BACKGROU D OF THE πr/E TrO 1- Prior Art

Orgaπosolv hydrolysis processes havs been success fully demonstrated on certain types of cellulosic material particularly iignoceliulosics. The easiest wood to delign fy by organosolv solutions is aspen while conifers such as hemlock, Douglas-fir and pines showed substantial resis¬ tance. Sugarcane rind was found to be relatively easy to hydrolyze. Cotton linters which are essentially cellulose especially the crystalline fraction; were very difficult to hydrolyze by prior art processes. The reasons for the hydrolysis differences are related to variations and heter geneity in structure and the chemical composition of cellu losic materials. Thus traditionally organosolv processes have been used primarily with cellulosic materials which are easy to deiignify. Cotton linters have been avoided especially in sacc arification work because of their resistance to hydrolysis and the harsher process condition required for their hydrolysis in rapid conversion of the polymeric glucan to monome ic sugars.

The prior art. has described various org-anosolv processes for deligni ication and/or saccharification of cellulosic materials and vegetable crops. In general such processes involve the use of a mixture of water and a solvent such as alcohols or "<e ones and sometimes other solvents of a non-polar nature along with an acidic

compound to facilitate the hydrolysis. In most instances there is a several hour treatment required to accomplish delignification and additional hydrolysis of the cellulosic residue, depending on the hydrolysis ^" power of the solvent system used and its ability to delignify the particular lig- nocellulosic material. Prior art processes have been charac¬ terized by poor deiignification ability, slow hydrolysis rates and extensive sugar conversion into non-sugars, mainly furfural≤ and organic acids. Hence the sugar recoveries were too low to be commercially attractive to develop such processes on a commercial scale. All of the prior art sac¬ charification processes, of which we are aware of, suffer to some degree ^' from one or the other of these disadvantages. It has long been thought that such were inherent in organosolv processes, particularly with difficult to hydrσlyse σellulo- . sic materials such as cotton linters an<? the conifers.

Thus ϋ. S. Patent No. 1,919,623 to Dre us (1933) describes pretreat ent of wood with concentrated acid in acetone-water carrier solvent mixtures mind after removal of. the organic solvent heating the acid-containing wood at low temperature for several hours to cause i _ situ hydrolysis of the carbohydrates without simultaneous dissolution of the lig- nin. The treated lignocellulose was reportedly practically insoluble in the acetone—ether water mixtures, on treatment of the prehydrαlysed material with the same solvent, only the excess acid was removed and used in further -treatments. De¬ composition of the pre-hydrolysed cellulose material to su¬ gars was effected on boiling in an aqueous weak acid solution, ϋ. S. Patent No. 2,022,654 also issued -co Dreyfus describes a similar approach for the production of cellulose pulp in that, wood chips are pre—treated with concentrated mineral acid carried in up to 80%'acetone in water to soften the wood and after substantially removing all the acid the chips are treated for 9 to 12 hours at 17Q ^a C to 230*C in a pressure vessel using 50 to 30% acetone water or mixtures of acetone and non-polar organic solvent. U- S. Patent No. 2,959,500 to Schlapfer et al describes a hydrolysis pro¬ cess with the solvent consisting of alcohols and water and

optionally of a non-polar solvent at 120 ^a C to 200°C in the presence of a small amount of an acidic compound which was claimed by the inventors as unreactive with the alcohols. The process as thought is relatively slow and limited in saccharification power and the sugar yields are much less than quantitative. ϋ. S. Patent No- 1,964,646 to Oxley et (1934) shows slow saccharification with strong acid. U. S. Patent No. 1,356,567 to Kl inert and Taventhal (1932) teac the use of aqueous alcohol at elevated temperatures for production of cellulosic pulp in a pressure vessel using small quantities of acids or bases as delignification aids. The treatment is described in steps of three hours each. Other prior art is described in ϋ. S. Patent No. 2,951,775 to Aoel in-aa ich wood is saccharified bv the use of muitioie applications of concentrated hydrochloric acid at 25°C to 30°C. ^* Objects of Invention

The main object of the present invention is to rapidly and quantitatively solubiiize and recover chemical components of cellulosic materials.

A further object, of the invention is to reduce the hydrolysis time and substantially increase sugar for¬ mation rates in hydrolysing cellulosic materials.

A further object of the invention is to reduce sugar degradation to non-sugars during high temperature hydrolysis of cellulosic materials.

A further object of the invention is to simul¬ taneously dissolve and then recover separately the chemical constituents of cellulosic materials to yield mainly xylose, hexose sugars and lignin if the material is ligno- cellulosic.

A further object of the invention is to, if so desired, convert the isolated centoses and hexoses into respective dehydration products such as furfural and hydrox methyl furfural, levulinic acid by re-exposure to high tem¬ perature and recover monomeric_ urfurals, levulinic acid.

A further object of the invention is to quanti¬ tatively hydrolyse cellulosic materials at such a rate that

when the organic volatiles are evaporated from the hydrol sis liquor and the lignin if any is separated from the aqueous solution, higher than 10 percent by weight sugar solids is obtainable from the solution. A further object of the invention is to substa reduce the concentration of acid reσuired to maintain and regulate a given hydrolysis rate and thereby substantiall reduce the catalytic effects of acids in degradation of sugars at high temperature. Alternately, the object of the present inventi is to reduce the reaction temperature required to achieve a certain desirable reaction rate during the hydrolysis process and thereby maximize the sugar recovery.

