VOLUME IBIOMASS CONVERSION WITH CELLULASES
INTRODUCTIONThis treatise deals with a novel, ecologically benign, technically effective and economically practical cellulose-processing applied science (wood refinery) and its associated biotechnology; as well as the production of sugars and alcohols, particularly as derived from cellulosic materials; and industrial enzyme engineering processes.It is clear that a large number of cellulose-, starch- and sugar-containing plants can be processed to produce sugars and alcohols. Sugar-containing plants such as sugarcane, sweet sorghum and nipa palm are the best candidates for the high-yield production of alcohol fuels. Similarly, the starch-containing crops such as cassava, sweet potatoes, yams, taro and tannia are also good candidates but require an additional step to break down starch to sugar. However, the emphasis of this report is on the major cellulose-containing biomass, which requires special treatment before it can be used to produce glucose and alcohols. To utilize cellulosic raw materials the following steps are necessary: (1) Growth, harvest and delivery of raw materials to processing plants; or, alternatively, the collection and delivery of cellulosic "waste" products. (2) Pre-treatment or conversion of the raw material by mechanical, physical, chemical or enzymatic methods to break down cellulose to sugars and to modify or remove unwanted side-products, usually lignin and hemicellulose. (3) Recovery and purification of sugars from reaction mixture. (4) Fermentation of sugars to alcohol and purification by distillation. (5) Treatment of process residues to reduce pollution and to recover potentially valuable by-products. From the considerable R&D work carried out in all areas of pre-treatment it appears that acid and enzymatic hydrolytic processes hold the most promise for developing countries. Though acid hydrolysis technology is more advanced, greater ultimate potential is expected for enzymatic hydrolysis which is, therefore, recommended. The economics of producing alcohol from cellulosic materials is generally considered at this time to be unfavorable, although the data upon which this conclusion is based are limited. Therefore, it is recommended that further R&D be undertaken in this area in order to evaluate one or more commercially attractive processes for producing alcohol from fermenting cellulose. The treatise consists of three major parts in six volumes: the first part presents the biochemical background of the enzymatic hydrolysis of cellulose and describes principles and approaches to the utilization of cellulosic raw materials (biomass) and principles of biotechnological processes; the second describes case studies of processes realized on a pilot scale (no industrial processes for enzymatic hydrolysis of cellulose exist as yet); the third describes the state-of-the art in enzyme engineering processes at the industrial level, some of which could be coupled with the novel cellulose-producing technology and associated biotechnology into an integrated industrial process(es), and some of which could be considered as case studies for courses in research and training. Volume I - Biomass Conversion with Cellulases - deals with applied principles of the process, starting with some of the potential biodegradable agricultural and agro-industrial cellulosic residues and their availability, and provides a thorough review of various methods for pre-treatment of raw materials (including mechanical, chemical, physical, biological, and combinations of the above, that make cellulosics more reactive for enzymatic conversion to valuable products): cellulase enzymes, their natural sources, and cellulase preparations that are commercially available in several countries; composition and activity of available enzyme preparations and a description of their important properties that affect their potential industrial applications. This volume also describes the technology of glucose production in enzyme reactors, inhibition of the hydrolytic reaction by its products, mainly glucose and cellobiose, and principal ways to overcome this negative feature of the biotechnological process. The principles of practice of enzyme recycling are discussed in this section, as well as state-of-the-art production technologies, the economics of extant processes, and summary implications of the foregoing for developing countries Volume II - Principles of the Enzymatic Degradation of Cellulose - covers the following principal biochemical areas of the enzymatic degradation of cellulose: adsorption of cellulases on cellulose; behavior of adsorbed cellulases on the cellulose surface; synergism between components of the cellulase system; mechanism(s) of degradation of cellulose by cellulases of varied origin; comparative consideration of cellulases of fungal and bacterial origin and design of cellulase systems with improved characteristics by protoplast fusion, enzyme fusion, protein engineering and genetic engineering in general. The overall aim of this consideration is to provide a contemporary basis for developing a strategy for industrial application of enzymatic conversion of cellulose to glucose (and then to ethanol, among other useful products). This section shows, in particular, how knowledge of some properties of cellulolytic enzymes, such as specific activity, adsorption capacity, thermostability, product inhibition, etc., can be translated into concrete requirements for cellulases for their potential industrial applications. Volume II also describes recombinant DNA technology that is essential for the generation of optimized enzymes for industrial utilization. Procedures which increase thermal stability and enhance catalytic efficiency are detailed. The text includes techniques which have been utilized on cellulases in particular as well as those which are applicable to the stabilization of enzymes in general. The protocols currently used for the generation of the engineered proteins include the polymerase