLE 12.4

Some Biotechnology Improvements in Bioethanol Production Contributing to Food security

prices between 2000 and 2007 (Rosegrant, 2008). The impact was 39% of the real increase in maize prices.

However, these estimations can not be very accurate if the complexity of the markets and the different interactions between parameters affecting the price of crops are not scientifically studied in detail. The key question to answer is: Why small or large farmers can choose to grow crops for biofuels instead of food? Many challengers would say that it is not a business for farmers, and only govern­ment subsidies are the reason these projects can exist.

The choice now faced by farmers is between two alternatives. They can choose either to produce energy crops or to do nothing at all if the actual crop they have is profitable. However, profits in biofuels depend on how many hectares will be used for bioethanol feedstock. Usually sugarcane in South America and wheat and sugar beet in Europe require more than 10 hectares to be profitable depending on the prices in their annual contracts. Small-scale producers below 10 hectares can make a profit only when they belong to strong cooperatives or associations. On the set-aside land, farmers can grow bioethanol crops for a supplementary income in small-size plantations.

In the case of Africa where food security plays a key role, increased demand for biofuels is a positive development for African farmers. They have been getting paid less and less for their products, but now prices for farm products are on the rise, and farmers’ incomes are rising. After years of suffering from falling sugar prices, African farmers are finally seeing prices on the upswing due to increased demand for sugarcane ethanol, which in turn makes it more profitable to grow the crop. Countries, such as India, Mexico, and the Sudan, are world produc­ers of sweet sorghum and they see the opportunity for farmers in biofuels. This crop provides food, livestock feed, and biofuel. It grows in dry conditions; toler­ates heat, salt, and waterlogging; and provides a steady income for poor farmers. To produce ethanol, the sorghum stalks are crushed to yield sweet juice that is
fermented. But that grain can also be used for food or for chicken and cattle feed. The grain, just in case of market fluctuations, is a source of starch that can be used for bioethanol. This integral use of the crop gives stability to the farmers, which makes the business more attractive for combined production of raw materials for food and ethanol.

But one of the more important driving forces in farmer incomes from bioetha­nol crops is the price of oil. Farmers will rejoice when seeing fuel ethanol prices rise. Lastly, fuel ethanol prices simultaneously increase as oil prices increase. Most of the countries producing fuel ethanol define the prices of this product based on subsidies, raw material price in the market, and oil prices (Bernardi, 2001; Cohen et al., 2008).

The generation of technological schemes includes the determination of the opti­mal (or near optimal) process flowsheet plus the alternative configurations. This implies that the task of process synthesis is a complex activity that can be divided into several subproblems (Gani and Kraslawski, 2000). In this book, these sub­problems can comprise the following jobs applied to the production of fuel ethanol:

• Generation of alternative processing pathways

• Identification of required unit operations

• Analysis of many sequences of unit operations to assemble an optimal or at least a feasible configuration

• Configuration of distillation columns for separation of different mix­tures (including azeotropic ones)

• Configuration of reaction-reaction and reaction-separation integration processes

• Other additional issues as heat integration

To undertake this task, a series of solving strategies has been proposed. They can be classified into two large groups: knowledge-based process synthesis and optimization-based process synthesis. The second group of strategies is oriented to the formulation of a synthesis problem in the form of an optimization problem. In turn, the first group of strategies is concentrated on the representation of the design problem as well as the organization of the knowledge required by this problem (Li and Kraslawski, 2004), i. e., it is oriented to the development of a representation that is rich enough to allow all the alternatives to be included, and “smart” enough to automatically ignore illogical options. In general, the proce­dure of generating multiple alternative process flowsheets is carried out through different strategies implying, for instance, the combination of heuristic rules with evolutionary strategies for process design. Heuristic methods are particularly use­ful when very large and complex problems are dealt with. Usually, these methods are often based on the observation of how many other problems of the same type are solved. Knowledge-based methods imply an evolution in the learning of the research subject as the proposed problem is being solved. This, in turn, leads to the generation of new and better alternatives. Nevertheless, it is impossible to ensure that the optimal structure can be achieved when heuristic methods are employed (Westerberg, 2004). Some approaches for knowledge-based process synthesis mostly used in process systems engineering, as well as the most innova­tive and perspective ones, are briefly described below.