A further object of the present invention is to reduce the energy required for hydrolysis by use of a major volume proportion or in excess of 7G percent of acetone which has heat capacity and heat of vapori¬ sation much, lower than that of water and thus can be easily volatilized to cool the hydrolysis liquor. A further object of the invention is to obtain substantially pure low DP cellulose on very short selective delignification and hydrolysis of cellulosic materials, which is useful as animal fodder, food additiv and as industrial filler and adsorbent. These and other objects will become increasing apparent by reference to the following description. GZ ZRAL DESCRIPTION

The present invention relates to an improvemen in a. process for the production of carbohydrate hydroly- sates as sugars from a comminuted cellulosic material which can contain lignin by treating the material in a pressure vessel with a solvent mixture of acetone and water containing a small amount of an acidic compound at elevated temperatures to form reducing sugars in a liquor the improvement which comprises:

Cal providing mixtures of acetone and water c taining at least about 7Q volume percent acetone and the

- _ -

catalytic acidic compound as the solvent mixture in the pressure vessel at the elevated temperatures with the cellulosic material;

Cb) treating the celLulosic material in the soL- vent mixture for a limited period of time a -he ele¬ vated temperatures until the cellulosic material is at least partially dissolved and such that at least 90 per¬ cent of the solubilized sugars from the cellulosic .-nateri are recovered without degradation to non-sugars in the liquor; and

(c) rapidly cooling the liquor as it is removed from the pressure vessel.

The present invention -also relates to an impro%**e- ent in a process for the prα-duction cf carbohy-crate hydroiysates as sugars and lignin from a comminuted cellulosic material which can contain lignin by treating the material in a pressure vessel with a solvent mixture of acetone and water containing a small amount of an acid compound at elevated temperatures to solubilize any ligni and to form reducing sugars in a liquor, the improvement which comprises:

(a) providing mixtures of acetone and water con¬ taining at least about. 70 volume percent acetone and the catalytic acidic compound as the solvent mixture in the - pressure vessel at the elevated temperatures with the cellulosic: material;

(b) treating the cellulosic material in the sol¬ vent mixture for a limited period of time sufficient to dissolve less than 50 cerceπt of the cellulose in one sta at the elevated temperatures until the cellulosic materia is at least partially dissolved and such that at least 90 percent of the solubilized sugars from the cellulosic material are recovered without degradation to non-sugars wherein the carbohydrates in the cellulosic material are dissolved and* ydrolyzed partially or subs antially completely?

(c) continuously removing the liquor from the pressure vessel;

(d) rapidly cooling the liquor by controlled f evaporation of acetone to retain aqueous solution; and (e) recovering the sugars and any lignins from residual aqueous solution.

Unexpectedly, it has been found that acetone in volume concentrations in water of greater than 70% with a catalytic amount of an acid greatly accelerated the hydrol sis rates in forming stable complexes with the sugars form the hydrolysis at elevated saccharification temperatures where there is limited retention time in the pressure vess Such phenomenon where decomposition of polymeric carbohydr is accelerated by sugar complex formation is not described the prior art and would not be predictable from prior art descriptions. Usually, such complexes have not been beliej to exist in aqueous acetone solutions especially at such high temperatures. The result of the present invention- is that at the selected conditions there is substantially no degradation of sugars during the saccharification process although the acetone complexes are fαun-1 to hydrolyse roug 500 times faster than the alkyl giucosides and polyglucan described in the prior art. Further benefit of the aceton sugar complexes is their facile separation into individual sugar species based on such simple processes as volatiliza tion, selective hydrolysis and liquid-liquid extraction. Complex formation of onomer!c sugars in anhydrous acetone in the presence of mineral acids at room temperature is described in Methods in Carbohydrate Chemistry, Vol. IX,pp. The term "cellulose material" includes material of vegetable and woody origin, generally in comminuted for

The acidic compounds can be of inorganic or org origin and should be inert with respect to the solvent. S inorganic acids as sulphuric, hydrochloric and phosphoric acids are preferred; acidic salts such as aluminum chlorid and sulphate, ferric chloride and organic acids such as tr fluoroacetic acid can also be used.

The elevated temperatures are between 145*0 to and most preferably between 160 ^β C to 210 ^β C.

The catalytic amount of the acidic compound is preferably between 0.05 to 0.5 percent by weight of the solvent mix¬ ture. Smaller amounts are effective especially when high temperatures are selected. A reaction time per treatment of less than required to dissolve 50 percent of the solid residue at the particular acid concentration and reaction temperature should be used and allows generally accept-ably high yield of reducing sugars in dissolved form. The suga exposure time to high temperature will regulate the rate o solvent feeding to the reactor and will generally depend on the acid concentration, amount of acetone and level of elevated temperature used. Thus for very rapid hydrolysis acid concentrations αf 0.04 to 0.06 Normal, acetone con¬ centrations of about 30% and temperatures over 2Q0 ^a C can be used. However, for near theoretical sugar yields, low acid concentration (0.02 Normal and lass) high acetone concentration (above 30 percent) and high temperature (abo 200 ^e C) are most suitable.

