\“v' I ‘ ‘ J \ 5:36!!!" OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY MATERIALS: Place in book return to remove charge from circulation records - E ii", 13"???” g A94 n.4, ‘ 1 \j" ‘ w . \_.‘ if, 9 O M a a 299:3 CHEMICALS FROM WOOD VIA HYDROGEN FLUORIDE By Susan Selke A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of NmSTER OF SCIENCE Department of Chemical Engineering 1981 / G 1/ 5’5 ABSTRACT CHEMICALS FROM WOOD VIA HIDROCEN FLUORIDE By Susan Selke We have produced evidence that HP saccharification of ligno- cellulosic materials when compared with hydrolysis saccharification techniques has the important features of (1) high conversion of cellulose and hemicellulose to simple sugars, (2) recycle of HF in the process thereby eliminating the need for major acid neutralization requirements, (3) a highly reactive lignin for chemical use and (4) no elaborate pretreatment requirements. Extensive sugar repolymerization can occur during fluoride removal, but post-hydrolysis for 1 hour at 140°C with 50 mM sulfuric acid converted 95% of sugar oligomers to monomeric sugarS. After neutralization with calcium carbonate and filtration, these sugars were fermented with yeast, yielding ethanol as the major distillation product. Several process flow alternatives are suggested, and laboratory investigations required for further process development are identified. To Barry, Lori, and Erik ii ACKNOWLEDGEMENTS The author gratefully acknowledges the support and encouragement of Dr. Martin Hawley, and also extends grateful appreciation to the other members of the "HF Project," Dr. Derek T. L. Lamport, Gary Smith, Haim Hardt, Rick Chapman, and Jim Smith. iii TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES INTRODUCTION THE NEED FOR CHEMICALS FROM WOOD BIOMASS SACCHARIFICATION PROCESSES HF SACCHARIFICATION PROCESS DESIGN PRELIMINARY ECONOMIC EVALUATION CONCLUSIONS REFERENCES APPENDIX - Survey of Saccharification Technologies Historical Background Concentrated Acid Processes Sulfuric Acid HCl - Liquid HCl - Gas Dilute Acid Processes - Sulfuric Acid Enzymatic Processes Enzymatic Saccharification Simultaneous Saccharification and Fermentation Miscellaneous Processes iv Page vi vii 15 22 25 26 27 27 30 3O 37 1+2 1+5 55 55 63 68 Page Economics 70 Summary and Comparison of Hydrolysis Processes 73 HF Saccharification of Wood 83 References 87 BIBLIOGRAPHY 90 LIST OF TABLES Comparison of Characteristics of Typical Lignocellulosic Hydrolysis Technologies Sugar Yields From Wood Chips Sugar Yields From Filter Paper Comparison of Chemical and Feedstock Cost Summary of Representative Concentrated Acid Hydrolysis Processes Summary of Representative Dilute Acid Hydrolysis Processes Summary of Representative Enzymatic Hydrolysis Processes vi Page IO 12 2A 7A 76 79 12. 13. 11+. 15. 16. 17. 18. 19. 20. LIST OF FIGURES HF-Saccharification Fluoride Retention as a Function of Evacuation Time Process Flow Diagram: Liquid Phase Solvolysis with Immediate Lignin Removal Process Flow Diagram: Liquid Phase Solvolysis with Lignin Removal After HF Removal or Post-Hydrolysis Process Flow Diagram: Gas Phase Solvolysis with Lignin Removal After HF Removal or Post-Hydrolysis Hokkaido Process Nippon Mokuzai Kagaku Process Northern Regional Research Laboratory University of Missouri Process Bergius-Rheinau Process Udic-Rheinau Process Noguchi-Chisso Process Scholler Process Madison Process TVA Process Brenner Process Georgia Tech Process Natick Process Wilke Process - Newsprint Wilke Process - Corn Stover vii Page 13 16 18 19 32 3h 36 38 39 #1 1+1. 1+7 1+9 50 52 52+ 57 59 6O 21. 22. 23. 2%. Indian Institute of Technology General Electric - University of Pennsylvania Gulf Process Hoch and Bohunek Process viii Page 62 6h 67 86 INTRODUCTION In September 1979 the MSU/DOE Plant Research Laboratory and the Department of Chemical Engineering at Michigan State University began a c00perative investigation into the use of hydrogen fluoride for the saccharification of lignocellulosic materials. HF saccharification of biomass was originally investigated in Germany during the 1930'3. However after World war II this research was virtually forgotten. The advent of cheap ethanol from ethylene meant alcohol from biomass was no longer economical. Recent rapid increases in oil prices and the certainty of a dwindling oil supply have now led to a resurgence of interest in fuels from renewable sources. HF had been used in Dr. Derek T. A. Lamport's laboratory as an agent for deglycosylating cell wall proteins since 1975. The idea of using HF for cellulose hydrolysis arose, and a search of the litera- ture uncovered the old German work, indicating that this was indeed a potentially useful reaction. This thesis is a result of some of our work in reinvestigating HF saccharification of lignocellulosic materials and comparing it to alternative hydrolysis technologies. The appendix contains additional information about alternative processes for the hydrolysis of biomass. THE NEED FOR CHEMICALS FROM WOOD we are all aware that petroleum is a finite resource. As sup- plies of oil dwindle and prices increase, it will become increasingly imperative to find alternative sources for the energy and chemicals we now obtain from petroleum. Combinations of nuclear, solar, hydro- electric and other energy sources may be able to substitute for most of the petroleum-derived energy used in our society, but it will be difficult or impossible to substitute such energy sources for two major areas of petroleum use - the production of liquid fuels and of raw materials for use in the chemical industry. Electricity can substitute for liquid fuels in some applications, but in agriculture, aviation, the military and similar situations, liquid fuels are likely to continue to be in demand. Coal has the potential of meeting some of these needs if the appropriate new technologies are successfully deveIOped, but coal is also a limited and non-renewable resource. Even when environmental considerations are neglected, as more demands are put on the world's coal supply, we accelerate the rate of coal depletion, and come closer to the inevitable time period when coal too will be a scarce resource. Biomass also has the potential to replace petroleum as a source of liquid fuels and chemical raw materials, and further, the use of biomass does not encounter this problem of a fixed and non-replenishable resource base. As long as prOper management is practiced, our supply of biomass is quickly replenished. In fact, proper management can significantly increase the rate of biomass production. However, our supply of biomass, though replenishable, is limited. A recent report by the U.S. Congress Office of Technology'Assessmentl estimates that with "high development" biomass could supply 12-17 Quads of energy per year by the year 2000, or 15-20% of current U.S. energy consump- tion. It is obvious, then, that other energy sources such as solar, nuclear, etc. must also be developed to replace petroleum and even- tually to replace coal as well. However, these alternative renewable energy sources cannot serve as chemical raw materials, and are not likely to totally replace liquid fuels. Therefore the obvious conclusion is that biomass will eventually become so valuable for its chemical and liquid fuel potential that it will be too valuable to burn as solid fuel. The implication for deveIOpment of biomass conversion technologies is twofold. First, the yield of sugars from the cellulosic portion of the biomass must be high. The sugars can be considered as chemical feedstocks or can be fermented to ethanol, which can be used as a liquid fuel or further converted to ethylene or butadiene for use as chemical raw materials. These conversion processes have already been commercially demonstrated. Second, the lignin fraction of the biomass must be recovered in a reactive form. It can then be used to produce phenol or a variety of other aromatic compounds by methods which have been shown to be feasible, but have not yet been refined due to lack of economic incen- tive in the past when oil was cheap. As an indication of the importance of these compounds, it should be noted that 95% of all synthetic polymers are derived from only three raw materials: ethylene, butadiene, and phenol.2 The process we are develOping at the MSU/DOE Plant Research Laboratory in cooperation with the Department of Chemical Engineering at Michigan State University - HF saccharification of wood — will meet these two requirements for effective biomass conversion: high yield of sugars, and a reactive lignin byproduct. Figure l is a schematic of the HF saccharification process. ZOHB¢mm mo Mooammmmh A¢UH2mmo (0") BIOMASS SACCHARIFICATION PROCESSES The ability of HF to dissolve cellulose was first noted in 1869 by J. Gore.3 In 1929 Helferich and BottgerLL investigated this reac- tion and determined that the products were polyglucans which could be converted to glucose by boiling with dilute acid. About 1933 Fredenhagen and Cadenbach5 discovered that the reaction mechanism involved the formation of glucosyl fluorides which, in HF solution, react with even small amounts of water to form glucose and regenerate HF. The glucose monomers can recombine to form polyglucans on pre- cipitation or evaporation of HF. Fredenhagen and Cadenbach went on to investigate the action of HF on wood, as did Hoch and Bohunek6’7 around 1937. Several patents were issued as a result of these inves- tigationS. A pilot plant was constructed in Germany and ran for six months using Hoch and Bohunek’s process for wood saccharification with gaseous HF. It appears that WOrld War II interfered with further develOpment of the process. With the exception of some investigation in the Soviet Union in the late 1950's, it appears that HF saccharifi- cation of cellulose was ignored until very recently. Alternative processes for saccharifying wood have been known for many years — in fact the first experimental plant for wood hydrolysis was constructed in 1894 and used dilute sulfuric acid.8 Germany had a substantial wood hydrolysis industry during World war II, and the Soviet Union has a number of wood hydrolysis plants in Operation today. These processes fall into three major categories - dilute acid, concentrated acid, and enzymatic. The existing Soviet plants and the German WOrld war II plants were based on dilute sulfuric acid. Con- centrated sulfuric acid was used in a 1963 Japanese plant, and concen- trated hydrochloric acid in a 1960 German plant. HCl gas was used on a pilot plant scale in Japan in 1958-1962.9 Current research on acid hydrolysis is directed primarily to dilute or concentrated sulfuric acid. The main thrust of current research in the U.S., however, seems to be directed at enzymatic hydrolysis, and various pretreatments attempted to increase yields. The main difficulties encountered are low sugar yields, pretreatment requirements, and high enzyme and/or chemical requirements, which render the processes uneconomical. In addition, the only processes with high yields - those utilizing concentrated acid - produce highly condensed non-reactive lignin. Thus none of these processes meet the requirements for effective biomass conversion - high sugar yield and reactive lignin - outlined above. Table l is a comparison of the characteristics of typical hydrolysis technologies. A further extremely important advantage of an HF saccharification process is that HF, which boils at 19.54°C and has a heat of vaporiza- tion of less than 1800 cal/mol, can be easily removed using low tem- perature heat to evaporate the HF from the reaction products for recycling. In addition, the only pretreatment of the wood that is required is drying. TABLE 1 COMPARISON OF CHARACTERISTICS OF TYPICAL LIGNOCELLULOSIC HYDROLYSIS TECHNOLOGIES Dilute Acid Conc. Acid Enzymatic HF Glucose Yield, % 50 85—90 50 85-95 Acid Consumption medium high none very low Lignin low- Reactivity low medium high high Enzyme & Chemical moderate high high low Requirements