In this chapter, the first processing steps for fuel ethanol production from dif­ferent feedstocks, especially starchy and lignocellulosic materials, are analyzed from the viewpoint of process synthesis. The importance of conditioning and pretreatment as decisive process steps for conversion of feedstocks into ethanol is highlighted. Main methods for conditioning molasses are presented as well as the enzymatic procedures for starch hydrolysis. The need of pretreatment of lignocellulosic biomass is analyzed considering the complexity of this type of raw material. Several methods for pretreatment and detoxification of biomass are briefly described.

1.1 CONDITIONING OF SUCROSE-CONTAINING MATERIALS

Sucrose-containing feedstocks for ethanol production contain concentrated solu­tions of sugars. In this sense, these raw materials are nearing a point for micro­organisms to convert them into ethanol. However, some substances contained in these solutions can have an inhibitory effect on fermenting microorganisms because the used cultivation media are complex, i. e., their composition is not completely determined unlike the synthetic media for which the concentration of each one of their components is exactly defined. Moreover, all the components are known in synthetic media. In complex media, the composition of such com­ponents, such as cane molasses, varies due to factors as the agronomic techniques for cane cropping, climate conditions, luminosity, type of employed fertilizers, water availability, and cane-cutting procedures. In this way, it is impossible to predict the exact concentration of fermentable sugars in molasses before the cor­responding analyses. Although, the presence of potential inhibitors is difficult to handle, but they should be removed or their concentrations should be reduced in order to improve the subsequent fermentation performance. In the particular case of molasses, they should be diluted and conditioned for it to be used as the main component of the fermentation media.

The molasses, obtained either from sugarcane or sugar beet, has a solids con­tent of about 80° Brix. This high concentration makes the molasses unviable for direct fermentation using yeasts. Therefore, it is necessary to dilute them to below 25° Brix (Murtagh, 1995). Yeasts start to ferment quickly at this solids concentra­tion. This limit is related to the osmotic pressure that molasses exert on the yeast cells because of their high content of sugar and mineral salts. Murtagh (1995) emphasizes that the calculation required to dilute the molasses should be accom­plished in units of percentage by weight because the degrees Brix are measured by weight and not by volume. Considering that the specific gravity of molasses is 1.416, a liter of molasses would have 1,413.95 grams. In addition, the sugar content of molasses to be diluted should be taken into account due to the varia­tions in compositions during each processing batch. Thus, the final concentration of sugars after the dilution of molasses down to 25° Brix is not always the same. For instance, the cane molasses in Colombian sugar mills contains 48 to 55% total sugars on average, whereas the North American molasses is poorer in sugars (46%). In general, diluted molasses containing sugar concentrations up to 18% can be used. If more concentrated molasses is employed, the ethanol formed during fermentation will make the microorganisms have a lower growth rate that leads to more prolonged fermentation times. Alternatively, a first portion of molasses diluted down to 18° Brix can be used to favor a faster yeast growth and, once the cultivation medium reaches 12° Brix, the second portion of molasses diluted down to 35° Brix is added to this medium (Murtagh, 1995).

The ash content of molasses is another important factor to be considered. If this content is greater than 10%, incrustation problems can arise in pipelines and distillation towers in the subsequent process steps. The incrustation represents the accumulation of certain salts, mostly calcium sulfate, which are formed as a consequence of using sulfuric acid for conditioning the molasses. The addition of sulfuric acid is intended to separate the sucrose hydrolysis into its two constituent sugars: glucose and fructose. Although the yeasts synthesize the enzyme required for such hydrolysis, the invertase, the previous breakdown of this disaccharide allows a faster fermentation. Furthermore, the acid addition allows adjusting the pH of the cultivation medium based on molasses in such a way that the yeast growth is increased and the development of other undesired microorganisms, mainly bacteria, is avoided. The calcium ions come from the clarification pro­cess of both cane juice and syrup. The incrustation can be prevented through sedimentation systems in the dilution tanks of molasses, fermenters, storage tanks for wine before distillation, and stillage storage tanks before its concentration. Special chelating agents can be employed in order to remove the solids causing incrustations from the molasses (Murtagh, 1995).

