TECHNOLOGY OF SOYMILK AND SOME DERIVATIVES

TAN BOE HAN

108201,249 TECHNOLOGY OF SOYMILK AND SOME DERIVATIVES

PROEFSCHRIFT

TER VERKRIJGING VAN DE GRAAD VAN DOCTOR IN DE LANDBOUWKUNDE OP GEZAG VAN DE RECTOR MAGNIFICUS IR W. DE JONG, HOOGLERAAR IN DE VEETEELTWETENSCHAP, TE VERDEDIGEN TEGEN DE BEDENKINGEN VAN EEN COMMISSIE UIT DE SENAAT DER LANDBOUWHOGESCHOOL TE WAGENINGEN OP DONDERDAG 2 OKTOBER 1958 TE 16.00 UUR

DOOR TAN BOE HAN CONTENTS Technology of soymilk and some derivatives

1. GENERAL INTRODUCTION: shortage of proteins 1 production and uses of soybeans 2 2. THE SOYBEAN: the plant — history, characteristics 6 the seed — structure, composition, proteins 7

3. MANUFACTURE of SOYMILK: 3.1 Introduction 13 3.2 Literature 15 3.2.1 Methods of manufacture 15 3.2.2 Dispersion of proteins from seeds 21 3.2.2.1 Particle size and other physical factors .... 21 .2 pH 23 .3 Salts 24 .4 Denaturation 24 .5 Germination 27• 3.3 Experimental part 28 3.3.1 Scope of work 28 .2 Design of experiments 30 .3 Technique of experiments 32 4 Results, discussion, conclusions ' . . .35 3.4 Summary 50 4. PRESERVATION of SOYMILK by HEAT 4.1 Introduction . 51 4.2 Literature . 51 4.3 Experimental part 52 4.3.1 Scope of work 52 .2 Design of experiments. 52 .3 Technique of experiments 53 .4 Results, discussion, conclusions .54 4.4 Summary 60

5. EVAPORATED SOYMILK 5.1 Introduction 60 5.2 Literature 61 5.3 Experimental part 67 5.3.1 Scope of work 67 .2 Design of experiments 68 .3 Technique of experiments 70 .4 Results, discussion, conclusions 78 5.4 Summary 98 6. SOYCURD 6.1 Introduction 98 6.2 Literature 10° 6.2.1 Methods of manufacture ...... 100 .2 Ripening of the cheese 101 6.3 Experimental part 103 6.3.1 Scope of work . . . • • • 103 .2 Design of experiments .'.'.'.' • • 103 .3 Technique of experiments 104 .4 Results, discussion, conclusions . . . • • • • • 105 6.4 Summary ...... • HI

7. SOYYOGHURT 7.1 Introduction HI 7.2 Literature . 112 7.3 Experimental part • 112 7.3.1 Scope of work 112 .2 Design of experiments 113 .3 Technique of experiments 114 .4 Results, discussion, conclusions . 114 7.4 Summary 119

8. GENERAL DISCUSSION 120

SAMENVATTING 127

REFERENCES 1 GENERAL INTRODUCTION.

It isa wellknow n fact that therapi d increase of the population inth eworl d createsdifficul tproblem sa sregard s the safeguarding of its food supply. Although anumbe r ofcountrie s have the dispo­ sal ofsufficien t sources towarran t a good nutrition of itspopu ­ lations,to oman y othercountrie sar eno t able tomee t the foodre ­ quirementsconsidere d asa minimu m forth esustainmen to flife . Promth e nutrients atpla y especially theprotein s are indispen­ sable for thebuildin g upan d maintaining ofth e human body. Ina s fara sth e underfeeding refers towar mclimate sth erol eo f fat is of lessimportance .

Although thispublicatio n isno tmean t asa nutritiona l study it may beusefu lt ocit e some figureswhic hma yserv e togai na nunder ­ standing of the background of thiswork . VISC O (1955)o f the In- .stitutNationa l de laNutrition , Rome, made some calculationsas ­ suminga protei n requirement of 1gram/da y per kg of body weight '^anda n optimal ratio of 1:1 betweenanima l andvegetabl eproteins . Calculated inthi swa y thedaily , individual requirement of total proteinamount st o6 0gram san d ofanima lprotei nt o3 0grams . Prom literature itappear s that theconsumptio n oftota l protein intro - 'pica l countries isnea r tothes e 60grams .However , from the most important sources of animal protein, meat and milk, the average dailyconsumptio n isonl y9 an d 9.3 gram respectively (worldavera ­ ge). According toVisc o this figure iseve nrathe rhig han d a bet­ terunderstandin gmigh t beobtaine d ifth emaximu man dminimu mcon - .sumptio n of animal proteins of the different countries are taken intoaccount . Maximum consumption .Minimu m consumption of animal proteins of animal proteins Asia 17.8 g/day 1.6 g/day Africa 26.1 5.7 Europe 69.5 7.7 South-America 72.3 16.1 North-America 60.5 20.- Prom these figures itma y be seen that even the maximum consump­ tiono fanima lprotei n inAsi aan dAfric a -wheremor e thanhal fo f theworld' spopulatio n isaccumulated - isbelo w thedail y required amount. ,A s itwil l not be easy to increase rapidly the production of animal proteins in underdeveloped countries the use of vegetable proteins of high nutritional quality should be emphasized. Oneo fth emos t important sources suited forthi s purpose is the soybean,o fwhic h theprotein s arehighl y nutritive. Besidesth e cultivationo fth esoybea nha sman y attractive points. First,th e soilan dclimat e requirementso fth ecro par eno thigh . Second,th e yieldpe runi to fare ai shigh ,whil e third theprotei n contento f the beans isver y high. Lastly, theamoun t oflabou r necessaryt o raiseth eprotein s islo wi ncompariso n with many otherproteins .

LAGER (1945)quote s figures fromth eU.S .Dep , ofAgric , Bureau ofEconomic s (R.P.Christensen):

AVERAGE OUTPUT OP PR OTEINS PER UNIT OP RESOURCES Poundspe r Poundspe r acre 100h ro flabou r Milk,whol e 39 89 Eggs 25 56 Beefcattl e 7 45 Wheat,whol e flour 90 1002 Potatoes 118 174 Peanuts 116 200 Soybeans |339 2821 Inorde rt oge ta notio no fth eimportanc e ofth esoybea nw efur ­ thermentio nsom e figureso nth eproductio n anduses . TheYearboo k ofPoo dan dAgricultur e (1956)give sth efollowin gdata . PRODUCTION OP SOYBEANS IN 1955 (1000 metric tons) Europe North America 10320 Latin America 120 Near East 10 Par East 10350 Africa 50 USSR ?

In North America the greater part was produced by the United States (10166),. while in the Par East China (9144), Japan (504), Korea and Indonesia produced jthe most soybeans.

According^ BURTISS (1950) from prehistoric times to 1908, soybean production and trade in soybeans, soybean cake, and soybean oil were confined almost exclusively to eastern Asia. A second sta­ ge, extending from 1908 to 1939, was marked by large exports of soy­ beans and soybean oil from Manchuria to Europe, where the soybean wason eo fth e leadingra wmaterial suse d byth e oilseed-processing industry. A third and present stage ismarke d by the pre-eminence of the United States inproductio nan d processing soybean foroi l andmeal .Thi sphas ebega n in194 0whe nwa rdisrupte d thetrad ebe ­ tweenManchuri aan dEurope . Itha s been estimated that inChin a 55%o f the crop is used in foodproduct ssuc ha sbea nsprouts ,soybea nmilk ,an d soycurd ("tofu"). These foodssuppl ya ver y considerable part ofth eprotei n inChi ­ nese diets. Soybean curds, prepared innumerou s forms, are roughly equivalent tomea t inoccidenta l menu.Te npercen t of the Chinese •output isfe dt oanimals ,8 %i suse d forseed ,an d most of there ­ maining27 %i scrushe d foroi lan dcake .Th eoil , inturn , islar ­ gelyuse d forfood ,whil e mosto fth ecak e isreturne d toth esoi l as fertilizer. Mosto fth e Manchurianexpor to f soybeancake swer e destined as fertilizers forth e rice fiels inJapan , and the sugar caneplantation s inTaiwa nan dSouther nChina . vAlso inJapa n foodproduct sderive d from thesoybea npla ya n im- portantpar t inth eJapanes ediet .Th e leadingsoybea nproduc tcon ­ sumed inJapa n is »miso" afermente dproduc tmad e fromsoybean san d rice."Befor e the lastwa rabou t 67%o fth esoybean sconsume d inJa ­ panwa s used for food products,22 %fo r oilan d cake,an d 11%fo r seed. In Indonesia soybeans have been grown since antiquity, and are widely used by the population for food preparations. However, it hasneve rbee ns o importanta s inChina ,Kore ao r Japan. Introduc­ tiono fselecte d high-yielding varieties fromTaiwa n increased the production ofsoybeans .