The prior art aqueous ^' weak acid and alcoholic or- ganosoiv processes are relatively slow and have limited hydrolysis power even with aasiiy hydrolysable ligno- cellulαsic αaterials such as. aspen and sugarcane rind (bagasse) . These woods usually take between 60 minutes to β hours to become hydrolysed where the sugars hydrolysed in a single step. The lignin is resinified to a dark refractory mass insoluble in alkali and most organic sol¬ vents. Shorter hydrolysis times between 30 to 90 minutes are specified for continuous percolation processes, how¬ ever the sugar yields rarely exceed 45 to 50 percent of the theoretical value by such, processing. Higher sugar yields are said to occur with enzymatic hydrolysis processes but these hav the draw back that only delignified cellulose can be hydrolyzed by enzymes and the hydrolysis times range between 4 hr for continuous to longer than 24 hr for ba _t ch fermentations. On the other hand, difficult to hydro lyze species such as cotton lintars -and Douglas-fir wood

- a-

can be easily treated by the present invention and diss within 40 and 20 minutes, respectively. The yields of reducing sugars and lignin are in excess of 95 percent theoretically available amounts and are obtained in hi 5 purity and very reactive form by the present invention. Reaction vessels with inert linings are used to eliminate the sugar degradation catalyzing effects of transition metal ions such as Ni, Co, Cr, Fe and Cu whi may be components of metallic vessel walls, tubing and Q other control elements wich which the hot liquor comes contact with.

Using the process of the invention, continuous percolation at predetermined rates, where there is a residence time of less than that required for hydrolysi 5 of 50 percent of the remaining solid residue at any ins at the prevailing temperature ^' and acid concentration selected in the reaction vessel, is preferred and resul _. in partial or total dissolution of the material depending on the extent the hydrolysis is allowed to Q proceed. In multiple step batch treatment partial hydr sis with dalignification, which occurs first, yields relatively pure cellulose. Continued hydrolysis with t same or different solvent mixture leads to total saccha fication and also allows stepwise separation of the var 5 wood components in high purity and high yield.

Notwithstanding these process options the recov of pentoses from the reaction mixture is generally by flash evaporation of the major fraction of the acetone with continued distillation under reduced pressure or b 0 steam stripping to yield the pentose sugar complexes in the distillate. Separation of pentoses and hexoses by simple means is made possible by the largely differing boiling points of their acetone-sugar complexes which f even in the presence of water during the high temperatu 5 hydrolysis step in the present invention provided the acetone concentration exceeds 70% by volume.

The dissolved lignin precipitates in the remaining aqueous sugar solution as relatively low molecular weight ^(■ M ^« 3200) granules which can be dried to a pcvier having spherical particulate sizes between 2 to 300 micrometer ' 5 on filtration or centrifuging and washing with cold water. Purification of the crude lignin is by repeated re-dissolu¬ tion in acetone, filtration to remove undissolved residues and re-precipitation into large excess of water or by spray drying the highly concentrated acetone solution. The re ai 0 ing aqueous solution after filtering off the lig in pre¬ cipitate is a clear solution of mainly hexose sugars of 10 percent or greater concentration and contains ^' other water soluble compounds.

The pentose distillate a he- se syrup when hydro- 15 lyzed by being acidified and boiled for at least 20 minutes yield the major sugar fractions in onosaccharids form and high purity. If so desired, on extended boiling of the separated sugar fractions in the presence of acid,' select- ; ive.conversion of sugars to appropriate ehydration pro- 20 ducts such as furfurals, levulinic acid and formic acid can be effected, as is known from the prior art.

After hydrolyzing the cellulosic material at ele¬ vated temperature for a limited period of time, it is very important that the temperature of the reaction mixture be 25 rapidly lowered to under 100°C to avoid unwanted degrada¬ tion of the sugars. This is best accomplished by controlled flashing off of the volatiles since sugar degradation was found to be insignificant below the boiling point of water even in the presence of dilute acids. Usually, the cool- 0 ing of the liquor can be continued to ambient temperatures or less (25 ^β C) before fermentation or further processing.

The above described process can be operated in continuous or semi-continuous manner using batch cooking principles for the latter. Semi-continuous saccharification 35 would employ a battery of pressure vessels ^' each at various stage of hydrolysis to simulate a continuous process. In

continuous operation, ail stages of hydrolysis are accom plished in a single pressure vessel and the pro.Juct r-.ix i always determined by the particular sacchari ication pro¬ gram set. Comminuted wood solids and the cco'-c .-.c iiq ur <are fed continuously to the pressure vessel at such a rit that the time elapsed between feeding and exit of the pro ducts would not exceed that determined earlier to obtain 50 percent hydrolysis of solid residue at any one stage considered for the process. Thus the. residence time would be always fitted to the cst sensitive stage in ord to provide sugar recoveries exceeding 90 percent for that particular stage. The three major stages of sacch n ica to be considered are:

(a) bulk deiignification and re-hy-lroiysis; luring this stage up to 75 percent of the lignin and 95 percent of the governing hemiceliuios.es. (xylose in hard¬ woods and mannαse in softwoods) may be removed. The soli residue yield is invariably above 5-0 percent of the starting material; (b) continued deiig i ication and cellulose purification stage; during this stage deiignification is largely completed and the rest of the hemicellulose sugar and some of the amorphous giucan are removed. The solid residue at this stage is generally less than 33 percent a is predominantly crystalline in nature;