chain reaction, site specific mutagenesis, random mutagenesis, cloning and expression of the individually altered proteins. Other genetic engineering techniques required for the achievement of the desired goals are outlined. Volume III - Economics of Enzymatic Cellulose Hydrolysis - indicates that, as yet, no country has developed an industrial process for the enzymatic conversion of cellulose into sugars or alcohol. Several countries have designed and built pilot or pre-pilot scale solid waste treatment plants. The advantages and disadvantages of the major specific technologies being developed in this area can only be evaluated by comparing them with each other. This volume describes processes being developed to a certain degree in the USA, Canada, Russia, India, Finland, Sweden, Italy, Japan, and South Africa. Eventually, an integrated set of commercially attractive industrial biotechnology processes based on biomass conversion and wood refinery will benefit developing countries, particularly:
Volumes IV, V and VI - Industrial Production with Immobilized Enzymes - describe the state-of-the-art of contemporary technological processes using immobilized enzymes and cells on the industrial level. Some of them could be combined with the enzymatic production of glucose from cellulose, described in Volumes IV (High Fructose Corn Syrups and Amino Acids for Food and Feed), V (Antibiotics, Food Additives and Dairy Products), and VI (Saccharose and Starch Hydrolysis), and/or with the subsequent microbiological conversion of that glucose into nutritional or other useful products. Others can serve as examples of modern industrial or semi-industrial processes based on practical application of principles of enzymology and biochemical engineering. Also, these enzyme processes can be considered as case studies for the course of research and training. Enzyme engineering is a relatively new field of science and technology that deals with the development of useful products or processes based on the catalytic action of isolated enzymes and/or non-growing cells, converted into heterogeneous (insoluble) catalysts. Immobilized enzymes/cells are readily separable from the product solution and can be reused, are often more operationally stable in fixed form than in solution, and typically are much less expensive considering their longer period of use without replenishment. Volumes VI, V and VI describe in detail eight processes that use immobilized enzymes or microbial cells and have found industrial application: - Production of high-fructose syrups by immobilized glucose isomerase, - Optical resolution of amino acids by immobilized aminoacylase, - Production of optically active D-amino acids by immobilized hydantoinase, - Production of L-aspartic acid by immobilized aspartase, - Production of L-malic acid by immobilized fumarase, - Deacylation of penicillin G and cephalosporin derivatives by immobilized penicillin amidase, - Hydrolysis of lactose in dairy processing (milk and cheese whey) by immobilized lactase, - Production of glucose-fructose syrups from saccharose by immobilized invertase. Each of these processes is considered in detail in terms of its background, available commercial preparations of respective immobilized enzymes, technological characteristics of the processes, economic estimations, and current (or previously reported) scale. BIOTECHNOLOGY OF THE ENZYMATIC HYDROLYSIS OF CELLULOSEA. BACKGROUND1. An interest in the enzymatic hydrolysis of cellulose began during the Second World War. Hordes of microscopic fungi inundated the American units stationed in the South Pacific. All cotton gear, tents, uniforms, knapsacks and cartridge belts, were ruined after a brief period in the humid tropical climate. Much cargo space was wasted just replenishing items that had become non-functional. 2. Since the situation was getting out of hand, solutions were called for. As a result, several laboratories within the Armed Services were established to investigate the nature of "rotting," to establish the causes, identify the invisible culprit and how it operated, and to find the ways and means to halt the damage. 3. The laboratories charged with these tasks carried out studies that included the investigation of more than 10,000 microorganisms. This enormous task included growing each one on textile strips and testing tens of thousands of strips for loss of tensile strength. It took several years to establish that the agent responsible was a special type of cellulose-destroying fungus. Therefore, it attacked cotton fabrics, since cellulose is the main component of cotton. 4. The chemical structure of cellulose is central to an understanding of these events. It is almost completely composed of "sugar", or more precisely, of glucose units. These are chemically bonded "head-to-tail" and form very long chains, containing hundreds and thousands of glucose units fixed into a rigidly ordered pattern rather like a crystalline lattice. The result is a stable structure because the chains are cross-linked by hydrogen bonds. In and of themselves the bonds are not very strong, but when blended with thousands of others they form a veritable monolith. As a result cellulose is not only insoluble in water, but its crystalline regions are virtually inaccessible to most chemical agents; neither water nor even strong acids easily penetrate the "crystallites". 5. However, the cellulose-destroying fungi achieved this by attaching themselves to cellulose-containing material and secreting a number of cellulose degrading enzymes. Several thousand different enzymes exist in nature and each is a catalyst for a specific chemical reaction. In the case under examination evolution has generated fungi which produce enzymes that degrade cellulose. Termed "cellulases", they act on cellulose in the immediate vicinity of the fungus, systematically cleaving the chemical bonds that link glucose molecules into polymer chains, until only glucose, the end product, remains. 