Acid and enzymatic hydrolysis of autohydrolyzed lignocellulosic substrates by David Allen Lamar A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Chemical Engineering Montana State University © Copyright by David Allen Lamar (1987) Abstract: Four biomass residues (barley straw, wheat straw, lodge pole pine, and Douglas fir) were analyzed for effects of an autohydrolysis pretreatment on lignin extractibility and cellulose hydrolysis, both acid and enzymatic catalyzed. A pure cellulose substrate, Chromedia, was also used for hydrolysis tests. The conditions used for the autohydrolysis were 205 °C and 10 minutes. These conditions were those found to be optimal for lignin extractibility from wheat straw during previous work at this laboratory. The extractibility of lignin (by an ethanol-water solvent) following pretreatment was very similar for barley straw and wheat straw. A total of about 75% of the lignin was removed during the autohydrolysis and subsequent solvent extraction. The amounts of lignin removed from the two woods were also very similar with about 30% of the lignin being removed. Experiments were performed to determine the effects of lignin content and substrate morphology on acid hydrolysis of lignocelluloses. These experiments involved hydrolyses on lignin-free substrates and substrates containing lignin. The delignification procedure resulted in a substrate that was no more hydrolyzable than a non-pretreated substrate with the lignin intact. Acid hydrolyses on ball-milled Chromedia revealed that as the amorphous content of the substrate increases the rate of hydrolysis also increases. Pretreated and non-pretreated substrates were hydrolyzed using both sulfuric and hydrochloric acids and enzymes to determine the pretreatment effect on degree of hydrolysis. Pretreatment of wood substrates resulted in only slightly increased carbohydrate conversion via acid hydrolysis over that observed for non-pretreated woods. No increase in hydrolysis rate was observed for pretreated straw substrates when acids were used as the catalytic agents. When mixed enzymes were substituted for acids in wheat straw hydrolysis, the cellulose conversions increased dramatically for pretreated substrates, with values in excess of 90% observed. A theory based upon the solubility of reaction products is presented to explain higher cellulose conversions with mixed enzymes versus acid hydrolysis results. This theory leaves open the possibility that autohydrolysis pretreatment renders all substrates investigated more subject to hydrolysis. A C ID AND ENZYMATIC HYDROLYSIS OF AUTOHYDROLYZED L I GNOC ELLULOSI C SUBSTRATES by D a v id A l l e n Lam ar A t h e s i s s u b m itt e d i n p a r t i a l f u l f i l l m e n t o f t h e r e q u ir e m e n t s f o r t h e d e g r e e of M a s t e r o f S c ie n c e in C h e m ic a l E n g in e e r in g MONTANA STATE U N IV E R S IT Y B ozem an, M o n ta n a A ugust 1987 P N main Ufa. JlfiIS APPROVAL o f a th e s is submitted by David A lle n Lamar This th e s is has been read by each member o f the th e s is committee and has been found to be s a tis fa c to ry regarding content, English usage, form at, c ita tio n , b ib lio g ra p h ic s ty le , and consistency, and is ready fo r submission to the College o f Graduate S tudies. / 7 D a te T j M S ? Chairperson, Graduate C d p itte e Approved fo r the Major Department Date; o Jinn Heajd, Major Department Approved fo r the College o f Graduate Studies Date Graduate Dean STATEMENT OF PERMISSION TO USE In presenting requirements fo r th is th e s is in p a r tia l f u lf illm e n t , o f a m a sters's degree a t Montana S tate U n iv e rs ity , the I agree th a t the L ib ra ry s h a ll make i t a v a ila b le to borrowers under ru le s of the L ib ra ry . B rie f quotations from th is th e s is are allow able w ith o u t special perm ission, provided accurate acknowledgement o f source is made. Permission f o r extensive quotation from o r reproduction o f th is th e s is may be granted by my major p ro fe sso r, o r in h is absence, by the Dean o f L ib ra rie s when, in the opinion o f e ith e r, the proposed use o f the m a teria l is f o r s c h o la rly purposes. Any copying o r use o f the m a teria l in th is th e s is fo r fin a n c ia l gain sh a ll not be allowed w ith o u t my perm ission. Signature iv ACKNOWLEDGEMENTS The fa c u lty and s t a f f o f the Engineering Department deserve and have my g ra titu d e f o r the help and support th a t they gave to me. I would e s p e c ia lly lik e to thank Dr. Daniel S h a ffer, my frie n d and advisor, f o r the advice and guidance th a t he gave me w hile I was a t Montana State U n iv e rs ity . My good frie n d s Steve J e tte and Ron Nakaoka were a great help in the completion o f th is e f f o r t . itu d e and love to my fa m ily , s a c rific e s Karen, Most o f a ll I owe eternal g ra t­ Frank and Nathan, w ith o u t whose and encouragement the completion o f th is degree would not have been p o s s ib le . Last, I would lik e to thank the Novo Corporation fo r g iv in g me the enzymes used during th is study. V TABLE OF CONTENTS APPROVAL ....................................... ii STATEMENT OF PERMISSION TO U S E ............................................................ i i i ACKNOWLEDGEMENTS................................................................................... IV TABLE OF CONTENTS................................................................................... v LIST OF T A B LE S ............................................................................................v i i LIST OF FIGURES..........................................................................................v i i i ABSTRACT................................................................................................... INTRODUCTION ........................................... Research O bjectives ix I ................................................................... 2 STRUCTURE OF LIGNOCELLULOSE............................................................... .3 C e l l u l o s e ...................................................................... 4 Hemi c e ll u l o s e ............................................................................... 5 Lig n in ............................................................................................... 6 S tru ctu re as I t Relates to H ydrolysis .................................. 6 PRETREATMENTS TO ENHANCE HYDROLYSIS ................................................ 11 EXPERIMENTAL ........................................................................................... 14 S u b s tra te s ....................................................................................... 14 C h a ra cte riza tio n o f Substrates ................................................ 15 A s h ........................................................................................... 15 M o is t u r e ............................................................................... 15 vi TABLE OF CONTENTS—Continued I ------------------------ E x t r a c t i b l e s ........................................................... ... 16 L i g n i n ................................................................... 17 C ellulo se and Hemic e llu lo s e ............................................ 19 A lpha-C ellulose A nalysis ........................................ 20 Beta- and Gamma-Cellulose A n a ly s is ................... 21 A utohydrolysis and Lignin E x tra c tio n .................................... 22 A u t o h y d r o ly s is ....................... 22 D e lig n if ic a t io n ............................................... 26 B all M illin g . ....................................................................... 26 Dry B all M i l l i n g ................................................................ 27 Wet B all M i l l i n g ............................................................... 27 Acid and Enzymatic H y d r o l y s i s ................................................ 27 Acid H y d r o ly s is ................................................................... 28 Enzymatic H ydrolysis ........................................................ 29 ....................................................................... 31 A utohydrolysis o f Substrates .................................................... 31 RESULTS AND DISCUSSION Barley Straw ........................................................................ 32 Douglas F ir and Lodge Pole P in e .................................... 34 Acid H ydrolysis E x p e r im e n ts .............................. 37 Barley S t r a w ............................... Douglas F ir and Lodge Pole P in e ............................ ... 37 . 37 Wheat S tr a w ........................................................................... 43 C hrom edia............................................................................... 46 Enzymatic H ydrolysis Experiments ............................................ 49 CONCLUSIONS....................................... 60 SUGGESTIONS FOR FUTURE RESEARCH ........................................................ 61 REFERENCES CITED 62 APPENDIX ................................................................................... ................................... 65 vi i LIST OF TABLES 1. Major Component Composition o f L ig noce liu lose .................... 2. Weight Percent Composition o f Substrates ............................ 31 3. Weight Percent Ash o f Ethanol-Benzene Extracted Substrates 32 4. Comparison o f A utohydrolysis Experiments f o r Barley Straw and Wheat S tr a w ............................................................................... 5. . 38 . ................................................ 39 Carbohydrate Conversion Results o f H2SO4 H ydrolysis o f Wood S u b s tra te s .................... ... 8. Comparison of Douglas F ir Lignin E x tr a c tib ility by Two S o lv e n t s ........................................................................................... 9. 36 Carbohydrate Conversion Results o f Acid H ydrolysis on Straw S u b s t r a t e s ................................................................... ... 7. 35 Summary o f A utohydrolysis Experiments on Wood S u b s t r a t e s ....................................................................................... 6. 3 40 Carbohydrate Conversion Results o f Autohydrolyzed Dioxane-Water Extracted Douglas F ir ........................ . . . . 10. C ellulo se A nalysis A fte r Acid H ydrolysis 41 ............................ 42 11. Carbohydrate Conversions o f Douglas F ir Substrates by HCl ............................................... 44 12. Carbohydrate Conversions o f Wheat Straw Substrates by HCl . . ........................................................................................... 45 13. Comparison o f Amorphous C ellulo se Content versus Amount o f C ellulose Converted ........................................... 47 14. C ellulo se Analyses on M ille d and Non-Milled Chromedia . . 48 15. Results o f Acid H ydrolysis on Chromedia ................................ 48 16. Results o f Enzymatic Hydrolyses on Wheat Straw 57 ................ 17. Results o f the Acid H ydrolysis Development Experiments . 65 v iii LIST OF FIGURES 1. Basic S tru ctu re o f a Plant C ell . . . . ................................ 4 2. Chemical S tru ctu re o f C ellulose . , ........................................ 5 3. S tru c tu re o f a P ortion o f L i g n i n ............................................ 7 4. An Acid Catalyzed H ydrolysis Reaction .................................... 8 5. Funnel Setup used fo r Lig nin Determination . . .................. 18 6. Sample Basket Apparatus ............................................................... 23 7. Setup used fo r A utohydrolysis and E xtra ctio n Experiments 24 8. E ffe c ts o f A utohydrolysis and Dioxane E x tra c tio n on Aspen 9. Woodmeal Lig nin ............................................................................... 33 A Proposed Mechanism fo r an Enzymatic H ydrolysis 51 . . . . 10. E ffe c ts o f C ellulase A c t iv it y fo r a 6 hour H ydrolysis . . 52 11. E ffe c ts o f Time on a C ellulase H ydrolysis ................. . . 53 12. E ffe c ts o f Cellobiase A c t iv it y on a 6 hour H ydrolysis . . 55 13. E ffe c ts o f Time on a C e llo b ia se -C e llu la se H ydrolysis . 