Status of commercial pilot pilot laboratory Technology in Soviet Union HF SACCHARIFICATION We have treated small chips of Bigtooth Aspen (Populus grandiden- papa) with liquid hydrogen fluoride in a Kel-F vacuum distillation apparatus. In typical experiments we treated I g samples of wood with 10 ml HF for one hour and then removed the HF by evacuation into a calcium oxide trap. Water was then added, yielding a water-soluble sugar fraction, and an insoluble lignin fraction, which were separated by centrifugation. Sugar recoveries were determined by preparing trimethylsilyl derivatives of the methyl glycosides and analysing them by gas-liquid chromatography. Fluoride retention in the sugar fraction was determined with a fluoride electrode. Fluoride in the lignin fraction was similarly determined after degrading the lignin via alkali fusion. Sugar yields have been found to be influenced by reaction tem- perature and amount of water present in the reaction system. Kinetics experiments are underway to clarify the time-temperature dependence. There appears to be an optimal water content in the range of HF containing 10% water. Sugar yields have ranged from A5-99%, depending on reaction conditions. In our early experiments at 0°C, sugar yields were calculated based on a single GC analysis. we have since deter- mined that one analysis is not sufficient. Therefore in Table 2 we report these yields in summary form indicating the ranges obtained. we have only recently begun the investigation of the room temperature 10 TABLE 2 SUGAR YIELDS FROM WOOD CHIPS* (l g samples) Temperature water Xylose Glucose Total Sugars Content (g) (g) (g) Yield 0 O .lu—.16 .uu-.5o .62-.65 70-74% 0 5 .ll-.l6 .47-.u9 .60-.75 68-85% 0 lO .13-.17 .50-.58 .67-.76 76-86% 0 25 .1u-.15 .27-.36 .uO-.M9 A5-56% 23 o .16 .uo .64 73% 23 h .17 .5A .73 83% 23 6 .15 .I5 .62 70% 23 9 .2u .62 .88 99.8% 23 ll .14 .h? .63 72% ll reaction, so these results are preliminary. we have also investigated the reaction of HF with filter paper. These results are summarized in Table 3. Near quantitative sugar yields were obtained from.HF-filter paper reactions at room tempera- ture. Yields at 0°C were slightly depressed. Fluoride retention, especially in the sugar fraction, has been found to depend on evacuation time, temperature, and water content. Considering only evacuation temperature, with time held constant at - one hour, the fluoride retention decreased exponentially with increasing temperature until reaching a limiting value. With tem- perature held constant at 100°C, fluoride retention decreased as time increased from 0.5 to 2 hours. Extending the evacuation time to 5 hours did not significantly improve the fluoride removal. With time and temperature constant, addition of water decreased fluoride retention up to 5%'water by volume. Further addition of water did not improve fluoride removal. The results are summarized in Figure 2. It should be noted that these evacuation times are a function of the reaction configuration in the laboratory, and would not apply to a commercial process. The lowest fluoride content achieved was A mg/g wood for the sugar fraction, and 0.1 mg/g wood for the lignin fraction. we believe most of this results from the formation of fluorides due to metals in the wood ash. The ash content of our'wood was 7 mg/g wood. During the process of HF removal, a large percentage of the sugar monomers recombine, forming oligomers. This process of reversion is favored by high temperature evacuation and the absence of water. 12 TABLE 3 SUGAR YIELDS FROM FILTER PAPER Temperature Water Content Glucose Recovery (00) (% of HF) 0 O 82.91 1 3.59 0 2.6 91.57 i 0.91 0 6.0 94.05 t 3.67 23 0 93.82 t 5.91 23 3.5 98.21 1 3.07 23 3.5 102.u7 t 3.15 23 7.7 92.21.L 1' 3.11.3 13 50 -\_ FLUORIDE RETENTION (mg F-/g wood) O \ 20 _ '\ "i \\ \O.\\ . \ h ‘ '\\ “\\\ lO ‘- {Xe- p V“ —«~:— -___"““ ~~~ .O_ _ . 0 ‘ ‘ ‘ 1 O l 2 3 4 EVACUATION TIME (hrs) Notes: Evacuation was at 100°C. x Anhydrous HF 0 HF containing 5% water * HF containing 10% water FIGURE 2 FLUORIDE RETENTION AS A FUNCTION OF EVACUATION TIME 11+ When the HF was removed by neutralization with calcium carbonate, 62-74% monomeric sugars were obtained. When HF was removed by eva- cuation, only 10-20% monomers were obtained. we therefore eXplored the conditions necessary for post-hydrolysis of the reversion products. We determined that over 90% of the sugars could be converted to monomers by post-hydrolysis for one hour at lHO°C, either with 50 mM sulfuric acid or with dilute hydrofluoric acid. We have begun to investigate the reactivity of the lignin produced by HF saccharification. The fluoride retention in the lignin fraction was reduced to 0.03 mg/g wood by crushing, washing and dialysing the lignin residue. Thus there is no significant degree of fluorination of the lignin. we have not seen any evidence of condensation reac- tions in the lignin. An experiment with dioxane acidolysis rendered the lignin 93% soluble, indicating that it is essentially noncondensed. Further investigation is in progress. we therefore have reason to believe that the lignin retains a high degree of functionality and thus will be valuable as a raw material for the production of aromatic hydrocarbons. we have also demonstrated that these wood sugars can be suc- cessfully fermented by the yeast Saccharomyces cerevisiae, yielding ethanol as the major distillation product. The yeast can grow and ferment wood sugars at fluoride concentrations as high as 100 ppm. PROCESS DESIGN There are several feasible process flow schemes for HF sacchari- fication of wood. They differ primarily in reaction phase - gas versus liquid, the point of lignin removal, and in the HF recycle scheme. Several of these variations are presented in Figures 3-5. All the processing schemes presented below are based on wood as the feedstock, with lignin and an aqueous solution of mixed monomeric sugars as the major products. Temperature and pressure are near ambient except where indicated. Figures 3 and 4 represent liquid phase reaction systems, and Figure 5 a gas phase reaction system. In Figure 3 the lignin is removed directly after the wood saccharifi- cation step, an Option not feasible for a gas-phase reaction. Figures 4 and 5 each indicate the Options of lignin removal following HF removal or following post-hydrolysis. Figure 3 uses wood as the feedstock in a reaction with liquid HF, providing for immediate lignin removal after the reaction. The wood is dried to an acceptable moisture content and then treated with liquid HF in the solvolysis reactor. Lignin is then removed, probably by filtration. Next the HF is removed from the liquid HF-sugar solu- tion, possibly by spray-drying. The HF, which will be slightly wet, is distilled to yield anhydrous HF and the HF-H20 azeotrope (35.6% HF). The wash water from the lignin washing, which will contain sugars and some HF, is added to the dried sugars along with the HF-water l5 l6 mopouasm ll A¢>Ozmm ZHZUHA HB¢HQHSSH mEHS mHmaHOEHom mm¢mm QHDUHA OPOPHQHOOH Omaha Roadhouomom sommm .odoo A.qdv mow hm mpOpfimfi00hm omnmmz MZ¢muoaom Ho>oaom -paom mm seawaq m - pos.aapnwaaa-mm wowo¢_+ o m mmOHPOmN¢ ommNHm soaetaaapman H L paaaaav Hovommm mm Ammonm — K Aeaaaaav mm moonohnq¢ Hmhhm .amm QSIOMOE posesmm no ampoadom Romano @003 .II . o . .1 ~ . . .. . . r . 1 3,.I13..- .1, 7 . . . .. x . I$r|1tII8av>11 . . a . OI, . . . u 1 l . 1 .\ . o . o n 1 1 F All yY A I: ll ('1‘ . . I I n . a .. . . ! i a a . . 'n . I . . . I? . . x. : l . 1 1 r . l 1 I n .l . 1 , _ 1 c. _ . _ _ l7 azeotrOpe from the distillation. Additional water and/or acid are added if necessary. The solution is then post—hydrolyzed at about 140°C to decompose sugar oligomers, resulting from reversion, into their constituent monomers. Next a base is added to neutralize the acid, and the precipitate removed, yielding a product of mixed sugars in aqueous solution. The fluoride precipitate is washed, and the wash water, which will contain sugars, added to the product solution. Next the precipitate is dried, and then it is treated with concen- trated sulfuric acid to yield anhydrous HF and a sulfate waste product. The lignin, after washing, is dried to yield the lignin byproduct. The anhydrous HF from the distillation and from the regneration are combined with a small amount of make-up HF and recycled to the reactor. Figure 4 is similar to Figure 3, also being a liquid phase solvolysis, but shows the Options of lignin removal following HF removal or following post-hydrolysis. In these Options, the initial steps are the same. Following the reaction, HF is removed, possibly by flash evaporation. Next, in Option (1), lignin is removed, by washing to dissolve sugars followed by filtration. Next the sugar solution is post-hydrolyzed at about lAOOC. In option (2), water is added and the post-hydrolysis completed. Then lignin is removed by filtration. In either case, the remaining steps, neutralization, distillation, lignin and precipitate washing, and HF regeneration are identical to those described for Figure 3. Figure 5 is a gas phase reaction with lignin removal after HF removal or after post-hydrolysis. An initial grinding step may be desirable to reduce the wood particle size. Vacuum drying may also l8 .mHquomnNmuamom mo Ad>ozmm mm mmam<.qdfiozmm 2szHA mHH3 mHmMHO>Hom mm¢mm medHQ a E¢mw¢Hm.30Hh mmmoomm : mchHh . mopmwdsm _ 1lemma novmnoqowom .oqoo mom ofldvfia ovmpflmfiomnm OOHHQ mm msOHOHQQ¢ * nmmdooqoo. mm ozono ng< Roman posoonmih Ni Adam campfimflooam dogma: II mm mm T _ Q as 3 0mm seamen coadeOm m u_a . mdoosum 5” . OOHPONHH W3 o weanmmz ,Tll. Awndwdw ooxfiz umhezwz nfiqwfiq 0mm owcmnoaopafi .hw_|dmwmmm.l.mw m mfimhaonean . Hm>oaom Hd>oamm owmmmmv D -paom saawaq am hoeoeom am y 7 m #03 . NUHUfi + O m uhHPHHwHHm lg F aw . Omahpomu<_o mwwm soaeoaaapaan nmaauaflp. hohhn pmdozmm no ampmdaom Romano ul hm QSOHOhAQ< hm gnuoxmz @003 . . . v .—u_._....a - . . O . . I . I -~ .. | . o c ' . D , ., l r" I . u —. . l I“ l9 mHquomQNmanom mo 443% mm E A<>Oeemm ZHZUHQ EH3 mHmHHokHom mag 20 “Ewan 309m mmmoomm pmsokmm .Ho ameoaaom Romano m mmDon pofiUOMth A I|+u mm m . seamed tease p Mann , :ommm Roman “Chan Hovoaoaowom .onoo Illllla we «pass a a 3 + Honda coapmo o m , mom IIIII l I I I quBHOm 9A6 w muoosum ma mm up nq< mammaonomm Hm>oaom oofipoufid m ooxfiz upmom nflnwfiq unannoz auuh.u.u..hw 7 owgnunopqfl “qoapmo 06.6% . _ Hd>oaom _ HO>QEOM mm mm _ mHMm “OOHpmo _ l ._ , mathOHOhm_rL amboaom “omega moov A~85508>v . a OoHHosHm hOPoOOm , wdfihn wdfienfia ammoosau mm mfimhdo>aom mm a u mmmwuboade—a o m ocm\mm soapnaaapaam a. Roomy mm muondhnq< L am as-oatz poor e 1 - Vi . l- 1 o. .1 4 I . if r 1 a l . l s . I 1 o it :1!!! . . rlv'i . . - t - Ill-I‘ll! . I . I .. - ll"- 1.1,..4- .__-..V .t.:.. llllnl’.