InEUROPE , except inth eDanubia ncountries ,soybea nproducti ­ onha sneve rattaine dcommercia l importance.Abou t 1935th eGerman s began aprogramm e ofencouragin glarge-scal e plantingso fsoybean s in theDanubia n countries,particularl y Rumaniaan d Bulgaria, by guaranteeing aspecifi c price for the production for contracted acreage. Thiswa spar t ofth eGerma neffor t tobecom e lessdepen ­ dent onoversea ssource s forfoodstuffs . The leadingworl d importero fsoybean sbetwee nth etw oworl d wars was,Germany . Beginning in,193 4th ene t imports intoGerman ydecli ­ ned materially, largely as a result of adriv e for greater self- sufficiency infats . Thesoybean s imported fromth eBalka nwer e of high quality andwer e used almost entirely in the manufacture of fullfatsoybea n flour and flakes for the German army. Afterth ewar,th eAmerica narm yshippe dsubstantia lquantitie so f soybean flour (mostly lowfat)an d moderate quantities ofsoybean s and soybean oil towester nGermany . Upt o189 8onl yeigh tvarietie sha s been broughtt oth e UNITED STATES. Intha tyea rth eDepartmen t ofAgricultur e begant oin ­ troducene wvarietie sfro mth eEast .B y 1937mor e than 10:000selec ­ tions,representin g250 0distinc t types,ha d been imported and these introductions formed the basis of a vigorous program of varietal researchan dselection . Varietalselectio nha stende d to producethre e distinct typeso f soybeans grown,fo rprocessin g for oil, forhay , and for use asa vegetable.Ha yvarietie sar eusuall ydistinguishe d byblac ko rbrow n seeds,an dar e finer-stemmed andmor e leafytha n theyellow-seede d varieties.Soybean sspeciall yadapte d fordirec t human consumption areknow na sedibl eo rvegetabl evarieties . The most suitableva ­ rietiesar ethos ewit hstrawyellow , greenish-yellow, or greenseed . Thesevarietie sar e generally easier to cook and more palatable thanothers .The yma yb euse deithe rgree no rmature .Vegetabl eva ­ rietiesusuall yhav elarge rseed san dcontai nmor e proteinan d less oil. However,soybea nmea lfo r livestock isth e major soybeanprotei n product inth eUnite dStates . In1942-1947 ,9 0t o 95%o f the total outputwa suse d for feed. Since 1941,soybea n flour has been the principalsoybea nprotei nproduc tothe rtha nmeal . Afterth e endo f thesecon dworl dwa rlarg eorder swer eplace d by theUnite dNation s Reliefan dRehabilitatio nAdministratio n for relief feeding inEu ­ ropean dAsia .

Promth e foregoing it isclea rtha t insom e East and SouthEas t Asiaticcountrie sth eprocessin go fsoybean s for food purposesha s beena wel lestablishe dhabi t fromancien ttimes . Infact , inChin a theus eo fderivative so f the soybean originates from about 3000 B.C. Inman y othercountries ,however , the use of soyproducts is littleknown .Especiall y forth eunderdevelope d countriesknowledg e ofth etechnolog y ofthes eproduct s isusefu l to meet the foodde ­ mandso fit sgrowin gpopulations . Regardingth e longhistor y of the soybean it isno t surprising thatman ysoybea nproduct shav ecom e intobeing . One of these pro­ ductsi ssoymilk ,a foodstuf fwhic hha salread y a very longhistory . Thismil ki sth estartin gmateria l forsom e otherproduct sderive d fromit ,suc ha sso ycheese ,soymil k powder,concentrate d soymilk, etc.Althoughthes eproduct scertainl yar eno tnew , theyhav ehardl y beenstudie d (atleas ti nEuropea n languagesno tman ydat ao nthes e productsar eknown) . As,moreover ,extensio n of the use of soymilk andderivative si sdesirable ,i twa sconsidere d important tounder ­ takea stud yo fth emanufactur e ofsoymil k and somederivative si n ordert oimprov eth equality ,th eyield ,an dt o lowerth ecos tprice . •fa '° b b _~ •: b 1

.• -•*•

to ^ 1 >! =' ? V r '^^^N. f%yf 11'. ID

~y>* •-] rt^'' 1 X

i O 1r 2~ • 1 • :l •- 1 ^~"~"l l~ —• i 1 "*~^i o 1 1 - 7 2 N -^1 1 —- 1 1 • f «D !VHL 1 1 3 \w '1 p— o •8

—~ b ^T\ \^Y

—_ b yp* ' 1 i — o 1 1 . A 1 I 'S&Mc 1 Ton J^ 1 19 1 o 1 N 1 1 1 i 1 r b i i—' » I u m Si !l 1 i 1 1 1 _/ • 1 t ? *^\ r*^ 1 ^£^2* \ *° 1 ~~1 , ' 9r~J J———" *ocA "^w A!/ •< %3K9K 1 \ "o 1 JL o f IV p I S-JJ 1 ^____ o 1 1 i I "° i 1 • . L——- - 1 • • o «. •i i • 1 - ^ ID i. . ir i 1 J"s - i. i . i ^5^—^- •* * « •o M 2 THE SOYBEAN plant: history/characteristics seed : structure, composition, proteins The early history of the soybean,like that of most important food crops, is lost in obscurity. According to MORSE (1950) ancient Chinese literature reveals that the soybean was extensively culti­ vated and highly valued as a food centuries before written records were kept. Many of the early writings record the advice of agricul­ tural experts on soil preferences, proper time of planting, methods and rates of planting, the best varieties to plant under different conditions and for different uses, time to harvest, methods of sto­ rage, and utilization of the many varieties for different purposes. Some of this expert advice goes back as far as 2207 B.C., indicating that the soybean was perhaps one of the oldest crops grown by man. The yellow- and green-seeded varieties were regarded as valuable foods for the use of mankind. In addition to the uses of the beansi many references were made during the centuries to the processes of preparing and the value of soybean sauce, soybean milk, soybean curd.soybean paste, and soybean sprouts, not only for food but also to treat various diseases and body ailments.

The soybean has been referred to as Glycine hispida (Moench) Maxim. The late Dr.Piper came to the conclusion that it must be named Soja max (L.) Piper. Other botanists still call it Glycine max (L.)Merrill„Th e name soya comes from the Chinese word "chiang- yiu" which means soy sauce, pronounced in Japanese as »show-yu" and contracted into "so-ya".

The soybean has its origin in eastern Asia and is believed tohav e been derived from Glycine ussuriensis, which grows wild throughout much of eastern Asia. Aplan t with characters between the wild and the cultivated species has been described by a Russian botanist as Glycine gracilis Skvortzov. There is little doubt as to the origin of the cultivated plant, since apparently none of the other wild plants found until now can possibly be its ancestor.

The cultivated soybean plant is an erect,branching summer legume, while the wild plant is a slender, twining variety. A complete se­ ries of forms connect both types of varieties. The varieties, which are very numerous,range in maturity from very early (about 75 days) to very late (200 days or more). With few exceptions, earliness is correlated with size, the tallest varieties being latest. Nearly all varieties are pubescent, that is the stems, leaves and pods are coveredwit h tawny (brown)o r grayhair .Severa l yellow-seeded va­ rietiesar e entirely glabrous,tha t islackin g inpubescence . Theplan t hastrifoliat e leaves, eachcompoun d leafbein gmad eu p of three leaflets. Occasionally varieties occurwit h fouro r five leaflets, especiallysom e ofth evegetabl e typesan d earlySiberia n types. Ingenera l the leaflet isovate ,lanceolate , but insom eva ­ rieties iti snarrowl y lanceolateo ralmos tlinear . Thesmall , inconspicuous flowers borne inaxillar y racemesa tth e nodesar e either white or purple.Th e flowers first appear at the baseo fth estem , thenprogressivel y toward the tip, theperio d of blooming not being overthre eweeks . The pods, onet o fourinche s inlength ,hav ecolour srangin gfro m straw-yellow over grayan d brownt oblack.Th ebean s inon epo dnum - be twot othree ;the yar eyellow , greeno f black incolour .