(c) proceeding to total saccharification, the residual cellulose of stage (b) is decomposed to monc eric sugars. This step may take mere than one liquo change to accomplish a better than. 90 percent sugar recov In continuous operation, liquors collected from the various stages of hydrolysis may contain sugars from all stages (a) to (c) which is the situation with an appa tus having no means of separating the top pre-hydrolysis liquor from the rest of the liquor pump d in with the chi _W ith the presen _t invention such separation for purificati of the sugars, is unnecessary because the sugars occur as complexes, pentoses having a different volatility than t

hexose sugars with which they cay be mixed. The lignin is separated on basis of its insolubility in water and i recovered outside the reactor on flash evaporation of th organic volatiles. Separation of the first and second stage liquors from the rest of the hydrolyzate would have particular significance on continued heating of the liquors to caus dehydration of especially the pentose sugars to produce ^' corresponding furfurals and levulinic acid. In this cas only minor amounts of hexose sugars would have to be saccharified. The sensible way to produce furfural from pentose sugars is following the flash evaporation stage and completion of the first reduced pressure separation the sugars according τ. their volatility. Alternately, steam stripping may also be used with good results and relatively pure pentose solutions be obtained in nearly quantitative yields. Such distillates when acidified C2. be reheated under highly controlled conditions and high purity furfu-ral be produced in better than 95% yields. In practical hydrolysis, based an the semi-contin ous process, five liquor changes would be required tα cau total saccharification and dissolution and provide mass recoveries better than 95%. The preferred liquor to wood ratio is 7:1 to 10:1. Due to the shrinking mass bed the total amount of liquor required for hydrolysis of 100 kg of aspen wood at a constant liquor to wood ratio of 7:1 is 1356 kg for an overall liquor to wood ratio of 13.56:1 Under these conditions the average sugar concentration in the combined residual aqueous phase (271 kg) is 30 percen (32.3 kg of recovered sugars) .

In continuous percolation, the liquor to wood ratio can be kept constant at 10:1 as by necessity succes ive additions both wood and liquor will carry hydrolyzat of the residuals already within the reactor. This also est-ablishes sugar concentrations to be in the order of 37 to 40 percent following flash evaporation of the volatiles. Such high sugar solids concentrations were

hitherto possible only with strong acid hydrolysis systems but not with dilute acid hydrolysis.

Discussion of the liquor to wood ratio is extreme important in organosolv and acid hydrolysis processes since

^■ 5 it directly relates to energy inputs during the hydrolysis and solvent recovery as well as during alcohol recovery fro the resulting aqueous solution following fermentation of the sugars to ethanol or other organic solvents. Thus the liquor to wood ratio will have a profound effect on the

10 economics of biomass conversion to liquid chemicals as well as the energy efficiency (energy gained over energy expanded in conversion) of the process.

Steaming of the comminuted cellulosic material before mixing with the Hydrolysis liquor can be used to

15 advantage to expel trapped air. Such treatment will aid rapid liquor penetration. Such practice is well known from the prior art.

EXAMPLE I Saccharification. power and s-igar survival were

20 compared for three competitive systems namely: acidified water (aqueous weak acid) , acidified aqueous ethanol and acidified aqueous acetone in the following example.

In every case purified cotton linters having TAPP

25 0.5 percent viscosity of 35 cP and 73 percent crystallinity index at 7 percent moisture content were used- Acidifica¬ tion was affected with sulfuric acid by making up stock solu tions of the various solvent systems each being 0.04 Normal respect to the acid. Hydrolysis conditions were as follows:

30 In a series of experiments one gram samples of cotton linters Coven dry weight were placed in glass lined sta.inless steel vessels of 2Q ml capacity along with 10 ml of the solvent mixture and heated at lSQ ^β C for various lengths of time and residual solids and detected sugars in

35 solution were plotted on graph paper. The times to obtain dissolution of -about US , 75, 50 -and 25 percent of the substrate were read from the graphs and show in Table 1. At the end of the reaction periods heating,-was

interrupted, the vessel chilled and its cold contents filtered through medium porosity glass crucible, the undissolved residue first washed with warm water followed by rinsing with several 5 ml portions of acetone and finally by warm water. The residue weight was determined gravimetrically after drying at 1Q5°C.

For comparative analytical purposes the combined filtrates were diluted to 100 ml with water and a half illiϋter aliquot was placed in a ^' test tube with 3 ml of 2.0 Normal sulfuric acid added and subjected to a second¬ ary hydrolysis at 100 ^β C by heating in a boiling water bath for 40 minutes. The solution was neutralized on cooling and the sugars present i_. the solution were determined by their reducing power. The results were thus uniform based essentially on the resultant onosaccharide≤ liberated during the hydrolysis process. Theoretical percentage of reducing sugars available after the hydrolysis of the sub¬ strate was determined by difference between the known chemical composition of the starting material and the weight loss incurred due to the hydrolysis. To account for the weight increase of the carbohydrate fraction due to hydration of the polymer on breakdown into moncmeric sugars, the weight loss is normally multiplied by i.iiil, the weight percentage (11.11%) of the added water to the cellulose- in hydrolysis to monomeric sugars.

As evidenced from TA3.LE I, hydrolysis rates im¬ proved constantly as the acetone concentration increased to 50 percent. However, significant improvements were observed only as the acetone concentration was raised above 70 per- cent by volume of the acidified solvent mixture. Very rapid hydrolysis rates were obtained with nearly anhydrous acetone solutions. The dissolved sugars were found to be most stable when using a solvent mixture of between SO to 90 percent acetone even though the relative half lives were relatively short. Sugar survivals over 90 percent are obtained as long as the reaction time at temperature is kept below that re¬ quired for hydrolyzing 50 percent of the substrate to dis¬ solved products. The time required to hydrolyte 50 percent