Different studies to neutralize the effect the osmotic pressure has on process microorganisms have been done. One approach is the development of special yeast strains with higher resistance to the salts contained in molasses, i. e., strains more tolerant to elevated osmotic pressures. Nevertheless, the most common approach to offset the negative effect of salts is the conditioning of molasses by adding dif­ferent organic and inorganic compounds. Some of the substances employed for conditioning sugarcane molasses (that can be applied to beet molasses as well) are presented in Table 4.1.

The presence of metal traces in beet molasses also affects the ethanolic fer­mentation using yeasts. The employed agents for cane molasses, for example, EDTA (ethylene diamine tetraacetic acid), ferrocyanide, and zeolites have dem­onstrated their usefulness during the conditioning of beet molasses (Ergun et al., 1997). In the particular case of zeolites, it has been suggested that they can act as a pH regulator, which has been verified in fermentations with high glucose

TABLE 4.1

Some Ways for Conditioning and Supplementation of Sugarcane Molasses for ethanolic Fermentation using Saccharomyces cerevisiae

Source: Adapted from Sanchez, O. J., and C. A. Cardona. 2008. Bioresource Technology 99:5270­5295. Elsevier Ltd.

contents using Saccharomyces bayanus. In this way, higher ethanol concentra­tions have been achieved (Castellar et al., 1998). When immobilized cells for ethanol production from molasses are used, different impurities (inorganic salts, nonfermentable sugars, sulfated ash, colored substances) are fixed onto the sur­face of the biocatalyst decreasing its productivity. One of the strategies to dimin­ish this negative effect is the removal of impurities by microfiltration before the fermentation process that can increase the amount of produced ethanol to 18.1% (Kaseno and Kokugan, 1997).

Several substances employed as nutritive supplements in media based on molasses are also presented in Table 4.1. In general, cane molasses has most of the nutrients needed for yeast growth. However, the addition of small amounts of the nitrogen and phosphorous sources is required. In this case, the addition of nitrogen in the form of ammonium sulfate is not recommended since this salt contributes to the generation of incrustation problems. The addition of gaseous ammonium is also not desirable because it can significantly increase the pH of the medium. Urea is the most used compound as a nitrogen source in alcoholic fermentations.

Diammonium phosphate is employed as a phosphorous source that implies the reduction in the required amount of urea (Murtagh, 1995). The usage of yeast extract as a nitrogen source has been proposed as well but this material is expen­sive (Cachot and Pons, 1991). Molasses supplementation can be enhanced with the addition of some hydrolytic enzymes converting some biopolymers and non­fermentable substances contained in molasses into compounds assimilable by the yeasts, such as monosaccharides or free amino acids (Acevedo et al., 2003).

During batch fermentation, a series of operating procedures are periodically repeated to ensure the growth and development of process microorganisms. These procedures can include the washing and disinfection of the fermenter, fill­ing up the fermenter with the culture medium and sterilization of such medium, inoculation of microbial cells, fermentation, and unloading of the bioreactor con­tent at the end of cultivation process. The simplest batch ethanolic fermentation comprises an initial step for yeast propagation in an aerated bioreactor (seed fer­menter) where the microbial cells are multiplied in order to attain the appropriate concentration for them to be inoculated into a larger fermenter. In this fermenter, the conditions needed for anaerobic cultivation of yeasts are ensured, favoring in this way, the production of ethanol. The main drawback of this process consists in the operating and feedstock costs needed for each fermentation batch to ensure the yeast propagation until it reaches a concentration high enough to allow for the appropriate cell growth and ethanol production rates. In addition, the yeasts are not reutilized, which indicates one should not employ all the potential of cell biomass formed during the process.