56 . . ix ABSTRACT Four biomass residues (b a rle y straw , wheat straw , lodge pole pine, and Douglas f i r ) were analyzed fo r e ffe c ts o f an a u to h yd ro lysis p re tre a t­ ment on lig n in e x t r a c t i b i l i t y and c e llu lo s e h y d ro ly s is , both acid and enzymatic cata lyzed. A pure c e llu lo s e su b stra te , Chromedia, was also used f o r h y d ro ly s is te s ts . The co n d itio n s used fo r the a u toh ydro lysis were 205 °C and 10 min­ utes. These co n d itio n s were those found to be optimal f o r lig n in e x tra c t­ i b i l i t y from wheat straw during previous work a t th is la b o ra to ry . The e x t r a c t i b i l i t y o f lig n in (by an ethanol-w ater so lve n t) fo llo w in g pretreatm ent was very s im ila r fo r ba rle y straw and wheat straw . A to ta l o f about 75% o f the lig n in was removed during the a u toh ydro lysis and subsequent solve nt e x tra c tio n . The amounts o f lig n in removed from the two woods were also very s im ila r w ith about 30% o f the lig n in being removed. Experiments were performed to determine the e ffe c ts o f lig n in content and sub strate morphology on acid h y d ro ly s is o f lig n o c e lIu lo s e s . These experiments involved hydrolyses on lig n in - fr e e substrates and substrates co n ta in in g lig n in . The d e li g n i f i ca tio n procedure re s u lte d in a substrate th a t was no more hydrolyzable than a non-pretreated sub strate w ith the lig n in in ta c t. Acid hydrolyses on b a ll- m ille d Chromedia revealed th a t as the amorphous content o f the substrate increases the ra te o f hyd ro lysis also increases. P retreated and non-pretreated substrates were hydrolyzed using both s u lfu r ic and h yd ro ch lo ric acids and enzymes to determine the pre­ treatm ent e ffe c t on degree o f h y d ro ly s is . Pretreatment o f wood substrates re su lte d in only s lig h t ly increased carbohydrate conversion v ia acid h y d ro ly s is over th a t observed fo r non-pretreated woods. No increase in h y d ro ly s is ra te was observed fo r pre tre a te d straw substrates when acids were used as the c a ta ly tic agents. When mixed enzymes were s u b s titu te d f o r acids in wheat straw h y d ro ly s is , the c e llu lo s e conversions increased d ra m a tic a lly f o r p retre ate d su b strates, w ith values in excess o f 90% observed. A theory based upon the s o lu b ilit y o f re a ctio n products is presented to exp lain higher c e llu lo s e conversions w ith mixed enzymes versus acid h y d ro ly s is re s u lts . This theory leaves open the p o s s ib ilit y th a t autoh y d ro ly s is pretreatm ent renders a ll substrates in v e s tig a te d more subject to h y d ro ly s is . I INTRODUCTION The re a liz a tio n th a t petroleum supplies are not lim itle s s has awak­ ened the U.S. to the need o f an a lte rn a tiv e hydrocarbon source fo r fu e l and chemical fee d-stocks. A search has been focused on determ ining a hydrocarbon source th a t does not compete w ith food supplies o r oth er valuable raw m a te ria ls . An ideal source o f hydrocarbon might be waste o r by-product biomass from a g ric u ltu ra l operations o r wood product in ­ d u s trie s . The u t iliz a t io n o f th is biomass would b e n e fit the above in ­ d u s trie s by converting low value m a te ria ls in to more valuable commodities. The energy requirement o f the U.S. is roughly 76 q u a d r illio n BTU's (Quads) per ye a r. The generation o f th is energy would re q u ire 40 m illio n b a rre ls o f o i l per day. A major goal is to replace expensive petroleum used in energy production w ith low cost biomass [ I ] . P rese ntly, two to three Quads o f the U.S. energy requirements are supplied by biomass u t iliz a t io n . Combustion o f fo re s t products is the prim ary source fo r th is energy. A conservative estim ate o f the energy th a t w i l l be supplied by u t iliz a t io n o f biomass by the end o f th is century is 15 Quads [ I ] . Biomass fo r energy could be supplied from d ir e c t and in d ir e c t sources. D ire c t sources might include farms developed to grow pla nts s o le ly fo r energy uses, w hile in d ir e c t sources would includ e a g ric u l­ tu ra l and fo re s t waste o r by-products mentioned above. The in d ir e c t sources are o f prim ary in te re s t in th is study. I t is estim ated th a t 278 m illio n dry tons o f a g ric u ltu ra l by-products and 108 m illio n dry tons o f unused m ill and logging residues are produced annually [ I ] . T h e o re tic a lly , i f these m a te ria ls were converted to glucose and the glucose fermented to a lc o h o l, approxim ately 30 b i l l i o n gallons o f ethanol could be produced per ye a r. This ethanol would meet the e n tire c u rre n t in d u s tria l demand and provide ethanol fo r gasoline blending 2 as w e ll. Most o f the present in d u s tria l grade ethanol is now produced from petroleum-based feedstocks. This use o f biomass would th e re fo re reduce the demand fo r petroleum [ I ] . The above estim ate o f ethanol from IignocelluT ose requires a 90 percent conversion o f the c e llu lo s e to glucose. B a rrie rs e x is t in I i g - n o cellulose s th a t prevent such high conversion o f c e llu lo s e to glucose. E x is tin g technologies on ly provide about a 50 percent c e llu lo s e conver­ sio n . A hig her conversion is d e sira b le to econom ically produce c e llu lo s e derived products. Research O bjectives The f i r s t o b je c tiv e o f th is in v e s tig a tio n is to te s t three lig n o c e llu lo s ic substrates fo r degree o f lig n in removal and enhanced hyd ro lysis o f t h e ir c e llu lo s e to glucose a fte r a novel pretreatm ent. The substrates to be in v e s tig a te d are ba rle y straw , lodge pole pine, and Douglas f i r . These m a te ria ls w i l l be pre tre a te d by au toh ydro lysis a t c o n d itio n s found to be optimum fo r lig n in removal from wheat straw during previous work a t th is la b o ra to ry . The second o b je c tiv e is to in v e s tig a te reasons th a t might explain the low acid h y d ro ly s is y ie ld s observed w ith p re tre a te d wheat straw . Meeting th is o b je c tiv e e n ta ils applying new h y d ro ly tic c a ta ly s ts and/or co n d itio n s to the hydrolyses in an attem pt to increase y ie ld s . 3 STRUCTURE OF LIGNOCELLULOSE Forest and a g ric u ltu ra l residues c o n s is t o f several prim ary compo­ nents. These components include c e llu lo s e , hemic e llu lo s e , lig n in , pro­ te in , and miscellaneous e x tr a c tib le s . Table I summarizes the percent composition o f in d iv id u a l components by w eight. Table I . Major Component Composition o f L ig n o ce llu lo se [2 ]. Component C ellulo se Hem icellulose Lignin E x tra c tib le and P rotein % Composition 45-50 20-25 20-30 0- 1:0 Together these components make the basic s tru c tu ra l u n it o f b io ­ mass, the p la n t c e ll. A c e ll, in simple terms, is composed o f two basic p a rts , the lumen and the c e ll w a ll. o f the c e ll. The lumen contains the liv in g m atter Once dead the c e l l 's lumen is e ith e r void space o r f i l l e d w ith e x tra c tib le s [2 ] . The c e ll w all serves as a mechanical d iv id e r between in d iv id u a l c e lls . R ig id ity o f a p la n t s ta lk is the d ir e c t re s u lt o f i t s c e ll w a lls . The c e ll w all also consists o f two p a rts , the prim ary w all and the secondary w a ll. The prim ary w all is very th in in comparison w ith the secondary w a ll. The secondary w all con sists o f three d is t in c t layers termed the o u te r (S1) , middle (S2) and in n e r (S3) laye rs [3 ] . Surrounding the prim ary w all and separating adjacent c e lls is the middle lam e lla. Figure I [1] is a re pre senta tion o f the basic s tru c tu re o f a p la n t c e ll. C ellulo se is m ainly in the m ic r o f ib r ils , shown as lin e s in the diagram. O rie n ta tio n o f the f i b r i l s is d iffe r e n t in the respective p o rtio n s o f the c e ll w a ll. The amount o f c e llu lo s e in the p la n t is highest in the secondary w all and decreases toward the middle la m e lla . Hemic e llu lo s e has i t s 4 • Secondary wall Primary wall Middle lamella Figure I . Basic S tru ctu re o f a Plant Cell [ I ] . highest percentage lumen. H em icellulose c e llu lo s e spaces in f ib r ils . between the middle and lig n in Lignin the lam ella and decreases form a m a trix and hemic e llu lo s e c r y s ta llin e regions of th a t toward the surrounds the found in the are also the m ic r o f ib r ils , the amorphous (n o n -c ry s ta llin e ) regions [2 ]. C ellulose C e llu lo se is a lin e a r serving as the monomer. bonds. Figure 2 is polymer of D-anhydroglucose molecules These monomers are lin ke d by /? -l-4 -g lu c o s id ic a schematic o f a c e llu lo s e m olecule. C ellulose degree o f p o lym eriza tion ranges from 3,500 to 14,000 glucose u n its when in a n a tiv e form. The average length o f c e llu lo s e molecules range from 2,500 nm to 5,000 nm [2 ,5 ]. The lin e a r molecules lay one on another forming bundles of molecules, f i b r i l s , th a t are held tog eth er by la te ra l hydrogen bonding. The la rg e number o f hydrogen bonds re s u lt in c r y s ta llin e regions o f about 60 nm in length th a t comprise 67 to 90% o f the c e ll w a ll. Since the c e llu lo s e molecule is longer than 60 nm, the molecules pass through several c r y s ta llin e and amorphous regions [2 ]. F ib r ils are surrounded by a sheath o f hem icellulose and lig n in [5 ]. The apparent morphology o f c e llu lo s e depends on the methods o f a n a lysis and also the source o f the c e llu lo s e . At present c e llu lo s e is 5 Figure 2. Chemical S tru ctu re o f C e llu lo s e . categorized in to fo u r d is t in c t types, c e llu lo s e I, II, m , and IV. Each group is based on the aggregation o f molecules in the c r y s ta llin e s o lid . Degree o f c r y s t a l li n i t y decreases w ith c e llu lo s e type ( i . e . , C ellulo se I is more c r y s ta llin e than C ellulo se I I , class and so f o r t h ) . The to which a p a r tic u la r c e llu lo s e belongs depends on the method used to produce the pure c e llu lo s e . The class is id e n tifie d by the xray d if f r a c t io n p a tte rn o f the sample [4 ]. C ellulo se in i t s n a tive form is classed C ellulo se I . is a c e llu lo s e th a t has been regenerated C ellulose I I from s o lu tio n at ambient temperatures o r one th a t has been mercerized w ith c a u s tic s o lu tio n in excess o f 15% sodium hydroxide. morphologies. several C ellulo se I and I I are the most common Treatment o f c e llu lo s e w ith anhydrous ammonia o r one o f d iffe r e n t amines produces c e llu lo s e III. Heat treatm ent o f c e llu lo s e I I o r regeneration o f c e llu lo s e from s o lu tio n s at elevated temperatures re s u lts in c e llu lo s e IV [4 ] . Hemicel Iulo se Hem icellulose is a polymer o f simple sugar molecules lik e c e llu ­ lose, though i t con sists o f more than one type o f sugar. The backbone o f the polymer is a lin e a r chain co n tainin g D-xylose sugar u n its lin ke d to g e th e r by /? - l- 4 - g lucosidic bonds. is not a lin e a r homopolymer. U nlike c e llu lo s e , hem icelIulose I t contains side chains branching from 6 the main chain o f D-xylose sugars. the xylose contain molecules v ia 1-3 The branches are u s u a lly bonded to g ly c o s id ic 1-4 and 1-6 g ly c o s id ic bonds. glucose, glucose xylose, galactose, and galactose. mannose, lin k s , they can also The side chains can contain arabinose, Composition but of and uronic acids o f a p a r tic u la r hemic e llu lo s e va rie s from source to source, not only w ith p la n t species, but also w ith clim a te and lo c a tio n o f the p a r tic u la r p la n t. za tio n o f hemic e llu lo s e exceeds 200 [ 1, 2] . Degree o f polym eri­ ranges from 100 to 200 molecules but ra re ly Hem icellulose does not form c ry s ta ls lik e c e llu lo s e and is found on ly in an amorphous s ta te . It is u s u a lly in in tim a te contact w ith p la n t lig n in s . I t is thought th a t the hemic e llu lo s e and lig n in chem ically bonded to g e th e r. are Lignin Lig n in is a h ig h ly complex, three-dim ensional polymer o f various phenolic acids connected by ether lin ka g e s. U nlike the o th e r compo­ nents discussed so fa r , lig n in has no set p a tte rn o f s tru c tu re . Figure 3 [1] presents a possible s tru c tu re o f lig n in . Lig n in to g e th e r. acts as the cement th a t holds the c e llu lo s e fib r ils I t is an in te g ra l p a rt o f a system th a t gives p la n ts t h e ir stre ngth and r i g i d i t y . This polymer not only acts as a cement but also as a p ro te c tiv e s h ie ld against elements th a t would otherw ise destroy the p la n t by a tta c k in g the c e llu lo s e . lig n in [6]. can l i m i t the m icro bial In v e s tig a tio n s have shown th a t degradation o f p la n t polysaccharides S tru ctu re as I t Relates to H ydrolysis A l o t o f work has been done to determine how the chemical physical s tru c tu re of carbohydrate components. IignoceT lulose in h ib its h y d ro ly s is of and the The fo llo w in g is an overview o f t h is work and some o f the conclusions th a t have been drawn. Many d iffe r e n t aspects o f p la n t s tru c tu re and chemical make-up 7 Figure 3. S tru ctu re o f a P ortion o f Lig nin [ I ] . have been id e n tifie d as detrim ental to the h y d ro ly s is o f biomass. are the hemic e llu lo s e - lig n in b a r r ie r , c r y s t a l li n i t y These o f the c e llu lo s e , surface area, degree o f polym erization o f c e llu lo s e , and pore size d is ­ t r ib u t io n . Some o f these p ro p e rtie s act s y n e r g is tic a lly to prevent h y d ro ly s is . here. The ra m ific a tio n s o f in d iv id u a l aspects w ill be discussed H ydro lysis o f c e llu lo s e can be catalyzed by e ith e r acids o r enzymes. Although chem ically the re s u lts are the same, the mechanisms are q u ite d if f e r e n t . A successful acid catalyzed h y d ro ly s is (see Figure 4) takes place when the oxygen atom o f the /M ,4 - g ly c o s id ic bond is attacked by a hydrogen io n . This a tta ck re s u lts in a p o s itiv e charge on the oxygen which then p u lls e le ctron s from the oxygen-carbon bond re s u ltin g in a p a r tia l p o s itiv e charge on the carbon atom. Non-bonding ele ctron s o f a water oxygen atom are a ttra c te d to th is p a r tia l p o s itiv e charge. Elec­ tro n s from the o rig in a l carbon-oxygen bond form a bond w ith the a tta ckin g hydrogen io n , and hydrogen-oxygen bonding e le ctron s from the water mole­ cule form a bond between the carbon atom and the a tta c k in g oxygen atom re le a sin g a hydrogen atom. This series o f events re s u lts in the breaking o f the /? -l,4 -g ly c o s id ic bond by the a d d itio n o f a water molecule in between two glucose monomers o f the c e llu lo s e molecule. A proposed mechanism fo r an enzyme catalyzed h y d ro ly s is is presented la te r in th is work. CH2OH Figure 4. CH 2OH An Acid Catalyzed H ydrolysis Reaction. Hemic e llu lo s e and lig n in are believed to be chem ically bonded in biomass. These two components p ro te c ts the c e llu lo s e [3 ]. form a sheath th a t surrounds and The I ig n in -h e m ic e llu lo s e b a rrie r prevents con tact o f h y d ro ly tic agents (such as enzymes and acid) and c e llu lo s e molecules, thus preventing h y d ro ly s is . Removing e ith e r hemicellu lo s e , lig n in , the or both, improves access to c e llu lo s e increasing the h y d ro !iz a b iI i t y o f the c e llu lo s e [7 ,8 ,9 ,1 0 ,1 1 ]. It has been proposed th a t c e ll u l y t i c themselves to lig n in molecules. enzymes ir r e v e r s ib ly attach This adsorption removes the a ffe c te d enzymes from the re a ctio n m ixture [1 2 ]. Increasing enzyme charge to the h y d ro ly s is m ixture only increases h yd ro lysis up to a p o in t. This im p lie s th a t enzyme adsorption is not the only in h ib ito r y mechanism o f Iig ni n . Several chem ically researchers have shown th a t some aromatic compounds can in h ib it enzymatic h y d ro ly s is Aromatic compounds having an in h ib ito r y from the breakdown lig n o c e llu lo s e . of lig n in molecules As an example, of hem icellulose a ffe c t are products derived or hot water e x tra c ts of wheat e x tra c t norm ally contains p- coumaric a cid , f e r u lic acid , and v a n illic acid [1 4 ]. have been shown to be detrim ental [1 3 ]. to d ig e s tio n , These compounds and hence enzymatic h y d ro ly s is , o f c e llu lo s e in rumenic animals [6 ]. The h e m ice llu lo se -1 ig n in b a rrie r is not the on ly c h a ra c te ris tic