The soybean seed isa typica l legumeseed. Itsweigh tvarie s from about 2g pe r 100seed st oove r4 0g pe r 100seed s incertai nvege ­ table varieties (WILLIAM S.Markle y 1950). Mostvarietie s incom ­ monus eweig hbetwee n1 0an d 200g pe r 100seeds .See dsiz ei squit e dependent onenvironmenta l conditions but relative sizesar e quite constant. Soybeansvar y inshap e fromalmos t sphericalt oelongate d orflat . Themos textensiv edat ao nth emicroscopica lstructur eo fth esoy ­ beancom e fromK O ND O (1912). Hestudie d theanatomica l structure of anumbe r ofvarietie sfro mChin aan d Japan,an dmad e thefollo ­ wingobservations .

The seed coat consistso f four layers,a palisad e layer, theco ­ lumncells ,th espong y parenchymaan d thealeuron elayer . Theepidermi s ismad eu po fpalisad e cells,40-6 0u inlengt han d 10-20u inwidth.Th ecellwall sar ethic kan d the lumenbecome snar ­ rower inth edirectio n ofth e surface of thesee d coat.Th e walls themselves are always colourless. The colour of the seed coat is located inth e lumen. Thecolumncell sfor mth esecon d layer.Thesecell sar eshape d like hourglasses, eachen d beinghexagonal ,an dar ea littl e longertha n the palisadecells. Owing toth e form of thecell s there are large intercellular spaces. The cellsan d the intercellular spaces are empty. The hourglass cellsseparat e readily from adjoining layers and fromeac hothe ran db ythei rcharacteristi c form furnisha mean s of identifying soybeanmeal. Thecolumncell sar e 27-70u long and 16-27u wide . Belowth ecolumncell s isa layero fspong y parenchyma consisting of about7- 8row s ofempt ycell s ranging inwidt h from40-6 0mi - crons. In a dry condition this layer is very thin, because the cells are much compressed radially. The aleuronelayer consists of asingl e series of moderately thick- walled cells with dense protein (aleurone) contents. Hi a tangen­ tial section the aleurone cells are seen to be rectangular or polyV gonal. The other rows of cells are compressed into a colour~less, fiberlike layer.

The hilum. is 2.5-4.0 mmlon g and 1.0-1.5 mm wide. It consists of a double layer of palisade cells, a fibrovascular bundle, a layer of starshaped cells, spongy parenchyma and an aleurone layer. The palisade layer consists of two rows of cells, differing in Colour. The cells of the outer layer are 59-67 u long and 7-10 u wide, those of the inner layer are 39-59 u long and 9-12 u wide. So the outer layer is somewhat thicker than the inner. Another diffe­ rence is that outer layer is usually yellow in colour, while the inner layer is brown, green etc. Beldw the palisade layer lies a layer of starshaped cells with thick walls. The cells contain a redbrown protoplasm. In the middle of the hilum lies the fibrovascular bundle surroun­ ded by the starshaped cells. The bundle consists of spindle-form tracheids with reticulated walls. The spongy parenchyma and the aleurone layer have the same struc­ ture as already described for the seed coat. It is important to nof tice that in the spongy parenchyma there are two fibrovascular bundles. The columncells are not present in the hilum, but near the hilum these cells are extraordinary long. In the vicinity of the hilum the columncells are 150-175 u long and 27-31 u in width and thus they are 2.5-4 times the lenght of the palisade cells. The further from the hilum the shorter the columncells become andatthe"fiat side of the seed the columncells measure only 0.7-1.0 times the palisa­ de-cells.

The embryo consists of two seedleaves (cotyledons), the primary leaves (plumule) and the hypocotyle, whose root end (radicle) is at the micropyle. The seedr. leaves usually are flat and elliptical, the two primary leaves are ovoid and free of starch, while the root­ let is a little bit curved and perpendicular to the primary leaves. The cotyledons ordinarily are yellow-coloured. The inner and outer epidermis consists of small cells, filled with small grains of aleurone. Below the inner epidermis are 2-6 rows of palisadecells, making up about one third of the cotyledons. The following rows consist of still longer cells and the last one or two rows are made up of isodiametricalcell sdirectl ybelo w theoute r epidermis. The palisade cells are 82-105 u long and 11-24u inwidth , but inth e middle ofth etissu e thecell smeasur e 94-188u i nlengt han d 24-42 u inwidth . These cells are filled with large grains of aleurone and fat. The aleurone grains are elliptical orpolyhedra l in form andver y big (5-16 u).Th eChines e varietieshav e biggergrain s than the Japanese.Th ecell s containmuc h oil,whic h islocate d between thealeuron egrains .

Accordingt o literature soybeans arestarchfree .A sregard s this much disputed point KO NDO , however, has found that some of the examined varietiesd ocontai nstarch .Th eChines e varieties and one Japanese variety were free of starch. The starch grains are small and irregular inform . Theouter ,abou ton ethir d parto fth e coty­ ledonsar efre eo fstarc han d it isth e innerpar ttha tcontai nth e starchgrains .

As to the chemical composition, Morse has quoted figures from LE CHAR T IER (1903), being themos t extensive of itskind . For theweigh t ofth estems , leavesan dpod s the ratio 25.45 :40.1 8: 34.37wa s found while the dry plant proved to contain 11.96% of protein, ofwhic h 6.24% ispresen t in the pods. Thedifferen tpart so fth ebean shav e the following composition.

Proportions Dry Nitrogenous Carbohy­ Pat Ash ofth esee d matter substances drates

Entiresee d ioa.% 90.18 38.06 12.06 17.80 4.44 Cotyledons 90 89.43 41.33 14.60 20.75 4.38 Embryo 2 87.99 36.93 17.32 10.45 4.08 Seedcoa t 8 87.47 7.00 21.02 0.60 3.83

Pa From these figuresi tma y besee n thatmos to f theprotei n4& -con ­ centrated inth ecotyledons .

Asregard sth echemica lcompositio n ofth edr ymatte ro fth esoy ­ bean, includingsee d coat,CARTTE R andHOPPE R (1942) madeex ­ tensive studieso nth e average compositiono Ti ovarietie s

Pig2 A.Structur eo fth esoybee nsee d (.$««• \A\\ha^),

10 AVERAGE CHEMICAL COMPOSITION OP SOYBEANS OP TEN VARIETIES. ON A MOISTURE-FREE BASIS,GROWN IN FIVE LOCATIONS IN FIVE CROP YEARS.

Crude protein 42.78% Phosphorus 0.659

Crude oil 19.63 Potassium i#g7 Crude fibre 5,52 Calcium 0.275 Sugars (as sucrose) 7.97 Ash 4,99

Besides the wellknown fact that the soybean is an oilseed, the bean also proves to contain a very high percentage of crude protein, as may be judged from these figures.

On comparison of the composition of the soyproteins with that of animal proteins (both in the form of milk) a marked resemblance is evident. We cite the following figures from the F.A.O. Nutritional studies No.8 (Brock and Autret, 1952).

Cow's milk Soybean milk gpe r litre g per litre

Arginine 1.26 2.82 Histidine 0.91 l.oi Lysine 2.625 2.40 Tryptophane 0.525 0.56 Phenylalanine 1.785 1.95 Methionine 0.840 0.56 Threonine 1.505 1.57 Leucine 3.434 2.93 Iso-leucine 2.625 2.09 Valine 2.52 2.00 Proteins 35.00 35.00

The authors conclude that the amino-acid content of soymilk is comparable to that of cow's milk, except in methionine. The methio­ nine content of soymilk is, however, double that of the milk of the European mother and three times that of the milk of the African, mother.

11 —Cross-section of hilum of a yellow soybean from China. PI, outer palisade layer; P, inner palisade layer; S, asteroid parenchyma; Sp, spongy parenchyma; F", fibro-vascular bundle; F, fibro-vascular bundle of the testa; A, aleurone layer; Sz, hour-glass cells. (After Kondo.)