The typical process for fuel ethanol production based on batch fermentation is the so-called Melle-Boinot process. In Brazil, it uses molasses or cane syrup and comprises the weight and sterilization of feedstock, followed by the adjust­ment of pH with H2SO4 and of the degrees Brix to values of 14 to 22. Yeasts ferment obtained wort. The produced wine is decanted, centrifuged, and sent to the separation stage of ethanol, whereas the yeasts are recycled to the fermenta­tion stage in order to reach high cell concentration during cultivation (Kosaric and Velikonja, 1995; Sanchez and Cardona, 2008), as shown in Figure 7.1. A simple and effective method proposed to maintain high cell concentrations is the propagation of yeasts every 13 batches in the fermentation medium, which leads

FIGURE 7.1 Schematic sequence of operations during batch industrial process for etha­nol production from sugarcane.

to a productivity higher than that obtained for the conventional process with yeast recovery (Navarro et al., 1986). Shojaosadati (1996) highlights that yeast reuse results in a decrease in new growth with more sugar available for conversion to ethanol and a corresponding increase in ethanol yield of 2 to 7%. For etha­nol production from beet molasses, this author obtained 8% lower consumption of the feedstock employing cell recycling. Traditional productivities reached by batch fermentation are in the range of 1 to 3 g/(L x h). However, if a significantly high concentration of yeasts (44 g/L) and a high supplementation of yeast extract (28 g/L) at low ethanol content (approximately 60 g/L) are employed in a glu­cose-based medium, ethanol productivity as high as 21 g/(L x h) can be achieved (Chen, 1981). This value is comparable to those obtained during continuous culti­vation, as shown in Table 7.1.

The recirculation of stillage from a previous batch to the current fermentation, as schematically illustrated in Figure 7.2, has also been proposed. The stillage represents one of the distillation product streams during the subsequent ethanol recovery step that contains a significant amount of water and a much-reduced amount of ethanol. The addition of stillage to the culture broth can lead to lower water consumption and the reduction of stillage volume to be treated.

Conversion of cellulose into ethanol can be carried out through SSF, as in the case of starch. For this conversion, several enzymes with cellulolytic activity (basically endolucanases, cellobiohydrolases, and P-glucosidase) are added to the suspension obtained by mixing water with the solid fraction resulting from the pretreatment step and that contains cellulose and lignin. In the same way, pro­cess microorganisms (yeasts) are added to this mixture in the bioreactor where SSF immediately converts the formed glucose into ethanol. Taking into account that sugars (glucose, cellobiose) are more inhibitive for the conversion process than ethanol, SSF can reach higher rates, yields, and ethanol concentrations in comparison with SHF (Wyman et al., 1992). The increased ethanol concentra­tion in the culture broth allows the reduction of energy costs during distillation. In addition, SSF offers an easier operation and a lower equipment requirement than the sequential process since no hydrolysis reactors are needed. Moreover, the presence of ethanol in the broth makes the reaction mixture less vulnerable to the action of undesired microorganisms (Wyman, 1994). Nevertheless, SSF is inconvenient in that the optimal conditions for hydrolysis and fermentation are different, which leads to a difficult control and optimization of process param­eters (Claassen et al., 1999). In addition, larger amounts of exogenous enzymes are required (Cardona and Sanchez, 2007).

Since the time in which the first introduction to SSF from biomass began, the duration of the batch process has decreased from 14 days required for conversion of 70% cellulose into ethanol with final concentrations of 20 g/L to 3 to 7 days needed for reaching 90 to 95% conversions with final ethanol concentrations of 40 to 50 g/L (Cardona and Sanchez, 2007; Wyman, 1994). The concept of the SSF process was first described by Takagi et al. (1977). Takagi, Suzuki, and Gauss (Gauss et al., 1976) had previously patented the SSF technology for bioethanol production by which the yeasts simultaneously metabolize the glucose into eth­anol in situ during the enzymatic saccharification of the cellulose. This patent expired in 1993 and it has been utilized for small-scale demonstrations, according to Ingram and Doran (1995), but until now, no commercial plants have been built at an industrial level (Cardona and Sanchez, 2007).