preventing conversion o f c e llu lo s e to glucose. Evidence o f th is can be 9 ascertained from the h yd ro lysis o f c o tto n . no lig n in Cotton contains v ir t u a lly and only small amounts o f hem icellulose [3 ]. The lack o f lig n in and low amounts o f hem icellulose should render cotton c e llu lo s e h ig h ly Several susce p tib le to h y d ro ly tic a tta c k by e ith e r enzymes o r acids. in v e s tig a to rs in I t a ly have shown th a t th is is not the case. A fte r a 24-hour enzymatic h yd ro ly s is they found on ly a 22% conversion o f c e llu lo s e to sugar [1 5 ]. This in d ic a te s th a t the hem icellulose- lig n in b a r r ie r is not always the major in h ib it o r o f h y d ro ly s is . Saddler e t a l . [ 8] and Fan e t a l . [16] have studied the e ffe c ts o f c e llu lo s e c r y s t a l li n i t y on degree of h y d ro ly s is . They separately reported th a t c r y s t a l li n i t y may represent a major block to c e llu lo s e conversion. They found th a t h ig h ly c r y s ta llin e c e llu lo s e re s is ts h y d ro ly s is , whereas amorphous c e llu lo s e is re a d ily hydrolyzed. experiments sub strate show th a t as c r y s t a l li n i t y a h y d ro ly s is increases. proceeds, This the suggests T h eir ,unhydrolyzed th a t amorphous regions o f c e llu lo s e are p r e fe r e n tia lly hydrolyzed lea vin g c r y s ta llin e sub strate in ta c t. As the c r y s t a l li n i t y increases, the ra te o f conver­ sion slows [1 0 ,1 6 ]. Although these researchers have placed c r y s t a llin ­ i t y high on the l i s t o f b a rrie rs to h y d ro ly s is , they agree th a t i t is not the g re a te st block to h yd ro ly s is [8 ,1 0 ]. T h e ir data c o rre la tio n s seem to in d ic a te th a t surface area has the la rg e s t a ffe c t on hyd roly­ s is . G re th le in [17] has suggested th a t the enzyme-catalyzed h yd rolysis o f c e llu lo s e is not dependent on the c r y s t a llin it y in c e llu lo s e . He belie ves surface area is the most im portant c h a ra c te ris tic c o n trib u tin g to the ra te of surface area is h y d ro ly s is . d ir e c t ly The enzymatic h y d ro ly s is dependent on the pore size fo r a given d is tr ib u tio n . There may be a larg e surface area, but pore size may l i m i t the amount o f surface accessible to enzymatic a tta c k . G re th le in discusses a pore size o f 51 X based on the size o f h y d ro ly tic enzymes. Acid molecules are g e n e ra lly very small in comparison w ith enzyme molecules. an e ffe c t Therefore, pore size lim ita tio n s should not have as great on the ra te o f acid h y d ro ly s is . Several researchers at Dartmouth College have found the ra te constants fo r acid h y d ro ly s is o f 10 c e llu lo s e to glucose fo r a wide v a rie ty o f su b strates. These constants are a ll w ith in the same order o f magnitude at a given temperature and acid co n ce n tra tio n . The substrates varie d in c r y s t a l li n i t y , surface area, and pore size d is tr ib u tio n [1 8 ]. Degree o f po lym eriza tion o f c e llu lo s e some b e lie ve has an e ffe c t on h y d ro ly s is . chain molecules are less s h o rte r chain molecules. o f c e llu lo s e re s u lts accessible to is a c h a ra c te ris tic th a t T heir b e lie f is th a t long h y d ro ly tic agents than are This lowered access to the g lu c o s id ic bonds in slowed h y d ro ly s is rates and less conversion. These researchers have shown th a t as chain length increases, hyd ro lysis ra te decreases [1 9 ]. There is a general consensus th a t the hemic e llu lo s e - lig n in b a rrie r slows the h y d ro ly s is o f c e llu lo s e , e s p e c ia lly when enzymes are used as a c a ta ly s t. A disagreement c r y s t a l li n i t y and purpose o f th is surface research e x is ts area p e rta in in g e ffe c ts study was to to c e llu lo s e in v e s tig a te the degree th a t h y d ro ly s is . The the e ffe c t o f a pretreatm ent on degree and ease o f conversion o f c e llu lo s e to glucose. This researcher was aware o f c r y s t a l li n i t y and surface area as in h ib i­ to ry fa c to rs , but the degree o f in d iv id u a l importance was not in v e s ti­ gated. 11 PRETREATMENTS TO ENHANCE HYDROLYSIS D iffe re n t treatm ents p r io r to a h y d ro ly s is have been studied to e ffe c t an increase in the s u s c e p tib ility o f lig n o c e llu lo s e to h yd ro ly­ t i c agents. Physical These pretreatm ents can be chemical o r physical in nature. pretreatm ents are p rim a rily th a t reduce p a r tic le s iz e . m illin g o r g rin d in g operations Chemical pretreatm ents in vo lve su b jecting the lig n o c e llu lo s e to a chemical agent. Some pretreatm ents invo lve a combination o f physical and chemical elements. A novel method o f using physical treatm ents sim ultaneously w ith h y d ro ly s is has been studied a t the U n iv e rs ity o f Montana. T h e ir process involved a simultaneous wet b a ll m illin g and enzymatic h y d ro ly s is o f various c e llu lo s e sub stra te s. This process continuously produces a c tiv e s ite s f o r h y d ro ly tic agents to a tta ck [7 ]. Although physical p re tre a t­ ments may enhance the h yd ro lysis o f lig n o c e lIu lo s e s , they are very energy in te n s iv e o p era tions. b ility o f th is Therefore, High energy consumption makes the economic fe a s i­ type o f pretreatm ent doubtful f o r com m ercialization. the prim ary work o f developing a pretreatm ent process has been focused on chemical pretreatm ents. combinations of chemical and physical Some studies are looking at d is ru p tio n of lig n o c e llu lo s ic s tru c tu re . As stated above, chemical pretreatm ents in vo lve s u b je c tin g lig n o ­ c e llu lo s e to a chemical agent. Some o f the more common agents are s o l­ u tio n s o f acids o r a lk a li, organic so lve n ts, and hot w ater, both liq u id and vapor. A prim ary advantage o f chemical pretreatm ents, over physical pretreatm ents, is the a b i li t y o f some chemicals to remove lig n in from the s u b s tra te . Chemical treatm ents can also swell the carbohydrate- lig n in m a trix , thus decreasing the mass tra n s fe r lim ita tio n s in h y d ro l­ y s is . A lk a lin e s o lu tio n s (o f which NaOH is the most common) are thought to increase h yd ro lysis by sw e llin g the s u b stra te . Although water can swell lig n o c e llu lo s e , i t does so p r im a rily by e n terin g regions between c r y s ta llin e u n its . Sodium Hydroxide expands the c e llu lo s e m a trix by 12 breaking bonds in the c r y s ta llin e u n its o f the c e llu lo s e re s u ltin g in an increase in the amount o f amorphous c e llu lo s e [2 0 ]. Gharpuray et a I • have shown th a t ca u stic removes both lig n in and hem icellulose in a d d itio n to i t s s w e llin g a ctio n [1 0 ]. This component removal allows g re a te r access to the c e llu lo s e molecules by in cre asing surface area and pore s iz e . One problem w ith c a u s tic pretreatm ents is the consump­ tio n o f a lk a li during the pretreatm ent. This loss o f reagent de tra cts from the economics o f a process using c a u s tic pretreatm ents [ 21] . D ilu te acid pretreatm ents a t elevated temperatures have been shown to increase the h yd ro lysis ra te o f the pretreatm ent re sid ue . This type o f pretreatm ent could be c a lle d a p re h yd ro lysis since i t hydrolyzes the hem icellulose fr a c tio n , hem icellulose e x is ts . is thus removing i t removed, from the m a trix . I ig n in -h e m ic e llu lo s e bonding Since the no longer This c o n d itio n allow s the lig n in to be removed by s o lv a tio n i f so d e sire d . G re th le in e t a l . noted however th a t d e li g n i f i ca tio n was unnecessary [1 8 ]. The d is ru p tio n o f the carbohydrate-1ignin m a trix by d ilu t e acid was s u ff ic ie n t to render the residue h ig h ly susce ptible to an enzymatic h y d ro ly s is . Acid pretreatm ents increase the pore volume a llo w in g g re a te r p re tre a te d pe n e tra tio n su b stra te s. of c e llu la s e enzyme molecules in to The increased access o f enzymes re s u lts in higher ra tes o f c e llu lo s e to glucose conversion. Engineers a t the U n iv e rs ity o f Arkansas have studied the use o f moderate tem perature, d ilu t e acid p re h yd ro lysis in an acid hyd ro lysis process. This s u b s ta n tia l product degradation. concentrated acid pretreatm ent h yd ro lysis hydrolyzed hem icellulose degradation product, used. hydrolyzes o f the if c e llu lo s e products o f glucose, re sid ue . would be la rg e ly w ith o u t follow ed by a The e a s ily converted to its on ly a concentrated acid step was F u rfu ra l and 5-hydroxymethyl fu r fu r a l degradation hem icellulose The p re h yd ro lysis is fra c tio n f u r f u r a l, the (HMF), the acid catalyzed are poisonous to yeast c e lls th a t would ferment sugars to ethanol o r o th e r products [ 22] . A j o i n t p ro je c t between the U n iv e rs ity o f Pennsylvania and General E le c tric Corporation has developed a solvent d e lig n ific a tio n p re tre a t­ ment. This pretreatm ent involves a hot aqueous ethanol treatm ent o f a 13 s u b s tra te . This re s u lts in a s o lid c o n s is tin g o f c e llu lo s e , p a r t ia lly degraded hemic e llu lo s e , and lig n in t ic a lly hydrolyzed and re s u lts [2 3 ]. This residue is then enzyma­ in c e llu lo s e conversions o f 80 to 90% [I]. The pretreatm ent schemes th a t seem most prom ising are those th a t in vo lve a steam o r high temperature water treatm ent. are known as autohydrolyses. These processes The term autohydrosis comes from the fa c t th a t in the presence o f water a t an elevated temperature acetyl groups in hemi c e llu lo s e (are thought to ) break down and form a c e tic a cid . Therefore, an automatic acid catalyzed h y d ro ly s is takes place in the pretreatm ent m ixtu re . During so lva ted. A b e n e fit o f autoh ydro lysis is th a t considerable amounts o f be removed hydrolyzed [2 4 ]. rendered w h ile the in to a form th a t hemic e llu lo s e is is are and the can is hemic e llu lo s e s hydrolyzed lig n in lig n in an autoh ydro lysis e a s ily sim ultaneously Rupture o f the p ro te c tiv e lig n in - c e llu lo s e linkage is s u ff ic ie n t to a llo w a r e la tiv e ly easy h y d ro ly s is o f the c e llu lo s e . re sid ua l c e llu lo s e from autoh ydro lysis can then be The enzym atically converted to glucose w ith y ie ld s o f 90 to 95% [2 5 ]. A pretreatm ent re la te d to a u toh ydro lysis is the steam explosion o f lig n o c e llu lo s e . In th is process a substrate is subjected to high pressure steam («500 p s ig ), cooked a t th is temperature fo r a sho rt time («5 seconds), and then the m ixture is flashed to atmospheric pressure. This explosive fla s h in g g re a tly d is ru p ts the s tru c tu re o f the c e llu lo s e in cre asing the pore volume and surface area o f the su b stra te . process th a t uses th is pretreatm ent is the Iotech process. A In th is procedure, the steam explosion is follow ed by an enzymatic h yd rolysis w ith c e llu lo s e conversions o f about 90% and hemic e llu lo s e conversions of. approxim ately 80%. The lig n in need not be e xtra cte d from the sub strate but can be l e f t to be removed a t the end o f the h y d ro ly s is . Of a ll the pretreatm ents and processes discussed here, the use o f a d ilu t e acid pretreatm ent, a u to h y d ro ly s is , o r the steam explosion technique o f the Iotech process seem the most com mercially prom ising. 14 EXPERIMENTAL This sectio n describes the substrates th a t were used and the ex­ perim ental procedures th a t were follow ed during th is study. Substrates Substrates used fo r th is in v e s tig a tio n were wheat straw , ba rley straw , lodge pole pine chips, Douglas f i r chips, and Chromedia. Wheat straw was a spring wheat o f the Pondera v a rie ty and the ba rle y straw was C la rk 's b a rle y . Two bales o f each o f these substrates were obtained from Larry Van Dyke o f Manhattan, Montana. in November o f 1982. The straws were harvested Lodge pole chips were supplied by the Brand-S Lumber Company o f L ivin g sto n , Montana. W illow Creek Lumber Company, also o f L iv in g s to n , provided the Douglas f i r chips. v a r ie tie s o f woods were sawmill tim b er in December o f 1983. The chips o f both residues from the harvesting o f liv e I t was approxim ately fo u r weeks between the tim b er harvest and the a c q u is itio n o f the chip s. A ll fo u r o f the above substrates were stored in sealed p la s tic bags u n til used. s u b stra te , Whatman Chromedia, was used in several experiments. A fifth This m a teria l was e s s e n tia lly a pure c e llu lo s e in powdered form designed fo r use as a column packing in chromatography. The Chromedia was manufactured by W & R B a!ston, LTD o f England. A ll o f the su b stra te s, except the Chromedia, were prepared fo r use as fo llo w s . hours. F ir s t, the m a te ria ls were a i r d rie d fo r approxim ately 48 Next, the substrates were m ille d in a W iley hammer m ill equipped w ith a I -mm discharge screen. The m ill was. stopped p e rio d ic a lly and allowed to cool to prevent overheating o f m aterial being ground so th a t sub strate would re ta in i t s natural in t e g r it y . The m ille d m a terial was then screened to is o la te the 35 to 60 mesh fr a c tio n . AU ASTM and TAPPI standard methods required th is size fra c tio n fo r a n a ly s is . was used in a ll This fra c tio n subsequent c h a ra c te riz a tio n s and experiments. A fte r g rin d in g and screening, the substrates were stored in containers open to the atmosphere. 15 C ha ra cte riza tio n o f Substrates Substrates used fo r th is research pretreatm ent experiments were performed. ized by r e la tiv e amounts of ash were characterized before The m a te ria ls were cha racte r­ content, m oisture, e x tr a c tiv e s , lig n in , c e llu lo s e and hemic e llu lo s e . Ash The ash content in the substrates was determined using a s lig h t v a ria tio n [2 6 ]. in the ASTM D 1102-56 Standard Test Method f o r Ash in Wood This procedure involved weighing about 2 grams o f the substrate in a ta re d , p re v io u s ly ig n ite d (a t 600°C) p o rce la in c ru c ib le and l i d . The m a teria l was weighed to the nearest 0.1 m illig ra m . and detachable I id were placed temperature below IOO0C. in The c ru c ib le an ashing oven w ith The oven was slow ly heated to a s ta rtin g the 600°C ashing temperature and allowed to remain there fo r 30 minutes. At the end o f th is period the l i d was placed on the c ru c ib le and the c ru c ib le was then tra n s fe rre d to a dryin g oven to co o l. (The dryin g oven temperature was maintained between IOS0C and IlO 0C.) were cooled fu r th e r in m illig ra m . a d e sicca to r and weighed to The c ru c ib le s the nearest 0.1 The percent ash was corrected fo r m oisture in the i n i t i a l sample and reported to the nearest 0. 