‘ 20 be desirable, as evacuation of the wood has been reported to increase the rate Of absorption of HF.7 The dried wood is mixed with gaseous HF in the reactor. Next HF is removed, probably by evacuation. water is then added to hydrolyze the glucosyl fluorides, releasing HF. In one Option, this HF is then removed, distilled to produce anhydrous HF and HF/H2O azeotrope, and the anhydrous HF recycled. Otherwise this step is skipped. The choice will be determined by the amount of HF present after the glucosyl fluoride hydrolysis. If it is large, the additional HF removal step will be necessary. Next, in one Option water, the HF/H20 azeotrope (if produced), and addi- tional acid if necessary are added to the sugars and the solution post-hydrolyzed at about 140°C. Next lignin is removed by filtration. In the other Option, water is added to dissolve the sugars and the lignin removed. Next the azeotrope, if produced, and additional water and/or acid, if needed, are added and the solution post-hydro- lyzed. In either case, neutralization, precipitate removal, and HF regeneration are done as in Figure 3. If lignin is removed prior to post-hydrolysis, the lignin wash water is added to the post-hydrolysis step; if after, to the neutralization step. Early removal of the lignin has two advantages. The lignin will be exposed to an acidic environment for a shorter period of time, and thus should undergo less degradation, making it a more valuable product. Also, removal of the lignin at this point means that streams for further processing will be liquid, rather than liquid-solid suspensions or slurries, which should make processing easier. Alternatively, separating the lignin after post-hydrolysis may lead to more complete 2l sugar recovery, as the oligomers created by reversion would be conver- ted tO monomers, which should be more easily solubilized. In the preceding diagrams, the extractives have been assumed to stay with the lignin. The ash will probably react with HF forming fluoride salts, the insoluble ones exiting with the lignin and the soluble ones with the sugars. Thus they will result in a small HF loss, unless this can be prevented. One possibility is that addition of a small amount of sulfuric acid may lead to the formation of sulfates in preference to fluorides, reducing the HF loss. Other process options are the use of sulfuric acid rather than HF in the post-hydrolysis stage, and alternative HF recovery-recycle schemes. Anhydrous HF can be recovered from the HF/H20 azeotrope by solvent extraction, precipitation with alkali fluorides and regenera- tion, neutralization and regeneration, or other schemes. These can also be adapted to fluoride removal from dilute HF solutions. PRELIMINARY ECONOMIC EVALUATION A full evaluation of the economics of HF saccharification of wood must await further specification of the design. However we have done a preliminary comparison of costs with some alternative hydrolysis technologies. In this analysis we assume a liquid phase reaction similar to Figures 3 or A, with an HF/wood ratio of 1.2:1 and a water content of 9% of the HF. An HF loss of A mg/g wood (8 lb/ton) is assumed. Costs used are $AO/dry ton for wood chips, $ .60/lb for HF, $20/ton for calcium carbonate, and $75/ton for sulfuric acid. A glucose yield of 90% is assumed. No byproduct credits are taken, and all costs are charged to the glucose. With these assumptions, the total feedstock and chemical cost per pound of glucose is estimated to be 5.4 cents. Of this, 3.4% is due to the cost of the wood. A similar analysis was performed for Grethlein's acid hydrolysis process, with the substitution of wood for newsprint as the feedstock. For his optimistic design with a 30% slurry, cost of feedstock and chemicals would be 5.5¢/lb of glucose. For what he considered a more realizable case with a 10% slurry, cost would be 6.2é/lb. In both cases, approximately 5.3% is due to the feedstock cost.10 The cost of feedstock and chemicals for Wilke's enzymatic hydrolysis design was also estimated, using the same assumption about feedstock. In this case Wilke's 1976 Chemical cost was 22 23 0.8356/lb of glucose.ll Feedstock cost would be an additional 8.06/lb, for a total cost, updated to 1981, of 9.2¢/1b. These costs are summarized in Table A. It can be easily seen that a major variable in the sugar cost is the cost Of the feedstock per pound of glucose product. Thus the high yield of the HF process plays a major part in keeping the raw material cost low. Two of the remaining areas with major impact on the economic viability of HF saccharification are energy and equipment costs. At present we can discuss these in a qualitative manner only. HF saccharification can utilize temperatures and pressures near ambient, in contrast to dilute acid hydrolysis, which requires high tempera- tures and pressures (200-500 psi, ISO-300°C). Due to the relatively low heat of vaporization of HF and its low boiling temperature, energy requirements for distillation will be moderate. In contrast to other acids, anhydrous HF can be handled in carbon steel equipment, so special acid-resistant materials will be required in only a few areas, such as the distillation reboiler. Therefore no unusually high costs are expected in these areas. The last area which will have a major effect on the economic viability of HF saccharification is the value of the byproducts. Xylose as well as glucose has significant potential as a chemical feedstock. Lignin may well be of even greater value. Investigations into its uses have barely begun to expose its potential. 2h TABLE A COMPARISON OF CHEMICAL AND FEEDSTOCK COST (per lb glucose) Process Total Chemical and Feedstock Cost Feedstock Cost (Azlh) (é/lb) HF 5.A 3.A Dilute H2SOu 30% slurry 5.5 5-3 10% slurry 6.2 5.3 Enzymatic 9.2 8.0 CONCLUSIONS Effective use of biomass as a replacement for petroleum as a source of chemicals and liquid fuels requires a conversion process which can provide a high yield of sugars and a reactive lignin byproduct. Biomass will eventually become too valuable to burn. HF saccharification can produce high yields of sugars and reactive lignin. A preliminary evaluation of the economics of the process is encouraging. There remains a great deal of work to be done before HF saccha- rification will be ready for commercialization, and even before the process economics can be realistically evaluated. We at the MSU/DOE Plant Research Laboratory and the Department of Chemical Engineering at Michigan State University plan to continue our investigation of HF saccharification. Our immediate Objectives are first to further define the optimal reaction conditions - speci- fically time, temperature, water content, and the Options of gas or liquid phase reaction. Next we will specify a process design, and undertake a more complete evaluation of the process economiCS. In addition, during the course of our investigations we will identify the process components requiring investigation on a larger scale. On the basis of our work so far, it appears that HF saccharifi- cation of lignocellulosic materials may be superior to other acid and enzymatic hydrolysis technologies. 25 10. 11. REFERENCES Congress of the United States, Office of Technology Assessment, Energy from Biological Processes, July, 1980. Goldstein, Irving 8., Science, 189, 847 (1975). Gore, J., J of the Chem Soc,lg§, 396 (1869). Helferich, B. and s. Bottger, &, fig, 150 (1929). Fredeghagen, K. and G. Cadenbach, Angewandte Chemie, E6, 113 1933 - Luers, H., Holz Roh und Werkstoff, l, 35 (1937). Luers, H., Holz Roh und Werkstoff, 1, 3A2 (1938). Sherrard, E. C. and F. W. Kressman, Ind & Eng Chem, 37, A (1945). Oshima, M., Wood Chemistry Process Engineering7Aspects, Chemical Process Monograph Series 11, Noyes Dev. Corp., New York, 1965. Grethlein, H., Biotech & Bioeng, 29, 503 (1978). Wilke, C. R., R. der Yang and U. Von Stockar, Biotech & Bioeng 2m. 6. 155 (1976). 26 APPENDIX APPENDIX SURVEY OF SACCHARIFICATION TECHNOLOGIES HISTORICAL BACKGROUND The discovery that cellulose could be converted to sugar by concentrated acid was made by Braconnot in 1819. Simonsen, in 1894, appears to have been the first to successfully use dilute acid to saccharify wood. His method utilized 0.5% sulfuric acid at a pressure of nine atmospheres, and led to the construction of a small experi- mental plant. Because the products were very dilute, the plant soon closed. Classen, in 1899, used sulfurous acid in a process which was widely patented. An experimental plant with a capacity of two tons per day of dry sawdust was constructed at Highland Park, Illinois, in the early 1900's by the Classen Lignum Company of Chicago, which had purchased the American rights to his process. A commerical-scale plant was built in Hattiesburg, Miss., but never operated successfully. The major difficulties encountered were the slowness of the process, the quantity of acid required, corrosion problems, and degradation of the product sugars. A later plant at Port Hadlock, Washington, was also unsuccessful. M. F. Ewen and G. H. Tomlinson modified the Classen process, and an experimental plant using their process was constructed in Chicago Heights, I11., by the Standard Alcohol Corp. Ewen and Tomlinson then 27 28 decided sulfuric acid was preferable as a hydrolyzing agent, and a plant was built in Goergetown, 3.0., by the Standard Alcohol Corp., using this, the American process. It and a similar plant at Fullerton, La., Operated successfully until sometime after WOrld war I, when curtailment of lumber production and consequent high transportation costs for wood waste left them unable to compete economically with alcohol from blackstrap molasses. Both plants produced about 5000 to 7000 gallons of alcohol per day, using southern yellow pine as substrate. Yields were approximately 22 gallons of alcohol per ton, with a 45 minute digestion time.1 In 1929 B. Helferich and S. Bottger investigated the ability of hydrofluoric acid to dissolve cellulose.2 About 1933, K. Fredenhagen and G. Cadenbach began to study the saccharification of wood using HF.3 Their work and that of others led to the patenting of a process using liquid HF to saccharify wood. Further investigation by Hoch and Bohunek led to the development and patenting of a process using gaseous HF.u A pilot plant was constructed using this process in 1937, and ran successfully. These processes appear to have been ignored outside Germany and Austria, with the exception of some investigation in the Soviet Union. World War II brought a resurgence of interest in saccharification of wood. These more modern processes fall into three broad catego- ries: dilute acid, concentrated acid, and enzymatic. The dilute acid processes generally use sulfuric acid in concentrations Of .h to 1%. The concentrated acid processes use either 85% sulfuric acid, 29 35-41% hydrochloric acid, or gaseous HCl. The enzymatic processes commonly employ Trichoderma viride (or reesei), or Thermoactinomyces. Details of these processes are presented in the sections following. CONCENTRATED ACID PROCESSES SULFURIC ACID Hokkaido Process (Dialysis Process) The Hokkaido Process was developed in Japan in 1948, and a one ton/day pilot plant established. Further pilot plants followed in 1957 and 1958, with a full-scale plant opening in 1963 with a capa- city Of 100 ton/day of dry wood. A prehydrolysis of wood chips is carried out with 1.2-1.5% sulfuric acid at 140-150°C to produce xylose, or alternatively by steam digestion at 180-185°C to directly produce furfural in yields of 65-75 kg per ton Of dry wood. The wood is then dried and crushed before the main hydrolysis, which consists of treatment with 80% sulfuric acid at room temperature. To minimize the formation of sugar degradation products, a low mixing ratio Of acid to wood is required. The Hokkaido process accomplishes this by a mixing process involving spraying the dried powdered wood and acid together in a thin film, and then immediately filtering the product under pressure and washing it. Mixing time is only 30 seconds. Sugar yields of 90-96% are reported, with mixing ratios of 0.9-1. After filtering under pressure and washing, the solution has an ’acid concentration of 30-A0%. It is then treated with diffusion 30 31 dialysis using an ion exchange resin. 80% of the total sulfuric acid is recovered as a 25-35% solution. It is recycled after evaporation to 80%. The sugar solution from the dialysis contains 5-10% sugar and 5-15% sulfuric acid. 1.8-2% of the total sugars are lost in the dialysis. The sugar solution is then post-hydrolyzed by heating at 100°C for 100 minutes. It is then neutralized with 1ime and the calcium sulfate removed by filtration. The pH is adjusted to 2.5 to remove calcium ions, and the solution then concentrated to 50% at 60°C. Sodium chloride is added to separate the sugar as a double salt. Next the precipitate is washed with cold water to remove the sodium chloride and isolate the crystalline glucose. The overall yield of glucose is reported to be 83-85% of theoretical, yielding 280—290 kg of crystalline glucose from one ton of dry wood.5’ Cost estimates and energy requirements are not available. For flow sheet, see Figure 6. Thorough-Drying Process The thorough-drying process is a modification of the Hokkaido process developed by J. Kobayashi. The wet pre-hydrolyzed wood is immersed in dilute sulfuric acid, drained, and then dried at A0-50°C by blowing with dry air. The drying process concentrates the acid and completes the main hydrolysis. Yields are expected to be nearly the same as for the Hokkaido process, with lower power consumption. Again, cost estimates and energy require- ments are not available. 