Embryo of a yel­ -Soybean. Cells of low soybean seed from Japan. the cotyledon filled with fat, pro­ I. Whole embryo from the ven­ tein and starch; P, protein; 8, tral side. II. Half of embryo starch. (After Kondo.) sejen from the inner side. III. The two leaves of the plumule. C, cotyledons; W, hypocotyl; B, leaves of plumule. (.After Kondo.)

Pig2 Structureo fth esoybea nseed .

12 3 MANUFACTURE OP SOYMILK

3.1 Introduction. The firststatement s inol dChines e literatureo nth eprocessin g of soybeansd ono tsee mt orelat e tosoymil ka ssuch ,bu tt o"to u fu"a produc t derived from soymilk.As ,however ,th efirs tneces ­ sary phase inmakin g toufu isth epreparatio n ofsoymil k inthi s chapterw edra w attention toth efirs t known reference tomakin g toufu. This first statement isalread y found inth eHa nDynast y

Taoistwork , HuaiNa nTsu r orwriting so fLi uAn ,a princ eo fHua i Nanwh odie d in12 2B.C.(MORCE , Markley, 1950). However,a sha sbee nmentione d before,th eplan tha sbee n known since about 2800B.C .an dbeside s ancient literature reveals that the soybeanwa scultivate d andhighl y valued asa foo d centuries before written records were kept.W e may therefore safelyassum e thatth eprocessin go fsoybean sdate sbac kt obefor eth eperio do f theHa nDynasty . Iti sver y likely thatChines ean dJapanes eworker s already made extensive studieso fth eprocessin go fsoybeans .Dat a inlanguage s accessiblet oEuropea nresearc hworker sdi dno tappea rearlie r than thebeginnin go fth e20t hcentur yhowever . Examplesar eth ewor ko f Inouye (1897),Katayam a (1906),L i(1911 )an dothers . The soybeanwa sfirs t brought toth eattentio n ofEuropean sb y the German botanist,Kaempfer ,wh ospen ttw oyear s (1691-92)i nJa ­ pan.Althoug hh ediscusse d indetai lvariou sfoo dproduct s prepared fromth esoybean , little interestwa stake ni nth ecro pan di twa s not untilth ebeginnin go fth etwentiet h century thatth eproduc ­ tiono fsoymil kan dderivative swa sstarte do na nindustria lscale . Thiswa sdon e about191 0near Parisan da tFrankfur t inGerman y (INST. INT.D'AG R IC. ,1936).Afte rtheSecon d World Wara soymilkpowderplan tbuil tb yMille r forth eChines e Governmentwa s completed (SMITH, 1949),whil ei nth eUnite d Statesanothe r plant was locatedi nOhi o (LAGER, 1945).Sincethe nsevera lothe r plants seemt ohav e beenbuil ti nAsiati c countries.

Soymi Ik mayb edescribe d asa stabl e dispersiono fwater-disper - sableconstituent so fth esoybean .Dependin g uponth era wmateria l andth emetho do fmanufactur eth edr ymatte rconten t isabou t10% . Thisdr ymatte rmainl yconsist so fproteins , fats,an dsugars . Inth efollowin gtabl esom e figureso nth ecompositio n ofsoymil k areshown .

13 Java Japan China India USA Human Cow Prinsen- Rata- Chlu De Markley milk milk Geerligs yama 1927 1948 1950 1896 1906

Proteins 3.13 % 3.02 % 5.72 % 4.2 % 2.79 % 1.25 % 3.50% Fats 1.89 2.13 2.44 3.4 2.22 2.50 4.00 Sugars - ' - 0.43 1.8 6.05 6.00 3.25 N-free extr. - 1.88 - - - " - Ash 0.51 0.41 0.73 0.75 0.94 0.25 0.75

Soymilk has a cream yellow colour and In a. raw'state"it smells of raw beans. This beany odour disappears in cooking and then a typi­ cal flavour of cooked soymilk develops which is much appreciated in countries where soymilk and derivatives form an integral part of the diet. According to LAGER (1945) the handicap of unpalatability has been conquered by physical rather than chemical methods,so that the milk is sweet and rather nutlike. Literature data concerning the pH of the milk are contradictory. According to BL0CH (1906) the milk is slightly alkaline, while PRINSEN GEERLIGS (1896) states that its reaction is mild acid. Since the milk can be prepared in many ways it is quite pos­ sible that the method of manufacture also influences the reaction of the milk. According to MATAG R I N(1939 )th e boilirig?point'is about 101°C; the milk rises in boiling, forms a film and may cause in­ crustation. The raw soymilk deteriorates rapidly. If raw soymilk is left at room temperature it separates into two parts. The lower part is whitish and resembles the curd of animal milk and the upper part has a transparent, greenish yellow colour (MONNIER, 1935).If the milk is boiled it can be preserved for a much longer time. It is only after 15 to 29 hours that symptoms of putrefaction appear. Concerning the tuberculose-bacteria, MATAGRIN (1939) states: "II garantit par rapport au produit animal unavantag e d'extreme im­ portance: 1»absence certaine de bacille de Koch, dont la presence assez frequente dans le lait de vache menage de contagion tubercu- leuse plus encore les animaux domestiques que les nourissons, puis- que la sterilisation ou la pasteurisation se generalisent aujour- d'hui dans 1'allaitement artificiel". On microscopic observation a large number of fat globules are vi­ sible in the milk. The smaller globules are in an animated Brownian movement. By far most of the fat globules are between 0.5 to 1 u in size; larger globules of from 10 to 12 u rarely occur. MATAGRIN (1939) states that no starch grains are present in the milk, while the milk does not react to colouring methods of iodine dissolved inKJ, methyleneblue or rutheniumred. Photo 1 is a microscopical photo of soymilk made by us.

14 3.2 Literature.

3.2A Methods of manufacture.

Datao nth emetho d ofmanufactur e ofsoymil ka sapplie d byth e populations ofsevera l countries inEas tan dSout h East Asiaar e very numerous.SMIT H andBECKE L (1946)hav ecompile dth eexis ­ ting literatureo nth epreparatio no fsoymilk .Th ereference s num­ ber 77,o fwhic h3 3ar epatents . To illustrate these original manufacturing methodsw ewil l cite someauthors . Accordingt oP RI N SEN-GE ER L I GS (1896 ) onJava ,th ebeans , after beingsoake d for3 hours ,ar egroun d betweentw ostones .Th e thick liquid isboile d afterwhic hth emas s isfiltere d througha cloth. BLOCH (1912)describe sho wi nIndo-Chin ath ebean sar egroun d after being soaked overnight.The grindingi sperforme db ytw omill ­ stonesan daddin ga constan t supplyo fwater . Thenth edispersio n is filtered througha cloth . LI YU-YING (1912)state s that inChin ath ebean sar esoake d for 18t o2 4hour safte rwhic h theyar egroun d withasuitabl e amount ofwater .Th egrindin g isperforme d bytw omillstone so ffro m0. 5 to 1mete ri ndiameter ,Th ethic k liquid isreceive d ina tan kan d then filtered througha line ncloth . Inth erepor to fSMIT H (1948),wh ovisite d China,Korea ,an d Japant ostud yth eprocessin go fth esoybea n into foodproducts ,n o principal changesar efoun di nth emethod so fmanufacture .H e states that inChin a thebean sar ecarefull y washed andsoake d inwate r for6 t o7 hour si nsumme ran d2 4hour si nwinter.The nth ebean sar e ground ina ston e millan dth egroun d mass isfiltere d througha filter cloth. Therati o between wateran dbean s iste nt oone . INOUYE (1884)state s that inJapa n thebean sar esoake dfo r about twelve hoursan dthe ngroun d betweenmil lstone sunti la uni ­ form mass Is obtained. Thismas s isboile d forhal fa nhou r with three timesa smuc hwater , after which iti sfiltered .

Inth efollowin g shortdiscussio nw ewil l considerth eresearc h work doneo nth emanufactur e ofsoymilk .

