bi.a1 Amyla.se Preparations coINVERSION AGENTS FOR ALCOHOL1C FERMENTATION LU CHENG HA0 AND
JOHN A. JUMP Joseph E. Seagram & Sons, h c .
Louisville, Ky.
Alcohol Recovery Still
T
HE use of amylases as saccharifying agenb in brewing and
ferment,ation is an old art, the beginnings of which are lost in antiquity. Takamine (6) first introduced mold bran, produced by growing Aspergillus oryzae on wheat bran, to the modern distillery, but its use did not find favor on account of the slight offflavor of the alcohol resulting from the use of mold. During recent years mold amylase preparations as saccharifying agents in industrial alcohol fermentation of corn have been studied by Underkofler, Fulmer, and associates (8, 3, 8). Underkofler (7) discussed briefly the relative merit of microbial amylases and malt, and drew the conclusion that “when malt and mold bran are used under optimum conditions for each, the alcohol yields obtained are esRentially the same, based on total dry matter going
The activity of a number of commercial bacterial and mold amylase preparations as conversion agents in alcoholic fermentation were determined. Three samples of bacterial amylase preparations were used to replace malt in premalting corn mash. Successful liquefaction was obtained in all cases. A number of mold amylase preparations in crude, extract, and precipitated forms were used in the saccharification of starch in corn mash. The weights of mold amylase preparations and that of malt required to produce equally satisfactory alcohol yields from the same weight of corn have the following ratios:
into thc process, with apparently a slight advantage for the mold-bran”. In previous studies most of the mold amylase preparations tested were prepared in the laboratory. These materials have proved to be efficient conversion agents, However, they are not available in quantities for large scale experiment. A number of mold and bacterial amylase preparations are on the market, but information is limited concerning the use of these preparation8 in the alcoholic fermentation of grain. It should be of interest to determine the relative efficiency of these materials as liquefying and saccharifying agents in alcoholic fermentation, because these determinations will show whether any of the commercially available amylase preparations can be utilized with advantage BS
Ma1t:crude mold amylase preparations = 8 : 5 Ma1t:extract of mold amylase =8:4to8:6 Malt: precipitated mold amylase = a:o.4toa:i.2 The alcohol yields obtained from the mold amylase preparations range from 89 to 94% of the theoretical yield, calculated on the basis of starch in corn. From these yields it can be concluded that nearly all of the commercial mold amylase preparations tested are good conversion agents in alcoholic fermentation. However, their use on a pommercial scale is dependent on their respective costs. 521
522
I N D U S T R I A L AND E N G I N E E R I N G CHEMISTRY
Vol. 31, No. 6
in terms of proof gallons per bushel of total grain. These ex-
TABLE 1. SOURCES OF MICROBIAL AMYLASEPREPARATIONS pressions are interpreted and calculated as follows: Name of Sample Producer CONVERSION EFFICIENCY. This term expresses the efficiency Bacterial amylase I D of conversion of starch in corn by the conversion agent. It is Bacterial amylase I1 Wallerstein Co., Inc. Bacterial amylase I11 n Mold bran I Fiederick Stearns & Co. Mold bran I1 Wallerstein Co., Inc. RHozyme Maltase No. 20 Rohm & Haas Co., Inc. RHosyme Intermediate Maltase Intermediate Joseph E. Seagiam &Sons,Inc. Converzyme No 503 Converryme No 505 (lab. preparation) Converzyme No. 511 Polidase-C Schwars Lab., Inc. Polidase-S Diatane type 1211 Jeffrey Lab. Diatane B-10 Taka-Diastase Parke, Davis & Co. Alase Takamine Lab. Crystazyms Frederick Stearns & Co. a Name of producer not revealed because of limited supply of product or other technicalities.