6. Glucose, the main nutrient, enables the fungi to multiply, grow and spread over new areas, secreting more enzymes until the available substrate, cellulose, is exhausted. 7. In nature this usually takes a very long time. Years elapse before a tree stump decays. In some cases the process can be accelerated greatly, as in the South Pacific example. However, the reaction can be sped up even further. If the enzymes - cellulases - are isolated from the microorganisms, pooled and then added to the cellulose, the resulting glucose will not be consumed by the fungus, but will accumulate in the reaction mixture. There is another advantage; if cellulose-containing waste, not pure cellulose, were to be used as substrate, the waste problem would also be solved. Glucose can be fermented to form ethanol and used as fuel instead of other chemicals, and additionally, the dehydration of ethanol produces ethylene - a major compound used in much of the modern chemical industry. Depending on its purity and the economic efficiency of the process, glucose thus obtained could be used widely by the pharmaceutical, chemical, food and bioscience-based industries. 8. Cellulose heads the list of the world's renewable resources. The worldwide production of cellulose by nature runs into billions of tons annually. Mankind's use of these resources results in the production of very large quantities of cellulose-containing waste. If even a fraction of this waste were to be turned into useful products by enzymatic treatment, renewable supplies of edible carbohydrate and gasoline substitutes would become a major share of the world market. Given the enormity of the potential rewards, it is not surprising to find that research in these areas in recent years has been assiduous. 9. Generally speaking, cellulolytic enzymes are not the only agents capable of breaking down cellulose to sugars. This can also be done by subjecting cellulosics to a variety of physical and/or chemical practices. Physical methods include the steaming of cellulose or its g-irradiation, though they are so far not technologically attractive (unless the "steaming" is performed in a very specific manner, see below). The best known chemical methods involve the treatment of cellulose with strong mineral acids, e.g., sulfuric and hydrochloric acids. 10. The use of acids in industrial application followed by fermentation of the resulting glucose to alcohol began 80 years ago. The first commercial application was in 1913 at Georgetown, South Carolina, where a plant was built to hydrolyze southern pine mill waste with 2% sulfuric acid at 175°C in rotary steam-heated digesters. The dilute sugar solution, produced at a 25% yield, was fermented to yield 5,000 gallons of ethanol per day. A second plant was built subsequently in Fullerton, Louisiana. Both plants produced alcohol profitably until the middle 1920s, benefiting from extremely low-cost sawdust waste and the lack of environmental controls. Eventually, a fall in the price of blackstrap molasses led to cheap supplies of sugar from this source and the two plants became unprofitable. 11. During the 1930s Scholler in Germany developed a process by which wood in a percolator was subjected to hydrolysis with successive batches of dilute acid. Wood was pretreated with 1% HCl and then digested with 0.5% sulfuric acid at 130-190°C for 18-24 hours in stationary digesters. By removing sugar as it formed yields amounted to roughly 50%, about twice the typical yield of earlier processes. Three such plants were built in Germany and used during the Second World War. Additionally, one was built in Korea, and one at Ems, Switzerland. In the USSR, a percolation plant was built in 1935, with a further 40 being built after the Second World War. 12. The Bergius process, also developed in Germany during the Second World War, used concentrated (42%) hydrochloric acid in special acid-resistant equipment. It resulted in cleaner syrup, but was expensive because the acid had to be recovered. Nevertheless, the sugars from the Scholler and Bergius processes were used for the production of alcohol and for the growth of yeast for human consumption. By the end of the Second World War, approximately 9,000 tons of food yeast was being produced per year in Germany using these processes. 13. The Forest Products Laboratory of Madison, Wisconsin, in co-operation with the Cliff's Dow Chemical pilot plant at Marquette, Michigan, investigated acid hydrolysis of wood to sugar during and after the Second World War. The work resulted in a modification known as the Madison Wood Sugar Process which accelerated the hydrolysis process and produced higher yields. In this process, chopped wood waste was continuously percolated with 0.4-0.6% sulfuric acid at temperatures gradually rising to 150-185°C. The process yielded 4-5% sugar syrup in 45-55% yields. After neutralization, the sugar syrup was concentrated to 50% molasses, which was fed to dairy cattle as a silage additive or used to grow brewer's, baker's and food yeasts. 14. Immediately after the Second World War a plant incorporating the Forest Products Laboratory developments was built at Springfield, Oregon to produce ethyl alcohol for the manufacture of synthetic rubber via butadiene. After its completion, the Springfield plant demonstrated its functional usefulness but in cost-effective terms it could not compete with petrochemical derived ethanol. 15. Only one of the above processes have succeeded as a commercial operation in the almost fifty years since the Second World War. The percolation industry in the former Soviet Union using multifunctional, raw cellulosic materials yields a production not only of alcohol but also of yeasts and furfurol (when lignin was used as fuel). Nowadays this industry consists of 40 full-scale plants with a maximum capacity of 1,000 tons of wood material per plant per day. The annual production from this technology is 400,000 tons of fodder yeasts and 195 million liters of ethyl alcohol. 