Thewor ko fCHI U (1927,Cornell )i smor econcerne d withth ephy ­ sicalan dchemica lpropertie so fsoymil k thanwit hit smanufacture . Thephysica lpar to fhi swor kdeal swit hspecifi c gravity, freezing point,electrica lconductivity ,viscosity , refractive indexo fse ­ rum, dialysis,pH ,an dkeepin g qualities* Thechemica l part deals

15 with the generalcomposition , effecto f soaking composition ofth e fat, kind ofproteins ,compositio n of the serum, and actiono ffer ­ ments, Chiuuse dtw ovarietie s of soybeans,Manch u and Black Eyebrow,and followed the Chinese method of preparing soymilk. The soakingpe ­ riodswer e2 ,4 ,6 , and 8hour srespectively . Twohundre d and fifty gramso fbean swer e soaked in 1.25 1o fwater . Then the beans were ground ina ston e mill, 1 1o fwate rwa s added, and the beans were ground againunti l theywer ever y fine.Themil kwas-filtere d through bagsmad e ofcloth .Afte r separating the solid particles from the liquid themil k was received incontainers . Thespecifi cgravit yprove d tovar y with the timeso f soaking: as thesoakin gtim e increased the specific gravity became higher. If thepercentag e ofdr ymatte rvarie d from 10.0t o 12.0%th e specific gravity ranged fromabou t 1.0295t o 1.0315. Therelativ e viscosity of themil kwa s calculated bydividin g the time of flow of a definite volume of milk through a capillary by that ofwate r atth e sametemperatur e (25°C), Viscosity ranged from 2.75 to3.6 5 for 11.0t o 12.0%o f dry matter. The variety Black Eyebrow consistently gave higher values than Manchu. Thep Hdetermine d at26° Cwa s6.4 3 for Manchuafte r 6an d 8hour s soaking.Fo r Black Eyebrow these values are 6.52 and 6.53 after 6 and 8hour ssoaking .H ecam e toth e conclusion that soymilk isonl y slightlymor eaci d thanco w milk. Soymilk colours lithmus red. Thecompositio no f themil kwil lvar ywit h the varietieso f beans used,th e methodso f manufacture, and the times of soaking. Accor­ ding to thisworke r there isa limit to the extraction, as itwa s not possible to extract more than 20%o f the solids of the beans withwater . Inmakin gsoymil k the beanshav e tob e soaked inwate r tofacili ­ tategrinding . Itha sbee n shownthat'long 'soaking removes the-un ­ pleasantodou ro f thebean s and increases thepercentag e of theto ­ talsolid s inth emilk . Ifth e water inwhic h the beans are soaked isuse d tomak eth emilk , the substances extracted by thewate r are retained inth emilk .Accordin g toChi u it isno t advisable tosoa k thebean s formor e than 8hours ,a sotherwis e the losso ftota lso ­ lidsextracte d byth e soaking waterwil lb e toogreat .Th e fatcon ­ tent increaseswit h longerperiod s of soaking. After 2and 8hour s the fat content of the milk fromBlac k Eyebrow is 2.31 and 2.56%, and from Manchu2.2 3 and2.44% .Fo r Black Eyebrow the proteincon ­ tent is5.7 3 and 5.90%an d 5.59 and 5.72%fo r Manchu. Manytrial swer emad e fordeterminin g fat insoymil k byth e Bab- cock method but they were not succesful.The,failur e seemed to be due to the nature of the solids not fat, and their reaction with

16 the sulphuric acid-Solid substances were formed which prevented the fat from going into the neck of the testbottle so that it was not possible to take the reading. Experiments have shown that the fat in the milk could be determined without difficulty by the Roese- Gottlieb method..

In Russia HORO W I TZ - W LA SS OWA , OBERHARD, andGUTER- MANN (1931) of the Moscow Central Biochemical Research Institute of the Pood Industry have worked out the following procedure. The ground soybeans are soaked for eight hours and treated with steam of 100°C for desodorisation.This is followed by the extraction with water and the separation of the residue (still containing 17.85- 19,25% fat, 0,53 -0,58% protein and 0,14% carbohydrates), The emul­ sion is cooked for five minutes and then 0,25% NaCl, 0,5% glucose, 2% soyoil, and 0,02% vanille or 0,2% almond essence is added. After mixing the liquid is sterilized. After addition of 10% sugar concentrated milk was obtained by evaporation at one fourth of the original volume. The composition was 11,76% protein, 7,64% fat, 42,12% sugar, 2,32% ash and 37.16% water, Prom the same institute BOG ATSK I I, STQROZHUK, and MU- ROMTSEV (1933) have done work onpreparatio n and grinding of soy­ beans, emulsifying, boiling, and desodorisation by blowing with hot air, The sterilisation was carried out for 15-20 minutes at 115- 120°C „ In Leningrad V OSK R ES ENS K 11 and D OBRUI N IN A (1935) made soymilk by soaking the beans for two hours and thoroughly emulsi­ fying; according to these workers longer soaking gives greater fat extraction but a less stable emulsion. In commercial production the material may be hulled, whole beans, thoroughly ground and homogen­ ized in water for at least one hour, then diluted with water and emulsified for at least 30 minutes. Pasteurization is at 80°C for 15 minutes. At the Institut des Recherches agronomiques Hanoi,- CASTAGNOL (1934) has developed a method of manufacture for preparing-soymilk on a large scale.He succeeded in getting a product of constant com­ position if the following operations were applied. 1. Soaking the beans 2. Grinding with a millstone arid using water 2% times the weight of the beans. . 3„ Dilution of the mass with water 2&time s the weight of the beans : .'. •; ' 4. Filtration of the thick liquid by means of a basket centrifuge 5. Centrifuging the milk.

17 Themil kprepare d inthi swa y isa whit e liquid resembling cowmil k inappearanc ean d containingabou t5.5-6 %dr ymatter . Castagnolha s successivelystudie d eacho fthes eoperations . Soaking. Itwa s found'thatwhe n the beans were soaked inwate r of 15°C20 0g o fbean sabsorbe d 215an d23 9g o fwate rafte r 12an d 48hour srespectively .Afte r6 0hour sth ewate rabsorptio nwa s23 9g . No significant differencewa s found when the soakingwa scarrie d' out inwate ro f30° Cdurin g 24 hours. Grinding. Ona n industrialscal e itwil lb eadvantageou s tous e a twinston e millwhic h permits a regular supply ofwate r during theoperation .Th eamoun t ofwate rwhic h givesth ebes t resultsma y dependo nth etyp eo fmill . Dilution. As itwa sno tpossibl et ofilte r the thick liquidob ­ tained aftergrinding ,anothe r50 0g o fwate rwa sadded . Basket centrifuge. Accordingt oCastagno l theus eo f filterpres ­ sesi sno tadvisabl ea sth ecapacit yo fth e filterpresse sdimini ­ shesver ysoo ndu et oth eaccumulatio no fresidua lmass . Thebaske t centrifuge ismuc hmor eefficien t as itca nb eeasil y cleaned.Be ­ sidesworkin gcondition sar emuc hmor ehygieni c thanwhe nusin gth e filterpress and large quantitiesca n be handled with an apparatus of comparatively smallcapacity , Centrifuging. Thepurpos e ofcentrifugin g ist odiscar d the cel­ luloseresidu ewhic hamount st oabou t 0.12%i nth e liquid. As regards the results obtained from 200 g of soybeans with a crudeprotein,conten to f33.80%;th eyiel d is95 4g 6 fmilk'o f2.91 % crudeprotein .Thi smean stha tth eoutpu to fth e most importantnu ­ tritivesubstance ,th ecrud eprotein ,i sonl y44% .Th e other56 %i s founda sa los si nth e industrialwast eproduct .S oi nthi srespec t theproces sneed simprovement .