1
I
(I
1
1
sources of amylase (Table I). I n addition, it is hoped that the data presented here will clarify the long-existing controversy between the merit of malt and that of mold amylase preparations as conversion agents. ~
FERMENTATION PROCEDURE
In genera& the fermentative determination Of the activity Of microbial amylase preparations as conversion agents of grain mash was conducted in the same way as that designated for malt conversion by Stark, Adams, Scalf, and Kolachov (6)with the exception that the conversion temperature was 130’ F. when
calculated on the basis of alcohol obtained from corn alone. The quantities of alcohol produced from yeast inoculum, premalt, and conversion agent are deducted from the total amount of alcohol. This net amount of alcohol produced from corn alone divided by that theoretically obtainable from corn gives the conversion efficiency. The theoretical yield is calculated on the basis of starch determined by the A.O.A.C. diastase-hydrochloric acid method (1). A detailed sample calculation was given in a previous publication (2). YIELDBASEDON CORN. The alcohol yield in proof gallons per bushel of corn is calculated from the grams of absolute alcohol obtained from each fermenter containing 56.3 grams of corn. The factor for converting grams of absolute alcohol from 56.3 grams of corn to proof gallons of alcohol per 56 pounds of corn is:
($$ + 6.60) X 2 = 0.301 where 6.6 = weight of 1 gal. of absolute alcohol, Ib. The yield of run 361 was therefore: 18.02 x 0.301 = 5.42 proof gal./bu. (wet basis) loo = 6.04 proof gal./bu. (dry basis) 5’42 100 10.4 YIELDBASEDON TOTAL GRAIN. A previous publication (5) gives the derivation of factors and equations involved. The alcohol yield in proof gallons per bushel of total grain is calculated as follows~ P.G./bu. = [ (% by vel. of alcohol as read in 100 CC. of dist.) (Initial sugar concn. of yeast mash per 100 ml. X 100 (ml. of dist.) concn. of 0’0208)1 175 (ml. of mash distd.)
-
-
mold amylase preparations were used. (The conversion temperature for malt was 145’ F.) The beer was distilled after 64 hours ‘I of fermentation. The specific gravity of the distillate was determash (gal. of mash/bu.) X mined with the immersion reflectometer at 17.5” C. (63.5O F.). According to the above example, the yield of run 361 was: Comparison of the results of “per cent alcohol” in distillate determined with the reflectometer and those determined with a 100 2 I12.01 (14 X 0.0208)] X F~ X 38 X P.G./bu. pycnometer showed that there was no appreciable difference = 5.08 (wet basis) between the two methods, and that the reflectomer is a reliable instrument for determining the percentage of alcohol in distillate The composite moisture of the total grain was lO.lS%, therefore: obtained from fermentations with mold amylase preparations as conversion agents. The yeast culture used a strain of SaccharoP*G*/bu. 5.08 X (100 loo10.16) = 5.66 (dry basis) myces cerevisiae labeled “DCL” a t this laboratory. All fermentations were run in triplicate. It is evident that, when calculations are based on alcohol from The quantities of microbial amylase preparation for liquefaccorn alone, the alcohol produced from the premalt and conversion agent is disregarded. The ‘Lconversionefficiency” and “yield in tion and for saccharification are estimated, on the basis of the potency of each, by a method employed a t this laboratory (4). The liquefying power is determined on the basis of the degradation OF ALCOHOL YIELDSFROM CORNMASHLIQUEFIED WITH TABLE 11. COMPARISON of soluble starch into dextrins, and the sacBACTERIAL AMYLASE PREPARATIONS charifying power is determined on the basis of Yieldb, P.G./Bu. Weight Ratio Conversion Baaed on: the production of reducing sugar from soluble Run Li uefying Liquefying Converzyme E5ciency, Total starch using Fehling solution as a testingagent. NO. daterial material No. 505 Corn“ % Corn Eraln The quantities of each amylase preparation 1.0 8 91 91.4 6.03 5.66 361 Malt (contro1)C are so chosen that they will give liquefying 6.04 6.72 91 91.5 362 Bacterial amylase I 1.0 8 6.14 5.77 91 92.9 363 Same and saccharifying activities equivalent to those 1 91 91.0 6.02 5.79 1 .. 5 0 58 342 Same 5.72 5.53 91 86.8 of one part and eight parts distillers’ malt, 343 Same 0.5 5 respectively. If the quantity of amylase 91 93.7 6.18 6.00 432 Bacterial amylase I1 1.0 8 91 93.5 6.17 5.99 433 Same 1.5 8 preparation thus estimated fails to give satis442 Bacterial amylase I11 1.0 8 91 92.9 6.14 6.78 factory activity, a higher concentration is 8 91 91.5 6.04 5.70 443 Same 0.5 generally employed to determine whether the 91 95.6 6.30 6.10 360 Concd. HCl 3.04 8 393 Same 3.0 8(malt) 91 84.0 5.64 6.44 same activity can be maintained with less Weight of corn in the mash is the same in all the fermentation runs; this corn has a starch conamylase preparation. tent of 61.8% and a moisture content of 10.4%. The efficiency of the microbial amylase All yields are oalculated on a dry basis. 6 This malt has m starch content of 50.9% ,and a moisture content of 8%. preparations is expressed as: conversion d The acidity,of the mash during cooklng is approxmately 0.06 N. The pH is adjusted to 6.6 before the addition of conversion material. efficiency, alcohol yield in terms of proof gallons per bushel of corn, and alcohol yield