16. In recent years, the Brazilians have taken advantage of the Soviet experience for their wood hydrolysis plants. The Brazilian plants are not only making alcohol but also are manufacturing charcoal for use in the Brazilian steel industry. COALBRA, the Brazilian company responsible for building plants to make alcohol from wood, took its name from its two products, coke and alcohol. According to recent data, COALBRA has a pilot plant under construction which is expected to produce 30,000 liters per day of alcohol from wood. This work is under the jurisdiction of the National Alcohol Programme (PROALCOOL), established in 1975 as a means to partially substitute ethyl alcohol for petroleum derivatives. Impetus for the PROALCOOL program arose from Brazil spending half of its export revenue - about US $10 billion - to buy petroleum from abroad. Currently, Brazil has the largest alcohol-fuel industry in the world, with sugarcane as the raw material (Rothman et al., 1983), and produces 12 billion liters of alcohol annually (International Energy Agency Newsletter, 1992). Lignocellulose is not as yet a raw material for commercial ethanol production on a significant scale (Lynd, 1990). 17. Currently conversion of corn and lesser amount of other grains to ethanol is the basis for a substantial industry in the US (Lynd, 1990). Ethanol produced from these sources is used as a gasoline additive that raises octane. Approximately 7% of the total motor gasoline used in the US is blended with ethanol (National Advisory Panel on Cost-Effectiveness of Fuel Ethanol Production, 1987), and each year more than 3 billion liters of ethanol is being produces in the USA (International Energy Agency Newsletter, 1992). The economics of corn-derived ethanol are currently dependent on federal subsidies and the value of processing residues as feeds (Venkatasurbramanian & Keim, 1985). 18. In recent years investigators, mainly in the United States, have been working on the development of continuous hydrolyzers of cellulosic materials. These include Grethlein and Converse at Dartmouth, Church at American Can, and Rugg at New York University, who have all been developing continuous dilute acid hydrolysis technology. The two-stage continuous hydrolysis has some important advantages over percolation, such as effective fractionation, improved by-product recovery, and higher sugar concentration. However, the prevalent thought today is that although such continuous saccharification may be ready for first level pilot scale investigations, it is not yet ready for commercialization. 19. It should be noted here that alcohol production through the acid hydrolysis of wood also presents a series of disadvantages: - to be economical, the scale of the process must be very large; in order to produce some 500,000 liters per day, an area of nearly 600,000 hectares is required to maintain a constant supply of raw material and fuel; - it is extremely capital intensive, requiring an investment per liter of industrial capacity which is double or triple those of other raw materials, while production costs are 50 to 100% greater; - the process requires a large amount of sulfuric acid (at the rate of one liter per three liters of alcohol produced); - the process calls for rather severe conditions, and as a consequence, requires expensive corrosion-resistant equipment; - the severe conditions of hydrolysis lead to a partial degradation of the glucose formed and, as a result, to a contamination of sugars by toxic by-products (e.g., furfural) which in turn are poisonous both for microorganisms, e.g., fermenting yeasts, and human beings. 20. Enzymatic hydrolysis is the most promising alternative to the use of dilute acid, but it is certainly not a replacement process. A future possibility is the development of a hybrid process which uses dilute acid to provide a pretreatment-fractionation which is coupled with enzymatic hydrolysis that converts the residual (resistant) cellulose to glucose. A choice between acid or enzymatic hydrolysis of cellulosics depends on certain circumstances, especially on the location of the respective plants, availability of raw materials and links with nearby chemical or biochemical factories able to supply mineral acids, culture broths, etc. 21. Another possibility, which is outlined below, is the use of a physico-chemical disruption of wood into a physical mixture of its major components with a subsequent fractionation of the mixture into individual components (wood refinery) using mild alkali and water. Clean, wet, and open to enzymatic attack cellulose will be susceptible to conversion into glucose. 22. Once the hemicellulose sugars and the lignin have been removed from raw materials such as wood, the remaining cellulose can be hydrolyzed. Both enzyme and acid catalysts are effective for this process. However, as a rule hydrolyzates after enzymatic conversion of cellulosics contain a substantially higher proportion of glucose than other products of cellulose degradation as compared to acid hydrolyzates. This may be particularly important when the aim is the production of pure glucose as food sugar, rather than its fermentation to alcohol, since glucose can be isomerized enzymatically into fructose (see Volume IV). B. OVERVIEW 23. Before a successful industrial process for glucose production from cellulose can be realized the problems that must be resolved are the following: (1) Selection of a readily available cellulosic raw material, the processing of which should be economically and technically feasible; (2) Development of an effective pretreatment process which could significantly increase the subsequent rate of enzymatic hydrolysis and final product yield; (3) Development of cellulolytic enzymes for glucose production for optimum conversion; (4) Optimization of the glucose production process from cellulosic materials; (5) Development of an optimal enzyme reactor for the most efficient conversion of cellulose into glucose in terms of a continuity of operation, insoluble cellulose residues and enzyme recycling; (6) Designing and operation of a pilot plant for the enzymatic production of glucose. 