of sugar survival. This criteria holds regardless of what stage of hydrolysis is considered. The solvent effect both on the hydrolysis rate and sugar survival for limited hydro sis times was the most surprising discovery of the present invention whereby maxima were found around 80 to 90 percent acetone concentration in the ^' reaction mixture. At higher acetone concentrations, the response of the hydrolysis rate to increase in temperature and acid concentration was observ to follow well known kinetic principles in contrast to both the aqueous dilute acid and acidified aqueous ethanol system in which the balance of increase in higher hydrolysis rates and sugar degradation did not improve with an increase in these parameters especially that of the temperature. The improved sugar survival with increase in acetone concentra- tion is attributed to formation of acetone sugar complexes which have improved stability at high temperature. The com¬ plexes are very readily and safely hydrolyzable to free sugars on heating with dilute acid, at 1QQ°C for a limited amount of time. In identical stationary acidified ethanol-water cooks, in. which the ethanol concentration was higher than 80 percent neither deiignification nor hydrolysis was ob¬ tained due to the fact that the acid catalyst was quickly co sumed by reaction with the alcohol by formation of ethyl hydrogen sulfate (C_,H_-0-SQ ₂ -OH) and formation of diethyl ether via condensation of two ethanol molecules. Ether for¬ mation was quite substantial under these conditions. Also alkyl glucosides formed in high concentration alcohol soluti are substantially more difficult to hydrolyse to free sugars than the corresponding acetone complexes, and alcoholysis results in. oligomeric sugars rather than monomers as is the case in acetone-water solutions. Thus alcohols prove to be largely unsuited for hydrolysis media due to the unwanted so vent loss and general danger from the explosive ether. With lignified materials the low deiig ification power of acidi¬ fied alcohol solutions is clearly a drawback. With 80:20 ethanol:water cooks in the presence of 0.19Q percent CO-04 Normal} sulfuric acid at 18Q ^β C the hydrolysis rate was 5.47 10 ³ i ^'1 and the half life of cotton linters decomposition

was 126.8 minutes. A maximum of 76 percent could be dissolved in 254 minutes, the crystalline residue showing substantial resistance to hydrolysis in the alcoholic solv .Residual acid concentration was found to be one fourth of originally applied, i.e., 0.01 Normal, the balance possibl consumed in the various side reactions.

It is evident from the data that under identica hydrolysis conditions excessively long hydrolysis times are required for complete dissolution of cotton linters both by acidified water and acidified aqueous ethanol medi An increase of the ethanol concentration from 50 percent t 80 percent did not improve the hydrolysis rate or improve particularly the sugar survival. The hydrolysis rate in ethanol water was only marginally better than in dilute acid in water.

These examples clearly show that a high acetone concentration over 70 percent is mandatory for high speed hydrolysis and high sugar survival. Under the conditions indicated for sugar recoveries better than 90 percent, reaction times (or high temperature exposure times of less than indicated for half lives are preferred) . Thus accord¬ ing to these data, total saccharification and quantitative sugar recovery would dictate a percolation or pass through process wherein the liquor residence time would not exceed lo minutes- when 80:20 acetone:water with 0.04 formal sul- furic acid is use as solvent mixture at 130°C temperature. The residence time would have to be substantially shortened when higher temperatures and larger acid concentrations are used as shown in the following examples. Solid residues less than 50% in yield show high degree of crystallinity (87%) and are pure white, have a DP Cdegree of polymerizationl of 130 to 35Q.

-BAD ORIGI

-16-

TASLΞ 1. FORWARD REACTION RATES IN ^' STATIONARY HYDROLYSIS OF COTTON LINTERS AS A FUNCTION OF ACETONE CONCENTRA ION.

Catalyst: 0.04 N H-SO^ Temp. 130 ^β C, Liquor/wαod=10/l.

ACETONE/ DISSOLVED REACTION R AT: REDUCING

WATER CELLULOSE IKE ^{• X} I _Q 3 i a— uAR

RATIO % min rain ^"* FACTOR YIELD, %

25 137 82

SO 330 46

0, H ₂ 0 2.1 75 660 16 99 2192

25 115 •a

•→ > _*

50 277

10/90 5 75 555 2. 1.2 O i- α 99 1842 ^* ■ w

J_ T mi tt

25 91 3 t

O C __:

— β ι_ a) 5

50 219 a. >

30/70 3.15 1.5 o 75 439 ■xJ α sr* s-. _ 99 1458

25 49 95 Zxc Recov . 50 118 64 Good Recov. 99 783 27 Recovery

25 12 98 Ξxc. Recov. 50 29 73 Good Reco

70/30 24.1 11.5 75 53 45 Poor 99 191 35 Recovery

25 5 99 Excellent 96 Recovery

30/20 50 13 52.7 25.1 75 26 73 Good Recov. 99 87 58 Poor Recov,

25 3 99 Excellent 50 6 94 Recovery

90/10 112.8 53.7 75 12 79 Good Recov. 99 41 56 Poor Recov,

EXA-PLE II The effect of acid concentration on the rate of hydrolysis and sugar survival in 80:20 acetone:water solve mixtures was studied at 130 ^β C temperature using cotton linters as substrate.

In stationary cooks one gram samples (oven dry) of cotton linters were hydroiyzed in glass lined stainless steel pressure vessels along with 10 ml of the appropriate hydrolysis liquor and heated until the original substrate mass was hydrαlyzed and dissolved. The levels of 25, 50, 75 and 99 percent of hydrolysis were determined by graphing as in Example I.

Work-up of the reaction products followed the same procedure as outlined in Example I. The results are indi- cated in TABLE 2.

Increased acid concentration resulted in higher hydrolysis rates within the range studied and a somewhat faster degradation of the sugars as the single stage hydrol sis times exceeded those indicated as half ^' lives for the solid residue- Ξqua.1 concentrations of suIfuric and hydro¬ chloric acid were found to give largely comparable results. The increased acid concentrations showed a substantial hydrolysis accelerating effect as evidenced by the rapidly decreasing half lives. Thus the hydrolysis rate can be readily controlled by limited acid concentrations, all other conditions being held constant.