-
xo
-
-
(1
INDUSTRIAL AND ENGINEERING CHEMISTRY
June, 1945
523
proof gallons per bushel of corn” are, therefore, of interest only for comparison of the activity of the different amylase preparations as conversion agents. On the other hand, the “yield i n proof gallons per bushel of total grain” is an index for comparing the total alcohol production of the process as a whole, including the amount of alcohol derived from the different conversion materials used, and is of p r a c t i c a l interest. BACTERIAL PREPARATIONS AS LIQUEFYING AGENTS
It is well known that bacterial amylase preparations p r o d u c e d from BaciUus subtilia or Centrifuge and Shriber Pressure Filters Bacillus mesentericua have high liquetying action on starch, and these enzyme preparations are Run 360 (Table 11)indicates that acid is still the most efficient Characterized by maximum activity in neutral and alkaline liquefying agent for the pretreatment of grain mash. Theoretical solutions and at high temperature. Bacterial amylases have not yields of 93-95oJo were consistently obtained with 5% of crude found utilization in distillery practice because they produce very mold amylase preparations. However, in our experience acid little sugar but large amountg of dextrin. w&s satisfactory only when mold amylase preparation was the It was thought that baoterial amylase preparations might be saccharifying agent. When malt is used, acid cooking is detriused for thinning cooked mash--in other words, to replace premental to the alcohol yield (run 393, Table 11). Possibly acid malt. TWO samples in extract form (bacterial amylase preparahydrolyzes starch to dextrins, a fraction of which can be saccharitions I and 11,Table I) and one in purified powdered form (bacfied by the amylases in mold amylase preparations although it is terial amylase preparation 111)were tested. The alcohol yields not attacked by the amylases in malt. from corn mash liquefied with these preparations and subsequently saccharified with Converzyme No. 505 (Table I) are MOLD PREPARATIONS A S CONVERSION AGENTS summarized in Table 11. The activities of sixteen samples of mold amylase preparations These results indicate that, when equal weights were used, as conversion agents in the alcoholic fermentation of corn mash bacterial amylase preparations gave either equal or slightly were determined. These samples include eight crude preparations higher alcohol yields than did malt. With the exception of bacin the form of mold bran, two intermediate preparationswhich are terial amylase 111, a copcentrated preparation, when the weight thewater extract of themold bran, and six precipitated and purified of bacterial amylase preparation was reduced to half that of preparations in powdered form. These samples are commercial malt, there was a considerable drop in alcohol yield. Although products except the Converaymes, which were prepared in the there was no significant increase in alcohol yield when bacterial laboratory by growing Aspergillus oryzae on wheat bran by the amylase preparation was used, the mash liquefied with bacterial aluminum pot method (8). The processes employed in producing amyIase had a considerably lower viscosity than that liquefied the commercial samples were not revealed by the producers. with anequa1 amount of malt. When tested with a IO-ml. pipet, the Data on the alcohol yields from corn mash converted with crude ratios of time of flow of 10ml. of each mash through the pipet were: mold amylase preparations are given in Table 111. BacterialThese data indicate that all of the crude amylase preparations water: : ‘ g + e d } = 3: 15 : 100 gave satisfactory conversion efficiencies. Among the samples mash tested at two concentration levels, there is no significant difference between 5 or 8 parts of amylase preparation employed. GenerTherefore, w far as liquefaction of starch is concerned, these bacterial amylase preparations are probably several times as potent ally speaking, any of the preparations tested gave higher contw malt. However, as far as alcoholic fermentation is concerned, version efficiency and higher alcohol yield based on corn than with the exception of bacterial amylase 111, a weight of bacterial those obtained by using malt. The alcohol yields, in proof galamylase preparation equal to that of malt is required to produce lons per bushel of total grain, obtained from mold amylase prepsatisfactory alcohol yield. aration were approximately the same as those obtained from malt.