24. In terms of their mechanism of action, cellulose-degrading enzymes can be subdivided into four groups regardless of the living organism from which they derive. These are one group of endoenzymes, two groups of exoenzymes and the fourth group consisting of cellobiases. "Endo" and "exo" are usually prefixed to the names of enzymes attacking polymer substrates in order to indicate whether they are acting on the internal or external portion of the polymer. For instance, if an enzyme attacks the chemical bonds far removed from the ends of a long polymer molecule it has the prefix "endo". It is "exo" when a short end group is cleaved off. 25. Hence, the longer the molecules making up the substrate (i.e., the higher the degree of its polymerization), the less pronounced is the role of the exoenzyme in the initial stage of the reaction since there only two ends to a linear chain but many internal bonds. On the other hand, an endoenzyme is most active when the conversion of the polymer begins and less so as the substrate molecules become shorter. That is why the degradation of polymer substrates in nature usually involves multi-enzyme complexes consisting of both "endo" and "exo", enzymes. 26. It can be seen that the enzymes of endo-wise mode of action begin the conversion process since the cellulose molecule consists of several thousand monomer units and the number of end glucose units in the native polymer are too few (compared to the number of intermediate glucosyl bonds in the cellulose) for the action of the exoenzymes to be noticeable initially. However, each successful attack of the endoglucanases breaks the polymer chain and forms two new ends in the shorter cellulose molecule, which can now be attacked by exoenzymes. Thus, the role of exoenzymes and the speed of their action increases as the endoenzymes degrade cellulose. 27. The exoenzymes acting on the partially split cellulose are of two types. One cleaves off glucose, the end-product of the hydrolysis, while the other, because of the specific structure of its active site, cleaves off cellobiose which is a dimer (a coupled glucose molecule). The first exoenzyme is an exoglucohydrolase, the second, exocellobiohydrolase. Recently experimental evidence has been obtained that some exoglucohydrolases in fact are endoglucanases, having a high transglucosidase activity. They attack long cellulose chains far from the ends, but are recycled from the reaction (as enzymes normally do) with the assistance not of water, as is usual for hydrolytic reactions, but apparently with the help of cellobiose, a final product of cellulose hydrolysis by endoglucanases. The resulting product of the "transglucosylation" reaction catalyzed by the endoglucanase is glucose. 28. Finally, the cellobioses are cleaved into two glucose molecules by the last of the enzymes in the cellulase complex - the cellobiases, i.e. - or are converted into glucose by some endoglucanases by means of transglycosylation, as described in the previous paragraph. More detailed descriptions of the dynamics of the enzymatic process are given in Volume II. 29. The reaction rate depends on the composition and quantity of the enzymes added to the cellulose, the state of the initial substrate (degree of its polymerization, degree of crystallinity, etc.), the quantity of native cellulose, the amount and the nature of constituents usually present in the native cellulose-containing material, the condition of the reaction, etc. 30. The formation of glucose and cellobiose during the course of the multienzymatic hydrolysis takes place at a regular rate when pure amorphous or crystalline cellulose is the starting substrate. Curves reflecting the yield or conversion at any given moment of the starting material to glucose and cellobiose have, therefore, been worked out along with the rate of accumulation of products. However, methods have not yet been developed to predict the dynamics of hydrolysis of heterogenous cellulosic material in an enzyme reactor. 31. There is another aspect to be considered. As noted above, in nature cellulose is mostly crystalline, and is rather resistant to all hydrolytic agents, including enzymes. To make crystalline cellulose amorphous and, therefore, more reactive toward cellulases, the very rigid, regular structure of its polymer chains must be broken. This can be done by subjecting the native materials to intensive milling, to treatment with phosphoric acid, or by dissolving it in special solutions. All these methods are technologically feasible. 32. But before deciding on the method, it is desirable to determine how enzymatic hydrolysis is affected by the native material's crystalline structure. The problem can be defined more comprehensively: what is the impact of the main structural characteristics of cellulose (i.e., specific surface, average particle size, degree of polymerization, and crystallinity) upon the rate of enzymatic hydrolysis. 33. This problem was resolved recently by a series of R&D projects in the U.S., Germany, Italy, and the former Soviet Union. Several dozen samples of cellulose, including native cellulose, were chosen for the experiments. All the pertinent structural parameters and the rate of enzymatic hydrolysis were registered and compared. Since the samples were of various origins, their structural factors also differ greatly. Thus, cotton was found to have the highest degree of polymerization - several thousand units - while that of cotton ground in a vibratory mill was lower - several hundred units (depending on the duration of treatment and types of mills). Cellulose samples treated with cobalt-60 isotopes or by electron accelerators had the lowest degree of polymerization - up to 20. However, as it turned out, samples of identical degrees of polymerization displayed different rates of hydrolysis. In other words, other structural factors were more important than polymerization in determining the rate of the enzymatic reaction. 