The overall process for ethanol production from lignocellulosic biomass by SSF is depicted in Figure 9.4. This process allows the recovery of energy along with ethanol production. The lignin, one of the major polymers present in ligno- cellulosic biomass, is not practically changed during the process and remains in stillage generated in the product recovery step. This lignin can be separated by

centrifugation from the liquid fraction of stillage and burnt in order to extract its high energy content. The obtained thermal energy is converted into process steam, which is employed within the same process. If co-generation is considered, part of the released energy is transformed into electricity that can supply all the needs of the plant, with a remaining surplus that can be sold to the grid. These energy recovery possibilities are an important feature of ethanol production pro­cesses when lignocellulosic biomass is utilized, and they allow the compensation of energy demands represented mostly by the steam requirements for the pretreat­ment reactor, distillation columns, and dehydration scheme.

Intensive research to carry out the SSF of lignocellulosic biomass has been car­ried out. Some examples of these efforts are presented in Table 9.1. Considering that enzymes account for an important part of production costs, it is necessary to find methods reducing the cellulases doses to be used. Thus, the integration of the cellulase production process using Trichoderma reesei with ethanolic fer­mentation has been proposed. As a small amount of enzymes remains entrapped in fungal cells producing cellulases, the addition of the whole culture broth of this process to the SSF reactor was proposed. Besides the fungal biomass and obtained cellulases, this broth contains residual cellulose and lignin. This allows the increase of the P-glucosidase activity, essential for the reduction of cellobiose levels, and a more complete utilization of sugars employed during the production of cellulases (Wyman, 1994). The addition of surfactants has been proposed for these purposes as well. Alkasrawi et al. (2003) showed that the addition of the nonionic surfactant Tween-20 to the steam exploded wood during the batch SSF using Saccharomyces cerevisiae has some effects: 8% increase in ethanol yield, 50% reduction in cellulases dosage (from 44 FPU/g cellulose to 22 FPU/g cellu­lose), an increase of enzyme activity at the end of the process, and a decrease in the time required for reaching the highest ethanol concentration. It is postulated that the surfactant avoids or diminishes the nonuseful adsorption of cellulases to the lignin. However, Saha et al. (2005) obtained marginal increases (3.5%) in sac­charification of rice hulls using 2.5 g/L of Tween-20.

In early works, Wyman et al. (1992) carried out a systematic evaluation of the ethanol yields obtained from several lignocellulosic materials pretreated with sulfuric acid and fermented with S. cerevisiae in batch SHF and SSF processes. Different dosages of cellulase supplemented with P-glucosidase were used. Obtained results show that corn cobs, corn stover, wheat straw, and weeping love grass, in that order, were adequately pretreated with acid, achieved high rates of enzymatic hydrolysis, and demonstrated high ethanol yields during 7 to 8 d of SSF. Switchgrass, in turn, exhibited the worst results. The different types of ana­lyzed pretreated woody crops showed similar cellulose-to-ethanol conversions though the yields were slightly less than corn stover, which was the material with the best values (92 to 94%). South et al. (1993) obtained 91% and 96% conversions for the batch SSF of dilute acid pretreated hardwood and poplar wood, respec­tively, using S. cerevisiae and cellulase supplemented with P-glucosidase. These values were attained during 3 d, but the authors did not report the concentra­tion of produced ethanol. Hari Krishna et al. (1998) also evaluated the optimal conditions of the SSF of sugarcane leaves, as they did for the SHF process (see Chapter 5, Section 5.2.2). These authors defined the best conditions for the 3-d cultivation process as 40°C and a pH of 5.1, which allows for achieving 31 g/L of ethanol from an initial substrate load as high as 15%. Nevertheless, the enzyme dosage was quite high (100 FPU/g cellulose). Similarly, the leaves of Antigonum leptopus, a weedy creeper abundant in the areas not employed for conventional agriculture, was evaluated for ethanol production by SSF using yeasts showing higher yields than in the case of SHF (Hari Krishna et al. 1999).