[~ ~ ~] { ’ $ ~ -
524
INDUSTRIAL AND ENGINEERING CHEMISTRY
Vol. 37, No. 6
tions. The authors were more interested to know the minimum quantity of each amylase preparation which will give an activity TABLE111. COMPARISON OF ALCOHOL YIELDS FROM CORN hi ASH CONVERTED WITH CRUDEMOLDAMYLASEPREPARATIONS comparable to that of eight parts of malt, because the quantity Yield P C./Bu., of fungal amylase required would have a definite bearing upon conversion B b e d on: the application from an economic viewpoint. Run Conversion Material Efficiency, Total Name Wt. ratio" % Corn Grain 5.87 5.74 8 89.0 Malt (control 5.93 5.58 90.0 Conversyme ho. 603 8 0.03 6.60 8 91.4 Conversyme No. 606 5.89 6.60 8 89.6 Mold bran I 6.93 6.73 90.0 Mold bran I 8 0 . 0 0 6 .03 91.0 8 Mold bran I1 5.98 6.70 90.8 M.old bran I1 6 6.93 6.74 6 90.0 Diatane T-1211 0.00 6.72 91.9 Polidase-C 8 0.15 5.90 93.0 8 Converzyme No. 811 The weight ratio of the other two components of the mash cornposition w a s 9 1 parts corn and 1 part liquefying material (malt) in each case.
No.
011 301 381 302 322 482 483 702 $02 I 03
MIXTURES O F MALT AND MOLD AMYLASE PREPARATION
Mixtures containing one part malt and five parts crude mold amylase preparation were used as saccharifying agents in additiaii to the one part of malt used for premalting. Diatane T-1211 antl Diatane B-10 were chosen for these experiments. Alcohol yield. obtained from corn mash converted with these mixtures antl those obtained from the Diatane control (without the extra on(% part of malt) are compared in Table VI. These results indicate that the addition of one part of malt to TABLE IV. COMPARISON OF ALCOHOL YIELDSFROM CORNMASH five parts of Diatane increased considerably the conversiou CONVERTED WITH INTERMEDIATE MOLDAMYLASE PREPARATIONS efficiency. Although the Diatane preparations were the only Yield, P.G./Bu.. Conversion Based on: samples used in this connection, evidence from other fermentation R U Conversion Material Efficiency, Total studies showed that other amylase preparations have the samc No. Name Wt.ratioa % Corn grain6 effect. Run 733 indicated that one part of malt used for conver5.85 5.73 8 88.6 681 Malt (control) 0.17 5.96 4 93.0 682 RHoayme Intermediate sion and one part used for premalting could be eliminated wheti 0.20 6.99 683 RHoiyme Intermediate 4 94.0 5.87 5.74 three parts of concentrated hydrochloric acid were utilized iii 8 89.0 661 Malt 0.17 5.88 8 93.6 562 Maltase Intermediate cooking the mash. The acidity of the mash during cooking wab 0.25 5.67 583 Maltase Intermediate 12 94.0 approximately 0.06 N . The mash cooked with acid had a low Other components of mash composition: 9 1 parts corn and 1 part liquefying material (malt) in each case. viscosity, and the time for completing the fermentation was b These intermediate mold amylaae reparations are in liquid form. shorter than that of a normal fermentation using premalt. The However in calculating total grain yiel8, the intermediate mold amylase preparatibns are considered as parts of the total grain bill. disadvantage of using acid cooking is that the mash must be neutralized before the addition of the conversion agent, which will be an added manipulation in plant practice as well as added The malt used for control fermentations was the same quality cost. as that used in the Seagram plant during 1943. The malt used in run 611 hss a Lintner value of 175O, which is inferior to the EFFECT OF PARTIAL ADDITION ON A L C ~ I ~ OYIELD L quality of prewar malt. The conversion efficiency (89.0%) is It has been reported (8) that, when mold amylase preparatioii approximately 1to 2% lower than that generally obtainable from is added at intervals, the alcohol yield can be improved. Apprewar malt using water control (without stillage) fermentation. proximately three fifths of the total amylase preparation is added Two samples of intermediate mold amylase