34. It was also found that the average size of the particles (all other factors being the same) of the initial cellulose was not crucial in enzymatic hydrolysis. Cellulose, amorphous cellulose in particular, is a porous material with a developed, sponge-like internal surface. Thus, it is not the average size of the particles that normally determines its accessibility to enzymes, neither is it crucial to the efficiency of hydrolysis. 35. The accessible surface area is the main factor in determining the efficiency of the enzymatic attack on insoluble cellulose. This has been demonstrated by the linear relationship between the rate of hydrolysis and specific surface (square meters of surface per gram of material). It is not, however, the only factor. The rate of enzymatic hydrolysis is inversely proportional to the degree of crystallinity (which can be determined readily by X-ray diffractometric analysis). 36. There is no surprise in the observed relationship between the degree of the crystallinity of cellulose and its specific surface area, since both affect the hydrolysis rate to the same extent. The reason is that the greater the crystallinity the more packed the polymer chains of cellulose and the less accessible the surface of the substrate. Conversely, when the crystalline structure is destroyed, the inner chains of the cellulose give way and its accessible surface increases. 37. Consequently, these data may be used to predict the rate of enzymatic hydrolysis for virtually any cellulose sample, which must not, however, contain many heterogeneous admixtures (lignin etc.). It is simply a matter of measuring the specific surface area of the cellulose or the degree of its crystallinity. That, in turn, permits a standardization of cellulose-containing materials in industry; an important accomplishment. Thus, another bridge has been built between the structure of a substrate and its capacity to react - and the practical problem of technology, i.e., the enzymatic production of glucose from cellulose. 38. This treatise emphasizes the problem of enzymatic conversion of cellulosics into sugars and of producing ethanol from the latter by microorganisms. The production of ethyl alcohol will be mainly considered in Part C, which deals primarily with scaling-up technologies for the enzymatic conversion of cellulose into fuels. C. RAW MATERIALS 39. Biomass in the form of agricultural and forest wastes accumulates every year in large quantities both in industrial and developing countries. This results in a deterioration of the environment and a loss of potentially valuable resources. Some of the potential biodegradable agricultural and agro-industrial cellulosic residues and their availability in several countries are listed below:
POTENTIAL AVAILABILITY OF SOME AGRICULTURAL
RESIDUES IN INDIA
ANNUAL PRODUCTION OF SOLID CELLULOSIC WASTES IN THE USA
POTENTIAL AVAILABILITY OF SOME AGRICULTURAL AND
AGRO-INDUSTRIAL
RESIDUES IN THE FORMER SOVIET UNION
40. If just a fraction of these materials were to be converted into sugars and alcohol as well as into gas and feed protein, a significant contribution could be made to the overall problem of resource recycling and conservation. However, before using virtually any of these cellulosics it is necessary to convert them so that they are more reactive toward cellulolytic enzymes. As indicated above (paragraphs 31-37), it is necessary to decrease their crystallinity and to increase specific surface area and, of course, to remove as much as possible the non-cellulosic matrix of the cellulose, which may otherwise sterically block the accessibility of cellulose to cellulase enzymes. D. PRE-TREATMENT 41. Natural cellulose is a crystalline polymer generally associated with hemicellulose and lignin in a matrix and is highly resistant to enzymatic attack. Therefore, pretreatment is necessary. Most approaches separate the different types of pretreatment into mechanical, chemical, physical (other than mechanical), biological, or a combination of these methods: For all these methods it is assumed that the cellulosic material is already in a form which can be processed readily. 42. Most pre-treatment processes have energy requirements that depend on the severity of the process. Severe mechanical and/or thermochemical processes have substantial energy requirements and when combined with those for product separation, can make some cellulosics bioconversion processes very inefficient. A special analysis performed recently by Datta, Exxon Research and Engineering Company, Linden, New Jersey, has shown that many of the proposed processes which use fine particles, severe conditions, etc., can consume a substantial amount of energy and lead to a generally inefficient process in terms of energy. Conversion processes that can use coarse particles, mild pretreatments, non-sterile conditions, etc., will have significant advantages in this respect. 43. Mechanical pretreatments (see paragraph 41) utilize shearing and impacting forces to yield a fine substrate possessing a low crystallinity index and high specific surface area, thus enhancing its susceptibility to enzyme action. Typically, mechanical pulp (stone groundwood pulp) is made by grinding logs or blocks of wood against a revolving abrasive stone in the presence of water. Chip groundwood, or refiner mechanical pulp, is provided by feeding chips or sawdust between a set of rotating, ridged plates (or disks) of a disk refiner. 44. An extension of the chip groundwood process is the thermomechanical process in which steam softens the chips prior to reduction in pressurized disk refiners. Mill and laboratory experiences regarding thermomechanical pulping at Smurfit Newsprint Corporation, Oregon, are described recently (Quick et al., 1991). Besides, among specially processed mechanical pulps steam defibrated and steam exploded pulps should be mentioned. Mechanical as well as thermomechanical pulps are practically identical in composition with wood.