32 Steanll Steam J Separation Wood Digestion Byproducts and Methanol > Pretreatment 1 Refining Acetic.Acid ~ 11 , Furfural Residue Saccharification Drying Grindingfl_+ Mixing and Dilution 80% Sugar solution and HQSON lignin Filtration Lignin byproducté Conc. 5 H2804 Glucose polymer 30-35% H2504 Water C .. tgzgign Diffusion Dialysis Gluco;e polymer . 5-10 H SO Lime ! 2 A H 0 f Milk Of 3:133:11" HOSE-l ' l<——Steam 2 _ 1ime 3' ° ySlS Neutralized Gypsum ! sugar solution to refining and crystallization D Figure 6 Hokkaido Process 31+ Lime Water Milk of Steam 1 11-1280)1L lime Sawdust Pretreat- Centrifugal Crude Neutrall- Gypsum ment Separation Xylos zatlon Separation Solution Drying Steaml H SO , Ri enin Sacchari- 1_§L_li___J Mix1ng I l p g fication zation Milk Of lime —'_'_l Decolor- Filtra— Neutrali- Filtration Centrifugal aste iZIng tion zation Separation l asses lflrude Ho- 1 Ion 0 Dr 1 Heat Solution EXchange ’§ y ng fi-—-— Refining c>c5 Steam; 1 l Centrifugal Crystal- 2nd Eva- Fluidized Second Steam Separation lization poration Roasting Concen. Crystal Waste Second Glucose Molasses washing [Cooling] Crystal- lization Centrifugal Separation lasses Crystal Xylose l Drying & Cooling Heat . . ‘ . Crystal gypsum to Iciziiiil L_lCrushing l Sleving Entive Carbog Xylose to ackaging ! I to Packaging fackaging Figure 7 Nippon Mokuzai Kagaku Process6 ‘ . ¢-.,--1_.__\r 1L“). .U 11!- .10 Mind *“‘+czmxgo ’ . b. . .7 . . -“ . . 'r v ‘—( rr‘5 3 1 \1 a r "'l —— 35 neutralized with lime and filtered. The solid residue is dried with hot air at a temperature of about 80°C to concentrate the remaining dilute acid to 72%. This residue is then impregnated with 85% sulfuric acid at 50°C for two minutes in an apparatus which compresses the residue to about 35% of its initial volume under a pressure of 175 psi. The Optimum amount of concentrated acid was found to be #8 parts per 100 parts of dried residue, or 27 parts per 100 parts of corncobs. The resulting material is then post-hydrolyzed in 8% sulfuric acid at l20-l30°C for 7-10 minutes. A solution of 10.5% glucose is obtained. The glucose solution is filtered to remove lignin, neutralized with lime, and filtered again, for a yield of 85.5% of theoretical.8 See Figure 8 for flowsheet. University of Missouri O. C. Sutton and others at the University Of Missouri, Rolla, Mo., recently deveIOped a process for hydrolyzing corn stover with sulfuric acid. The corn stover is ground and then prehydrolyzed with A.h% sulfuric acid at 100°C for 50 minutes. It is then filtered. The xylose-rich liquid is processed by electrodialysis to recover the acid. The solids are dried and then impregnated with 85% sulfuric acid, next diluted with water to an 8% acid concentration, and then 36 Dilute Production Residue (85% moisture as dilute ‘ H so ) Dewatererl 2 A Residue (47% moisture) Lime Xylose Crystal- lization OI‘ Hydrolyzer folvents (8% HESOM) Filter Dextrose solution Fermenter Figure 8 Northern Regional Research Laboratory8 -. . l , . .1 , .7 ,1. . ‘.' I, .\ a- I c. _ ' u h ‘ ‘ u. .5. ‘ ‘ .J U ' v ' V or . ~. ) U x . - - _. -v‘ . , .. ._ , 1 1 ... . ~ ‘. . I I i 2,. .. . . l Vii " I ._ . T -'"" —'. , —- _ - . ._ . . ' .' —o_ ‘ -_.—.—...—————~._ . . . . . . .. . 0 ~- I bl -.' t \ t 4- — a ' . —— -—- —-_-——--—. . . 4 a 4 . v . . n. n I. 5“ \- t I I ‘1- . I 'l r - l ' o . . 4 . . - 1 ..’. '. . .. t, A. _ _-_.-_._.._.._.-—-—-———— ‘ v - ‘1..__.._fi . . _ 1...- . '. '. . .. . . v n. 1 ——-—————.—_— a . I ‘c r . - u L-. -- y... -~- , 1A- I o.:,\4 I . .~l . . . . .. _ -u..—q-.—-_——'-- ' O A . .1 7 R -.. u l- . ‘ n ‘ ‘ - . l - . .. . , I I ‘3 . ‘ c 4 I -11--. .1 1 l . . ¢‘ - ' rt \ ‘» ..; r 1 . .._ . \ 1 O a , . . . ‘ I *. O - I 1 t ' .ll . . . . a , 1 - ' . ," .., ~ . , 1 I I I- _ 1 . ‘1, - ‘-‘ v x‘ , ~ ‘ . 11.1 ...,.J 7 f . _ . , . .. g , a O ~' — — £u;4_—--——-—‘-——~— ._ H _.—‘——.— . \. .1 .- - A ., , , . ’ I 37 hydrolyzed at 110°C for 10 minutes. The glucose-rich liquid is also processed by electrodialysis to recover the acid. Overall combined yields are 94% of xylose and 89% Of glucose.9 See Figure 9 for flowsheet. HCl - LIQUID Bergius-Rheinau Process The Bergius-Rheinau Process was the first commercial hydrolysis process using hydrochloric acid. A pilot plant was constructed in 1920, an intermediate scale plant in 1925, and a large-scale plant in 1932, with another following. Both plants closed after World war II. A A0% hydrochloric acid solution is used, and recycled several times until the sugar concentration reaches about 40%. Acid required is about 7 times the weight of the wood. A unit of 8-10 reactors is used, with an intermittent, multistage counter-current flow system. This solution is then distilled at 36°C under reduced pressure to recover about 80% of the HCl. The solution then contains 55% sugar and a small amount of acid. The sugar solution is next spray-dried with hot air to expel the residual hydrochloric acid. The lignin from the reactor is washed with water to recover sugar and hydrochloric acid. Hot oil is used for heating. During the Operation Of the plants, corrosion problems were the most difficult problems encountered. ’ For flow sheet, see Figure 10. 38 Recycle water HASOL (13,470 1b/hr) (10056, *5 1* 35,140 gal/hr Electro- Cornstalks Pre- Filter 0.7h%1Xylose dialysis0.80% x los (220 ton/day) hydrolysis .18%G1ucose U it 3.19% Glucoée P-El ------- (3)4”le0 gal/hr) .+.35 Gluco e . 12,500 gal ‘55, hr) Rotary Dryer 80°C ‘ h.0h% Glucose 110°C 8 % H 50 (Impregnator (recycle) 1' ‘ Recycle gggrglggii L.89% Glucose Filter Water 2 A .1 . __‘ (12,6h0 gal/hr) Solids . 4630 lb/hr) Figure 9 University of Missouri Process9 u ‘_ — 1 . .. I 1 .-. n..- WOod HCl 39 Crushing ,Alcohol Pre- hydrolysis Drying %iggin ‘ yproduct water Main hydrolysis H01 Gas to ‘Recycling 40% HCl Steam washing Drying Alcohol FermentationL Figure 10 Bergius-Rheinau Process 55% sol. Steam ‘ Distilla- tion sugar Post- -————-3 hydrolysis ,Dry sugar “Product l Desalting L Concen- tration V7 Crystalli- zation .__.in.._ Drying l_#r Udic-Rheinau Process The Udic-Rheinau Process was a modification Of the Bergius- Rheinau process. A semi-works plant with a capacity of 1200 metric tons of wood per year began operation in 1960 using this method. The wood is chipped to 15 mm and then prehydrolyzed at 20°C with 35% HCl to remove the hemicellulose. Next it is hydrolyzed with 111% HCl at 20°C in the same digester. The lignin remaining is washed with HC1 solution and then with water. A multistage counter-current flow extraction system is used, with a large quantity of HC1. HC1 is recovered from the sugar solution by vacuum distillation and reconcentrated in two rectification columns Operating at different pressures. 94-95% of the acid is recovered. The remaining solution contains 60-65% sugar. It is diluted to 10% sugar and post—hydrolyzed at a temperature over 100°C. The remaining acid is then removed by deionization using ion exchange resins. The xylose solution from the prehydrolysis step is treated similarly. The solutions are then concentrated and the sugars crystallized. Next the mother liquors from the crystallization containing xylose, mannose, and glucose are hydrogenated at 30 atmospheres to produce polyalcohol solutions. The lignin is used as a raw material for plastics. The products are crystallized dextrose, xylose, xylitol, poly- alcohol solutions, and lignin products?-7 For flow sheet, see Figure 11. Al ,Conc. H01 ‘ HC1 Gas '1 11 1 Absorption gfimos. Column ‘essure Column [—— 21% HC1 Recoficen- S‘ 3%? tration 32% HC1J. 111% HC1 HC1 Wood Pre- ( Main Acid‘ ( 100 hydrolysis ’ Hydrolysis Evaporation kg) , , (low P) _ Hemicellue Lignin 2% H01 50% sugar lose 10% solution (22 k ) Evapora- ‘EHEO Pbst- H 0 tion A Washing hydrolysis I [LigninF 2 kg Crystalli- Nether 4g; zation Liquor I Pngystalline__A xylose (7 kg)’ 12% sugar sol. Ion Exchange ._L' Evapora- tion 70% sugar sol. 1 C - 13.5 kg sugar rystalli v zation in mother liquor Mether liquors hydrogenated to polyalcohols Figure 11 Cr stalline Glucose (31 kg) Udic-Rheinau Process6 A2 HC1 - GAS Prodor, Darboven, Hereng The Prodor process (1925); Darboven process (1930), and the Hereng process (1952) were all early processes using hydrogen chloride 5,6 gas or combinations of liquid and gaseous HC1. Noguchi-Chisso Process The Noguchi-Chisso process is a more modern process using HC1 gas. .A pilot plant consuming one metric ton of dry wood per day and producing 300 kg of dextrose crystals per day Operated from 1958-1962. Wood particles of sawdust size are permeated with 3-5% HCl in a quantity 0.5-0.7 times the amount of wood. Steam is introduced at loo-130°C to carry out the prehydrolysis. The wood is then extracted by a counter-current flow of water, yielding a 10-15% solution of xylose. Next the wood is flash dried with hot air and permeated with an HC1 solution of less than 38% concentration. It is then fluidized with cold HC1 gas, absorbing HC1 to a concentration of 52%, or alter- natively is conveyed along pneumatically in the presence of HC1 gas. Next it is rapidly heated to AO-A5°C while being fluidized with hot HCl gas (or, again, being conveyed pneumatically), to complete the main hydrolysis. The product, a solid, is next dried at a high temperature with HC1 gas to vaporize the HC1 in the material. A counter-current A3 extraction with water yields a 10% sugar solution, which is then post- hydrolyzed by heating it in the presence of the remaining acid. The sugar solution is deionized and decolorized by counter- current flow with ion exchange liquors, and then concentrated and crystallized. The prehydrolysis solution is treated in the same manner. The HC1 gas is recycled after the removal of water. An anti-sticking agent is used to reduce the stickiness of the intermediate product, which is essentially a mixture of a viscous, thick sugar solution and lignin. The agent used is finely pulverized dry lignin, which is used to coat the surface of the wood particles.5-7 See Figure 12 for flow sheet. Lawrence Berkeley Laboratory The Lawrence Berkeley Laboratory is investigating a high pressure HC1 gas process in a special high surface/volume ratio reactor. Dry hydrogen chloride gas is contacted with average natural-state moisture wood. Yields are reported to be at least 80% at a minimum hydrogen chloride and water concentration of 66 wt.%. Problems have been encountered in maintaining the reaction temperature below that at which degradation occurs. No information on HC1 recycling, if any, is presently available.10 AA 3- 5% HC1 Steam I Acetic Crush- Pre-hy: .Acid Aceticp_ ing “drolySls Separ. .Acid . (45 kg) Hot air Drying Ion Exchange (38% HC1 HC1 Concen- Penetration tration ydresenetion o Polyalco- HC1 HC1 Gas Absorption lization h2%.Acid Crystalline} IXylose-7(1OO Heat to SOCCh‘f‘ri‘ I kg) LI’O-LI'SOC fication I lto Furfural Crystalline 5 Hot HC1 Drying HC1 t Glucose Gas (300 kg) Saccharified solids Crystal- . . . MOlasse , Vllzatlon ~Cpfiijgi} H 0 : washing. Lignlg 10% sugar sol. Concen- Ion Post- Heat tration 'Exchange hydrolysis Figure 12 Noguchi-Chisso Process6 ' 5“ -"-->- r- u ~ v .1. - U . 1 1 . _ ———————— ______..___1,1 1.1 . n . 1‘ ‘L I \ , , \ - . . ) I Q .1 ‘ ' I. i ‘ I J ' . | . . - ' ‘ v I I .. ~-—D - I ‘ I C ‘ , l . ‘ . .1 a . , . s . - , . ‘ y -- - . ,~ . - .J 1 , 7- c . . . . I ' - . V v ' , . I __ u . v —.__.-——_——._ . -__,‘ ‘ . . —‘—-. rt. 1 . . . ~ . n _____ _.__fi_ .——-—————— ' l .‘ ” ‘ A I ~.y Q ' ’ . \‘ ' 4 t |-" I ‘ - - ‘ . 1' .- 1 ' l .H —- v- . . l I. . I . ._ _~—~————————-—-—— . .. . _. ‘ . F v . . .- 1 . ~ , l‘ 1 . . .. . l I ‘ - ‘. -~‘_..A._1 — ‘.