Accordingt oCastagno lther ear etw o importantpoint stha t from an industrialpoin t ofvie w need proper consideration. 1.Th efermentatio no fth emateria l 2.Th e improvemento fth eyiel d ofcrud eprotein . Fermentation isa n important factor. Inorde rt oavoi d thisth e soakingperio dmus t beshor tan d the temperature of thewate rlow . Duet ofermentatio n theacidit y increasescausin g acoagulatio no f theproteins . Promth efermentin gliqui d a lacticaci dbacteri acoul d beisola ­ ted.A counto fth ebacteri ai nasampl eo ffres hlm'il kshowe dtha tther e are15.00 0t o20.00 0bacteri ape rml .Th e freshmil k hasa n initial acidity of 12.0m lNaO H 0.1 n,,an d after 24hour s theacidit y in­ creasedt o44. 5ml .A tth een do f3 6hour sa stron godou rdevelope d duet oa decompositio no f theproteins ;th e acidity diminished to

18 8.6 ml.Th eprotei nwa sseparate d ina curdlik e forman dsubmerge d ina serum . Itwa sfoun d thatth ecur dcontaine dth emajo rpar to f the protein,onl yabou ton etent hbein g lefti nth eseru m inth e formo famino-acids . Castagnolconclude stha tther ear etw okind so ffermentation .Du e toth efirs t fermentationth emil k becomessour ,bu tno tenoug ht o preventth eproteolyti c fermentationwhic hultimatel y prevails.A t theen dther ei sa nincreas e inamin oacid san da lessenin go fth e aciditybecaus eo fit sneutralisatio nb yth edecompositio nproduct s ofth eprotei n (ammoniai nparticular) . Theaci d fermentationha s precipitated thegreate rpar to fth eproteins ,whil eth enitroge n remainingi nth ewhe yoriginate s fromdecompositio nproduct so fth e proteines (pepton,aminoacids , etc,). The improvement of the protein output. Asha sbee nshow nonl y about44 %wa sextracte d withth eproces suse db yCastagnol .A bet ­ terkin do fextractio nwoul dreduc eth ecos tpric eo fth emilk.Cas ­ tagnol investigated theinfluenc eo fthre e factors which canb e controlled underactua l factoryconditions . 1.Th eacidit yo fth ewate r 2.Th einfluenc eo fNa-ion si ndifferen tform s 3.Th einfluenc eo fth esoakin gperiod . He concludes thatth eacidit yo fth ewate rha slittl e influence, whileth eNa-ion shav emuc hinfluenc eo nth eoutpu to fprotein .Ho ­ wever, objectionsmus tb emad ea st oth eaccurac yo fhi swork . He himselfstates :"Pa rsuit ed umaterie l touta .fai tinsuffisan tpou r obtenirun eexperimentatio n digned evaleur , leschiffre s obtenus ne peuvent fournir qu'une indicatione ti lserai t indispensable d'opererave cu nmaterie l travaillantd efaco n reguliere etse - rieuse pour obtenirde srenseignement splu s precis". The influenceo fth esoakin gperiod .I norde rt ofin dou twhethe r transformationstaking.plac e inth ebean sa sa resul to fsoakin g could improveth eextractio nsevera lbatche so f10 0g o fbean swer e soakedfo r12 ,24 ,36 ,an d4 8hours . Withth epurpos eo finhibitin g fermentation,whic hmigh ttak eplac edurin gsoaking ,th etemperatu ­ reo fth ewate rwa smaintaine da t15° C Theoutpu to fprotei nwa s 34.5, 26.4,26.7 ,an d9. 9respectively .Soakin gfo r4 8hour sgav ea milkwit hcasei nwhic hwa scentrifuge d outo fth e liquidfo ra larg e part. Thisma yexplai nth elo wconten to fprotei n inthi smilk . The amounto faminoacid sa sdetermine d bytitratio n with NaOH N/10i s28 ,16 ,26 ,an d2 5m lrespectively ;S oth eamin oacid spas s throughaminimu ma t2 4hour ssoakin gan dafterward sincrease sagain . Notwithstanding all precautions takenth ep Hdiminishe swhe nth e durationo fsoakin gwa slengthened^Therefor eCastagno lconclude stha t forhi sworkin gcondition stwelv ehour ssee mt ogiv eth ebes tresults .

19 InIndi asoymil kwa sno tmuc hknow nunti lth e greatBenga lfamin e of 1943whe nsoymil kwa sproduce db yGUH A andD E from the Cal­ cuttaUniversit yfo r feedinga fe whundred so fstarvin g infantsan d toddlers. Amajorit yo fth esubject ssurvive d thecrisi s ona die t mainlyconsistin go fsoymilk . In194 4wor ko nsoymil kstarte do na systemati c basis at theIn ­ dianInstitut eo fScience ,Bangalore .Th emetho d ofmanufactur ewa s as follows.Thebean swer esoake dovernigh t inwate ran dthe nsprea d out intray st o germinate for two days to enhance the nutritive valueo fth eprotei na swell'a sth evitami ncontent.Afte rgerminatio n theski nwa sremoved .Th ekerne lwa sthe nheate d forabou t tenmi ­ nutesa t70° Cwit ha dilut esolutio no f bakingsod a (0.08%)o rgly ­ cerol (1.08%)whic hextracte d thebul ko f thecolourin gmatte ran d thebitte rprinciple .Th eextracte dwate rwa sremove d and theker ­ nelwa swashe dwit hwate ran d groundt o afin epast ewit hth e addi­ tiono fcalciu mhydroge nphosphate .Th epast ewa sthe ntreate dwit h about three timesth e volume ofwate r and heated to the boiling point for1 5minutes .Th emil kwa sfiltere dthroug h thesieve .Tw o percentcan esuga ran da smal lamoun to fsal twer eadde dt o improve, thetaste .A sa majo rpar to fth emil k isuse d assou r curd itwa s found advantageous toad d invertsuga r byhydrolyzin g cane sugar withdilut eacid .Abou t25 %o fth eorigina lbea n iswaste da sresi ­ duetha tca nb euse da scattl efee do rwhe ndrie dca nb emixe dwit h wheatflou rfo rpreparatio no fbrea dan dbiscuits . Thecompositio no fth emil kwa sa sfollows . Totalsolid s 10.15 % VitaminA 750 I.U./litre Protein 4.2 % Thiamine 0.82 mg /litre Pat 3.4.;. % Riboflavine 1.1 mg /litre Carbohydrates 1.8 % Nicotinic acid 2.49 mg /litre Salt 0.75 % Ascorbicaci d 21.6 mg /litre Asma y besee n frontth e foregoing examples the methods ofmanu ­ facture incountrie swher esoymil k ismad ediffe rregionnally .Ther e are howeverman y operations incommon,an d these operations mayb e designated asth epretreatmen to f the material, the extraction of theproteins ,th eseparatio n andoperation s for the improvement of thequalit yo fth emilk . Ast oth eresearchwor k done,th ewor k ofC H IU doesno tgiv ein ­ formation onth e process of manufacture, as itmainl y deals with thephysica l and chemical propertieso fth emilk .CASTAGN0L ' s work,althoug h dealingwit h the process ofmanufactur edoe sno tre ­ presentareliabl estud y ofho wth eproces sca nb e improved.Thewor k done inRussi a isdifficul t to evaluate, due to lack of informa­ tion,whil e lastlyth ewor ko fD E ismor eo nth enutritiona lside .

20 Ingeneral ,i tca nb esai dtha tespeciall ydat ao nth eoutpu to f themanufactur eo fsoymil kar ever yscare .

3.2,2 Dispersion of vegetable proteins from seeds.

Inadditio nt oth eliteratur e referred to,ther eexist sextensiv e literatureo nth eisolatio no fprotein s from seeds ingeneral .I n thecompilatio n ofth eliteratur eo nisolatio nan dutilizatio no f vegetable proteins (SMITH, 1954)th enumbe ro freference si s 161. This literature mainly deals withth eisolatio no fvegetabl e pro­ teinsfo rindustria lpurpose san dha sbee npartl yrea d ina nattemp t to find useful indicationsfo rth emanufactur eo fsoymilk . Itappear s thatespeciall y toth efollowin g five factorsatten ­ tionwa spaid . 1.particl e sizean dothe rphysica l factors 2.H-ion s 3.salt s 4.denaturatio n 5.germinatio n Toou rwor kth efirs tan dth esecon d factorar eimportant .Th e literature onth erol eo fsalt s forextractio n ofvegetabl e pro­ teins isvoluminous ,bu tfo rou rpurpos e itha sles svalu ea sth e useo fsalt si nth emanufactur e offoodstuff s ingenera l willno t be considered,Denaturatio no fth eprotein sha sa nimportan t influ­ enceo nth e choiceo fth estartin g material. Lastly, germination mayb eo finfluenc e onth enutritiv e value ofth emilk .

3,2.2.1 Particle size and other physical extraction factors.