preparations, a t the time of conversion, and the balance is put into the ferHHozyme Intermediate and Maltase Intermediate, were used as menter after 20 hours of fermentation. Alcohol yields obtained conversion agents. These intermediate preparations are extracts of crude preparations and are in liquid form. The alcohol yields obtained from corn mash, converted with these intermediate mold amylase preparations, are summarized in Table IV. TABLE V. COMPARISON OF ALCOHOL YIELDSFROM CORNXASH These data show that RHoxyme and Maltase Intermediate CONVERTED WITH PURIFIED MOLDAMYLASEPREPAR.4TIONS preparations are excellent conversion agents. The mashes conConversion Material ~~~~~~~i~~ Yield, P.G./Bu., Based 011: verted with these preparations were considerably thinner than Run WL Efficiency, Total NO. Name ratio" % Corn grain those converted with malt or any crude amylase preparation. 231 hralt (control) 8 8 . 0 5 . 8 7 5.06 8 . 0 With four parts of RHoxyme Intermediate or six parts of Maltase 252 89.1 5.87 0.375 5.88 RHozyme 0.25 88.0 5.81 5.80 233 RHozyme Intermediate preparations, conversion efficiencies and alcohol 0.00 5.98 232 91.0 Maltase, No. 20 0 . 2 7 5 yields (based on total grain) were considerably higher than 1.50 89.4 5.90 5.88 Taka-Diastase 552 5.95 0.375 603 90.4 5.96 Polidnse-5 those with eight parts of malt. 1.15 91.3 6.04 5.98 542 Alase 1 . 9 0 543 9 2 . 5 6 . 1 1 6.00 Alase Samples of six purified amylase preparations rere tested for 6.14 0,376 742 93.0 0.15 Crystazyme their efficiency as conversion agents. All of these preparations 4 Amounts used were estimated on the basis of potency of these materials, are in powdered form. The alcohol yields from corn mash condetermined according to an unpublished method. for determining amyloclastic and saccharogenic powers used a t Seagram laboratories. Other verted with these purified mold amylase preparations arc comcomponents of mash composition: 91 parts corn and 1 part liquefying pared in Table Vi. material (malt) in each case. These data show that nearly all of the samples tcsted were preparations of high amylase concentraOF ALCOHOLYIELDSFROX CORN-MASHTABLE VI. COMPARISON tion. As little as 0.25 or 0.375 part gave higher conCONVERTED PREPARATIONS AND 1% MALT version efficiencies and alcohol yields than did the malt Mash Compn., Wt. Ratio' Yield, P.G./Bu., control. However, in this limited number of deterLiquefying Conversion Baaed on: minatiom, the respective quantities of the different K~~ conversionconversion material ~ f f i ~ i ~ ~ ~ Total ~ , material (malt) % Corn grain samples tested may not be the optimum concentraNo. Material. Because some of tpe samples of mold as well BB bacterial amylase preparations used in thia investigation are intended for other uses (for instance, Taka-Diastase is designated as a digeatant of starch foods), the data presented here do not justify any conclusion drawn on the relative merit8 of these samples 8 8 liquefying or Saccharifying agents outside of those in the alcoholio fermentation of grain under specified conditions. The authors hope they did not discredit any of the samples. 1
Malt 8 5 Diatane T-1211 (control) Diatane T-1211 i31 M,alt Diatane B-10 741 kait 733 istane B-10 5 a 91 parts corn in each case. b Concentrated hydrochloric acid. 031 702
1 1
88.0 90.0
5.80 5.93
5.73 6.74
1
94.0
0.20
5.94
94.6' 90.5
0.24 6,37
3h
5.98 6.15
June, 1945
INDUSTRIAL AND ENGINEERING CHEMISTRY
from corn mssh cooked with acid and subsequently converted with Converayme No. 505, with and without the partial addition method, are compared in Table VII. These results confirm that there is an increase of 1to 2% in conversion efficiency as well as in alcohol yield when the mold amylase preparation is added at intervals.
ACKNOWLEDGMENT
94.0 6.20 6.88 96.2 6.35 6.10 96.0 6 . 2 7 e.06 96.2 6.35 6.10 a The mauh oompooition oontained 91 psrta by weight of oorn in eaah aam.