— . A ' ' ‘ __ .v-‘__ u-n-‘m . " . ‘ I - -. — 111.11.111.11.“ .- A ———~———.——— _ - _ ~ - . __—_ . " . ,.-.. ...... A -———r I ...o... 1 . M 1 a . . . . .-. . 1 . 1 ’ ' .a, I I . ———.__..L___._.~ ‘ a . . I ] .. »: . . . ., . - - m. .J ‘. ‘ n . .._ . ..- - - . -- 1 .¢-___.._. ————_—. ‘ i l I - l 1 l u—__-1‘.._.1 . . -__.-—.h———.—..-—l ! r' ’ - I . a «. . .. , . '~ ’ C. . 1 —.--- _-_.1_... o——.———.—.——-. _. . I '.' ' _. .‘ _ — ‘ I A O r - n ' . - . >_—___.__—'_ - x . . m.—-.p . - . w . v a . , - -r‘aa m ———_——.—A DILUTE ACID PROCESSES SULFURIC ACID Scholler Process The Scholler process was used in plants in Germany, Italy, Manchuria, and Korea beginning in 1931. By 19Ml at least 20 plants using this process were believed to be operating in Germany. The process had been attempted on a pilot plant scale in the United States before WOrld War II, but it had been abandoned as uneconomical. The procedure involves the compression of sawdust and wood chips in brick—lined steel percolators to ll lbs per cubic foot by 30-45 psi steam, and then preheating with live steam to 265°F. Next suc- cessive batches of 0.5% sulfuric acid are forced through the wood, beginning at 165 psi and reaching 180 psi by the final batch. Each batch remains in the percolator for approximately 7-10 minutes before being expelled by steam. The wood chips are kept at the required temperature for 30 minutes by introducing steam, and then the next batch of acid is added, at a temperature 15-30°C lower. The pressure difference induced by the lower temperature acid causes the expulsion of the saccharified solution from the pores of the wood, and is known as the "cold shove" process. Up to 24 batches of acid per charge of wood can be used, though ll is typical. The dilute sugar solution, average concentration 3%, from the percolator is cooled, neutralized, 45 M6 and prepared for fermentation. Hydrolysis temperatures are typically 130-150°C for the first three cycles, ISO-160°C for the nth-9th cycles, and 180°C for the 10th and llth cycles. Total hydrolysis time is about 20 hours. In addition to alcohol, the plants produced glycerol and feed yeast. Yields were about 50 gallons of 95% alcohol per oven dried 6,11,12 ton of wood chipS. For flow sheet see Figure 13. Madison Process The Madison process is a modification, or series of modifica- tions, of the Scholler process develOped by the Forest Products Laboratory at their Marquette pilot plant in 19h1-1945. Sugar yields were approximately 50%, with alcohol yields averaging 50-65 gallons of 95% alcohol per oven dried ton of wood chips. It was estimated that a plant consuming 220 tons of dry wood per day would produce 65-75 tons per day of lignin to be used as fuel, consume 26,500 lb of sulfuric acid per day, and require 1,106,000 lb/day of 190 psi steam, 900 hp of electricity, 1,700 ft3/ min compressed air, and 1,900,000 gal/day of water. It would also use 2.2 lb of lime per gallon of alcohol produced. The process involves pressurization of wood chips or sawdust in the digestor to 50 psi, increasing to 150 psi for the final batch of acid. Acid concentrations range from .85% to 0.4%, and 8-15 batches are run per charge of wood. Temperatures are l35-lSO°C for 30 minutes for prehydrolysis, and 150-190°C for 3 hours, raising at Batch reaction Wood chips Intermittent dilute (0.5%) Hesou alternated with steam Pressure approx. 8 atm. packed in and compressed l Saccharification L——- Percolator l+7 Intermittently released ‘ Lime After reaction, pressure suddenly reduced, lignin discharged as fine particles ‘prystalline ICrystal- ‘Ulucose Llization N sugar solution (ave. 3%), Neutrali- zation J, Filtra- tion l Concen- tration 40% sugar sol. Concen- 28% sugar tration solution Figure 13 Scholler Process6 Filtra- tion #8 5°C per minute, for the main hydrolysis. The total hydrolysis time is about six hours. The saccharified solution is discharged conti- nuously. The acid in the sugar solution is neutralized with lime under pressure and the solution filtered. A sugar solution of 5-6% 6,12—14 concentration is produced. See Figure 14 for flow sheet. TVA Process The TVA process is a further modification of the Madison process, resulting from the Tennessee Valley Authority pilot plant. Dilute hydrolysis solution from a previous batch is pumped into the t0p of the hydrolyzer containing the wood chips. Then hot water and sulfuric acid are added, increasing the temperature to 385°F. The acid concentration varies, averaging 0.53%. The final pressure is 200 psig. Total percolation time is 145-190 minutes. The hot acid hydrolyzate is neutralized, and the calcium sulfate sludge concentrated to 50% solids and disposed of in a landfill. The sugar solution is fermented by Saccharomyces cerevisiae, and the pentose sugars collected in the bottoms from the beer strip- ping tower. They are then concentrated to a 65% solution for sale as a feed supplement or for conversion to furfural. The estimated cost in 1975 dollars was $1.90/gal of ethanol for a 25 million gallon per year facility, based on a 15% after tax return on investment and a $34 per ODT cost for wood waste. See Figure 15 for flow sheet. #9 Batch reaction Steam to heat and pressurize Hogged wood First cycle: 1.2% H sou, 1.5-2 times wood or sawdust, weight. Rest 30 mgn. (Pre-hydrolysis) £30k:28::da:g Subsequent cycles: 0.h% H SOA’ 50—75% lEmIb'wood/ft3 weight of wood, decreasing. Rest 5 min., discharge. Heat & pressurize with steam, 10-20 cycles, 6-9 hrs, final T 195°C. Sugar $01.: 3-6% sugar, .l-.2% wood oils & lignin residue, .4-.8% sulfuric acid, .l-.3% acetic acid, .1% furfural Stationary Digester lime Sugar Flash ‘ Neutra- . Fermen- Lignin discharged Sol. Tank lization tation under pressure at J ———————» end of last cycle Alcohol Figure 1% Madison Processlu IQ'. ‘1'. 50 Semi-continuous Furfural> Furfural Tower wood |Flash Sludge Slud Chips1 DigeSter | ’ Tank Clarifier W I Fermen- tation water Extractivel Alcohol F IYeast Yeasta j Tower Stripper Separation Rectifying I Ethanol3 Tower Compres- . 515' sor Evapo- Evapo- I rator . rator Steam .entose concentrate Figure 15 TVA Processl2 51 Grethlein Process Dr. Hans Grethlein of Dartmouth proposed a process for the hydrolysis of a refuse slurry based on an isothermal plug flow reactor, Operating at 230°C with a residence time of .19 minutes and an acid concentration of 1%. After hydrolysis the slurry is neutralized with calcium hydroxide and filtered. The hydrolyzate contains glucose, decomposed sugars, and other organic impurities, but can be successfully fermented. Yields of 56% are predicted for this process.5’15 Brenner Process Dr. waiter Brenner of New York University has built a one ton per day demonstration plant for the hydrolysis of sawdust or pulped newspaper. The material is cram-fed into a twin barrel extruder, heated to A50°F with superheated steam, while being compressed to 500 psi, and water is expelled by the action of the screws. Near the barrel outlet, 0.5% sulfuric acid is injected. After approximately a 20 second reaction time the mixture is quenched by sudden cooling on leaving the extruder. Glucose yields are 60%. Energy consumption is approximately 1600 Btu/lb of cellulose, or 2700 Btu per pound of sugar produced.l6 See Figure 16 for flow sheet. 52 Newspaper or 0.5% H2SOu Sawdust pulp Ti Cram Sulfuric Acid Feeder ‘ Injector Sgperheated Twin-screw Glucose Syrgp and\ Steam. Extruder l unreacted materiaf' Expressed water Figure 16 l Brenner Process 53 Georgia Tech. (G.I.T. Process) Investigators at Georgia Tech. are developing a hydrolysis process for wood or other biomass using dilute sulfuric acid and employing a steam explosion pretreatment and delignification by solvent extraction, along with prehydrolysis. Ethanol or other suitable solvents will be used for delignifi- cation, and will be recycled. The prehydrolysis is carried out with 0.5% sulfuric acid at 150°C. The main hydrolysis takes place at 190°C. The acid is neutralized with calcium hydroxide or other alka- li. A 5% sugar solution results, which is concentrated by vacuum evaporation to 13—20% before fermentation. Enzymatic saccharification will also be investigated. Yields are reported to be 80-85%, with an estimated selling price for 95% ethanol of $1.69 per gallomlo’l7 See Figure 17 for flow sheet. Dartmouth College At Dartmouth College the hydrolysis of cellulose with dilute sulfuric acid and live steam at 600-700 psi is being investigated. Acid concentrations from 0.5-2.0% have been used, with corn stover and Solka floc as feeds. Experiments with pOplar are planned. The main aim of these experiments has been the development of a kinetic model for the hydrolysis of cellulose.lo 51+ Solyent Cellulosic Steam Filtration ‘ h Delignification Li Biomass Explosion and Washing by Extraction extract I Hemicellulosic tdet Hexoses & Pentoses Prehydro- 0. H SO 2 h lysis I a Hemicelluloéic Hexoses & Pentoses Sludge Clarifi- Neutrali- Acid Dilute H 80h cation zation Hydrolysis Concen- Carbon Sterili- tration Treatment zation Single-cell lYeast Fermen- rotein ISeparation tation (Yeast) l, [----j1 Recycle Yeast , 60-85% ‘lAlcohol .Acid ‘ Ethanol qurification Wash Alkali Process design includes provision for omission of various steps, such as solvent extraction. Provision for enzymatic hydrolysis is also included. Figure 17 Georgia Tech Processlo 55 ENZYMATIC PROCESSES The enzymatic process for cellulose saccharification fall into two major groups - those involving saccharification alone, and those involving simultaneous saccharification and fermentation. ENZYMATIC SACCHARIFICATION Natick The Natick process was developed at the U.S. Army Natick Labora- tories in Natick, Mass., and is the oldest and most develOped of the enzymatic hydrolysis processes, having reached a pilot plant stage. The organism used for enzyme production is a mutant strain of Trichoderma, Trichoderma viride (reesei) QM9u14, and more recently MCG77. The organism is grown on either of two types of Kraft pulp for an average fermentation time of 100 hours. The productivity is estimated at 40 IU/l/hr of cellulose for QM9414. The saccharification in the pilot plant work has been done primarily using a delignified pulp similar to paper mill waste. The hydrolysis is conducted at 40°C (50°C in other experiments) with pH controlled between 4.79 and 4.89. Conversion using 10% substrate with an enzyme concentration of 1.5 IU/ml is 85% in 72 hours. A 15% glucose syrup is obtained. Sugar costs are estimated at 35-40fi per kilogram of glucose. 56 For use on substrates other than Kraft pulp, the process relies on physical pretreatments, of which ball milling was found to be most effective, though two-roll milling has also been used. The pre- treatments are designed to increase the surface area, decrease cellu- lose crystallinity, and increase the substrate bulk density. Enzyma- tic hydrolysis of lignocellulosic materials gives extremely poor yields without pretreatment. Ball-milled newspaper yielded glucose syrups of 2-10% concen- tration, with a 60% yield after A8 hours. Numerous other substrates have also been investigated.lo’18-24 See Figure 18 for flow sheet. Wilke Processes Dr. C. R. Wilke of the University of California at Berkeley has developed a process for the enzymatic hydrolysis of newsprint. Shredded and hammer-milled newsprint at a 1:20 solid/liquid ratio is hydrolyzed for 40 hours at “5°C. Cellulose conversion is estimated at 50%, and a A.0% sugar solution is obtained. 