01%. M oisture The m oisture a n alysis used ASTM D 1102-56 Standard Test Method fo r Ash in Wood [2 6 ]. it This te s t s p e c ific a lly analyzes f o r ash content but also describes a procedure fo r determ ining m oisture content. The standard method e n ta ile d weighing a sample o f substrate to the nearest 0.1 m illig ra m using a tared glass weighing v ia l. The v ia l containing the sample was placed in an oven c o n tro lle d between IOS0C to IlO 0C. Two d iffe r e n t procedures were used from th is p o in t. The f i r s t m a teria l procedure follow ed the standard method by drying the fo r 2 hours and then tra n s fe rrin g the v ia l d e s ic c a to r. unstoppered to a The v ia l was allowed to cool and was then stoppered and reweighed to the nearest 0.1 m illig ra m . The v ia l was unstoppered and 16 returned to the oven f o r one hour, then cooled in the d e s ic c a to r and reweighed. obtained. It This procedure was repeated u n til was found th a t a ft e r about fo u r samples had reached a constant w e ig h t. hours of oven drying the Therefore, f o r convenience a second method o f m oisture content was used. was kept in the oven ove rnig ht a constant weight was For th is method, the v ia l (e ig h t to twelve hours), fe rre d from the oven to a d e sicca to r f o r c o o lin g . then tra n s ­ When the v ia l was co o l, i t was stoppered and weighed to the nearest 0.1 m illig ra m . With e ith e r c a lc u la te d samples. as method the the Weight amount o f loss m oisture between the in d rie d the and sample was the undried The m oisture is reported to the nearest 0.1%. E x tra c tib le s The e x tr a c t!bles were defined as the components in the lig n o c e llu loses so lu b le in benzene-ethanol, e th a n o l, and w ater. These components are gums, re s in s , waxes, and in some substrates, ca te ch o l. fo r e x tr a c tiv e s used TAPPI T 12 os-75, Preparation The te s t of Wood fo r Chemical A nalysis (In c lu d in g Procedures o f Removal o f E x tra c tiv e s and D eterm ination o f M oisture Content) [2 7 ]. Approxim ately 80 grams of, the sub strate was weighed to the nearest 0.01 gram. The substrate was then placed in to a 250 ml e x tra c tio n thim ble w ith a coarse f r i t t e d glass d is c . placed in a Soxhlet e x tra c tio n apparatus and the Soxhlet This thim ble was solve nt re s e rv o ir charged w ith 350 ml o f benzene and 175 ml o f 95% e th an ol. v o ir was placed in a V a ria c -c o n tro lle d heating mantle. The re se r­ The Variac was adjusted so the b o il-u p ra te o f the e x tra c tio n solve nt re s u lte d in 6 to 7 re flu x e s per hour. The e x tra c tio n was continued fo r 6 hours. At the end o f th is p e rio d , the substrate was suctioned to remove as much o f the so lve n t as p o s s ib le . The substrate was then washed w ith fresh ethanol and suctioned d ry. A second e x tra c tio n was performed by charging the re s e rv o ir w ith 95% ethanol w h ile m aintaining 6 to 7 re flu x e s per hour as before . This 17 e x tra c tio n on ly proceeded f o r 4 hours. The sub strate was then suc­ tion ed dry and washed w ith deionized w ater. was s p l i t in to two s im ila r p o rtio n s . At th is time the sample The two p o rtio n s were placed in 500 ml Erlenmeyer fla s k s added. They were kept a t SO0C fo r one hour w ith fre que nt s t ir r in g v ia a glass rod. and 500 ml o f b o ilin g deionized water was The substrate was f ilt e r e d allowed to a i r d ry . from the w ater e x tra c t and A m oisture a n alysis was performed on the a ir d rie d e xtracte d sub strate (e x tra c t-fre e s u b s tra te ). bles was determined m a teria l by and e xtracte d a simple weight m a te ria l. The amount o f e x tra c ti - loss between E x tra c tib le s were the s ta r tin g reported on a m o is tu re -fre e basis to the nearest 0. 1%. Lignin The next c o n s titu e n t analyzed f o r was lig n in . The method used was a v a ria tio n o f ASTM D 1104-56 Standard Test Method f o r H oloce llulose in Wood as described by Browning [2 8 ]. The method involved weighing about 2.5 grams o f the e x tra c t-fre e substrate to the nearest 0.1 m illig ra m . The sample was placed in a 30 ml coarse p o ro s ity f r i t t e d glass f i l t e r ­ ing fu n n e l. c h lo rin e The funnel gas was slow ly was placed in passed through a bath o f ic e the (see sample water w hile Figure 5 ). A fte r 3 minutes, the c h lo rin e gas flo w was in te rru p te d so the sample could be s tir r e d w ith a glass rod. continued fo r 2 more minutes. funnel to cover the sample. dioxane suctioned o f f . The c h lo rin e gas flo w was then Enough 1,4-dioxane was added to the The ic e water was then drained and the Next the sample was washed tw ice w ith a SO0C s o lu tio n o f 5% monoethanol amine in 1,4-dioxane. Each wash was allowed to remain f o r 2 minutes before being suctioned o f f . The sample was then washed tw ice w ith room temperature 1, 4-dioxane. The sample was tra n s fe rre d to a 60 ml fr itte d glass Buchner f i l t e r i n g funnel o f coarse p o ro s ity and washed tw ice w ith room tempera­ tu re deionized w ater. o r ig in a l clo g g in g . fu n n e l. This The sample was then tra n s fe rre d back to the procedure prevented the fr itt e d discs from The funnel was reassembled in to the c o n fig u ra tio n necessary fo r use o f the c h lo rin e . The gas was passed through the sample fo r 3 18 6 m m GLASS TUBE rPM RUBBER STOPPER 52 mm (ID ) GLASS TUBE IO cm FRITTED GLASS CRUCIBLE ■26 mm ^ ll RUBBER STOPPER RUBBER RFS RUBBER CEMENT STOPPER 6 mm GLASS TU B E 12 mm GLASS TUBE S U C TIO N Figure 5. FLA S K Funnel Setup used fo r Lig nin Determination. 19 minutes and the washing, procedure repeated. This sequence o f c h lo rin a tin g and washing was continued u n til the sub strate no longer underwent a c o lo r change upon monoethanol amine s o lu tio n a d d itio n o r u n til the substrate was w h ite . is th a t associated w ith the 5% ethanol amine wash. The c o lo r change The sample was washed w ith dioxane u n til neutra l («7 pH) and then washed tw ice w ith 30 ml a liq u o ts o f d ie th y l e th e r. A fte r the eth er wash, the sample was washed w ith deionized water and tra n s fe rre d to a tared P e tri dish fo r a i r d ry in g . The lig n in weight loss sample. between the content was determined as the m o istu re -fre e e x tra c t-fre e substrate and the d e lig n ifle d C ellulo se and Hem icellulose The percentage o f to ta l c e llu lo s e was determined by assuming th a t the mass not c e llu lo s e . accounted fo r by the lig n in , m oisture, and ash was The procedure used to analyze fo r the r e la tiv e amounts o f c e llu ­ lose and hem icellulose was TAPPI standard method T 203 os-74, Alpha-, Beta-, and Gamma-Cellulose in Pulp [2 9 ]. It was assumed th a t the a lp h a -c e llu lo s e fra c tio n was c r y s ta llin e c e llu lo s e , the b e ta -c e llu lo s e was amorphous c e llu lo s e , and the gamma-cellulose was hem icellulose. This method suggests use o f 1.5 gm o f d e lig n ifle d pu lp, weighed to the nearest 0.1 m illig ra m . However, i t was not p o ssible to obtain 1.5 gm o f pulp during an actual experimental run due to the small amounts o f sample th a t could be generated. sm a lle r amount o f sample used. The reagents were adjusted fo r the The fo llo w in g procedure was based on 1.5 grams o f sample. The sample was placed in a 125 o r 250 ml Erlenmeyer fla s k w ith 100 ml o f 5.21 ± 0.005 N carbonate-free sodium hydroxide (NaOH). was placed on a magnetic s tir r e r suspension o f the pulp in the NaOH. insure th a t no a ir was drawn in to and s tir r e d to insure The fla s k complete S tir r in g was done in a manner to the m ixtu re. The fla s k was then placed in a water bath w ith the temperature maintained a t 25 ± 0.2°C. 20 When 30 minutes had passed, 100 ml o f deionized water was added to the pulp-NaOH m ixtu re . The d ilu te d m ixture remained in the water bath fo r 30 m inutes. The m ixtu re, w hile in the bath, was s tir r e d o cca sio n a lly w ith a glass s t ir r in g rod. A fte r the 60 minute p e riod , the mixture, was f ilt e r e d coarse p o ro s ity Buchner f i l t e r i n g fu n n e l. through a The f i r s t 10 to 20 ml o f the f i l t r a t e was discarded and the re s t o f the f i l t r a t e was c o lle c te d fo r la t e r a n a ly s is . A lpha-C ellulose A nalysis ten m i l l i l i t e r s Ten m illit e r s o f the c le a r f i l t r a t e and o f 0.5 N potassium d ichromate (K2Cr2O7) s o lu tio n were placed in a 250 ml Erlenmeyer fla s k . T h irty m illit e r s o f concentrated s u lfu r ic acid (H2SO^) were slow ly added to the m ixture w h ile the fla s k was s w irle d . The re s u ltin g F if t y m ill i t e r s s o lu tio n was kept hot f o r 15 minutes. o f deionized water were then added and the fla s k was cooled in a room temperature water bath. A blank was made using 12.5 ml o f deionized water and 12.5 ml o f the 5.21 N (17.5% by weight) NaOH. F if t y m illit e r s o f concentrated s u lfu r ic acid were slow ly added to the blank. The pulp filtr a te ammonium s u lfa te and blank were t it r a t e d C F e rrio n ') p o te n tio m e tric end by N ferrou s A phenanthroline (C12H8N2) -fe rro u s s u lfa te in d ic a to r s o lu tio n was also used to confirm the p o in t. s o lu tio n Ferrous ammonium s u lfa te unstable and was standardized before each use. accomplished 0.1 (Fe(NH4) 2 (SO4) 2) s o lu tio n w ith the end p o in t d e te r­ mined p o te n tio m e trical I y . (FeSO4) w ith a p o te n tio m e tric titr a tio n S tandardization was w ith 0.1 N potassium d ichromate s o lu tio n . The a lp h a -c e llu lo s e percentage was ca lcu la te d as fo llo w s : A lp h a -c e llu lo s e % = 100 - where is 6.85 (V2__=_Vi) x N x 20 V1 = T itr a tio n o f pulp f i l t r a t e , m i l l i l i t e r s V2 = Blank t i t r a t i o n , m i l l i l i t e r s N = Exact n o rm a lity o f the fe rro u s ammonium s u lfa te s o lu tio n 21 A = T itr a tio n o f the pulp f i l t r a t e used in th e o x id a tio n , m illilit e r s W = Oven-dry weight o f the pulp specimen, grams Beta- and Gamma-Cellulose A nalysis F if t y m ill i t e r s o f the pulp f i l t r a t e were p ip e tte d in to a 100 ml graduated c y lin d e r equipped w ith a ground glass stopper. F if t y m illit e r s o f 3 N s u lfu r ic acid were added to the c y lin d e r and w ith the stopper in place the c y lin d e r was in ve rte d several times to insure complete m ixing. The c y lin d e r was placed in an BO0C water bath fo r several minutes to aid the coagulation o f betac e llu lo s e , which p re c ip ita te s e ith e r allowed to p r e c ip ita te . upon a c id if ic a tio n . The c y lin d e r was stand ove rnig ht o r was c e n trifu g e d to s e ttle the F if t y m ill i t e r s o f the c le a r supernatant liq u id was p ip itte d o f f , being c a re fu l th a t none o f the p re c ip ita te was removed. placed in a 250 ml Erlenmeyer fla s k . d ichromate s u lfu r ic were added to the acid were c a r e fu lly fla s k . The liq u id was Ten m illit e r s o f 0.1 N potassium Then 90 ml poured in to the fla s k of concentrated w h ile s w irlin g . This s o lu tio n was kept hot f o r 15 minutes and then cooled in a room temperature w ater bath. A blank was prepared in the same manner except a s o lu tio n o f 12.5 ml o f deionized w ater, 12.5 ml o f 17.5% NaOH, and 25 ml o f 3 N s u lfu r ic liq u id . acid was s u b s titu te d fo r the c le a r supernatant These s o lu tio n s were t it r a t e d by the same method used du ring the a lp h a -c e llu lo s e d e term inatio n. The r e la tiv e fra c tio n s o f beta- and gamma-cellulose were ca lcu la te d as fo llo w s : Gamma-cellulose % = 100 - — ^ ^ x 20 Zo where X W V3 = T itr a tio n o f the s o lu tio n a ft e r p r e c ip ita tio n o f b e ta -c e llu lo s e , m i l l i l i t e r s V4 = Blank t i t r a t i o n , m i l l i l i t e r s B e ta -c e llu lo s e % = 100 - ( a - c e llu lo s e % + 7-c e llu lo s e %). 