Dale and Moelhman (2001) tested different commercial cellulases during batch SSF processes for ethanol production from biomass. Best results corresponded to Validase TR of Valley Research and CEP of Enzyme Development Corp. These authors state that Amano and Genencor cellulases were not adequate for this pro­cess. However, this information should be corroborated for each type of biomass since these authors do not clarify what material was utilized.

Softwood is more difficult to degrade by SSF than hardwood (Sanchez and Cardona, 2008). Stenberg et al. (2000a) employed the resulting slurry of the steam pretreatment of SO2-impregnated spruce in SSF tests using S. cerevisiae and determined that the best initial load of substrate was 5% (w/w) obtaining 82% yield of the theoretical based on the cellulose and soluble hexoses present at the start of the SSF process. The productivity was doubled related to SHF. Higher dry weights negatively affected the fermentation process because of the inhibitors present in the slurry (the pretreated material was not washed). The load of cel — lulase was in the range of 5 to 32 FPU/g cellulose.

One of the most relevant factors in the alcoholic fermentation is the pos­sibility of infection by acidolactic bacteria. Stenberg et al. (2000b) studied the influence of lactic acid formed during fermentation on the course of batch SSF of steam exploded spruce chips (softwood) using baker’s yeast. It is interesting that the contamination was higher when washed substrate was used, in com­parison with the utilization of the whole suspension resulting from pretreat­ment. This indicates that acidolactic bacteria are more sensitive to inhibitors’ presence than yeasts.

As mentioned above, one of the main disadvantages of SSF processes using lignocellulosic biomass lies in the different optimum conditions of enzymatic hydrolysis of cellulose and fermentation. Cellulases work in an optimal way at 40° to 50°C and pH of 4 to 5, whereas the fermentation of hexoses with S. cerevisiae is carried out at 30°C and pH of 4 to 5, and fermentation of pentoses is optimally performed at 30° to 70°C and pH of 5 to 7 (Olsson and Hahn-Hagerdal, 1996; Sanchez and Cardona, 2008). Varga et al. (2004) proposed a nonisothermal regime for batch SSF process applied to wet oxidized corn stover. In the first step of SSF, small amounts of cellulases were added at 50°C in order to obtain better mixing conditions. In the second step, more cellulases were added along with the yeast S. cerevisiae at 30°C. In this way, the final solid concentration in the hydrolyzate could be increased up to 17% dry matter concentration achieving 78% ethanol yield.

In general, increased cultivation temperature accelerates metabolic processes and lowers the refrigeration requirements (Sanchez and Cardona, 2008). Yeasts such as Kluyveromyces marxianus have been tested as potential ethanol produc­ers at temperatures higher than 40°C (Ballesteros et al., 2004; Hari Krishna et al., 2001). Kadar et al. (2004) compared the performance of thermotolerant K. marx — ianus and S. cerevisiae during batch SSF of waste cardboard and paper sludge not finding great differences between both microorganisms at 40°C, although cellulose conversions (55 to 60%) and ethanol yields (0.30 to 0.34 g/g cellulose) were relatively low. Ballesteros et al. (2001) carried out several fed-batch SSF tests at 42°C during 72 h using K. marxianus in the case of by-products of olive oil extraction (olive pulp and fragmented olive stones). Their results showed 76% ethanol yields of theoretical for olive pulp and 59% yield for acid-catalyzed steam-exploded olive stones. With the aim of increasing ethanol yields from olive pulp, Ballesteros et al. (2002) employed liquid hot water (LHW) pretreat­ment reaching an 80% yield of the theoretical value and recovering potentially valuable phenolic compounds.