3h% of the enzyme is recovered for recycling. The organism used for enzyme production is the fungus Trichoderma viride QM9A14. Utility requirements for the production of 238 tons/day of glucose from 885 tons/day (plus 66 tons/day for the fermentation system) of newsprint are estimated at 8060 kw electricity, 36000 1b/ hr of steam, and 83000 gal/hr of water. Estimated sugar cost is 5.2é/1b (1975 dollars), exclusive of newsprint costs. A similar process has been deveIOped for the hydrolysis of 57 Cellulose Recycle Nutrients Mycelia l L Inoculum Production vertical Vessel Vessel Filters f water I Sterile Air Water Enzyme Cellulose Hydrolysis Pulp Vessel Evapora— Polishing (Coarse tion Filters Filtration Concentrated_ Waste; Waste Sugar Syrups Solids Solids Figure 18 Natick Process18 (Pilot Plant) 58 wheat straw using a prehydrolysis with dilute sulfuric acid in place of the milling. A five-stage reactor with a capacity of 1439 tons/ day using 1% sulfuric acid at 100°C with a residence time of 1.1 hour is employed, with recycling of the filtrate to produce a stream containing 2.1% xylose. This solution is then neutralized with lime. Yield is estimated at 500 lbs of total sugars per ton of wheat straw, at an average cost of 4.7fi/1b, exclusive of straw costs. Ethanol is produced by fermentation of the glucose. The remaining xylose and cellobiose are converted to single cell protein by a Torula yeast. An alternative concept for utilization of these sugars is anaerobic digestion for methane generation after removal of the yeast. It was estimated that 384 ft3/day of methane could be genera- ted by a plant utilizing 1376 tons/day of corn stover. For this process, glucose costs of 10.0¢/lb (1978) were estimated. Ethanol cost was estimated at $1.60 per gallon, both exclusive of corn stover costs. 58% enzyme recovery was assumed, but further experiments indicated this might be too high.25'29 The fermentation of the xylose solution from the prehydrolysis has also been investigated. Fusarium oxysporum will ferment xylose to ethanol with a yield of .41 g ethanol per g xylose fermented. However the growth rate is low and is inhibited by ethanol concen- trations greater than 1.5%, stopping completely when ethanol concen- trations reach 4.1%. In addition, growth stOps before all the 30 xylose is utilized. See Figures 19 and 20 for flow sheets. 59 water Wash _ IFilter Solids to > I ‘ furnace l Rec cle‘ 3 Stage Counter- Milled Current.Mixer- Solids Hydrolyzer’LEnzyme Solution Newsprint Filter LSugars V Medium Induction Solids L . Growth 3 Induction Centri- Medlum -Fermentors Fermentor (Series) fuge Cell recycle AIMEE“ Nbcelium and Residual Solids Figure 19 Wilke Process - Newsprint2 l ‘\. ,‘-- 6O Steam Acid Water Corn . l Acid Pre— a Filtra- Eylosez other sugars, acid§ Stover Milling treatment tion tc., to xylose utilization £01165 1 l Sugarsz waterz enzyme Water Hydro- Filtra- Counter-current Sugars, Mixer-Filters lysis tion water Solids ‘ generation Enz e water Wat 2 Nut ' nt A1 C0 Wat r er rle s r Ei ‘ e Deli 'fied Steri-‘ 2—Stage a Centri- Filtra- cellulose llzer Fermentor fuge tion TCell recycle 1 celium Cellulose, Water to Water reuse after Nutrients condensation Air 95% Ethanol . . r L4 Multi—effect Steri- Fermen- 3 Centri- Distil- Evaporator llzer tor Evaporator & -§fl§EEE—-¥ ggrzz Anaerobic y Digester Yeast ' 7 Water Air, CO2 Figure 20 Wilke Process - Corn Stover 9 ML..- - I , l}. ' - o r . - --» - A -'- I -n .. ’ -\ - ‘ ‘ . . ' I ___.._ - I » - . _ l .o 1 ‘ >— . «. - -' ~ . . \ . ' . .. . . ——_——————_——————’ h- ._._._. law“ : .— . | a L k ._ . —— ———-—————.——_ ' V ' -r‘ ' u -‘ I. _. . _ a. « a u —0 a, . ’- " ~ .4 \ . .V c I , v i C-' . - . ~ -\ —A— 1 \ . _ - , . I I. I. _ .. \ v . . . \ II ' 0 ' -' ‘ ‘ ”'7~~ k‘ .———-——-—-——|-_ 1 I #- H — . ‘ o “ . r- . . _. - - “,v . - . —- _ g —— 0' -_ - s -. \ . ~— 0 \ ‘ I -- —- .—_—_—_————- _——v—.__- -ea-_- 1 ‘ \ 7 ‘ ~ . u‘ — ’ l ' ..\ _._... . ‘ h _ .- n__ J , 11., u . \ . x . _ . x V. 6 _——‘I" 61 Indian Institute of Technology The Indian Institute of Technology in New Delhi has developed a modification of the Natick process for the hydrolysis of newsprint in which the sugar is recovered by dialysis followed by fractional crystallization to separate the glucose from the cellobiose. See Figure 21 for flow sheet. Rutgers The work in this field at Rutgers University is primarily directed toward producing hypercellulolytic mutants of T. reesei which are able to grow at elevated temperatures on inexpensive carbon and nitrogen sources.lo’3l Swedish Forest Products Laboratory The Swedish Forest Products Laboratory is investigating a white rot fungus, Sporotricum pulverulentium, which decomposes lignin as well as cellulose and hemicellulose.5 Miyazaki University Peracetic acid was used as a delignifying agent in research at Miyazaki University, Japan, for a hydrolysis process using T. viride and A. niger, but the process was found to be uneconomical and research halted.5 62 A. Cellulose waste Cellulose Preliminary Heater Pulveriser Waste crusher & grinder 200°C Steam Air Mediumi Steri-‘ Fermen- . Polymembrane‘ lizer tor Filter Ultrafilter Drying ‘ i Hydrolysis ‘ Reactor LSolids V L Fractional Polymeric membrane Crystallizer MOlecular sieve Cellobiosegg Compressed Air Glucosegg B. Milled newsprint ,Product sugars Sugars & Enzyme Solution Wash water . 1 T I Milled 3 Stage Counter“ Ti HYdrOlySIS Filter Newsprint Current Mixer-Filter! lid Reactor T Enzyme solution Induction solids I Water .1 .1: f Medium & Growth Induction Centri- , . Mycelium, ‘ - Filter Sugars Fermenter Fermenters, 3 fuge thell recycle Figure 21 Indian Institute of Technology5 63 SIMULTANEOUS SACCHARIFICATION.AND FERMENTATION General Electric - University of Pennsylvania The G.E.-Univ. of Penn. process involves the delignification of biomass (pOplar chips or urban refuse) by n-butanol, and subsequent simultaneous anaerobic saccharification by Thermoactinomyces and fermentation to ethanol by Clostridium thermocellum. The alcohol is recovered under low pressure. At an early stage of development ethanol yields were low, but a later report indicated that a yield of 1.02 g ethanol/1 was obtained, with about 90% sugar utilization.5’lo’32 See Figure 22 for flow sheet. Massachusetts Institute of Technology The procedure being deve10ped at the Massachusetts Institute of Technology also involves anaerobic saccharification and simultaneous fermentation to acetic acid and ethanol, all by Clostridium thermocellum at 60°C. Studies using this organism on ball-milled corn residue yielded cellulose degradation of 74%. Acetic acid and ethanol are produced in equal mole ratios. The product solution is 3.A% reducing sugars, 1.2% acetic acid, and 0.5% ethano1.5’33’31+ In later studies, a new strain of Clostridium thermocellum was isolated which had a higher ethanol tolerance and a favorable ethanol-to-acetic acid production ratio of 5/1. Studies are also being done with g. thermosaccharolyticum and with mixed cultures of 6h H20 & Na2CO3 L . _ Exploded wood Butanol' Separationl Cellulose Solids chips Delignification l - Butanol _ Butanol phase , VLignin ,Aqueous phase, I Recovery Xylan ‘ ‘ L1 in Prec1pltatlon I Hemicellulose X lanas Enzymatic Simultaneous Hydrolysis Extractive Fermentation & ‘jgflgygim . Saccharification Butanol SISolvent lRecovery Thermoactinomyces cellulase enqyme & C1. Acetobutylicum g? i Thermoactinomyces‘ Enzymatic Acid Simultaneous Vacuum Mutant Production Hydrolysis Hydrolysis Fermentation and - Saccharification t20% Glucose Acid‘ » 1 Syrups Sugarp Thermoactino- Syrups, myces cellulase enzyme & Cl. _ Thermocellum alcohol tolerant 119-20%.Alcohol IVacuum mutant lCondenser Figure 22 General Electric - University of Pennsylvania5 65 the two organisms. However, the favorable ethanol to acetic acid ratio which is obtained on Solka floc does not occur when the feed- stock is changed to corn stover. It is suspected that this is due to the effects of other water-soluble materials in the corn stover, in particular low molecular weight and aromatic compounds. An investigation of Q. thermohydrosulfuricum, which is reported to ferment both C6 and C5 sugars to ethanol, is also in progress.10 Gulf The Gulf 0i1 Chemical Co. process involves the simultaneous saccharification of cellulose by T. reesei QM9414 and fermentation to ethanol by Saccharomyces cerevisiae ATCCA132, Candida brassicae IF01664, and Saccharomyces carlsbergensis. The experimental medium is purified cellulose. The saccharifi- cation is carried out at 40°C at a pH of 5.0, using whole culture enzyme and a 5% v/v yeast broth. Ethanol solutions of 25-30 g/l were obtained after 2A hours. A 7.5-15% slurry of pretreated cellulose is used, and conversions up to 90% are reported. Ethanol is recovered by stripping with steam, followed by rectification. The process has also been tested on waste cellulose such as sawdust, bark, and pulp waste liquors. Production cost from waste was estimated at $1.16/ga1 of l90-proof alcohol for a 1000 ton/day plant, with a projected alcohol selling price of $1.32/ga13O or 10 $1.44/ga1. Byproduct molasses is produced for use as an animal feed. Corn stover and poplar are also being considered as feedstocks. 66 A 50 ton/day pilot plant is in Operation at Shawnee Mission, Kan . . 10.35-37 sas, with a 2000 ton/day commerCial plant scheduled for 1983. See Figure 23 for flow sheet. Natick Recent studies at Natick investigated the saccharification of cellulose by T. viride QM941A coupled with fermentation by Candida utilis QM82A0. The substrate used was Solka Floc, pretreated by differential speed two-roll milling. Saccharification is carried out for 20 hours at 45°C, then the temperature lowered to 40°C and yeast cells added. About 3.5 hours later when the dissolved oxygen concentration decreased to a negligible level, the temperature is further reduced to 30°C. pH is controlled at 4.0. The result was 22% more saccharification when yeast was added than without it. Production stopped when A5% of the cellulose had been utilized, probably due to ethanol inhibition.38 67 Nutrients and Water ___§L_ 1 Pulp Mill Raw Material 1 Enzyme waste Preparation I Production Recycle Enzyme —j;Nutrients and Water Si ultan “13%?‘> Municipal Raw Material I S m h .eous , Solid waste Preparation acc arification &.Fermentation Recycle Steam yeast waste Liquor AlCOhOl Anhydrous & Industrial Treatment Solids Recovery grade alcohol , (Sol‘n) ‘Iwaste a Solids Figure 23 Gulf Process 37 MISCELLANEOUS PROCESSES SOLVENT EXTRACTION George Tsao at Purdue University has deve10ped a rather unique process for the hydrolysis of biomass. The hemicellulose is first removed by a prehydrolysis with dilute acid. Next the lignin-cellulose residue is treated with a cellulose solvent. Cellulose is reprecipitated by adding water, and is then highly reactive to acid or enzymatic hydrolysis. The cellulose solvents that have been examined are Cadoxen (25- 30% ethylenediamine and 4.5-5.2% cadmium oxide or hydroxide in water) and CMCS (20 g sodium tartrate, 15.5 g ferric chloride, and 14.5 g sodium sulfite dissolved in 1000 g of 5% sodium hydroxide solution). Cadoxen has the disadvantage of being toxic, while CMCS is nontoxic. Cadoxen will dissolve up to 10% by weight of cellulose at room tem- perature, and CMCS up to 4%. Addition of water reprecipitates the cellulose, and further washing of the precipitate with water recovers the solvent, which is restored to its original strength by evapora- tion under partial vacuum for CMCS, or by a somewhat more complicated process for Cadoxen. Experimental results with corn residue and CMCS gave 80% yields of glucose after 40 hours of enzymatic hydrolysis, compared to a 30% yield without solvent extraction. 68 69 Crude cost estimates for the production of sugar using this process are 3¢/lb for glucose and xylose combined. The pentose sugars are fermentable to ethanol and butanediol using strains of Klebseilla and Aeromonas genera. According to 39 recent statements by Dr. Tsao, xylose can be converted to xylulose by the enzyme Glucose isomerase, and the xylulose can then be fermented to ethanol by Saccharomyces cerevisiae, which is commonly used in fermentation of glucose. 