- 13-

TABLΞ 2. EFFECT OF ACID CONCENTRATION ON FORWARD HYDROLYS RATES IN STATIONARY HYDROLYSIS OF COTTON LINTERS

Temp.: 180°C, Solvent: Acetone/Water ^* =80/20,L/W-1

25 32.3 99 50 77.9 90

Q.Q1 0.047 75 155.8 8,9 67

99 517.0 57

25 12.2 99 50 29.4 95

0.02 0.095 75 58.8 23.6 2.65 71 99 195.0 57

25 5.0 99

0..04 0.190 50 13.0 52.7 5.92 96 75 26.0 73 99 87.7 53

25 3.5 99

50 3.5 87

0.06 0.285 75 17.0 82. Q 9.2 63 99 56.2 52

25 2.3 99

5Q 5.6 88

0.10 0.475 123.8 13.9 75 11.2 60 99 37.3 50

25 12.8 98 5Q 30.1 92

0.02 0.07 21.7 2.44

61.2 69 HC1 75 99 2Q4.3 60

EXA+PLE III Temperature effects on hydrolysis of cotton linters were studied with acidified aqueous acetone solu¬ tions containing 0.04 Normal sulfuric acid in 80:20 acetone water at different hydrolysis times so that weight losses of 25, 50, 75 and 99 percent could be determined as in Example I. All cooks were preconditioned to 3.5°C before being placed in the oil bath to minimize the effect of heating-up time at the various temperature levels studied. Work-up of the products and analysis followed the same procedure as described in ΞXA-HPLS I .and the results are summarized in TABLE 3.

The data indicate that increased temperature had the most profound accelerating effect of the hydrolysis rate and generally in such single stage batch cooks reaction times exceeding sugar dissolution half lives at any stage of the hydrolysis increased somewhat the rate of sugar degradation at the higher temperature regirses used. How- ^' ever, it was learned that such high temperature hydrolyses afford practically instantaneous high-yield hydrolysis to be carried out on even such difficult to hydrolyza substrate as cotton linters. The rate of sugar degradation can be offset somewhat by lowering the acid concentration and by increasing the liquor to wood ratio whereby the forward reaction rate (k. } ixx hydrolysis remains unaffected but the sugar degradation rate Uc^) is lowered. Thereby sugar sur¬ vival, which depends on the ratio of k^/^ ^is la*-3e ^. lγ improved especially if high acetone concentrations are used.

- -

TABLE 3. EFFECT OF TEMPERATURE ON HYDROLYSIS RATE OF COTTO LINTERS AND SURVIVAL OF SUGARS IN ACIDIFIED 3QΪ2Q ACETONE WATER.

Catalyst: 0.04 Normal S,SO., L W=10/l.

REACTION DISSOLVED REACTION REDUCING TEMP. CELLULOSE TIME R RATE. SUGARS m:t_n 10- am FACTOR ♦ * i

25 40 73 50 96 65

145' 75 193 7.2 3.42 53 99 640 40

25 19 91

50 49 64

160 75 93 21.6 10.3 48 99 329 37

25 5 99 50 13 96

130 52.7 75 25.1 25 - 73 99 87.7 58

25 1.0 99

200 50 2.3 301 .43 93

75 4.6 78 99 15.2 53

25 0.39 99

50 0.93 92

210 75 1.85 45 354 30 99 6.17 53

Acetone/water = 90:10, 0 10 Normal H.2SO4,

** kwat _ _. er* 1.0 (fc1, » 2.1; TAB S 1}

EXAMPLE -IV Cooks reported in this ex-ample explore the hithe to unobserved relationship of increasing the sugar surviv at reduced acid concentration and increased reaction temp tures without any reduction in the high hydrolysis rates disclosed herein. This unusual discovery is demonstrated the data of TABLE 4.

The effect of reduced acid concentration but hig reaction temperature is demonstrated by cooking one gram samples of cotton linters (oven-dry weight) in glass line stainless steel pressure vessels along with 10 ml of 30 : 20 acetonerwater cooking liquor containing 0.01 and 0.005 No H-SO. with respect to the solvent mixture, and heated unt 50 percent and 75 percent dissolution of the substrate was obtained at 190 to 220 ^β C reaction temperature.

Cooling and wor -up of the reaction products to determine sugar survival and reaction rates were performed as outlined in EXAMPLE I.

The data indicate that acid concentration can be successfully reduced and traded by increasing the reaction temperature without loss in reaction rate with a concotsitt increase in sugar yield (survival) whe hydrolysis liquors of at least 80 percent acetone content are used. Such a trend is clearly against all previously published scienti- fie results (Seamen, J.F. , ACS, Honolulu 1979; Bio-Energy, Atlanta 1930) where the increase in hydrolysis rates and sugar survival was a function of both increased acid con¬ centration and higher temperature. The surprising solvent effect of the acetone water system has never been observed or reported in scientific literature or the prior art befo

-_\£-

TABLE 4. EFFECT OF HIGH PEACTION TEMPERATURE AND VERY LOW ACID CATALYST CONCENTRATION ON SURVIVAL OF SUGAR ON HYDROLYSIS OF COTTON LINTERS IN 80:20 ACETONE: WATER SOLVENT.

L/W=10/l.