Thepretreatmen tdetermine sth eparticl esiz ewhic hi ntur nha sa n important influence uponth eamoun to fdispersabl e protein. Obser­ vationso fthi skin d weremad eb yBISHO P (1929)wh ostate s that barleymea l ground througha i-m msiev eyield s 25.0%o fth eorigi ­ nalnitroge no nextraction . Ifth esiev eopenin gwa sreduce dt o0. 5 mmth eoutpu t amountedt o31.2% , whilea coffe e mill increasedth e yieldagai nt o36.2% .STAKE R andGORTNE R (1931)als orealize d the importanceo ffin e grindingan dth eimportanc e ofa standard particlesiz efo rcomparativ e studies.The y ground themea lo fva ­ riousseed st opas sthroug ha 10 &mesh-siev e beforeextraction . NAGEL, BECKEL,an dMILN ER (1938)considere dsom e physical factors which influenced theamoun to fnitrogenou s constituents dispersed inwate r from fat-freesoybeanmeal .Th efactor sinvesti ­ gatedwer eth esiz eo fth eparticles,th esolvent-mea lratio ,tempe ­ rature,time,an dseparatio no fth emea lresidu e fromth esuspensoid .

21 Thesoybean suse dwer ea n Illinivariety.The ywer ecracked , flak­ ed,an d thenextracte d withpetroleu m ether,Th efat-fre e mealwa s airdrie d atroo m temperature, and ground. Inorde r toavoi d the possibility ofhea t denaturation the temperature ofth emea lwa s alwayskep t below40°C , Itwa sobserve d that thehull s were much hardert ogrin d andtha t particles retained onth ecoarse r screens largely consisted ofhul l fragments. The extractionwa scarrie dou ta sfollows. Aweighe d amountusual ­ ly 2.5 go fth e fatrfree meal wasshake nwit h distilled water at roomtemperatur e for3 0minute so na mechanica l shaker.Th emixtur e wasthe ncentrifuge d for6 minute sa tabou t300 0r.p.m .an dth eli ­ quidwa scarefull y decanted or filtered intoa Kjeldahl flaskan d analyzed fornitrogen . Theresidu e wasagai n shaken with another portiono fwate rfo ra nequa l lengtho ftim e andth eliqui danaly ­ zeda sbefore .Thi s processwa srepeate d asman y timesa swa scon ­ sidered necessary (usually three). Grinding todifferen t particle sizeswa sobtaine d bymean s ofa Wiley millwit hsieve so f2 ,1 ,an d 0.5-mm openings. Inadditio n two other kindso fmea l were obtained bydr y grinding the flakes witha porcelai n ballmil l with flint pebbles until allth emate ­ rialpasse dsieve so f10 0an d20 0mesh .Als oi nth ebal l mill 10g o f meal fromthroug h 100mes h sieveswa sgroun d with40 0m lo fwate r for4 &hours .Th eothe rmeal s (2.5g )wer eextracte d with 100m lo f water. Itrequire d about 24hour st ogrin d asampl e until itwoul d allpas sthroug hth e10 0mes hscree nan d4 8hour s toreduc et oth e20 0 meshsize ,O ntripl e extraction thesolubilit y increasedi nthi sor ­ derwit ha decreas ei nparticl esiz ea swoul d beexpected . Theamoun t ofnitrogenou smateria l dispersed increased from 80.9t o92.6 %ex ­ cept thatth e200-mes hmea l isslightl y lesssolubl e thanth e100 - mesh meal.Th erathe r unexpected result with the200-mes h mealca n be explained byth efac t that thelon g pounding inth ebal l mill (48hours) ,necessar yt oreduc e themea l to200-mesh , caused ade ­ naturationo fth eprotein . Onth ebasi so fth egreates t amount ofnitroge n extracted with a minimumo flabour ,th emetho d ofwe tgrindin g ina bal lmil l isth e mostefficient .Althoug h thisprocedur edoe sno tsee m tohav e been employed inanalytica l studieso nth eextractio no fsee dproteins , it appears tooffe r distinct advantages for this purpose. A very thorough grinding isobtaine d and theliqui d prevents localover ­ heatingan dmechanica l shock, botho fwhic h apparently cause dena­ turation. Ofth emea lparticl esize sstudie d the100-mes h mealwa sth emos t satisfactory. Itno tonl yyielde da large amount ofprotei n indis ­ persionbu ti twa sals omuc h easier tothro wdow n incentrifuging ,

22 thuspermittin ga clea rseparatio nb ydecantation.Thi s sizeo fmea l wasuse di nsubsequen twork . If2. 5g o fmea li sextracte db yincreasin gamount so fwate r there isnothin gcritica labou tth emeal-wate rrati oa st oth esu mo fth e successive extractions.However ,fo rth efirs t extractions the ex­ tracted nitrogen ranges from59. 4t o81.0 %i fth eamoun to fwate r is increased from2 5m lt o10 0ml .I ti sconclude d that awater - mealrati oo f10 0t o2. 5b yweigh ti ssatisfactor yfo rthi sextrac ­ tion. Inth etemperatur e rangeo f1.5 °t o45.0° Cth edisperse d amount ofnitrogenou sconstituent sincrease sb y0.25 %pe rdegre ecentigrade . Thetim eo fextraction , liketh emeal-wate rratio ,i sno tat"al l critical. Theresult s indicate that 1minute' s shaking foreac h of three extractionswil l dispersea smuc hprotei na s3 hour sfo r eachextraction .However ,th efirs textraction s range from71. 9t o 81.1%i fth eextractio ntim e isincrease d from1 t o3 0minutes . Iti sclea r thatal lth efactor s considered exert theirow nin ­ fluence uponth eextraction . This influenceapplie s eithert oth e sumo fthre e extractionso rt oth efirs t extraction.

3.2.2.2 H-ions concentration.

SMITH andCIRCL E (1938)investigate d theeffec t ofH-ion s concentration ofhydrochlori c acid, sulphuric acid, oxalic acid, sodium hydroxyde,an dcalciu m hydroxydeo nth edispersio no fpro ­ teins fromth esoybea n (variety Illini). Extracted flakeswer e ground ina Wiley mill through a0. 5m m screenan d2. 5g o fthi smea lan d10 0m lo fth edispersin g solution wereplace di na 25 0m lbottl ean dshake nmechanicall yfo r3 0minu ­ tes.Th edispersion swer ecentrifuge dfo r6 minute si na centrifuge developinga maximu m relative centrifugal forcea tth ebottl eti p of197 5time sgravity .Th etemperatur eo fth eextractin g solutions varied from24 °t o26°C .Th enonprotei nnitroge nwa sno tdetermine d but reportedb yanothe rworke rt ob e5.55 %o fth etota l nitrogen. Analytical dataar ereporte d without correction formoisture . Ingenera li tca nb esai dtha tth edispersio ncurv eobtaine dcon ­ sistso fthre eparts :a maximu ma tlo wpH ,a minimu ma ta highe rp H (3.5-5.0),an da slightl yslopin gpar tabov ep H6.5 . Withhydrochlori caci da maximu mdispersio no f83 %i sreached,an d at pH4.1-4. 2th eminimu m dispersion isabou t 8%.A tp H0. 5th e amountdisperse dwit h trichloroaceti caci di sver y low,5.5% ,an d onlyi nver yconcentrate d aciddoe si tris eagain . Theaci dan dbas efor mon econtinuou scurv eo nwhic hth eamoun to f nitrogenousmatte rdisperse d bywate r isa singl epoint .Th emaxi -

23 mumamoun to fnitroge ndisperse db yaci d isslightl y lesstha ntha t dispersedb ywater ,whic hdisperse sabou t 84pe rcent . The most efficient dispersion appearst ob eproduced bysodiu m hydroxide.Thedispersio ncurv efo rthi sbas ei sslightl yabov etha t forcalciu mhydroxid ewhic ha thighe rconcentration spoin tdownwar d quite sharply; thisbehaviou r indicates that larger quantitieso f limemigh t markedly decrease theamoun t ofnitroge n dispersed. Similar studieswer e undertaken with cottonseed (OLCOTT and FONTAINE,1939) . peanut (FONTAINE andBURNETT , 1944),an d sunflower (SMITH andJOHNSON , 1948).Th ecurv efo rpeanut sre ­ semblesver ymuc h thatfo rth esoybean ,whil eth edispersio n curve ofcottonsee d ischaracterize db ya ver ybroa dminimu m (pH3.0=7.0 ) precededb ya maximu man dfollowe db ya nalmos t horizontalpart.Th e curvefo rpeanut sha sa minimu ma tp H3 , followed bya slopin gpar t andthe na maximu ma tp H11.1 . .Itappear sfro mthes egource stha tth ep Hha sa mos t importantin ­ fluenceo nth eextractio no fsoybea nprotein sa swel la ssom eothe r vegetable proteins.Thereforei twil lb enecessar yt ohav e thisfac ­ torwel lunde rcontro lwhe nmanufacturin g soymilk.