22 A uto hydrolysis and Lig nin E xtra ctio n The c a rrie d a u to h yd ro lysis out in ca p a c ity . an Autoclave lig n in e x tra c tio n pretreatm ents Engineers ZipperClave w ith were a one l i t e r The autoclave has two 500-watt heating bands and was rated to 2000 psi a t 450°F. pressure vessel Temperature Model and to in s id e 8530 d ig it a l An a ir driven a g ita to r s h a ft extended in to the s tir the the re a ctio n m ixture during autoclave was monitored an experiment. using a Cole Palmer thermometer and a Type K (Cromel-Alumel) thermo­ couple in s ta lle d in a s ta in le s s stee l therm owell. Substrate was held in e ig h t small packets th a t were made o f 200 mesh s ta in le s s stee l w ire c lo th fastened w ith f l a t pieces o f aluminum (Figure 6) . These packets were held in a ra d ia l metal d is c s . The aluminum closures were f i t t e d the d is c s . arrangement by two in to holes d r ille d in The two discs were separated by a 1/2 ID s ta in le s s steel pipe (Figure 6) . This pip e, w ith the e ig h t packets o f su b stra te , was mounted on the a g ita to r sh a ft o f the autoclave. The packets could then be ro ta te d during the run to minimize mass tra n s fe r lim ita tio n s th a t might a ffe c t e ith e r the autoh ydro lysis o r the lig n in e x tra c tio n . A 500 ml Parr bomb was plumbed to the autoclave and served as a high pressure steam generator. Plumbing from the bomb to the autoclave was low pressure s ta in le s s steel tub ing w ith Swagelock f i t t i n g s . This lin e had an Autoclave Engineers low pressure valve w ith high tempera­ tu re packing (model number 6V81U4TG) to c o n tro l the in je c tio n o f steam. The steam lin e was heated w ith heat tape to prevent co o lin g o f the steam during steam in je c tio n s . Cooling o f the autoclave was a tta in e d by venting to a cold tra p submerged in an ice bath. Venting was con­ tr o lle d by a valve o f the same type used in the steam lin e . Figure 7 is a diagram o f th is experimental set up. Autohydroly sIs The tared w ire mesh packets were f i l l e d s tra te and weighed to 0.1 m illig ra m . w ith e x tra c t-fre e sub­ Each packet held about one gram; th e re fo re , approxim ately e ig h t grams o f substrate were pre tre a te d per 23 Agitator shaft L " I Wire cloth "envelopes" \ Figure 6. / . ■- , I \ ■ . ........J Sample Basket Apparatus. Air driven agitator Valve Pressure gauge Valve Parr bomb Thermocouple Sparge tube Sample basket Dewar's Ilask Figure 7. Bomb heater I liter autoclave Setup used fo r Autohydrolysis and E xtraction Experiments 25 run. The packets were assembled and mounted on the a g ita to r s h a ft. E ith e r 300 o r 350 ml o f deionized water were charged to the Parr bomb and heated to a pressure o f 1200 psig (1500 psig f o r some ru n s). autoclave was charged w ith ISO0C. 600 ml o f deionized w ater and heated to This volume was required to thermowell were submerged. The insure th a t the sample and the A temperature o f ISO0C was chosen because s im ila r work by Waymen and Lora in d ica te d th a t a t and below 150°C the su b strate does not undergo any s ig n ific a n t change [3 0 ]. Once the bomb was a t the desired pressure and the autoclave at 150°C, the a g ita to r was s ta rte d and adjusted to 300 rpm. then released in to the autoclave. The in je c tio n o f high pressure steam allowed fo r very ra p id heat-up o f the s u b stra te . was desired so a sce rta in e d .) th a t accurate Steam was (This ra pid heating tim e-at-tem perature e ffe c ts could be When the autoclave reached operating temperature, the steam in je c tio n was discontinued. Temperature o f the autoclave was maintained w ith in ± 2°C o f the chosen operating value by venting vapor to cool o r adding steam to heat. When the predetermined time had elapsed the autoclave was vented to the cold tra p , being ca re fu l not to a llo w any uncondensed vapors to escape the autoclave tra p . and This procedure allowed its contents. Average fo r ra p id heat-up and co o lin g o f the cool-down times observed were 141 ± 65 seconds and 75 ± 15 seconds, re s p e c tiv e ly . Once the autoclave had been p ro p e rly vented i t could be opened and the assembly o f packets removed. f i ca tio n te s ts , the o th e r fo u r packets were opened and the contents q u a n tita tiv e ly removed. copious Four packets were saved fo r d e lig n i- The autohydrolyzed substrate was washed w ith amounts o f deionized w ater, w ater, and then allowed to a ir d ry . suctioned to remove the excess When the sub strate was a ir dry i t was cha racte rized as described e a r lie r . However, the ethanol-benzene e x tr a c tiv e s , te s t was not run. The amount o f liq u id in the autoclave, cold tra p , and Parr bomb was measured to tra ce the liq u id 's movement during the a u to h yd ro lysis. A ir d rie d te s tin g . m a teria l was sealed in a sample b o ttle fo r subsequent 26 Delig n if ic a tio n The next step in the pretreatm ent process was an a lc o h o lic e x tra c tio n to remove lig n in th a t was rendered solu ble by the autohy­ d ro ly s is . A 50:50 volu m etric s o lu tio n o f 95% ethanol water was used. and deionized The fo u r remaining packets from the a u toh ydro lysis were reassem­ bled in to t h e ir holder and placed in the autoclave. The cold tra p was reconnected to the system and as before placed in an ic e bath. Six hundred m il l i t e r s o f the 50:50 ethanol-w ater s o lu tio n were added to the autoclave, the autoclave was heated to ISO0C1 and the a g ita to r adjusted to 300 rpm. Temperature was c o n tro lle d by the methods used during the a u to h y d ro ly s is . The autohydrolyzed substrate was l e f t in the alcohol m ixture fo r one hour. At the end o f the hour the autoclave was vented to the cold tra p , the autoclave was opened, and the contents removed. The sub strate was q u a n tita tiv e ly washed from the pouches w ith deionized w ater, vacuum f ilt e r e d to remove the excess w ater, and then a ir d rie d . Once the sub strate was a ir dry i t was cha racterized using the proce­ dures p re v io u s ly described. mined. Again, the e x tr a c tiv e s were not d e te r­ The a i r d rie d sample was then sealed in a sample b o ttle fo r fu tu re experiments. B all M illin g Chromedia, one o f the substrates used, was b a ll attem pt to produce an amorphous c e llu lo s ic m a te ria l. m ille d in an The b a ll m illin g was c a rrie d out w ith both wet and dry su b stra te . The i n i t i a l Chromedia su b stra te , the dry b a ll m ille d Chromedia, and the wet m ille d Chromedia were analyzed to determine the re la tiv e amounts o f the various c e llu lo s e types present. were performed on a ll three m a te ria ls . Acid h y d ro ly s is te s ts Since Chromedia is e s s e n tia lly pure c e llu lo s e , lig n in determ inations were unnecessary. 27 Dry B all M illin g Steel b a ll bearings o f various sizes (1/4 inch diameter to inch diam eter) were used to m ill the dry Chromedia. 1/2 The b a ll bearings were loaded in to a p o rce la in g rin d in g j a r and Chromedia was added to fill the void spaces between the b a lls . The to ta l the j a r was approxim ately one t h ir d o f i t s volume. and placed on a tum bler ro ta tin g m inute. space occupied in The j a r was sealed between 30 and 40 re v o lu tio n s per The temperature was approxim ately 20°C. Total m illin g time was 96 hours w ith p e rio d ic down times to scrape the sub strate from the corners o f the j a r . The dry b a ll- m ille d substrate was stored in Nalgene sample ja r s . Wet B all M illin g Wet b a ll m illin g was c a rrie d out using p o rce la in c y lin d e rs (20 mm by 20 mm) in the same j a r used in the dry m illin g o p e ra tio n . Chromedia was mixed w ith deionized water to form a s lu r r y . The s lu r r y was vacuum f ilt e r e d in to remove excess w ater, then placed the ja r w ith the c y lin d e rs . Chromedia s lu r r y was added to ju s t cover the porcelain c y lin d e rs . The j a r was f i l l e d , as before to about o n e -th ird cap acity. The j a r was ro ta te d w ith the same v e lo c ity and the same length o f time as the dry m illin g o p e ra tio n . The m illin g operation was again stopped p e r io d ic a lly to scrape the m a terial from the corners o f the j a r . Once the 96-hour period had passed the sub strate s lu rr y was washed in to a fla s k and c e n trifu g e d . The water was decanted o f f and the Chromedia tra n s fe rre d to glass sample ja r s so i t could be stored wet fo r fu tu re a n a ly s is . Acid and Enzymatic H ydrolysis P retreated substrates were tested to asce rtain the r e la tiv e ease o f h y d ro ly s is among e x tra c t-fre e , autohydrolyzed, and autohydrolyzed.r a lc o h o lic The e xtracte d m a te ria ls . substances several d iffe r e n t means during the hydrolyses. were hydrolyzed by One o f the methods was a two step acid h y d ro ly s is using e ith e r s u lfu r ic o r h y d ro c h lo ric a cid . The o th e r methods were various enzymatic hydrolyses using e ith e r s in g le 28 enzymes o r mixed enzymes. Acid H ydrolysis A v a ria tio n o f ASTM D 1106-56 Standard Test Method f o r Lignin in Wood [31] was used f o r the acid hydrolyses. This te s t uses a two stage acid h y d ro ly s is in which the co n d itio n s are strong enough to hydrolyze a ll of the behind. polysaccharrides S e ve rity fra c tio n of the of the leaving co n d itio n s c e llu lo s e would the lig n in and were lessened hydrolyze. ash fra c tio n s so th a t only These co n d itio n s a were a rriv e d a t using a t r i a l and e rro r experimental procedure. C onditions were sought th a t would re a d ily hydrolyze lig n in - fr e e su b stra te , but would not hydrolyze e x tra c t-fre e sub strate to any great e x te n t. These co n d itio n s were not fu lly achieved (a t le a s t w ith the c o n d itio n s th a t were attem pted), so co n d itio n s were chosen where the experimental work-up o f the sample was e a s ie s t. co n d itio n s increased in s e v e rity , c o llo id a l These suspensions were very d i f f i c u l t work-up d i f f i c u l t . to As the acid hyd ro lysis suspensions would form. filte r , thus making sample The co n d itio n s chosen f o r s u lfu r ic acid were: 1. 100:1 weight r a tio acid s o lu tio n 2. 30°C fo r one hour 3. D ilu te to 0.7 N acid strength 4. SO0C f o r two hours Acid hydrolyses were o f actual run in acid to sub strate using 16 N a 250 ml approxim ately 300 m illig ra m o f s u b s tra te . h y d ro ly s is , sub strate th a t fla s k w ith When 2 hours had lapsed fo r had not hydrolyzed was tra n s fe rre d tared Gooch c ru c ib le w ith a coarse f r i t t e d m a teria l ErTenmeyer d is c . to a The non-hydrolyzed was washed w ith copious amounts o f deionized water and the c ru c ib le placed in a dryin g oven w ith the temperature between IOS0C and IlO 0C. The c ru c ib le was kept in the oven ove rnig ht before weighing. The exte nt o f h y d ro ly s is was reported as the m o is tu re -fre e weight Toss o f the sample. 29 Enzymatic Hydr o ly s is Three methods o f h y d ro ly s is c e llu la s e used enzymatic a c e llu la s e and a c e llo b ia s e , h y d ro ly s is enzyme, were another and the th ir d used. used One method a m ixture of a used a m ixture o f three enzymes, a c e llu la s e , a c e llo b ia s e , and a hemic e llu la s e . The enzyme s o lu tio n s were made by mixing the desired amounts and types o f enzymes w ith a 0.1 M c it r a t e b u ffe r s o lu tio n . B u ffe r s o lu ­ tio n s o f the desired pH, 4 .8 , were mixed from 0.1 M stock s o lu tio n s o f c i t r i c acid and sodium c it r a t e . To ob tain a s o lu tio n th a t buffered to a pH o f 4 .8 , 46 volume p a rts o f c i t r i c acid s o lu tio n were mixed w ith 54 volume p a rts o f sodium c it r a t e s o lu tio n . About 100 m illig ra m s o f the substrate were placed in a 200 x 25 mm te s t tube and wetted w ith one m i l l i t e r o f the c it r a t e b u ffe r . w e ttin g was complete, fo u r m illit e r s were added to the su b stra te . water bath and s tir r e d s t i r r e r and s t i r ba r. fo r Once the o f the desired enzyme s o lu tio n The te s t tube was submerged in a SO0C the time of h y d ro ly s is v ia a magnetic At the end o f the desired time p e riod , the contents o f the tube was tra n s fe rre d to a 100 x 13 mm te s t tube and c e n trifu g e d . The c le a r supernatant liq u id was frozen fo r fu tu re reducing sugar a n a ly s is . Reducing sugar was analyzed using a method described by M ille r [3 2 ]. This te s t used a reagent re fe rre d to as M ille r 's reagent. The reagent was made from the fo llo w in g re c ip e : 77.75% Deionized water 20% Rochelle s a lt (Potassium sodium ta r tr a te ) 1% 3 ,5 - d in it r o s a lic y I ic acid 1% Sodium hydroxide 0.2% Phenol 0.05% Sodium S u lfite For the M ille r method, one m i l l i t e r o f te s t sample and three m il l i t e r s o f M ille r 's reagent were placed in a 100 x 13 mm te s t tube. The te s t tube was then placed in a b o ilin g water bath fo r fifte e n minutes to develop the c o lo r f o r an o p tic a l d e n sity te s t . A blank was 30 generated using two m illit e r s o f the b u ffe r s o lu tio n and s ix m illit e r s o f the M ille r 's reagent. The o p tic a l d e n sity o f the re s u ltin g s o lu ­ tio n s was found using a Varian DMS 90 U V -V isible Double Beam Spectro­ photometer th a t was in te rfa c e d w ith an Apple I I Plus computer. wavelength was set a t 575 nm and the s l i t w idth a t 4 nm. photometer was c a lib ra te d using standard s o lu tio n s The The spectro­ of D-glucose; th e re fo re , the concentration u n its obtained were m illig ra m o f apparent glucose per m illit e r s sugars. s o lu tio n and not absolute amounts o f reducing The f in a l u n its reported were m illig ra m apparent glucose per gram o f dry sub strate o r m illig ra m o f apparent glucose per gram o f to ta l c e llu lo s e in the sample. The M ille r 's reagent would not accu rate ly in d ic a te a sugar concentration higher than one mg/ml o f s o lu tio n . Due to th is fa c t, the spectrophotometer was glucose about 0.35 mg/ml. s o lu tio n u su a lly c a lib ra te d w ith s o lu tio n s of Although a t times a less concentrated c a lib ra tio n had to be employed, because a less concentrated hyd ro lysis s o lu tio n needed to be evaluated. L in e a rity o f the re la tio n s h ip between li g h t absorbance and sugar concentration existed only in a narrow range around the c a lib r a tio n co n ce n tra tio n . Therefore, samples to be tested were d ilu te d (w ith b u ffe r s o lu tio n s ) and evaluated u n til t h e ir concen­ tr a tio n s were close to the c a lib r a tio n con centratio n. The enzyme preparations themselves contained sugar before contact w ith any su b stra te . This sugar le v e l was accounted f o r by preparing a standard enzyme s o lu tio n subjected to SO0C fo r the standard h yd rolysis time o f 6 hours. The amount o f apparent glucose was then determined and subtracted from subsequent values o f hydrolyzed samples. value was the apparent glucose achieved from h y d ro ly s is alone. The fin a l 31 RESULTS AND DISCUSSION A utohydrolysis o f Substrates The f i r s t goal o f th is study was to evaluate h y d ro ly s is character­ is t ic s o f three sub stra te s. The three substrates were b a rle y straw, lodge pole pin e, and Douglas f i r . Degree o f lig n in removal a fte r an au to h yd ro lysis and ethanol-w ater e x tra c tio n was determined. C ellulose conversion amounts v ia an acid h y d ro ly s is and an enzymatic hyd ro lysis were also determined. A utohydrolysis co n d itio n s were 205 °C and 10 minutes. used f o r th is work These c o n d itio n s were those found to be optimum f o r the d e li g n i f i ca tio n o f wheat straw during previous work at th is la b o ra to ry [2 4 ]. Table 2 contains re s u lts o f the substrate c h a ra c te riz a tio n s . Wheat straw composition (generated by Nakaoka [2 4 ]) was included fo r comparison purposes. The ash percentages reported were those from te s ts performed before the ethanol-benzene e x tra c tio n o f the su b stra te s. Subsequent ashings o f ethanol-benzene extracted substrates revealed th a t some ma­ t e r ia l th a t te ste d as ash was removed during the e x tra c tio n procedure. Table 3 shows the ash percentages a fte r the ethanol-benzene e x tra c tio n . Al I subsequent mass balance c a lc u la tio n s were made using the Table 3 ash re s u lts . Table 2. Weight Percent Composition o f Substrates (M o istu re -fre e b a s is ). Substrate C ellulose Hemi c e ll Ulose Lignin Ash B a rley straw 43.6 15.6 23.9 7.7 9.2 Wheat straw 39.3 14.1 28.9 8.3 9.4 Douglas fir 50.8 21.2 26.1 0.2 1.7 48.3 21.3 24.6 0.4 5.4 Lodge pole pine . E x tra c tib le 32 Table 3. Weight Percent Ash o f Ethanol-Benzene Extracted Substrates. Substrate Barley straw Wheat straw Douglas f i r Lodge pole pine The desired hydrolyze most r e s u lt of the of the Ash 4.2 5.8 0.1 0.2 autoh ydro lysis hemic e llu lo s e and solu ble in an ethanol-w ater s o lv e n t. pretreatm ent was to render the lig n in h ig h ly The c e llu lo s e residue from the au to h yd ro lysis ( i f the above goals were met) should be in a form th a t allow s easy h yd ro lysis to glucose. The theory o f an a u to h yd ro ly s is , as discussed be fore , is th a t an acid environment groups. re s u lts from decomposition o f hemic e llu lo s e acetyl The acid co n d itio n s catalyze the h y d ro ly s is o f hemic e llu lo s e and some c e llu lo s e in the su b stra te . This h y d ro ly s is produces water solu ble products th a t are e a s ily removed from the lig n o c e llu lo s e m a trix during the a u to h yd ro lysis. A utohydrolysis co n d itio n s hemic e llu lo s e - lig n in fa s t. During bonds. can also break lig n in - lig n in and These re action s are thought to be very slower re a c tio n s , th a t fo llo w these carbon-carbon bonds are formed between lig n in fa s t re a ctio n s, monomers. The f i r s t re a ctio n s produce products th a t are solu ble in solvents th a t u su a lly w ill not d isso lve Once the carbon-carbon becomes h ig h ly se rie s lig n in , in s o lu b le such as water o r ethanol-w ater s o lu tio n s . bonds in are formed, the the above solvents re a ctio n mechanism, t i m e-at-temperature is lig n in s o lu b ilit y . at-tem perature repolymerized [3 3 ]. lig n in Due to th is very im portant fo r Wayman and Lora observed th is a u toh ydro lysis t i me­ e ffe c t on residual lig n in in aspen woodmeal using a dioxane-w ater solve nt (see Figure 8) [3 0 ]. Barley Straw The a u toh ydro lysis and a u to h y d ro ly s is /e x tra c tio n o f ba rle y straw produced re s u lts s im ila r to those observed w ith wheat straw [2 4 ]. 