REACTION DISSOLVED REACTION REDUCING TEM . CELLULOSE TIME ,R RATE, SUGARS % 3 X —1 min 10 min % o.αi o aaltH ₂ S0 ₄ ] 490 pp

50 43.1 87.7

18Q 75 96.3 14.4 64.3

50 18.8 9G.4

19 Q

75 37.7 36.8 70.5

50 7.4 91.5

20Q 75 14.8 94.2 73.2

50 2.0 91.5

210 7 241.4 75 5. 75.7 o.oα 5 Normal [H ₂ SO _Λ ] -245 oc

-t ^**

50 45.3 92.0

190 15.3 '

75 90-6 73-3

50 17.7 93.0

200 39.2 75 35.5 74.4

5 6.9 94.0

210 100.4 . 75 13.3 73.4

50 2.7 36.3

220 257.8 75 5.4 81.0

50 0.25 93.0

230 659.9 75 0.36 87.5

E AMP E V One gram samples of several wood species were hydrα- lyzed in 30:20 acetonerwater containing 0.04 Normal sulfuric acid at 180°C. Hydrolysis rates were calculated only for the crystalline cellulose fractions to avoid the confound¬ ing effect of easily hydroiyzable lignin and hemiceiluloses . Times to mass losses of 25, 50, 75 and 99 percent of the original oven dry mass along with the calculated reaction rates are recorded in TA3LΞ 4. Work-up of the products followed the same pro¬ cedure as indicated in EXAMPLE I except that after removal of the volatiles by distillation it was necessary to remove the precipitated lignins by filtration αr centrifuging.

It is quite evident that under identical conditions the hydrolysis rates for wood are roughly twice chat of cotton linters. Due to the increased forward reaction rates sugar recoveries became quite impressive indeed.

The rate of Douglas-fir hydrolysis was somewhat slower than that of aspen and sugarcane rind. However, when hydrolysis in a purely aqueous system was attempted under otherwise exactly matching conditions (same temperature and acid catalyst content) a hydrolysis rate of 0.5 x 10 min was obtained and only 6 percent weight loss was recorded for a 280 min long cook at 130 ^a C the usual dilute acid hydrolysis temperature. Thus the high acetone content hydrolysis liquor allowed at least IG0 times fas-er hydroly ^¬ sis of Douglas fir by simultaneous dissolution of the Lignxn than possible in purely aqueous systems.

Among the products of partial saccharification of wood,solid residues of about 30 to 35 % yield are pure white, devoid of residual lignin. This cellulosic fraction has a crystallinity index of 80% from aspen wood and a degree of polymerization (DP) of between SO to 280. Similar results are obtained with the other wood species.

TABLE 5. HYDROLYSIS RATES OF SELECTED WOOD SPECIES IN 80: ACETONE:WATΞR MIXTURES AT 180 ^α C IN THE PRESENCE 0.04 NORMAL SULPHURIC ACID AS CATALYST.

CLiquor/Wood=10/l}

WOOD DISSOLVED REACTION R RATE REDUCING SPECIES CELLULOSE,% TIME, in lO-^ in ^*1 'ACTOR SUGARS,*

25 2.1 99 50 5.0 98

ASPEN 75 10.3 135.2 — 96 99 34.5 92

25 2.2 99

SUGARCANE SQ 5.0 93

RIND 75 10.4 96 99 34.5 92

25 3.0 99 50 7.0 97

DOUGLAS-FIR 75 14.0 93 92 99 46.1 36

EXAMPLΞ VI It is found to be a further advantage o f the pre¬ sent invention that the high acetone concentration clearly favors formation of relatively s tabla acetone-sugar com- plexes in spite of the presence of ter . The be tter s ta¬ bility of the sugar complexes at high temperature profound¬ ly affects survival of the dissolved -sugars . The improve¬ ments are quite evident from the data in TABLE 1.

Further due to the differences in vo latility and solubility, of the various sugar complexes the invention allows facile segregation and nearly quantitative isola¬ tion of the five major wood sugars , if so desired. How¬ ever, due to the mixed nature of the sugar derivatives in aqueous hydrolyzates , if such thorough and detai led separa- tion is desired , it is always necessary to neutrali ze the recovered aqueous sugar wort after removal of the volatiles and concentrate the wort to a syrup . The syrup is then redissolved in anhydrous acetone containing 3 percent cid , allowed to stand at least 6 hr unti l all sugars formed their respective di-acetone comp lexes before attempting the detaile separation as described below . The separated sugar comp lexes are readily hydrolyzed in dilute acid on boiling at least 20 to 40 minutes .

Thus 10 g COD) coarse aspen wood sawdust (passing a 5 mesh screen) was charged with 100 ml o f hydro l zing liquor made up to 80 : 20 acetone -.water and 0 . 04 Normal sul¬ furic acid as catalyst . The bomb was brought to ^' 130 ° C temperature by immersing it into a hot glycero i bath within 9 min and heating was continued until the required reaction times were reached.

In another larger bomb 450 ml of hydro lysis liquor containing 80 : 20 acetone : water and 0 . 04 Norma l sulfuric acid was also preheated and connected through a syphon tube and shut-off valve to the reaction vessel . Fo llowing three minutes at reaction temperature (9+3 ^* =I2 min total ⁾ the reaction liquor was drained into a small beaker containing 75 g crushed ice. The reaction vessel was immedia tely recharged with hot liquor from the stand-by ves sel and the reaction was allowed to proceed for an additional 3 min