If, when thermotolerant yeasts are used, the microbial cells can also assimilate pentoses, the SSF process can become more perspective (Cardona and Sanchez, 2007). Yeasts such as Candida acidothermophilum, C. brassicae, S. uvarum, and Hansenula polymorpha can be used for these purposes. In this case, the addition of a larger amount of nutrients to the medium is required. Alternatively, the uti­lization of higher cell concentrations could be implemented for obtaining better results (Olsson and Hahn-Hagerdal, 1996; Ryabova et al., 2003). The difficulty lies in the fact that higher temperatures enhance the inhibitory effect of ethanol. Therefore, the isolation and selection of microorganisms that could be adapted in a better way to these hard conditions should be continued. Kadar et al. (2004) make reference to several reports about the utilization of thermotolerant microor­ganisms for ethanol production.

One approach to accomplish the SSF of biomass without the addition of cel — lulases consists in the utilization of mixed cultures in such a way that the hydro­lysis and fermentation of lignocellulosic biomass be carried out simultaneously (Cardona and Sanchez, 2007). This procedure was applied to sweet sorghum stalks employing cellulase — and hemicellulase-producing fungus Fusarium oxysporum along with S. cerevisiae (Mamma et al., 1995). Considering that sweet sorghum stalks contain several carbohydrates, such as sucrose, glucose, hemicellulose, and cellulose, the obtained yields in this process were higher than the theoretical yield from only glucose (0.51 g EtOH/g glucose); this is explained by the additional bioconversion of cellulose and hemicellulose into ethanol (Mamma et al., 1996). However, the final ethanol concentrations in this type of SSF process were quite low considering the separation process. Logically, the mixed culture presents a high complexity during its implementation at the industrial level and has the additional disadvantage that optimal growth conditions for two or more different microorganisms are not the same. Besides, part of the substrate is deviated for the growth of the enzyme-synthesizing microorganism. Panagiotou et al. (2005a) carried out the SSF of cellulose with F. oxysporum demonstrating the produc­tion of ethanol under anaerobic conditions. In addition, the metabolite profiling of the microorganisms cultivated in different media was achieved through the measurement of the intracellular concentration of key metabolites (Panagiotou et al., 2005a, 2005b, 2005c).

The SSF process also can be done in continuous regime. In the same work of South et al. (1993) cited above, the behavior of a continuous-stirred tank reactor (CSTR) for continuous SSF of hardwood using the same microorganism and the same enzymes was investigated. For a residence time of 3 d, 83% conver­sion was achieved, i. e., less than in the case of batch SSF, which is explained by the decrease in the reactivity when conversion in the biomass hydrolysis is increased. Ethanol concentration reached 20.6 g/L for a residence time of about 2 d. Obtained results showed that enzyme and substrate concentrations in the feed within studied ranges did not influence cellulose conversion. This indicates the existence of mass transfer restrictions related to cellulose, besides inhibi­tory effects caused by the substrate and products. The fact that the enzymatic hydrolysis rate decreases with the course of hydrolysis has often been reported in the literature. According to Zhang and Lynd (2004), this decreased reactiv­ity of residual cellulose can be due to less surface area, fewer accessible chain ends, and/or adsorption of inactive cellulase on the surface of lignocellulosic particles. The addition of fresh substrate can stimulate the release of more sol­uble sugars indicating the loss of cellulose reactivity at the end of hydrolysis or the increase of reactivity for the “new” encounters enzyme-substrate compared with the “old” ones.

9.2.2.1 Modeling of SSF of Cellulose

In Chapter 5, Section 5.2.3.2, the importance of cellulose hydrolysis description was discusses, taking into account the development of suitable simulation tools to be employed during process synthesis procedures. The difficulties that arose due to the several factors affecting the kinetics of cellulose hydrolysis were also high­lighted in that section. Evidently, these difficulties are also present when model­ing of cellulose SSF is performed. One of the most comprehensive mathematical models for this process in both batch and continuous regimes corresponds to the description provided by South et al. (1995). As mentioned in Section 5.2.3.2, these authors developed a model based on experimental data obtained in their previous work using commercial fungal cellulase (South et al., 1993) employing pretreated wood as the feedstock. The kinetic model includes the cellulose con­version, the formation and disappearance of cellobiose and glucose, the forma­tion of cells, and the biosynthesis of ethanol structure, i. e., the main enzymatic and microbial phenomena taking place during the SSF process. In addition, a Langmuir-type model taking into account the adsorption of cellulases on the solid particles of cellulose and lignin is also considered. Moreover, this math­ematical description contemplates a population model for the residence time of solid particles entering the bioreactor in the case of the continuous regime. Thus, expressions describing the dependence of cellulose conversion on the residence time of nonsoluble, solid particles of biomass were derived. This last description confers great validity to the model because it provides a better approximation to real processes, which cannot be suitably explained by the traditional models for CSTR with soluble substances.