33 195° C 5 10 -175’ C autohydrolysis time at temperature (minutes) Figure 8 . E ffe c ts o f A utohydrolysis and Dioxane E xtraction on Aspen Woodmeal Lignin [3 0 ]. 34 Since these two substrates are c lo s e ly re la te d species, th is tency o f re s u lts would be expected. consis­ Table 4 presents a comparison o f au to h yd ro lysis re s u lts fo r wheat and b a rle y straw. Douglas F ir and Lodge Pole Pine Table 5 is a summary of autoh ydro lysis e x tra c tio n runs on the Douglas f i r and a u to h y d ro ly s is / and lodge pole p in e . The o v e ra ll weight losses from the autoh ydro lysis and the a u to h y d ro ly s is /e x tra c tio n runs fo r the components two woods were w ith in such as c e llu lo s e , 2% o f each o th e r. hemic e llu lo s e , and lig n in Removal were of also reasonably close f o r these two su b strates. Table 5 shows th a t removal o f lodge pole hemi c e llu lo s e from the lig n o c e llu lo s e This m a trix continued during the ethanol-w ater e x tra c tio n . continued Douglas f i r . d is s o lu tio n of hemic e llu lo s e was not observed w ith Lig nin analyses seem to in d ic a te th a t more lig n in was removed from Douglas f i r du ring the a u to h y d ro ly s is /e x tra c tio n than from lodge pole during the pretreatm ent. But an inconsistency in the lodge pole lig n in data makes th is conclusion suspect. C a lcu la tio n s in d ica te d th a t more lig n in was removed from the lodge pole during the autohydrol­ y s is than was removed during the a u to h y d ro ly s is /e x tra c tio n , which was not p o s s ib le . The e rra n t data was probably due to s u b je c tiv e judgments th a t have to be made during the lig n in determ ination procedure. Due to th is apparent inconsistency and the time involved in repeating the run, fu tu re work w ith wood was performed using Douglas f i r as the sub strate. Approxim ately tw ice as much lig n in was removed from the straws than the woods (on weight percent basis) during the pretreatm ents. percent s im ila r . of hemic e llu lo s e ba rley straw and woods were Given the in tim a te re la tio n s h ip o f these two c o n s titu e n ts in the lig n o c e llu lo s e m a trix , lig n in removed from The may have tog eth er these two fa c ts suggest th a t the repolymerized before its d is s o lu tio n could occur. Lig nin re polym e rizatio n would in d ic a te th a t e ith e r the tim e-at-tem peratu re was too long o r the temperature too high. Table 4. Comparison o f A utohydrolysis Experiments fo r Barley Straw and Wheat Straw. % Substrate Weight Loss % Lignin Removed % Carbohydrate Removed % C ellulose Removed % Hemicellulose Removed Autohydrolyzed Barley Strawa 43.9 66.2 37.6 20.6 85.1 Autohydrolyzed/ Extracted Barley Straw3 46.7 75.7 38.0 19.8 88.5 Autohydrolyzed Wheat Strawb [24] 42.8 59.3 37.4 no analysis no analysis Autohydrolzed/ Extracted Wheat Strawb [24] 48.6 78.8 39.0 no analysis no analysis a - m o istu re -fre e basis b - a ir dry basis Table 5. Summary o f A utohydrolysis Experiments on Wood Substrates (Moisture-Free Basis). % Weight Loss % Lignin Removed Autohydrolyzed Douglas F ir 35.2 28.2 Autohydrolyzed/ Extracted Douglas F ir 37.1 Autohydrolyzed Lodge Pole Pine Autohydrolyzed/ Extracted Lodge Pole Pine Substrate % Carbohydrate Removed % C ellulose Removed % Hemicellulose Removed 37.8 16.1 89.8 34.6 38.0 16.8 88.7 34.8 29.4 36.8 18.5 78.3 36.1 27.2 39.3 17.2 89.3 37 Acid H ydrolysis Experiments Barley Straw The low percentages o f lig n in and hemic e llu lo s e in ba rley straw a ft e r pretreatm ent should have re s u lte d in a substrate whose c e llu lo s e was h ig h ly su sce p tib le to h y d ro ly s is . When p retre ate d b a rle y straw was subjected to the s u lfu r ic acid h y d ro ly s is co n d itio n s less than 10% o f the remaining carbohydrate was hydrolyzed to water solu ble product. Table 6 shows th a t weight losses due to acid hydrolyses were very s im ila r fo r both wheat and b a rle y straw . Carbohydrate weight losses from the o v e ra ll process (acid h y d ro ly s is plus various pretreatm ents) were s im ila r fo r both p re tre a te d and non-pretreated straws although the la tte r are in d ic a te (s u rp ris in g ly ) the existence of somewhat h ig her. a carbohydrate hydrolyzed by an a c id ic environment. seem to have l i t t l e e a s ily hydrolyzed. These re s u lts fra c tio n th a t seem to is e a s ily Pretreatments as performed here e ffe c t on the carbohydrate fra c tio n th a t is not (Another possible reason fo r the low weight losses in p re tre a te d m a te ria ls w i l l be discussed la te r in th is w ork.) Douglas F ir and Lodge Pole Pine Table 7 contains the weight loss re s u lts from the acid hyd ro lysis o f the woods. woods reveal The o ve ra ll-p ro ce ss weight loss c a lc u la tio n s fo r the an in te re s tin g r e s u lt. Carbohydrate w eight losses were about 70% higher fo r p retre ate d substrates in comparison w ith those o f non-pretreated e x tra c t-fre e woods. This suggests th a t pretreatm ents have a g re a te r e ffe c t on wood c e llu lo s e than on straw c e llu lo s e . though the data in d ic a te d th a t an au tohydrolysis Even enhances the acid h y d ro ly s is o f wood c e llu lo s e , the h y d ro ly s is y ie ld s were s t i l l less than 50%. Therefore, another serie s o f experiments was performed to t r y to exp la in the low c e llu lo s e conversions. A p o ssible reason f o r low h yd ro!yza biI i t y was the high amounts o f lig n in m a trix a ft e r pretreatm ents. s till o f the wood c e llu lo s e present in the lig n o c e llu lo s e On a r e la tiv e b a sis, the weight percent Table 6. Carbohydrate Conversion Results o f Acid H ydrolysis on Straw Substrates (Weight %). Substrate Total Carbohydrate Converted During Acid H ydrolysis Total Carbohydrate Converted During Overall Process (Pretreatment + H ydrolysis) E xtract-Free Barley StravA 48.0 — Autohydrolyzed Barley StravA 8.4 42.9 Autohydrolyzed/ Extracted Barley Straw3 5.8 41.6 E xtract-Free Wheat Strawb [24] 54.9 Autohydrolyzed Wheat Strawb [ 24] 11.2 43.5 Autohydrolyzed/ Extracted Wheat Strawb [24] 5.4 39.6 a - Moisture-Free Basis b - A ir Dry Basis — Table 7. Carbohydrate Conversion Results o f H2 SO4 H ydrolysis o f Wood Substrates (M oisture-Free Basis; % Weight Loss). Substrate Total Carbohydrate Converted During Acid H ydrolysis Total Carbohydrate Converted During Overall Process (Pretreatment + H ydrolysis) E xtract-Free Douglas F ir 24.3 Autohydrolyzed Douglas F ir 4.5 40.6 Autohydrolyzed/ Extracted Douglas F ir 3.6 40.2 E xtract-Free Lodge Pole Pine 24.3 Autohydrolyzed Lodge Pole Pine 5.3 40.2 Autohydrolyzed/ Extracted Lodge Pole Pine 4.6 42.1 — — 40 lig n in in the pre tre a te d woods was higher than th a t found in the nonpre tre a te d woods. Wayman and Lora's re s u lts w ith aspen suggested th a t dioxane-water so lve n t may be more e ffe c tiv e water s o lv e n t. w ith f o r d e li g n i f i ca tio n than the ethanol- Therefore, the ethanol-w ater e x tra c tio n was replaced a 9:1 dioxane-water e x tra c tio n tio n s . f o r p o st-a u to h yd ro lysis e x tra c ­ See Table 8 fo r a comparison o f the before and a ft e r e x tra c tio n lig n in percentages and the percent lig n in removed fo r the two solve nts. Although more lig n in was fin a l removed by the dioxane-water e x tra c tio n , the lig n in percent in the residue was higher than th a t found in the ethanol-w ater e xtracte d sample. This r e s u lt was a ttr ib u ta b le to the fa c t th a t the a u toh ydro lysis was not as e ffe c tiv e removing lig n in as th a t observed during autohydrolyzed f i r yzed samples Total a previous Douglas fir run. Therefore, the samples contained more lig n in than the autohydrol- used during the amounts o f carbohydrates ethanol-w ater e x tra c tio n experiments. removed during the pretreatm ents and acid hydrolyses were about the same fo r both runs (See Tables 7 and 9 ). This in d ic a te s th a t the small increase in lig n in content between th is run and the previous run had l i t t l e conversion. Table 8. Comparison o f Douglas F ir Lig nin E x t r a c t ib ilit y by Two Solvents (M oisture-Free Basis; Weight %). E xtra ctio n Solvent To e ffe c t on the to ta l carbohydrate % Lignin Before E xtra ctio n % Lignin A fte r E xtra ctio n % Lignin Removed Ethanol-Water 29.5 27.6 9.0 Dioxane-Water 34.7 31.6 12.1 a sce rta in the extent of the e ffe c t of lig n in on an acid h y d ro ly s is , two hydrolyses were performed using an e x tra c t-fre e and a d e lig n ifle d e x tra c t-fre e sample o f Douglas f i r . The d e lig n ifie d sample was prepared using the c h lo rin e gas bleaching procedure o f the Jig n in d e term inatio n. The carbohydrate conversions fo r the e x tra c t-fre e sample and the lig n in - fr e e sample were 33.4% and 27.7%, re s p e c tiv e ly . Table 9. Carbohydrate Conversion Results o f Autohydrolyzed Dioxane-Water Extracted Douglas F ir (M oisture-Free Basis; % Weight Loss). Pretreatment Total Carbohydrate Converted During Acid H ydrolysis Total Carbohydrate Converted During Overall Process (Pretreatment + H ydrolysis) Autohydrolyzed 4.1 44.6 Autohydrolyzed/ Extracted 4.9 44.4 42 The comparison o f these two re s u lts seems to in d ic a te th a t lig n in does not have a detrim e ntal e ffe c t on c e llu lo s e h y d ro ly s is . However, Saddler e t a l . have shown th a t as lig n in is removed, enzymatic conver­ sion of c e llu lo s e increases [3 4 ]. If an acid h y d ro ly s is can be compared to an enzymatic h y d ro ly s is , the lower conversion o f a d e lig n ifie d sample suggests a problem w ith the acid h y d ro ly s is te s t. C e llu lo se analyses were performed on the residues from the acid hydrolyses of the two Douglas fir samples ju s t discussed. These analyses were done to a sce rtain the e ffe c ts o f acid catalyzed h yd ro ly­ s is on the various c e llu lo s e types. the various Table 10 contains a comparison o f carbohydrate amounts before and a fte r the h y d ro ly s is . Acid h y d ro ly s is o f the d e lig n ifie d sample re s u lte d in the conversion o f 19% o f the fr a c tio n . c e llu lo s e In fra c tio n and the e x tra c t-fre e only 48% o f sample a ll the hemic e llu lo s e o f the hemic e llu lo s e was hydrolyzed and on ly 6% o f the c e llu lo s e fra c tio n was hydrolyzed. The re s u lts fo r The removal o f lig n in c e llu lo s e these experiments on c e llu lo s e were as expected. increased access to c e llu lo s e conversion. The fa c t th a t a ll re s u ltin g o f the in higher hemi cel lu lo s e was hydrolyzed in the e x tra c t-fre e f i r and only 48% was hydrolyzed in the d e lig n ifie d reverse, fir higher was not expected. hemic e llu lo s e The expected re s u lt would be the conversion fo r d e lig n ifie d samples. Lig n in and hemic e llu lo s e are bonded tog eth er in untreated" Tignocellu ­ lose . The removal o f a ll the lig n in may cause changes in the lig n o - c e llu lo s e m a trix and the chemical and physical nature o f the hemic e llu ­ lo se . These changes in hemic e llu lo s e may render i t less susceptible to h y d ro ly s is . Table 10. C ellulo se A nalysis A fte r Acid H ydrolysis (B asis: Ig E xtract-Free Wood, M oisture-F ree). C ellulo se Type a - C ellulo se /? - C ellulo se 7 - C ellulo se E xtract-Free Douglas F ir Before A fte r 0.525 0.321 0 0.173 0.216 0 Deli g n i f i ed-E xtract-F ree Douglas F ir ' Before A fte r 0.525 0.302 0 0.121 0.216 0.113 43 F u rthe r acid hydrolyses were performed s u b s titu tin g h yd ro ch lo ric acid fo r the s u lfu r ic acid o f previous te s ts . fo r these hydrolyses were: H ydrolysis con dition s SN HClr 3 hours, and 80°C. The purpose o f these experiments was to examine the e ffe c ts o f HClr ra th e r than H2SO4 , and a more severe h yd ro lysis environment. The substrates hydrolyzed were e x tra c t-fre e Douglas f i r , d e lig n ifie d Douglas f i r , and autohydrolyzed eth an ol-w ate r extracte d Douglas f i r . these hydrolyses. Table 11 contains the re s u lts o f The HCl h yd ro ly s is o f e x tra c t-fre e f i r re su lte d in a carbohydrate conversion about 25% higher than the 33.4% conversion ob­ served fo r the previous H2SO4 h y d ro ly s is . The h y d ro ly s is conversion o f the autohydrolyzed/extracted substrate w ith h yd ro c h lo ric acid was essen­ tia lly unchanged from th a t observed when s u lfu r ic acid was used. When HCl was used fo r the h y d ro ly s is , the o v e ra ll process carbohydrate con­ versions o f e x tra c t-fre e f i r and p re tre a te d f i r were e s s e n tia lly equal, u n lik e th a t observed fo r the H2SO4 . w ith the straw s, This suggests, lik e th a t observed a c e rta in c e llu lo s e hydrolyzed and th is fra c tio n fra c tio n e x is ts th a t is is not increased by pretreatm ent. e a s ily Like th a t observed w ith H2SO4 , a lower c e llu lo s e conversion was observed fo r d e lig n ifie d wood in comparison to e x tra c t-fre e wood fo r HCl hydrolyses. Wheat Straw F urther hydrolyses were performed w ith h y d ro c h lo ric a c id , but wheat straw was used instead o f Douglas f i r . The substrate change was made to a sce rta in the e ffe c ts o f an HCl h y d ro ly s is on a straw su b strate. C onditions f o r these hydrolyses were as fo llo w s : one hour, 3 N, 80 °C fo r two hours. 