45. The use of fine substrates yields higher slurry concentrations, or higher bulk density, thus reducing the reactor volume. Using milling as a pretreatment method has the advantage of being relatively substrate insensitive but suffers from the major disadvantage of being energetically unfavorable (see below) and rendering the remaining lignin accessible as a significant inhibitor of the enzyme.
46. The estimated cost of mechanical pretreatment of wheat straw may vary from $0.01/kg for Fitz milling (model D comminutor, Fitzpatrick Company, Elmhurst, Illinois) to $2.24/kg for roller milling (based on the weight of the substrate).
47. Of the mechanical pre-treatments shown above, ball milling gives the most promising results in terms of hydrolysis rate and sugar yield. This pre-treatment is clean and easy to operate, but the long pre-treatment time makes its large-scale operation impractical. 48. The main difficulty with size reduction of cellulose is the fibrous nature of the material. This is evident from the comparatively low energy requirements for the size reduction of a brittle material such as coal, which takes only 5-7 KwHr/ton to reduce the particle size to 100-200 mesh by means of simple pulverization. For this reason, the energy content of feed is as low as 0.2-0.3% compared with 25-100% for fibrous cellulose.
56. Of the chemical pretreatments shown above, caustic pretreatment is a potential candidate for large-scale process development based on pretreatment cost, hydrolysis rate, and sugar yield. Chemical pretreatments, however, have disadvantages which cannot be ignored. These include the required use of specialized corrosion-resistant equipment, the need for extensive washing, and the difficulty of disposing of chemical wastes. 57. Physical pretreatments of cellulosics involve primarily g-irradiation and high-pressure steaming, either with or without fast decompression. Actually, steaming with a sharp reduction in pressure couples physical and chemical pretreatments; it will be considered in the last section of this section (paragraphs 66-70). 58. Steaming of biomass in the 150-200°C temperature range leads to increased enzymatic digestibility as a result of increasing pore size and the partial hydrolysis of hemicelluloses. The residence time at higher temperatures should be kept low to minimize reactions which produce by-products of a noncarbohydrate nature. According to some data and calculations, to reduce biomass containing 50% solids to a product containing 25-40% solids by steam treatment at 150-200°C, approximately 450-1300 kg of steam would be required per ton of dry biomass. This method has not yet proven effective in greatly increasing enzymatic hydrolysis. 59. γ-Irradiation was not effective as a pre-treatment of relatively pure cellulose (see below), but it slightly accelerated (up to 1.5-3 fold) the enzymatic hydrolysis of lignocellulose substances such as wheat and barley straw and bagasse. Other mechanical or chemical pretreatments, on the other hand, accelerated the conversion of bagasse 7- to 10-fold.
70. Steam explosion has proven to be one of the most energy efficient methods of pretreating wood substrates as well as one of the most efficient methods for enhancing subsequent enzymatic hydrolysis of hardwood species.
74. The most widely used substrates for the determination of total cellulolytic activity include filter paper, microcrystalline cellulose, cotton fibers and soluble carboxymethyl cellulose. Each method usually involves measuring the reducing sugars formed in the course of the enzymatic hydrolysis of the substrate. Filter paper hydrolysis, the so-called Mandels-Weber method, has been generally accepted for this purpose. This method determines cellulase activity in units of micromoles of reducing sugar (measured as glucose) liberated per one minute under standard assay conditions (filter paper units, FPU, or international units, IU). 75. For the last 30 years Trichoderma viride (= Trichoderma reesei) is known as the best source of "complete cellulase", i.e., cellulase which contains all the components necessary for the hydrolysis of crystalline cellulose. Using this microorganism in conjunction with the optimization of biosynthesis and mutation techniques, investigators at the U.S. Army Natick Research and Development Command, Massachusetts, and at Rutgers University, New Jersey have greatly increased enzyme yields and titers (in terms of total cellulase activity). The highest titer obtained to date, 14.8 I.U./ml, utilized fermentation of compression-milled cotton with the NG14 (Rutgers) mutant. Maximum productivity of 167 I.U./l/hr was recorded at Natick through the use of newer mutants. Recently it was reported that a commercial cellulase "Laminex" (T. reesei strain) manufactured by Genencor (San-Francisco based company) has a total activity of 84 IFPU/ml (International Filter Paper Units) (Philippidis et al., 1993). 76. With regard to cellobiase activity in a culture broth, 16.7 I.U./ml of cellobiase from Aspergillus phoenicis (QM 329) was produced at Natick with a maximum productivity of 135 I.U./l/hr over 119 hours, in a prepilot plant fermentation using a chemical pulp hydrolysis syrup. Commercial cellulase "Laminex" (see above) has beta-glucosidase activity of 91 I.U./ml (Philippidis et al., 1993). 77. To render these figures more meaningful, it should be noted that to produce 11% sugar syrup in 24 hours from compression-milled newspaper, the charge to the hydrolysis vessel would be 25% of cellulose containing the Trichoderma at an enzyme/substrate ratio of 10 I.U./gm. In this case the level of reducing sugars is high enough to permit direct practical fermentation to ethanol without an additional sugar concentration step. The production of glucose in the enzymatic hydrolysis of ball milled newspaper increases by 25-35% when cellobiase from Asp. phoenicis is added to the cellulase (T. reesei) broth at up to 3 I.U./ml.