and Fermentation (SSF) of Corn Starch

New tendencies in the corn-to-ethanol industry are aimed at dry-milling pro­cesses. The increase in ethanol production capacity in the United States is mainly represented by dry-milled corn ethanol plants (Sanchez and Cardona, 2008; Tiffany and Eidman, 2003). In this regard, the simultaneous saccharification and fermentation (SSF), a reaction-reaction integrated process, has been proving its potential for improving the overall process, as was discussed in Chapter 9, Section 9.2.2.1. This process only differs from the SHF process in that the saccharifica­tion and fermentation are carried out simultaneously in the same unit at 33°C (Figure 11.7). The remaining steps are similar for both processes.

Currently, worldwide production of fuel ethanol is carried out by using sugar — containing or starch-containing materials. In this chapter, main feedstocks for ethanol production (sugarcane, sugar beet, corn, wheat, cassava) are discussed in reference to their production, advantages, and drawbacks. Lignocellulosic biomass as feedstock for bioethanol production is also analyzed due to the recent demand for second-generation biofuels. Finally, some evaluation data related to these feed­stocks are highlighted from the viewpoint of process systems engineering.

3.1 SUGARS

Bioethanol can be produced from raw materials containing fermentable sug­ars, especially sucrose-containing feedstocks, such as sugarcane or sugar beet. Sugarcane is the main feedstock for ethanol production in tropical countries like Brazil, India, and Colombia. This feedstock can be used either in the form of cane juice or cane molasses. About 79% of ethanol produced in Brazil comes from fresh sugarcane juice and the remaining fraction corresponds to cane molas­ses (Wilkie et al., 2000). Sugarcane molasses is the main feedstock for ethanol production in India (Ghosh and Ghose, 2003). Cane molasses and different sugar- containing streams like syrup and B-type cane juice are used for producing bio­ethanol in Colombia. Beet molasses is the main feedstock for ethanol production in France along with wheat (Decloux et al., 2002).

Enzymatic detoxification is one of the new methods tested for inhibitors removal. For this, the phenoloxidase laccase is employed. This enzyme oxidizes the phe­nolic compounds derived from lignin. Martin et al. (2002) have evaluated the fermentation behavior of bagasse hydrolyzates pretreated by steam explosion and treated with commercial cellulases or acids. The hydrolyzates were treated with phenoloxidase laccase and then underwent alkalinization. The specific productiv­ity (measured as g EtOH/g bagasse x h) of the detoxified hydrolyzate was almost four times greater than the specific productivity of the nondetoxified hydrolyzate. The fact that the inhibitory effect was remarkably reduced when the phenolic compounds were specifically removed by the laccase suggests that this enzyme could be used to neutralize the inhibitory effect of low molecular weight soluble phenolic compounds instead of using nonspecific chemical methods (Palmqvist and Hahn-Hagerdal, 2000a). Other alternative biological methods involving microorganisms (microbial detoxification) have been proposed. For instance, in the case of dilute solutions resulting from biomass pretreated by pyrolysis, a bio­film reactor with a mixed culture of aerobic bacterial cells naturally immobilized on a plastic composite support has been used (Khiyami et al., 2005). The most employed biological methods of detoxification are shown in Table 4.8.

Biological Methods for Detoxification of Pretreated Biomass

Source: Adapted from Sanchez, O. J., and C. A. Cardona. 2008. Bioresource Technology 99:5270-5295. Elsevier Ltd. Note: hz = hydrolyzate.