12.4N HCl, 30 °C fo r These c o n d itio n s were s im ila r to x those used f o r the s u lfu r ic acid hydrolyses, except the number o f hydrogen ions in s o lu tio n was «50% higher fo r the f i r s t hour and «200% higher fo r the la s t two hours. The re s u lts o f these h y d ro ly s is co n d itio n s fo r e x tra c t-fre e wheat, autohydrolyzed wheat, and autohydrolyzed/extracted wheat are presented in Table 12. The e x tra c t-fre e wheat and the auto­ hydrolyzed wheat showed a 30% increase o f conversion f o r the whole pro­ cesses (pretreatm ent and acid h y d ro ly s is ) over th a t obtained w ith H2SO4 . Table 11. Carbohydrate Conversions o f Douglas F ir Substrates by HCl (M oisture-Free Basis; % Weight Loss). Substrate Total Carbohydrate Converted During Acid H ydrolysis E xtract-Free Douglas F ir 41.1 Lignin-Free Douglas F ir 34.4 Autohydrolyzed/ Extracted Douglas F ir 8.4 Total Carbohydrate Converted During Overall Process (Pretreatment + H ydrolysis) 43.2 Table 12. Carbohydrate Conversions o f Wheat Straw Substrates by HCl (M oisture-Free Basis; % Weight Loss). Total Carbohydrate Converted During Acid H ydrolysis Total Carbohydrate Converted During Overall Process (Pretreatment + H ydrolysis) Extract-Free Wheat Straw 72.4 — Lignin-Free Wheat Straw 31.0 56.5 Autohydrolyzed/ Extracted Wheat Straw 32.2 58.8 Substrate 46 C e llu lo se conversion fo r the autohydrolyzed/extracted sample was 23% higher fo r the HCl h y d ro ly s is . Chromedia The observation th a t a fte r three hours o f acid h y d ro ly s is less than 50% o f the amorphous carbohydrate in autohydrolyzed Douglas f i r was hydrolyzed by H2SO4 and only about 60% by HCl seemed u n lik e ly . S im ila r re s u lts were also observed w ith o th e r sub strates tested Table 13). (see Therefore, a serie s o f experiments was run to te s t the acid h y d ro ly s is o f a pure c e llu lo s e substrate w ith a high fr a c tio n o f amorphous c e llu lo s e . To produce the h ig h ly amorphous c e llu lo s e a pure c e llu lo s e sub strate, Chromedia, was b a ll m ille d . Both dry and wet m illin g s were performed. B all m illin g the dry Chromedia fo r 96 hours doubled i t s amorphous c e l l ­ ulose con ten t, w h ile wet m illin g reduced the amorphous c e llu lo s e fra c tio n to less than 1% o f the to ta l c e llu lo s e . This was an unexpected r e s u lt. However, researchers a t American Viscose Corporation observed th a t by w e ttin g an amorphous b a ll m ille d c e llu lo s e , the c r y s t a l li n i t y could be increased to the degree o f c r y s t a l li n i t y found in the non-ball m ille d su b strate [3 5 ]. The r e c r y s ta lliz a tio n o f wet c e llu lo s e may explain the decrease in the amorphous nature o f the wet b a ll m ille d samples. Because o f the apparent r e c r y s ta lliz a tio n o f wet m ille d Chromedia, the dry m ille d sample was chosen fo r acid h yd ro ly s is comparison w ith non-mi I led Chromedia. Table 14 contains a summary o f the r e la tiv e amounts o f the c a r­ bohydrate fra c tio n s found in the three samples (n o n -m ille d , wet m ille d , and dry m ille d Chromedia ). The acid h y d ro ly s is (w ith 5 N HCl, 80 °C fo r 3 hours) o f non-mi lie d and d ry -m ille d Chromedia produced the re s u lts presented in Table 15. I t is evident th a t doubling the amorphous nature o f the substrate had an e q u iva le n t e ffe c t on the c e llu lo s e conversion. s till The conversion was very low f o r samples th a t contained approxim ately 40% amorphous c e llu lo s e . These low conversions o f amorphous c e llu lo s e support con­ clu sio n s concerning the inadequacies o f the acid h y d ro ly s is te s ts . Table 13. Comparison o f Amorphous C ellulose Content versus Amount o f C ellulose Converted. Substrate Acid Grams o f Amorphous C ellulose Present Total Grams o f Carbohydrate Hydrolyzed Autohydrolyzed Barley Straw H2SO4 0.125 0.064 Autohydrolyzed/ Extracted Barley Straw H2SO4 0.106 0.046 Autohydrolyzed Douglas F ir H2SO4 0.080 0.031 Autohydrolyzed/ Extracted Douglas F ir H2SO4 0.104 0.026 Autohydrolyzed/ Extracted Douglas F ir HCl 0.104 0.060 Autohydrolyzed Lodge Pole Pine H2SO4 0.131 0.038 Autohydrolyzed/ Extracted Lodge Pole Pine H2SO4 0.107 0.032 48 Table 14. Table 15. C ellulo se Analyses on M ille d and Non-Milled Chromedia (Weight %). M illin g Alpha C ellulose Beta C ellulose Gamma C ellulo se None 81.2 18.4 0.4 Dry 62.6 36.7 0.7 Wet 90.0 0.4 9.6 Results o f Acid H ydrolysis on Chromedia (M oisture-Free B a s is ). C ellulo se Converted (Weight %) Mi l l i n g C ellulo se were higher None 4.1 Dry 7.3 conversions observed during acid catalyzed fo r a ll non-pretreated observed fo r p re tre a te d su b strates. substrates versus hydrolyses conversions These experimental re s u lts suggest th a t the ease o f c e llu lo s e h y ro ly s is is not increased by the p re tre a t­ ments performed during th is tio n s presented conversion could in the be in v e s tig a tio n . lit e r a t u r e increased by have While several suggested pretreatm ents th a t s im ila r in v e s tig a ­ c e llu lo s e to those performed during th is study [2 5 ,3 6 ], these c ita tio n s and the experimen­ ta l re s u lts from th is la b o ra to ry in d ic a te th a t acid h y d ro ly s is may not be in d ic a tiv e of the e ffe c t of autoh ydro lysis as a pretreatm ent. Therefore, a se rie s o f experiments was performed using enzyme catalyzed hydrolyses conversion. to a sce rta in the e ffe c ts o f a u toh ydro lysis on c e llu lo s e F urther discussions on the low acid h y d ro ly s is re s u lts are presented in the next se ctio n . Enzymatic H ydrolysis Experiments / As a fin a l d ro ly s is se rie s o f experiments to te s t the e ffe c ts o f autohy­ on rendering a substrate more conducive to h y d ro ly s is , an enzymatic h y d ro ly s is was performed. The sub strate used f o r the enzymatic hydrolyses was an o p tim a lly autohydrolyzed and ethanol-w ater e xtracte d wheat straw . The sample was l e f t wet a ft e r the e x tra c tio n to avoid adverse e ffe c ts o f drying on enzymatic hydrolyses observed at Colorado State U n iv e rs ity w ith lo b lo lly pine (although subsequent experiments showed th a t drying had no detrim e ntal e ffe c t on wheat straw conversion) [3 6 ]. fo r th is h y d ro ly s is were as fo llo w s : fo u r m i llit e r s The condition s o f enzyme s o lu ­ tio n , one m i l l i t e r o f c it r a t e b u ffe r s o lu tio n , a t SO0C fo r s ix hours. The enzyme s o lu tio n consisted o f 3.6 gm o f a c e llu la s e enzyme dissolved in c it r a t e b u ffe r to make 100 ml o f s o lu tio n . o f an i n i t i a l h yd ro lysis Reducing sugar analysis liq u o r revealed 561 m illig ra m s o f apparent glucose per gram o f dry su b stra te , which was a conversion o f about 74% o f the c e llu lo s e in the sample. sample ( a ir d rie d ) S u lfu r ic acid h y d ro ly s is o f a s im ila r re su lte d in only a 5.4 % conversion o f c e llu lo s e , and a 12.4 N h y d ro ly s is w ith h y d ro c h lo ric acid y ie ld e d a 32.2% c e llu ­ lose conversion. Comparing the enzyme and acid hydrolyses in d ica te s th a t an acid h y d ro ly s is te s t was a poor measurement o f the e ffe c tiv e ­ ness o f a u toh ydro lysis as a pretreatm ent. A po ssib le reason f o r the low conversions obtained w ith acid is th a t enzymes are very s p e c ific in the way th a t they a tta c k c e llu lo s e . A complete c e llu la s e enzyme a c tu a lly consists of three separate enzymes: an endoglucanase, an exoglucanase, and a c e ll obi ase. glucanohydrolase, an endoglucanase, randomly hydrolyzes 1,4 - c e llu lo s e in s id e the m olecular chain, which re s u lts in a ra p id reduction in the degree o f p o lym e riza tio n . The exoglucanase, /? - l,4 - c e llu lo b iohydrolase, a tta cks c e llu lo s e polymers from the reducing end producing c e llo b io s e , a dimer o f glucose which is water s o lu b le . Exoglucanase can a tta ck the ends found in the n a tive c e llu lo s e o r an end produced by a endoglu­ canase. The la s t type o f enzyme, /?-glucosidase (ce T lo b ia se ), hydro­ lyzes c e llo b io s e and o th e r water solu ble glucose oligomers to glucose. 50 Figure 9 is h y d ro ly s is . of a ll a schematic of th is proposed mechanism o f enzymatic Due to th is method o f c e llu lo s e a tta c k , a high percentage in d iv id u a l h y d ro ly tic Acids a tta c k c e llu lo s e re action s in a t o t a l l y soluble produce a solu ble random fa sh ion. h y d ro ly tic re action s y ie ld products. h y d ro ly s is by acids was measured by weight Therefore, Since lo s s , product. the few degree o f products must be so lu b le to be measured as conversion. Clausen and Gaddy obtained data showing a 90% conversion of c e llu lo s e could be obtained w ith 14 N HCl a t room temperature in a CSTR (no time frame was given in t h e ir lite r a tu r e ) th a t 8 N HCl converted 90% o f the c e llu lo s e using a se rie s o f CSTR1s. [3 7 ]. They also found in 30 minutes a t IOO0C Therefore, another p o ssible explanation fo r low conversions o f substrate v ia acid hydrolyses could be mass tra n s fe r lim ita tio n s , m ixtu res. re s u ltin g from low shear s t ir r in g of the h y d ro lysis Acid h y d ro ly s is m ixtures were s tir r e d on ly p e rio d ic a lly w ith a glass s t i r r in g rod, w h ile enzymatic h y d ro ly s is m ixtures were s tir r e d throughout the h y d ro ly s is . Since in itia l enzymatic hydrolyses re s u lte d in high c e llu lo s e conversions, the use o f enzymes to evaluate the e ffe c ts o f autohydrolyses was adopted. Enzyme strengths th a t gave the highest conversion in 6 hours were evaluated. C e llu c la s t 1.5 L, a c e llu la s e supplied by Novo C orporation, was the f i r s t enzyme whose strength was optim ized. C e llu c la s t 1.5 L is an enzyme prepa ratio n made from the submerged ferm entation o f a selected s tra in re s u lts o f the fungus Trichoderma re e s e i. obtained from the c e llu la s e charge v a rie d . various Figure 10 is 6 hour hydrolyses, in a p lo t o f which the I t can be seen from th is graph th a t a charge higher than 2.8 gm o f cel lu la s e prepa ratio n per gram o f dry substrate would not produce much improvement in c e llu lo s e conversion. Therefore, th is concentration o f c e llu la s e was chosen fo r fu tu re experiments. o p tim a lly autohydrolyzed wheat straw was used fo r th is series Wet of experiments. Experiments eva lu a tin g the e ffe c ts o f h y d ro ly s is time on conver­ sion using C e llu c la s t 1.5 L were also performed (see Figure 11). 51 P — I , 4— gJ L i-icran^JLucanoliyclr-oI a C e I l u l o a e e l JLulo C c e p t I b l e P ~ I „4 — glucan f$— I , 4— glucan — glucanobyd r oIae glucoe glueoeld Figure 9. A Proposed Mechanism fo r an Enzymatic H ydrolysis [5 ]. mg apparrent glucose/g dry fiber 600 - - 500 - - 400 - - 300 - - 200- - 100 - - g cellulase/g dry fiber Figure 10. E ffects o f C ellulase A c tiv ity fo r a 6 hour H ydrolysis. 800 L. 7 0 0 -600 - 500 - 400 - 300 - - 200- - 100 --j 0.0 4.0 8.0 12.0 16.0 Time (hours) Figure 11. Reaction Time E ffects on a C elluiase H ydrolysis. 20.0 24.0 54 Increasing h y d ro ly s is conversion by re s u lte d in from Lengthening 28% more conversion h y d ro ly s is . h y d ro ly s is 19%. time 6 to 12 hours the h y d ro ly s is than increased time was observed to at c e llu lo s e 24 hours 6 hours of Even though higher conversions were obtained w ith longer tim es, p ra c tic a l experimental considerations h y d ro ly s is tim e o f 6 hours fo r sub strate e va lu a tio n . used f o r these wheat straw . experiments was again wet, C e llu c la s t 1.5 L contains l i t t l e o p tim a lly d ic ta te d a The substrate autohydrolyzed c e llobiase a c t iv it y . Therefore, the m a jo rity o f the h yd ro lysis products would be c e llob io se and not glucose. Since in a reducing sugar te s t, a c e llo b io s e molecule would appear as on ly in d ic a te d . one molecule, a low h y d ro ly tic conversion would be To a lle v ia te th is problem a second enzyme, /?-glucosidase, was added to the h yd ro lysis m ixtures. supplied by Novo C orporation, Cellob ia se 250 L ) . It its This enzyme prepa ratio n was also tra de name being Novozym 188 (o r was produced by submerged ferm entation o f a selected s tra in o f the fungus A s p e rg illu s n ig e r. The c e ll o b iase preparation was added to the optim ized c e llu la s e s o lu tio n s biase. in varying amounts to ob tain the optimum amount o f c e llo ­ Figure 12 is a graph o f the re s u lts from varyin g ce llo b ia s e amounts. Charges o f c e llo b ia s e higher than 0.28 gm o f p repa ratio n per gram o f dry substrate had l i t t l e conversion. Therefore, subsequent hydrolyses. ce llo b ia se e ffe c t on c e llu lo s e the 0.28 gm amount was chosen fo r charging Wet o p tim a lly autohydrolyzed/extracted wheat straw was the substrate used fo r these experiments. Time stud ies were also performed w ith the mixed enzyme s o lu tio n . As the time increased from 6 hours to 12 hours, c e llu lo s e conversion increased 14%. A h yd ro lysis time of 24 hours d id not re s u lt in increased conversion over th a t observed during the 12 hour h yd rolysis (see Figure 13). The substrate used fo r these experiments was an o p tim a lly autohydrolyzed and ethanol-w ater extracted wheat straw . Using the optimum concentrations determined by the above e x p e ri­ ments, p re tre a te d and non-pretreated e x tra c t-fre e samples o f wheat were evaluated fo r c e llu lo s e conversion. The p retre ate d samples were wet mg apparrent glucose/g dry fiber 9 0 0 -800 - 700 - 600 - 500 H 300200 - ■ 100 - - g cellobiase/g dry fiber Figure 12. E ffects o f Cellobiase A c tiv ity on a 6 hour H ydrolysis. d) 11OO - - 900 - 800 - 7 0 0 -600 - 500 - 4 0 0 -- Time (hours) Figure 13. Reaction Time Effects on a Cellobiase-Cellulase Hydrolysis. 