40-42 The glucose can be fermented to ethanol using the usual methods. WEAK.ACID HYDROLYSIS F. H. Snyder patented a process in 1958 involving the impreg- nation of woody material with an alkaline or buffer reagent to keep the end pH at 3.1-3.4, and then heating to 250-300°C at 600-1250 psi for 60 seconds to 5 minutes. Approximately 30% of the pentoses are converted to furfural during this process. Subsequent post-hydrolysis yields 80-90% of the xylose and glucose. It appears that no further development of this process has been done. ECONOMICS It appears that considerable confusion exists about the relative economic merits of the various hydrolysis processes. No thorough economic analysis has been done for most processes, and in addition the existing analyses are often based on very different assumptions. An important exception is the analysis done by Charles R. Wilke on the enzymatic hydrolysis of newsprint, corn stover, and wheat stover in l975-78.25-29 Hans Grethlein carried out an equivalent analysis for the weak sulfuric acid hydrolysis of newsprint, based in large part on similar assumptions and using similar methods of cost estimation. The analysis was done for two base designs: a slurry of 10% solids and one of 30% solids. An estimate of 1.75- 2.45 cents per pound of sugar was obtained for acid hydrolysis, compared with Wilke's estimate of 5-205 cents per pound for enzyma- tic hydrolysis. All costs were exclusive of substrate costs. Irving S. Goldstein has stressed the necessity of basing a viable process on utilization of all the components of wood, rather than just on cellulose. Many process designs are based on burning lignin as fuel, but that is not the most efficient use of that raw material. Goldstein and others have suggested hydrogenation of lignin to phenol as one alternativefm’u5 Hemicellulose contains primarily xylose in hardwoods and mannose in softwoods. Mannose can be fermented to ethanol. Xylose 70 71 can be utilized in several ways. It can be converted to furfural by acids or fermented to ethanol, butanediol, acetone, etc. The acid conversion to furfural appears to be well established, but the fer- mentation processes are in need of further development. Goldstein also suggests that ethanol can be dehydrated to ethy- lene or converted to butadiene, basic building blocks for the syn- thesis of other chemical products. Goldstein paints a very optimistic picture for biomass utilization, based on acid hydrolysis with a 50% yield. He felt the production of ethanol from biomass was economically sound even in 1976, and predicted that ethylene from biomass might soon become economical. Other investigators have been more pessimistic, but all seem to agree that the economic viability of hydrolysis processes is intimately related to the price of oil. One of the most pessimistic estimates, also in 1976 by Paul Becher, was that the cost of petroleum would have to triple (at then-current ethylene prices) for ethylene from biomass to be economical.1+6 This assessment was based on enzymatic saccharifica- tion with a 50% yield. Arthur E. Humphrey appears to take a more middle ground. In 1975 he estimated that glucose costs would have to be 1-3 cents per pound of fermentable sugar for economic conversion to alcohol.1+7 In 1976 he estimated that alcohol from biomass would not be econo- mical until the price of gasoline rose to $1.60 per gallon. Today that price does not seem far away. Furthermore, that analysis did not include byproduct creditS. Humphrey emphasizes the importance of develOping cheaper pretreatment methods, greater cellulose 72 utilization, and new fermentation techniques to yield a product ethanol solution with less water. He also estimates that a 20-25% sugar syrup is most economical for fermentation. In summary, the basic requirements for the economic viability of a process for manufacturing alcohol (or other chemicals) from biomass appear to be: inexpensive pretreatment large yield little byproduct contamination of products sugar obtained in concentrated solutions efficient utilization of hemicellulose and lignin efficient fermentation techniques to minimize alcohol distillation costs. SUMMARY AND COMPARISON OF HYDROLYSIS PROCESSES CONCENTRATED ACID PROCESSES See Table 5. Advantages: Higher yields are obtained than in dilute acid or enzymatic processes. The reaction time is short. The processes can Operate at essentially ambient conditions. The xylose is obtained separately from the glucose. Disadvantages: Consumption of acid is high and recycling is expensive as no efficient methods of acid recovery have been deve10ped. Corrosion problems require special construction materials. Prehydro- lysis and post-hydrolysis are generally required in addition to the main hydrolysis. The material must usually be dried between the prehydrolysis and main hydrolysis steps. Products are obtained in dilute solutions. Lignin is left in a rather nonreactive form due to the harsh hydrolysis conditions. Detailed energy and cost analyses are not available. DILUTE ACID PROCESSES See Table 6. Advantages: Dilute acid hydrolysis appears to be cheaper than 4 enzymatic hydrolysis. 3 Reaction time is shorter than for enzymatic processes. The substrate need not be dried. Xylose can be separated 73 74 TABLE 5 SUMMARY OF REPRESENTATIVE CONCENTRATED ACID HYDROLYSIS PROCESSES SULFURIC ACID Process Hokkaido Nippon U. of Missouri Status full scale plant, 5 ton/day pilot lab 1963 plant Feed hardwood sawdust, waste cornstalks timber, charcoal Yield 83-85% as crys- probably same as 89% of glucose talline glucose Hokkaido Byproducts furfural, xylose molasses, gypsum 94% of xylose Reaction prehydrolysis, 1.2 same as Hokkaido Erehyd, 50 min, Conditions -l.5% acid, 140-150 .4% acid, dry, im- °C, dried, crushed, pregnate with 85% main hyd. 80% acid acid, dilute to 8% rm. temp., post-hyd l0 min, 110°C 100°C, 100 min Acid ion exchange mem— none - neutralized electrodialysis Recovery brane, 80% recovery as 25-35% soln., conc. by evap. Other NaCl, lime lime lime Chemicals Energy 86.1 mil BTU/hr required, 106.1 mil available (to ale) water Costs 7,531,000 cap inv 4.4 mil/yr oper. incl. $381200/ utility cost, $?97/ gal alc, 4.5 mil gal/yr (1979) SULFURIC ACID 75 TABLE 5 (cont.) HCl-LIQUID HCl-GAS Process N Reg Res Lab Udic-Rheinau Noguchi-Chisso Status lab semiiworks plant, pilot plant, 1960 1953-59 Feed Ground cornstalks logs & wastewood, logs, wastewood, chipped to 15 mm reduced to sawdust Yield 85-90% glucose mixed dextrose & 90%, 300 kg dextrose xylose /metric ton wood Byproducts 25% xylose, 69% lignin, polyalcohol xylose furfural from re- sol., xylitol, xy- maining xylose lose Reaction prehyd 100°C, 50- dry, prehyd'w 35% prehyd dilute acid, Conditions 185 min, 4.9-9.8% acid, 20°C, hyd w cool, HCl gas ads. acid, impreg. w. 41% acid, 20°C, heated Off, post 85% acid, 40°C, evap. acid, post hyd dilute to 8%, 45 hyd over 100°C min at 110°C Acid none-neutralized 94-95% recovery, heat to high temp, Recovery vacuum dist, re- eionization after conc, deionization ost hyd after post hyd Other lime Chemicals Energy Water Costs 76 TABLE 6 SUMMARY OF REPRESENTATIVE DILUTE ACID HYDROLYSIS PROCESSES DILUTE SULFURIC ACID Process Scholler Madison Georgia Inst of Tech Status plants - 1931, Marquette pilot lab Europe plant, 1945 P: Feed 'wood chips, sawdust wood chips, 220 ‘wood, other biomass tons/day . Yield 50 gal 190 proof 52-65 gal 190 proof 80-85% alcohol/CDT wood alc/ODT wood, 50% of sugar Byproducts furfural 65-75 t/d lignin lignin Reaction compr & preheated preheated w steam, steam explosion Conditions w steam 265°F, 0.5% 50 psi incr. to lignin extr w etha- acid, 180 psi, 150 psi, total time nol, recycled, pre— forced through in 6 hrs, .4-.85% acid hyd 0.5% acid, 150°C up to 24 batches, hyd 190°C, 5% sugar 45 min/batch cone to l3-20%'by vacuum evap Acid none - consumes 2.3 none - consumes none - neutralized Recovery lb 76% acid/gal alc 26500 lb/d acid Other lime, 2.2 lb/gal lime CaOH or other alkali Chemicals alcohol Energy 190 psi steam, 96 1,106,000 lb 190 lb/gal alc, 0.08 hp psi steam/day, 900 elec/gal, 213 cu ft hp elec, 1700 cu compressed air/gal ft/min compr air water 42.3 gal/gal alc 1,900,000 gal/day Costs $2,247,000 cap inv est selling price (1945) for 11,500 gal/day alcohol for 95% ethanol $1.69/gal 77 TABLE 6 (cont.) DILUTE SULFURIC ACID Process T-V.A. NYULBrenner Grethlein Status pilot plant 1 t/d demo plant, lab 1979 Feed ‘wood chips newspaper pulp, refuse slurry sawdust Yield 60% glucose max. 56% glucose predicted Byproducts Reaction 0.53% acid, ave., 0.5% acid, 450°F 1% acid 20 sec. Conditions percolated through Lwith steam, 500 psi residence time, 240°C wood chip bed, max in twin screw ex— isothermal plug flow T 385°F, 200 psig, truder reactor 145-190 min Acid none none none Recovery Other .lime lime Chemicals Energy 1600 Btu/1b cellulose (1700 Btu/lb sugar) Water Costs $1.90/ga1 alc for 3.46/1b sugar, 250 25 mil gal/ r, 15% ROI, $3Z/0DT rwood waste (1975) ODT/d plant; 2.27 /1b for 1000 CDT d (1974) 78 from glucose under proper conditions, though this is often not done. Disadvantages: High temperatures and pressures are commonly required. Products are obtained in dilute solution. Maximum yield of glucose is approximately 55%. The product streams are badly contaminated with sugar degradation products, including furfurals and organic acids. ENZYMATIC PROCESSES See Table 7. Advantages: The products are obtained in a pure form, as the reaction is specific. No high temperatures or pressures are required (unless they are utilized in the pretreatment process). Xylose is separated if a dilute acid pretreatment is used. The substrate need not be dried. Disadvantages: Extensive pretreatment is required, and is expen- sive and energy intensive. Yields average only about 50%, even after most pretreatments. Long reaction times are required. Enzyme require- ments are high, and only a low percentage of the enzyme can be re- cycled. The cost of production of the enzyme is high, Often amounting to 50% of the total sugar cost. The products are obtained in dilute solution. Sugar costs are high. SOLVENT EXTRACTION Advantages: Yields are higher than in enzymatic or dilute acid hydrolysis without extraction. 79 TABLE 7 SUMMARY OF REPRESENTATIVE ENZYMATIC HYDROLYSIS PROCESSES ‘ENZYMATIC HYDROLYSIS Process Wilke Wilke Natick Status lab lab pilot plant Feed newsprint (885 t/d corn stover newsprint, * 66 for enzyme (wheat stover) delignified pulp production) 1376 t/d Yield 50% conversion, 40% conversion 60% yield from news- 238 t/d glucose, print, in 15% soln, soln. 85% yield from pulp Byproducts lignin for fuel 384 cu ft/d methane from corn stover, single cell protein from wheat Reaction shredding, hammer- prehyd 1% acid, 100 ball milling for Conditions milling, 40 hr at C, 1.1 hr, 5 stage newsprint, hyd 48 45°C, sol/liq ratio hyd over 40 hrs, hrs for news, 72 hrs 1/20 45°C for pulp, 40-50°C, 1 atm, pH 4.79-4.89 hyd soln recycled Enzyme & T. viride, 3.5 T viride, 3.5 FPA T. viride Recovery FPA, 34% recov 58% recov est, high Other Enzyme ferment. lime, ferm nut., formalin, ferm. nut. Chemicals nutrients 4.5 t/d delignified cellulose Energy 8060 kw elec, 9875 kw elec, 36000 lb/hr steam 206000 1b/hr steam (corn stover) Water 83000 gal/hr 196000 gal/hr, corn stover Costs 5.2¢/lb sugar + 4.7p/1b sugar (wh) 35-40¢/kg sugar substrate cost, $23,390,000 fixed cap (1975) (1975), 10.0¢/1b 1c, 5.1p/1b xyl, %corn) (1978) + substrate cost from pulp $29,407,000 cap. 80 TABLE 7 (cont.) SIMULTANEOUS SACCHARIFICATION AND FERMENTATION Process G.E.