|
95. Cellulase complexes from various sources differ substantially in their composition in relation to the individual cellulolytic enzyme components present (see paragraphs 71 and 72). Consequently, the dynamics and yields of glucose formation in the course of the enzymatic hydrolysis of cellulose vary for different cellulase preparations. If a complex is deficient in cellobiase, for example, as is the case for most Trichoderma reesei preparations, cellobiose accumulates in the reaction mixture at the expense of the end-product of the hydrolysis, i.e., glucose. It was reported for some T. reesei preparations that, after a reasonable time of enzymatic hydrolysis, roughly 80% of the soluble reaction product consisted of cellobiose. It was thus recommended that the cellulase broth be supplemented with cellobiase from other sources in order to increase glucose yield and decrease cellobiose inhibition of cellulolytic enzymes. A substantial increase in yield of glucose from the hydrolysis of cellulose can therefore be realized.
96. By using T. reesei cellulase with filter paper activity of 2-5 I.U./mi with a pretreated substrate present in concentrations of 5-25% by weight, a total sugar concentration of between 4-10% is obtained in 24 hours. In this case, the glucose production level is high enough to permit direct practical fermentation of ethanol without an additional sugar concentration step.
97. It is difficult, and at present impractical, to raise the total sugar concentration in the Trichoderma hydrolysates above 6-8% since cellulases from this source are inhibited by the reactor products, i.e. glucose and cellobiose. Thus, the rate of hydrolysis by T. longibrachiatum cellulase is halved in the presence of 1.1% glucose or 0.8% cellobiose, when the amount of cellulose in the reaction mixture is small. With an increase of the substrate concentration the critical level of products also increases although it hardly exceeds 6-10%, since beyond this the inhibition of the process is substantial. This leads to a lower efficiency of the enzymatic hydrolysis (in terms of g/l/hr) as a result of a higher sugar concentration in the reaction mixture. For example, during alkali pretreated cotton stalks hydrolysis in the continuous type reactor, hydrolytic efficiency is equal to 5 g/l/hr for a steady state level of glucose of 1.8%, but drops to 2.5 g/l/hr after the flow through the reactor is reduced in order to increase the glucose concentration at the outlet to 5.5%. A further increase of glucose content in the resultant syrups leads to an additional fall in the reactor productivity. Clearly, more research is required to find a cellulase-producing strain which is less susceptible to product inhibition. Otherwise a cost evaluation must be made between glucose production and the additional sugar concentration step.
98. During the hydrolysis process, an enzyme can be lost in any of three ways. The enzyme can be adsorbed strongly onto unhydrolyzed cellulose; some of it may remain in the solution containing glucose; and some may be inactivated during the course of the hydrolysis (as a result of the heat or of the shearing effects produced when an agitating reactor is used).
99. Methods of enzyme desorption from cellulose, such as those using phosphate ion gradients to adjust the pH to neutrality, may permit greater enzyme recovery. Fresh solids can be brought into contact with the hydrolysate to adsorb some of the enzyme remaining in the solution. However, scientists at the Natick Laboratories found that this method of enzyme recovery may not be economical. Nevertheless, other investigators believe that methods must be found for each particular substrate in order to achieve a satisfactory balance of activities in the blend of fresh and recycled cellulase.
100. Another approach to enzyme recycling is the use of adsorbed cellulase in the continuous conversion of cellulose into glucose in a column reactor. This technology has been scaled up in Russia. As cellulose is digested, the released enzyme is readsorbed on excess or newly added cellulose and retains activity. Adsorption results in a marked economy of enzyme reutilization in the continuous process. The substrate can be retained in the column reactor until a high-percent conversion (96% in the case of pretreated cotton stalks hydrolysis, or nearly 100% for paper-making wastes) has been achieved. The sugars are recovered in a clear aqueous solution free of enzyme, cellulose, and any insoluble impurities. The adsorbed enzyme is sufficiently active to digest cellulose for a relatively long time; only requiring replenishment once every 1-2 months in the form of a commercially available culture fluid.
101. Economic evaluations of a number of specific scale-up technologies for producing sugars and ethanol from cellulose are provided in Vol. II. The economic assessment performed at the University of California, Berkeley, is provided as an example. In light of the discussion below as well as general considerations pertaining to economic issues, some reference costs should be identified (see following table).