LITERATURE CITED
525
' The authors gratefully acknowledge the helpful advice and suggestions of 5. L. Adams, H. C. BIankmeyer, W. D. Gray, and W. H. Stark. They also wish to express their appreciation for the assistance of D. K. Dubois and P. D. Raths. Thanks are due.to Jeffrey Laboratories, Pabvt Brewing Company, Rohm & Haas Company, Inc., Schwarz Laboratories, Inc., Frederick Stearns & Company, Takamine Laboratory, Inc., and WallerTABLEVII. EFFECT OF PARTIAL ADDITIONOF MOLDAMYLME stein Company, Inc., for supplying samples of mold and bacterial PBEPUTION ON ALCOHOL YIE~LD amylase preparations, The photographs used to illustrate this Yield P Q./Bu., Wt. Rstio of Converryme Convenion B L on:~ article were kindly loaned by Frederick Stearns & Company. Run No. 5060 EMaienay, Total No. let addition 2nd addition % Corn grain
412
cia
402 408 0
8 5
5
'8'
..2
SUMMARY AND CONCLUSIONS 1. The three bacterial amylase preparations teated gave satisfcletory liquefaction of cooked mash when a weight of bacterial amylase equal to that of the malt (or half that of malt in case of the concentrated bacterial amylase preparation) was used. 2. AU of the crude mold amylase preparations studied were efficient convexxion agents. When mold amylase preparations were used, the conversion efficiency was higher than that obteined with malt. The alcohol yielh, in proof gallons per bushel of total grain, resulting from crude mold amylase preparation and those from malt are generally the m e . 3. Nearly all the intermediate and purified mold amylase preparations tested were excellent convemion agents. The concentration of the different preparations required to give satisfactory conversion efficiency varied from 0.25 to 1.6 parte in the case of the purified preparations, and from four to six parts in the case of the intermediate preparations. 4. A mixture of one part malt and five parts Distane T-1211 or Diatane B-10 gave considmble increase in conversion efficiency as well aa in alcohol yield. Yields of alcohol over 5.9 proof gallona per bushel of total grain (dry basis) were succesafully obtained. [The Diatane preparations were the only samples tried for the mixed amylase (malt-mold) effect on this investigation.] 6. Cooking the maah with three parts of concentrated hydrochloric acid seemed to be the most efficient method for liquefying mash. Conversion efficiencies aa high as 96.6 and alcohol yields of 0.16 proof gallons per bushel of b t a l gain were obtained with acid cooking end Diatme. 0. Slight improvement of alcohol yield of 1 to 2% by the addition of mold amylase at interv& was confumed. 7. From the study of these sixteen samples of crude, intermediate, and purified mold amylase preparations, it can be concluded that nearly all of the commercial mold amylase p r e p arations teated we good conversion agents in alcoholic fermentation. However, their use' on a oommercial scale depends on their cost. Any of the mold amylase preparations studied in this investigstion can be used if it is svailable st B price proporbionally comparable to that of mdt.
( 1 ) Aesoc. Official Agr. Chem., Official and Tentative Methods of Analysis, 5th ed., p. 559 (1940).
(2) Hao, L. C., Fulmer, E. I., and Underkofler, L. A., IND.ENO. CHSM., 35, 314 (1943). (3) Schoene, L., Fuler, E. I., and Underkofler, L. A., Ibid., 32, 544 (1940).
(4) Seagram Laborstory,Tentative Method for Estimating Amylases
Activities (unpublished). (5) Stark, W. H., Adams, 8. L., Scalf, R. E., and Kolachov, Paul, IND.ENO.Cnsu., ANAL.ED.,15, 443 (1943). (6) Takamine, J. O., J. IND.ENO.CHBM.,6, 824-8 (1914). (7) Underkofler, L. A., Brewers Digest, 17, No. 12,29 (1942). (8) Underkofler. L. A., Fulmer. E. I., and Schoene, I,.. IND.RNG. CHEM.,31, 734 (1939). PBB~EXTED before the Division of Sugar Chemistry and Technology at the CHEMICAL SOCIETY in Cleveland, Ohio. 107th Meeting of the AMERICAN
Countercurrent Extraation Percolators