-U. of Penn. M-I.T. Gulf Status lab lab lab Feed Solka floc, ure cellulose corn residue Avicel & Solka floc) Yield low low; .42-.65 g low product/g cell. degraded Byproducts acetic acid Reaction delignification w anaerobic anaerobic, 40°C Conditions n-butanol, alc re- cov under reduced p, anaerobic Enzyme & Thermoactinomyces, Clostridium T. reesei Recovery Clostridium ther- thermocellum S. cerevisiae mocellum Other nutrients nutrients nutrients Chemicals Energy water Costs 81 Disadvantages: Recovery and recycling Of the solvent is an added process. This is essentially a pretreatment process. Other advantages and disadvantages are contained in the summaries of dilute acid and enzymatic processes, with the above modifications. HYDROGEN FLUORIDE SOLVOLYSIS Advantages: Yields are high (90-95% of total sugars, based on laboratory date). High temperatures or pressures are not required. The products are not degraded. The HF can be recycled without great difficulty. The mild hydrolysis conditions should leave the lignin in a reactive form suitable for further use. Few corrosion problems are expected, as carbon steel can be used with anhydrous HF for most applications. The only pretreatment required for wood is chipping and drying. No prehydrolysis is necessary. The products can be obtained in a concentrated form. Disadvantages: Xylose and glucose are Obtained in a mixed solution. The wood must be thoroughly dried. Post-hydrolysis may be required as a substantial protion of the sugars Obtained are not monomers 0 SUMMARY Because the HF process appears to have distinct advantages over the alternative methods for wood hydrolysis, particularly in yield and in mildness of Operating conditions, it seems worthy of further 82 investigation after being nearly ignored for forty years. This inves- tigation is currently being carried out at Michigan State University. The next section will present a more detailed description of the HF processes deve10ped in Germany in the 1930's. HF SACCHARIFICATION OF WOOD EARLY HF PROCESSES As mentioned earlier, B. Helferich and S. Bottger were the t first to investigate the dissolution of cellulose by HF, although it had been noted by J. Gore as early as 1869 that HF transformed paper, L cotton-wool, and other cellulosic materials into "glutinous substances" and dissolved them.1+9 Helferich's investigation showed that the products of the reaction of anhydrous HF with filter paper (cellulose) were poly- glucans which could be converted to glucose by boiling with dilute acid.2 About 1933 K. Fredenhagen and G. Cadenbach began investigating the action of HF in dissolving filter paper, and discovered that the reaction mechanism involved the formation of glucosyl fluoride. The glucosyl fluoride in HF solution reacts with even small amounts of water to produce glucose and regenerate HF. On precipitation or evaporation of HF, the glucose monomers can recombine to form polyglucans. Next Fredenhagen and Cadenbach investigated the action of HF on wood. They suggested saccharification of wood by HF as a quick method for determining the lignin and carbohydrate contents Of woods, straws, reeds, and similar materials. 83 84 The use of gaseous HF was also investigated, and found to saccha- rify cellulose if the temperature was low enough for a liquid adsorp- tion phase to be formed. Yields up to 95% were obtained with one part wood to one part gaseous HF (by weight). The HF was released 3 by vacuum or flue evaporation at a temperature of 100°C. HOCH AND BOHUNEK Around 1937 two Austrian investigators, Hoch and Bohunek, developed a process using gaseous HF under reduced pressure, in which a liquid adsorption phase was believed not to be required. A pilot plant was constructed in Germany using their process, and ran for six months. The wood was first chipped and dried to a 2-3% water content, evacuated, and then treated with 40 kg of 95% HF to 100 kg of dry wood at a pressure of 30 mm Hg, and a temperature of 35-40°C. Because of the low pressure, the HF penetrated deeply into the wood. After about 30 minutes the reaction was complete. The HF was re- covered by first further reducing the pressure and next raising the temperature to 62°C, while keeping the wood in continual motion. The HF recovered was recycled after contaminating acetic acid was removed by condensation. The reaction products, containing about 1% HF, were hydrolyzed in water or in 2% sulfuric acid under pressure for 30-45 minutes to separate the sugars. The lignin was removed by filtration. About 90% Of the cellulose had reacted to form sugar, of which they were 85 able to recover 80% for further processing. The method was also used on hemp chips, which are rich in pen- toses, with favorable results. It was stated that a fractionated second hydrolysis of the reaction products could satisfactorily separate the pentose sugars 4,50 from hexose sugars. See Figure 24 for flow sheet. RUSSIAN INVESTIGATIONS In the late 1950's Z. A. Rogovin and Yu. L. Pogosov at the Textile Institute in Mescow investigated the action Of 80-100% HF on cellulose in cotton lint at 10-30°C. It was found that 5-10 minutes was sufficient for complete reaction with 95-100% HF.51’52 WOOd Dry Lignin g Figure 24 86 HF Recycle 1 Chipping Drying Reactor HF Cooler fAcetic acid], Lime V Extraction Neutrall- Filtra- Sugars to Fermentation zation * tion , Gypsum ‘ Drying Hoch and Bohunek Process3 {Ii 1 1 Ah.___ REFERENCES l. Sherrard, E. C. &.F. W. Kressman, Ind & Eng Chem, 31, 4 (1945). 20 Helferich, Burckhardt & Stilfrid Bottger, Ann, 4 6, 150 (1929). 3. Fredenhagen, K. & G. Cadenbach, Angewandte Chemie, 46, 113 (1933). 4. Luers, H., Holz Roh und Werkstoff, 1, 342 (1938). 5. O'Neil, Daniel J. et a1, "Design, Fabrication and Operation of a Biomass Fermentation Facility," DOE Report DSE-3060-Tl (1979). 6. Oshima, Motoyoshi, "WOOd Chemistry Process Engineering Aspects," Chem Process MOnograph Series 11, Noyes Dev. Corp., New York, 1965. 7. Locke, Edward G. & Eric Garnum, For Prod J, 1;, 380 (1961). 8. Dunning, J. W. & E. C. Lathrup, Ind & Eng Chem, 31, 24 (1945). 9. Sitton, O. C. et al,.QEP, Dec. 1979, p. 52. 10. Biomass Refining Newsletter, Summer 1980. 11. Faith, N. L., Ind & Eng Chem, 31, 9 (1945). 12. Hokansen, A. E. & R. Katzen, 9E2, Jan. 1978, p. 67. 13. Plow, R. H. et a1, Ind & Eng Chem, 31, 36 (1945). 14° Harris, Elwin E. et a1, Ind & EnggChem,.31, 12 (1945). 15. Grethlein, Hans, Biotech & Bioeng Symp 2 (1975). 16. C a E News, Oct. 8, 1979, p. 19. l7. O'Neil, J. et al, paper presented at International Conference on Energy from Biomass, Brighton, England, Nov. 4-7, 1980. 18. Allen, Alfred L., Curtis R. Blodgett & J. M. Nystrom, AIChE Symp Ser 184, 20 (1979). 19. Brandt, Dixon, Lloyd Hontz & Mary Mandels, AIChE Symp Ser 133: 127 (1973). 87 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33- 34. 35. 36. 38. 88 Gallo, Benedict J. et a1, Biotech & Bioeng Symp 8, 89 (1978). Mandels, Mary, Lloyd Hontz & John Nystrom, Biotech & Bioeng, 16, 1471 (1974). Mandels, Mary, Sheila Dorval & John Medeiros, Proc. Second Annual Fuels from Biomass Symp. (1978). Mandels, Mary & David Sternberg, J Ferment Tech, 54, 267 (1976). Spano, L. A., J. Medeiros & M. Mandels, Conf. on Capturing the Sun Through Bioconversion, Proc., p. 5417(1976). Brandt, Dixon, Biotech & Bioeng Symplé, 275 (1975). Wilke3 Charles R. & Gautam Mitra, Biotech & Bioeng Symp 5, 253 (1975 . Wilke, Charles R., Ren Der Yang & Urs Von Stockar, Biotech & Bioeng Symp §. 155 (1976). Wilke, Charles R., Fuels from Biomass Symp., Proc., U. of I11., p- 115 (1977). Wilke, Charles R. et al, Proc. Second Annual Fuels from Biomass §ymp- (1978)» Batter, T. R. & C. R. Wilke, Report 1977, LBL-6351, “A Study of the Fermentation of Xylose to Ethanol by Fusarium Oxysporum," Lawrence Berkeley Lab, Univ. of Calif., Berkeley, Calif., 1977. MOntenecourt, B. S. & D. E. Eveleigh, Proc. Second Annual Fuels from Biomass Symp; (1978). Su, T. M., AIChE Symp Ser 181, 75 (1978). wang, Daniel I. C. et al, Proc. Second Annual Fuels from Biomass sypp. (1978). Cooney, Charles L. et a1, Biotech & Bioeng Symp 8, 103 (1978). Blotkamp, P. J. et a1, AIChE Symp Ser 181, 85 (1978). C & E News, April 16, 1979, p. 38. Emart, George, paper presented at AIChE 72nd Annual Meeting, San Francisco, 1979. Meyers, Steven G., AIChE Symp Ser 181, 79 (1978). 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 89 Tsao, George, personal communication. Tsao, George, Proc. Second Annual Fuels from Biomass Symp. (1978). Ladisch, Michael R., Christine M. Ladisch & George T. Tsao, Science, 201, 743 (1978). Ladisch, M. R., M. C. Flickinger & G. T. Tsao, Ener ,.4, 263 (1979). Grethlein, Hans, Biotech & Bioeng, 29, 503 (1978). C & E News, Dec. 6, 1976, p. 4. Goldstein, Irving 3., AIChE Symp Ser 177, 111 (1978)° C & E News, April 12, 1976, p. 12. Humphrey, Arthur E., Biotech & Bioeng Symp 2, 49 (1975). Humphrey, Arthur E., Advances in Chem. Ser. 181, 25 (1976). Core, J., J of the Chem Soc, 22, 396 (1869). Luers, Heinrich, Holz Roh und Werkstoff, l, 35 (1937). Rogovin, Z. A. & Yu. L. Pogosov, Gidroliz. i Lesokhim. Prom. igg, No. 1, p. 4 (1958). Chem Abs, 2, 8453i. Rogovin, Z. A. & Yu. L. Pogosov, Nauch. Doklady Vysshei Shkoly, Khim. i Khim. Tekhnol., 1959, No. 2, p. 368. Chem Abs, 23, 22912h. BIBLIOGRAPHY BIBLIOGRAPHY Allen, Alfred L, Curtis R. Blodgett, & J. M. Nystrom, AIChE Symp Ser, 184 (1979). Batter, Thomas R. and C. R. Wilke, "A Study of the Fermentation of Xylose to Ethanol by Fusarium Oxysporum," Lawrence Berkeley Lab., LBL—6351, Berkeley, Calif., 1977. Biomass Refininngewsletter, Summer 1980. Blotkamp, P. J., et a1, AIChE Symp Ser, 181, 85 (1978). Brandt, Dixon, AIChE Symp48er, 133, 127 (1973). , Biotech 8. Bioeng Symp, _5_, 275 (1975). Chem & Eng News, April 12, 1976, p. 12. _____J Dec. 6, 1976, p. 4. _____, April 16, 1979, p. 38. _____, Oct. 8, 1979, p. 19. Congress of the United States. Office of Technology Assessment, Energy from Biological Processes, 1980. Cooney, Charles L., et a1, Biotech & Bioeng Symp, 8, 103 (1978). Dunning, J. W. and E. C. LathrOp, Ind & EnggChem,I3I, 24 (1945). Emart, George, paper presented at AIChE 72nd annual meeting, San Francisco, 1979. Faith, N. L., Ind & Eng Chem, 31, 9 (1945). Fredenhagen, K. & G. Cadenbach, Angewandte Chemie,'&§, 113-117 (1933)- Gallo, Benedict J. et a1, Biotech & Bioeng_Symp,‘§, 89 (1978). Goldstein, Irving 8., Science, 189, 847 (1975). , AIChE Symp Ser, 177, 111 (1978). 90 91 Gore, J., J of the Chem Soc, gg, 396 (1869). Grethlein, Hans, Biotech & Bioeng Symp, 5, 303 (1975). , Biotech & Bioeng, 29, 503 (1978). Harris, Elwin E., et a1, Ind & Eng Chem, 31, 36 (1945). Helferich, B. & S. Bottger, App, filé, 150 (1929). Hokanson, A. E. & R. Katzen, 9E3, Jan. 1978, p. 67. Humphrey, Arthur E., Biotech & Bioeng Symp, 2, 49 (1975). , Adv in Chem Ser, 181, 25 (1976). Ladisch, Michael R., Christine M. Ladisch, and George T. Tsao, Science, 201, 743 (1978). Ladisch, M. R., M. C. Flickinger, and G. T. Tsao, Ener ,.4, 263 (1979). Locke, Edward G. & Eric Garnum, For Prod J, 11, 380 (1961). Luers, Heinrich, Holz Roh und Werkstoff, 1, 35 (1937). , Holz Roh und Werkstoff,‘l, 342 (1938). Mandels, Mary, Lloyd Hontz and John Nystrom, Biotech & Bioeng, 16, 1471 (1974). Mandels, Mary & David Sternberg, J Ferment Technol, 54, 267 (1976). Mandels, Mary, Sheila Dorval & John Medeiros, Proc Second Ann Fuels from Biomass Symp (1978). Meyers, Steven G., AIChE Symp Ser, 181, 79 (1978). Montenecourt, B. S. & D. E. Eveleigh, Proc Second Ann Fuels from Biomass Symp (1978). O'Neil, Daniel J. et a1, "Design, Fabrication and Operation of a Biomass Fermentation Facility," US DOE Report DSE-3060-Tl, 1979. , International Conference on Energy from Biomass, Brighton, England, Nov. 4-7, 1980, in press. Oshima, MOtoyoshi, WOOd Chemistry Process EngineeringyAspects, Chemical Process MOnograph Series 11, Noyes Dev. Corp., New York, 1965. 92 Plow, R. H., et al, Ind a. EngChem, 37, 36 (1945). Rogovin, Z. A. and Yu. L. Pogosov, Gidroliz. i Lesokhim. Prom., 11, No. 1, p. 4 (1958). Chem Abs, 5_2, 8453i. , Nauch. Doklady Vysshei Shkoly, Khim. i Khim. Tekhnol., No. 2, p. 368 (1959). Chem Abs, 2, 22912h. Sherrard, E. C. & F. W. Kressman, Ind & Eng Chem, 37, 4 (1945)- Sitton, O. C., G. L. Foutch, N. L. Book, & J. L. Gaddy, CEP, Dec. 1979: P° 52- Spano, L. A., T. Medeiros & M. Mandels, Conf. on Capturing the Sun through Bioconversion, p. 541, 1976. 511- T- M: AIChE §YIHD Ser. 181, 75 (1978). Tsao, George, Proc. Of 2nd Annual Fuels from Biomass Symp. 1978. Wang, Daniel I. C., et al, Proc. Second Annual Fuels from Biomass RITE: l978° Wilke, Charles R., Ren der Yan , and Urs von Stockar, Biotech & Bioeng Symp. Q. 155 (1976?. Wilke, Charles R. & Gautam Mitra, Biotech & Bioeng Symp, 2, 253 (1975). Wilke, Charles R., Fuels from Biomass Symp - Proc, 115 (1977). Wilke, Charles R., et al, Proc. Second Annual Fuels from Biomass gXTE: 1978- "1111111111i“