57 autohydrolyzed wheat, wet autohydrolyzed and ethanol-w ater extracted wheat, dry autohydrolyzed wheat, and dry autohydrolyzed and ethanolwater e xtracte d wheat. tio n s . Table 16 contains the re s u lts o f these evalua­ Comparison o f the e x tra c t-fre e wheat c e llu lo s e conversion w ith those o f the p re tre a te d samples demonstrates th a t the pretreatm ents were very e ffe c tiv e in increasing the c e llu lo s e conversion. Also, the high conversions obtained w ith substrates th a t had not been extracted to remove lig n in suggests th a t e x tra c tio n was an unnecessary p re tre a t­ ment step. It is also evident from the data th a t d ryin g had l i t t l e e ffe c t on the enzymatic h yd ro ly s is o f wheat straw . Colorado S tate U n iv e rs ity , Murphy e t a l . , at also had good conversions o f c e llu lo s e in autohydrolyzed wheat w ith dry samples [3 8 ]. They d id no evaluations on the e ffe c ts th a t dryin g o f samples had on enzymatic h y d ro ly s is . Table 16. Results o f Enzymatic Hydrolyses on Wheat Straw. Substrate______ % C ellulose Hydrolyzed Dried E xtract-F ree 12.8 Dried E xtract-F ree Hem icellulase Added to Enzyme S olutio n 17.4 Dried Autohydrolyzed 94.5 Dried Autohydrolyzed/ Extracted .89.4 Wet Autohydrolyzed 96.5 I Wet Autohydrolyzed/ Extracted An experiment hem icellula se, 98.0 was performed was added to the in which h y d ro ly s is a t h ir d enzyme type, m ixtu re. Although a the researchers a t Colorado State found th a t the Novo c e llu la s e enzymes had hem icellulase a c t iv it y , more hem icellulase s ig n ific a n t hem icellulase a c t iv it y [3 6 ]. p repa ratio n was Gamanase 1.5 L. was added to assure a The trade name o f th is enzyme Although the Novo Corporation suggests 58 a dosage o f 0.0005 gm o f Gamanase 1.5 L per gram o f dry s u b stra te , 0.25 gm o f enzyme per gram o f dry substrate was used fo r th is experiment to assure an excess o f enzyme fo r the h y d ro ly s is . The purpose o f th is experiment was to determine i f an enzyme could remove the hemic e llu lo s e from a non-pretreated c e llu lo s e . ment. substrate and thus increase access to the H ydrolysis time was increased to 12 hours fo r th is e x p e ri­ A carbohydrate conversion o f 17.4% in d ic a te s th a t the use o f the th ir d enzyme did not a tta in the e ffe c ts rendered by an autohydrolysis pretreatm ent. Murphy e t a l . suggested th a t f o r a process to be economically fe a s ib le , the enzyme charge ra te could be no more than 10 lU/gm o f su b strate [3 6 ]. (An IU is an in te rn a tio n a l accepted u n it o f measure fo r enzyme s tre n g th s . enzyme u n it and is the The u n it is defined as the release o f I pmol o f glucose per u n it.) The fin a l two-enzyme s o lu tio n used in th is work was evaluated fo r enzyme stre ngth using the Mandels e t a l . procedure and was found to contain 0.94 lU/ml o f enzyme s o lu tio n [3 8 ]. This tra n s la te s in to about 38 lU/gm o f s u b stra te , which is high fo r a commercial process. An enzyme s o lu tio n (using the same Novo enzyme preparations) prepared by Bertran and Dale had a stre ngth o f 0.55 lU/ml [3 9 ]. This s o lu tio n contained 3 gm o f C e llu c la s t 1.5 L and 1.5 gm Novozym 188 per 100 ml o f s o lu tio n . The enzyme used fo r the fin a l evaluations by th is la b o ra to ry contained 7.2 gm o f C e llu c la s t 1.5 L and 0.72 gm o f Novozym 188 per 100 ml o f s o lu tio n w ith a stre ngth o f 0.94 lU /m l. Comparing the strengths o f these two enzyme s o lu tio n s suggests th a t most o f the enzyme stre ngth was the re s u lt o f the c e llu la se and not the c e ll o b iase. The steep i n i t i a l slope o f a p lo t o f a c t iv it y versus c e ll o b iase charge supports the conclusion th a t the dominant v a ria b le determ ining enzyme stre n g th is c e llu la s e charge (see fig u re 12). Reducing the amount o f c e llu la s e by 75% (to «0.7 gm cellulase/gm dry fib e r ) would re s u lt in an enzyme stre ngth o f about 10 lU/gm o f s u b stra te . Figure 10 in d ic a te s th a t the decreased reduce the 6 hour conversion by about 30%. Data presented in enzyme stre ngth The would only c e llu la s e time studies in d ic a te th a t incre asing h yd ro ly s is time to 24 hours (a reasonable time 59 period fo r a commercial process) could increase the conversion 20% over th a t observed in 6 hours (see fig u re 11). Based on these conclusions, adequate conversions o f c e llu lo s e could be obtained w ith reduced enzyme charges by in cre asing the h yd ro ly s is tim e. Therefore, a g ric u ltu ra l an o v e ra ll residues process follow ed by using a u toh ydro lysis enzymatic h y d ro ly s is to p re tre a t and yeast ferm entation could be a fe a s ib le route to obtain useful liq u id hydro­ carbons. A great deal of fu r th e r work is needed to confirm the economic p o te n tia l o f such a process, but the basic technology looks q u ite prom ising. 60 CONCLUSIONS 1. When enzymes were used to catalyze the h y d ro ly s is o f pretreated wheat straw su b stra te , higher conversions were observed versus those f o r non-pretreated straw . However, s im ila r re s u lts were observed w ith both acid and enzyme catalyzed hydrolyses fo r d e lig n ifle d and non-delig n ifie d substrates whether o r not the substrates were p retre ate d before d e li g n i f i c a tio n . This in d i­ cates th a t among the e ffe c ts o f pretreatm ent on the lig n o c e llulose m a trix , changes in c e llu lo s e morphology have a g reate r e ffe c t in incre asing c e llu lo s e conversion than does a decreased lig n in con ten t. 2. Low conversions o f c e llu lo s e observed during acid catalyzed hydrolyses versus conversions observed fo r mixed enzyme hydro­ ly s is re su lte d from d iffe re n c e s in th e ir re sp e ctive re action mechanisms. c e llu lo s e Enzyme systems a tta c k s p e c ific molecule w ith lo c a tio n s on a a high percentage o f these attacks producing solu ble re a ctio n products. W hile, acid c a ta ly s ts randomly a tta c k c e llu lo s e bonds producing few so lu b le products per in d iv id u a l re a c tio n . c e llu lo s e Thus, using the w eight loss from a residue to measure extent o f re a c tio n , y ie ld s low conversions when acids are used to catalyze the re a c tio n . 3. S im ila r to ta l c e llu lo s e conversions observed fo r acid hyd roly­ s is only versus acid h y d ro ly s is plus a u to h yd ro lysis suggests the existence o f a carbohydrate fra c tio n th a t is e a s ily hydro­ lyzed by acid c a ta ly s ts . Conversion o f th is fra c tio n o f c e l l ­ ulose does not seem to be increased by pretreatm ent. 4. Straw lig n in was rendered more e x tr a c t!ble by pretreatm ent than wood lig n in fo r the same autoh ydro lysis c o n d itio n s . Ti me­ at-tem perature e ffe c ts may have caused wood lig n in to re p o ly ­ m erize, thus reducing i t s s o lu b ilit y in ethanol-w ater s o lv e n t. 61 SUGGESTIONS FOR FUTURE RESEARCH 1. I f the use o f acid h yd ro ly s is to determine the e ffe c ts o f c e l l ­ ulose pretreatm ents is to be continued as an adjunct to enzyme hydrolyses, the fo llo w in g suggestions should be adopted. The hydrolyses should be s tir r e d continuously and v ig o ro u s ly during the experiments, and the reducing sugar a n alysis method should be adopted to determine the degree o f c e llu lo s e conversion in a ll cases. 2. The tim e -at-tem p era ture c o n d itio n s used fo r three substrates (b a rle y straw , lodge pole pin e, and Douglas were those fir ) found to be optimal fo r the removal o f lig n in from wheat straw . These co n d itio n s may not be optim al fo r the o th e r substrates o r the optimal co n d itio n s f o r c e llu lo s e conversion. Therefore, experiments should be performed to determine the optimal con­ d itio n s fo r c e llu lo s e conversion fo r each su b s tra te . 3. The liq u o rs from a u toh ydro lysis experiments should be analyzed to determine i f sugars produced by au to-catalyzed hyd ro lysis re action s are in ta c t in the liq u o r o r i f they are destructed to furfuraTs.. 4. Wood substrates o f in te re s t should be tested to determine i f t h e ir enzymatic h yd ro ly s is ra te is increased by pretreatm ent as observed fo r wheat straw . 62 REFERENCES C ITE D 1. 2. Bungay, Henry R. E n e r g y , In te rs c ie n c e , New York (1981). Thomas, R. J . F e e d , F u e l s W o o d a n d a n d T h e A g r i c u l t u r a l C h e m i c a l s . E. B io m a s s Wiley O p t i o n s . R e s id u e s R e s e a r c h o n U s e f o r J . S oltes ( e d .) , Academic Press, New York (1983). 3. 4. Cowling, E. B. and K irk , I . (1976). A t a lla, R. H. F e e d , F u e ls W o o d a n d a n d K. B i o t e c h n o l . A g r i c u l t u r a l C h e m i c a l s . E. B i o e n g . R e s id u e s J . S oltes 6:95 S y m p . R e s e a r c h o n U s e f o r ( e d .) , Academic Press, New York (1983). 5. Blanch, H. W. and Wilke C. R. 1:1 (1983). 6. Gordon, A. C a r b o h y d r . 7. Nei I son, B i o e n g . 8. H., R e s . R e v ie w s Hay A. J ., Dinsdale, 57:235 (1977). M. J ., Delsey, 24:293 (1982). i n C h e m ic a l D., E n g i n e e r i n g . and Bacon, R. G. and Shafizadeh1 F. H. S. D. B i o t e c h n o l . Saddler, J. W., Brow nell, H. H., Clermont, L. P., and L e v itin , N. B io e n g . 24:1389 (1982). B i o t e c h n o l . 9. G re th le in , H. E., A lle n , D. C. and Converse, A. O'. 26:1498 (1984). B i o t e c h n o l . B i o e n g . 10. Gharpuray, M. M., Yon-Hyun Lee and Fan, L. T. 25:157 (1983). 11. C arr, M. E. and Doane, W. M. 12. David, C., B i o t e c h n o l . 13. B i o t e c h n o l . B i o t e c h n o l . B io e n g . Rornasie r, R., G re in d l-F a llo n , 27:1591 (1985). B io e n g . 26:1252 (1984). C. and Vanlautem, N. B io e n g . Sharma, A ., Mi I s te in , O., Vered, Y ., G ressel, J. and Flowers, H. B i o e n g . 27:1095 (1985). B i o t e c h n o l . 14. Chesson, A ., Stew art, C. S ., and Wallace, R. J . 44:1597 (1982). A p p l . E n v i r o n . M i c r o b i o l . 15. Focher, B., M a rz e tti, A ., Cattanaeo, M., Beltrame, C a r n iti, P. J. A p p . P o l y m e r S c i e n c e . 26:1989 (1981). 16. Fan, L .T ., Yong-Hyun Lee and Beardmore, D. H. 22:177 (1980). 17. G re th le in , H. E. B i g / T e c h n o l o g y . P. B i o t e c h n o l . 155, February (1985). L. and B io e n g . 63 18. G re th e lin , H. E. and Converse, A. 0. "Continuous Acid H ydrolysis fo r Glucose and Xylose P rod uction ," paper presented at I n t e r n a t i o n a l S y m p o s iu m o n E t h a n o l f r o m B io m a s s . Royal Society o f Canada, Ottawa, Canada (1983). 19. P u ri, V. P. 20. B i o t e c h n o l . 26:1219 (1984). B i o e n g . M ille t , M. A ., Baker, A. J . and S a tte r, L. D. 6:125 (1976). B i o t e c h n o l . B io e n g . S y m p . 21. 22. Grohmann, K., Himmel, M., Riaard, fo r g e t, R. and Graboski, M. B i o t e (1984). C., Tucker, c h n o l . B i o e n g . Shah, R. B., Clausen, E. C. and Gaddy, J. L. 79:1 (1984). M., Baker, J ., 14:137 S y m p . C h e m ic a l E n g i n e e r i n g P r o g r e s s , 23. Worthy, W. 24. Nakaoka, R. K. "A utohydrolysis and D e lig n ific a tio n o f Wheat S traw ," M aster's Thesis, Montana State U n iv e rs ity (1985). 25. Murphy, V. G., Linden, J. C., M oreira, A. R. and Lenz, T. G. Report DE-81023338 (1981). 26. ASTM Standard Test Method D 1102 - 56, "Ash in Wood," ASTM (1972). 27. TAPPI Standard Procedure T 12 os-75, "P reparation o f Wood fo r Chemical A nalysis (In c lu d in g Procedures fo r Removal o f E xtra ctive s and Determ ination o f Moisture C ontent)," TAPPI (1975). 28. Browning, B. L. M e t h o d s Sons, New York (1967). C h e m ic a l E n g i n e e r i n g o f December 7, 1981. N e w s , W o o d Volume 2. C h e m i s t r y . DOE Wiley and / \ 29. TAPPI Standard Method T 203 os-74, "A lpha-, Beta-, Gamma-cellulose in P u lp ," TAPPI (1974). 30. Lora, J. H. and Wayman, M. 31. ASTM Standard Test Method D1106 (1977). 32. M ille r , G. L. A n a l y t . Sarkanen, V. ^33. F o r m a t i o n , K. C hem , T a p p i . a n d 56, "L ig n in in Wood," ASTM 31:426 (1959). and . Ludwig, S t r u c t u r e s , 61,6 (1978). C. H. R e a c t i o n s . L i g n i n s : W iley O c c u r r e n c e , In te rs c ie n c e , New York (1971). N b4. Saddler, J. N., BrownwlI , H. H., Clermont, L. P., and L e v itin , N. B io e n g . 24:1389 (1982). B i o t e c h n o l . 64 35. Howsmon, J . A. and Marchessaultl 1:3,313 (1959). R. H. J. o f A p p . P o ly m e r S c i e n c e . 36. Murphy, V. G., Dockrey, K., Linden, J . C., and M oreira, A. R. "Enzymatic H ydrolysis o f Pinewood Pretreated w ith Aqueous Ethanol S o lu tio n o f Aluminum S u lfa te ," paper presented a t AIChE Annual Meeting, Los Angeles, C a lifo rn ia (1982). 37. Shah, R. B ., Clausen, E. C. and Gaddy, J. L. Progress. January (1984). Chemical Engineering I 38. Mandels , M., A n d re o tti, 6:21 (1976). R. and Roche, C. B i o t e c h n o l . B io e n g . S y m p . 39. B e rtra n , M. (1985). S. and Dale, B. E. B i o t e c h n o l . B i o e n g . 27:177 Table 17. Results o f the Acid H ydrolysis Development Experiments Substrate Acid Concentration (°c) Time (hr) Temperature % Hydrolyzed ION 40 1.0 4.2 Autohydrolyzed Wheat Straw ION 40 1.0 3.7 Autohydrolyzed Wheat Straw 12N 40 1.0 3.4 Delig n in fie d Wheat Straw ION 40 1.0 1.9 D e lig n in fie d Wheat Straw 14N 70 3.2 19.0 Autohydrolyzed Wheat Straw 14N 70 3.5 12.5 D e lig n in fie d Autohydrolyzed Wheat Straw 18N 70 3.0 80.2 D e lig n in fie d Autohydrolyzed Wheat Straw 18N 80 3.0 89.0 E xtract-Free Wheat Straw 18N 80 3.0 56.2 APPENDIX E xtract-Free Wheat Straw TABLE 17. Results o f the Acid H ydrolysis Development Experiments (continued) Substrate Acid Concentration Temperature (°c) Time (hr) Hydrolyzed % Delig n in fie d Wheat Straw 18N 80 3.0 98.7 E xtract-Free Wheat Straw 18N 0.7N «20 80 3.0 4.0 62.8 E xtract-Free Wheat Straw 18N 0.7N «20 80 1.5 4.0 60.6 Extract-Free Wheat Straw 18N 0.7N «20 80 1.0 2.0 59.2 D e lig n in fie d Autohydrolyzed Wheat Straw 18N 0.7N «20 80 1.0 2.0 23.8 E xtract-Free Wheat Straw 16N 0.7N 25 80 1.0 2.0 48.9 Extract-Free Wheat Straw 18N 0.7N 25 80 0.5 2.0 56.0 Autohydrolyzed/ Extracted Wheat Straw 18N 0.7N 25 80 0.5 2.0 15.1 MONTANA STATE UNIVERSITY LIBRARIES 3 762 10024284 9