before again discharging the reactor contents as above. In all, five liquor changes were effected and the liquors collected for analysis. The chilled reactor contents were analyzed as follows: Hydrolysate No. 1 and 2 were combined before evaporation of the low boiling volatiles. Flash evapora¬ tion of the acetone at low temperature (50 ^β C) and reduced pressure resulted in precipitation of a floccula.it lignin which aggregated to small clusters of granules on standing. ^" The lignin was carefully filtered off the mother liquor, washed with two portions of water and dried in _^ vacuo to constant weight as a powder. The lignin powder collected weighed 1.57 g and had a weight average molecular weight of 2300.. The combined filtrate (127 mi) was neutralized and subjected to steam distillation in an all glass apparatus and approximately 35 ml distillate was collected. Both the distillate and residual solution were made up to 100 ml and 0.5 ml portions of each were acidified with sulfuric acid to 3 percent acid and boiled for 40 min on a water- bath. The solutions were neutralized and the sugar reducing power determined by the Somogyi method. The yield of sugars was 1-89 g in the distillate and 1.96 g from the residual liquor. Gas chromatographic determination of alditol acetates of the sugars from the steam distillate indi¬ cated mainly xylose and arabinose whereas from the residual solution glucose, mannose and galactose with only minor traces of xylose were indicated. Hydrolysate o- 3 contained only traces of lignin after evaporation of the acetone solvent too small to collect and determine graviaetricaily. It was removed by centrifuging. The aqueous residue (97 al) was acidified to 3 percent acid with sulfuric acid, boiled for 40 rain and after neutralization filtered and made up to 100 ml. The reducing sugar content of the filtrate was determined by the Somogyi method to be 1-33 g. GC analysis of the alditol acetates determined on an aliquot sample indicated

mainly glucose with traces of annose and galactose. Hydrolysate No. 4 and 5 were processed and analyzed in the same manner as No. 3. H-4 yielded i. 3 g reducing sugars and H-5 yielded 1.40 g sugars both b.-ϊlng composed only of glucose as evidenced by -2 analysis of an aliquot sample.

The undissolved residue was 0.12 g following 2 h drying ^" in an oven at 105 ^β C.

The recoveries summarize as follows: Lignin powder 1.67 g

Total pentose sugars 1.39 g

Total hexose sugars 5.92 g

Undissolved residue (99% glucose) 0.12 q

SO g -ASS BALANCE :

1. LIGNIN RECOVERY : 98 . 2%

2. SUGAR P ΞCOVERY : 97 . 8%

EXAMPLE VII In a similar hydro lysis arrangement to E A P E VI

10 g OD Douglas-fir sawdust (to pass a 10 mesh screen) , ore-extracted wi th dichlorom hane and air dried to 3 percent mois ture content in a contro lled humidi ty room , was hydro lyzed with 30 : 20 aceton jwater - solvent contain- ing α . Q5 Normal Hydrochloric acid in f ive consecutive steps . Each reaction step cons isted o f three minutes at a reaction temperature of 200 °C . The heating up time

_* was 7 minutes. Again Hydrol sate No. 1 and 2 were combine whereas the subsequent fractions were analyzed separately. The combined liquor of H-l and H-2 yielded 2.39 g lignin on low temperature evaporation of the volatiles and 135 mi of aqueous liquor was collected on filtration of the powdered lignin. The dried lignin had a weight average molecular weight of 3200. The filtrate was neutralized to pH 8 and subjected to steam distillation in an all glass apparatus. The 28 ml distillate which was collected contained 0.62 g pentoses which after passing the filtrate through a cation exchange resin in the acid form -and repeated steam distillation of the filtrate

- 2 3-

yielded 0.58 g xylose as determined by GC analysis.

The residue remaining ^* behind after the above steam distillation C123ml) was neutralized on an ion exchange column, the filtrate concentrated to ^' a syrup, seeded with some crystalline mannosβ -and left standing overnight. The crystalline material was collected by filtration and recrystallized from ethanol-petroieum ether. The crystals were re-di-ssolved in water, acidi¬ fied to 3 percent-acid and boiled for 40 min to liberate the free sugars. After neutralization with silver carbonate the solution was analyzed by GC alditol acetates to determine the sugar concentration- the only sugar detected by GC was annose and the yield was cal¬ culated as 1.00 gr. The ethanol-petroieum ether solution was extracted with 5 ml portions of water and the collected aqueous layer combined with the syrup removed from the crystalline product above.. The solution was briefly heated to expel the alcohol, made up to 3 percent acid with hydrochloric acid, boiled for 40 min, neutralized with silver carbonate and alditol acetates were prepared for GC analysis. The combined syrup and filtrate contained a total of 58 g sugars of which 0.29 g was galactose, 0.25 g was glucose and 0.04 g* was mannose. Hydrolysate No. 3 gave 1.89 g pure glucose with

0-4 g of lignin precipitate on removal of the volatiles.

Hydrolysate No. 4 gave 1.6S g of pure glucose with only very small traces of lignin, whereas H-5 gave 1.35 g of glucose and no lignin- The undissolved residue was Q.lS g and was composed of 99 percent glucose. The recoveries summarize as follows: H-1,2S3: Lignin 2-79 g

Xylose 0.58 g

Ara inose (by difference) 0.04 g Mannose 3-- ^Q 0 g

Hexoses 0.53 g

E-3: Hexoses 1.39 g

H-4: Hexoses 1.66 g

H-5: Hexoses 1.35 g

- -

Uπhydrol zed residue 0.18 σ

10.57 g TOTAL SUGAR RECOVERY: 7.60 g = 95.95 % (of theoretical) LIGNIN RECOVERY: 98% Under large scale industrial conditions chilling of the recovered sugar solutions is best accomplished by controlled flash evaporation of the volatiles. Cooling of the liquor samples outside of the pressure vessel in EXAMPLES VI and VII with -crushed ice was adapted as matter of convenience for small scale treatments.