3.2,2.3 Salts.

Severalworker sstudie dth erol eo fsalt si nconnectio nwit hth e isolationo fprotein s from seed.W ementio nth ewor k of STAKER andGORTNE R (1931)wh ostudie dth edispersio n effecto faqueou s solutionso finorgani c saltso nprotein s from2 8differen t seedsan d grains.Als oSMITH , CIRCLE andBROTHE R (1944)investigate d theeffec to f1 2neutra lsalt so nth epeptizatio no fnitrogenou scon ­ stituentso fsoybea nmeal ,whil eFONTAIN E andBURNET T (1944) studiedth esal t influenceupo nth epeptizatio no fpeanu tproteins . Ingenera li tma yb esai dtha ta definit ean dsystemati c effecto f thesalt swa sfound .Stake ran dGortne rprove d thatth epeptizatio n behaviouro fvariou s groupso f related seeds appearst ob erathe r distinct andcharacteristi cfo reac hgroup ,

3.2.2.U Denaturation.

Theoi lindustr y produces large amountso flow-fat , proteinrich waste productswhic hmigh t become available forth eextractio no f proteins.Insuc ha cas ei tshoul db etake n intoaccoun t thatth e va­ rioushea t treatmentst owhic hth emateria l issubjecte d mayhav e causeda denaturatio no fth ematerial . Iti sclea r thatsuc ha de ­ naturation,measure d interm so fwater-solubl e proteins,wil l effect theyiel do fprotein sunfavourably .

24 Asha sbee npointe d outb yBECKEL , BULL, andHOPPE R (1942) therear eman ystage si nth eproces swher edenaturatio nmay ,occur : a. Duringstorag ei fth ebean scontai nmor etha n13 %wate ro rdurin g drying toreduc e thewate rconten t below 13%.b .Durin gheatin gt o increase theeas eo fflaking , c.Durin gth eextractio ni nconnecti ­ on with the extraction temperature,d .Durin g desolventizing by meanso f steamdryers ,e .Durin g heating before the beans are fed to theexpelle rwher e theheatin g isconsiderable . Theseworker shav e studied the denaturationo fdefatte d soymeal heated from60 °t o 127°Ca trelativ e humiditiesfro m 0t o 100%fo r halfa nhou r to twoan d ahal f hours.Abov e 80°Ca t 100%relativ e humidityth e change inrat e ofdenaturatio n israpi du p to 100°C, butafterward sth erat eseem st oapproac ha limit.Lowerin gth ere ­ lativehumiditie scause sth edenaturatio nt odecrease . Whilethi swor kwa sdon eunde rlaborator ycondition sBELTE R and SMITH (1952)studie d thedenaturatio n of flakes obtained from fivesoybea n oilplants.The ymentio nth e following unit operations ina solven tplan toperation:weighin gth ebeans ;cracking ,cleaning , and dehulling; conditioning the grits (heating to predetermined temperature and adjustment ofth e moisture content); flaking;ex ­ tracting the oil;desolventizing ; deodorizing; toasting or other heat treatment ofth e residual flakes. These manufacturing steps aredivide d intothre egroups ;a .preparatio no fth ebean s forex ­ traction,whic h involvesth e first fouroperation s b.extractiono f the oil from theprepare d flakes and c, recovery of the residual meal ina usabl e formwhic hrequire sth ethre eremaining'steps . The dispersibility of the protein from the original beans may vary considerable,a factorwhic hmus tb e recognized inth e inter­ pretation of the results of an investigation of this nature. The nitrogendispersio nresult sar ereasonabl yconsisten t fromon eplan t to anotheran d showtha tonl y amino rportio no fth etota l denatu­ ration effect inprocessin g can be attributed't o the preparation steps. Infac tmor evariatio n iseviden t betweenth eorigina lbean s thanbetwee nth evariou spreliminar y operations.Experienc e hasin ­ dicated that,unde rcondition snormall yexistin g insolven textrac ­ tionplants ,th eextractio nwit hhydrocarbo nmixture ssuc ha sthos e used commercially doesno tappreciabl y denature theprotein . Iti s evident fromthi sinvestigatio ntha tth emajo rportio no fth edena ­ turationoccur s inth efina lmea ltreatin gsteps . Afterollextractio nth edispersibilit yrange sfro m7 8t o81% ,whil e after deodorizing thisma yb ea s low as43% .Toastin gma y cause a further reduction inth edispersibilit y toabou t7% . Afterth edesolventize rdischarg eth edispersibilit y isstil lve ­ ry high* Theauthor s state that'i fthi s type of.mea l is"required

25 for industrial use,th esnai lamoun t ofresidua l solvent canb ere ­ moved by proper useo fstrippin g gaseswithou t further significant denaturation. Thepeptizabilit y ofth e totalnitroge n contained insevera lcom ­ mercial hydraulic-press peanut meals was determined by BURNETT and FONTAINE (1944). They investigated the effect of pH upon theamoun t ofnitroge n peptized incompariso n with similar data of solvent-extracted meal. The meals prepared under relatively mild, controlled conditions aresimila r to those obtained forsolvent-ex ­ tracted meals. However, slightly smaller amounts of nitrogen are peptized inth ep Hrang eo fminimu m peptization and inth e alkaline range than isth ecas ewit h solvent-extracted meal. Incontras t to this, the representative meal ofunknow n history exhibits a signi­ ficantly lowerdegre eo fpeptizatio n overth ep Hrang e4 to 10tha n doessolvent-extracte d meal.Slightly morenitroge n seemst o bepep ­ tized onth eaci d side of theminimu m peptization point inhydrau ­ lic-pressed mealstha n insolvent-extracte d meals. They conclude that hydraulic-pressmeals , because of variations inth emethod s ofprocessing , varywit h respect to thepeptizabili ­ ty of their nitrogenconstituents; some are nearly equivalent to solvent extracted mealbu t others are definitely inferior inthi s respect. Iti sconsidere d possible,b y theus e ofproperl ycontrol ­ led hydraulic-pressing procedures, to remove oil effectively from the peanutunde r conditions which result ina minimu m ofchang e in the peptization properties of thenitrogenou s constituents of the resulting meal. The influence of processing factors on the denaturation of pro­ teins from peanut meals was also studied by FONTAINE, SA­ MUELS, and IRVING (1944). Theseworker ssubjectedfullfa tflake s and defatted peanut meal to temperatures from 80° to 118°Ca t 100 and 0%relatiev ehumiditie sfo rhal fa nhou rt otw oan d ahal f hours. The peptizability of the nitrogenous constituents of theheat - treated solvent-extracted mealdoe sno tdiffe r materially from that ofheat-treate d fullfat flakes. It issignifican t that heating for as longa s2. 5 hoursa t 100°Can d 118°Can d 0%R.H. , ora t 80°Can d 100%R.H. , resultsi nonl ya sligh t change inth epeptizability . At highertemperatures ,howeve r (118°Can d 100%R.H.) , the peptization values indicate that a complete denaturation of peanut protein is attained rapidly. Incontras t tothis , soybean protein showsconsi ­ derable denaturationafte r 2ft-hoursheatin g atl20°C and 0%R.H.an d alsoafte r2Mrhour sheatin ga t 80°Can d 100%R.H . Itma y be seen from these sources that if the residue from oil factories ist ob euse d serious denaturation of the protein should betake n intoaccount .However , itshoul d benote d that theworker s

26 in this field consider itquit e possiblet oextrac t theoi lfro m the beanso rth eflake si nsuc ha wa ytha tn osignifican t denatu- rationoccurs .Thi sapplie st oth esolven textractio na swel la st o thehydrauli cpressin gprocesses .I nvie wo fth ever y large amounts of soybeanmea l fromo'ilfactorie swhic hi seithe rfe dt oanimal so r returnedt oth esoi la sfertilize r these residuesca nb econsidere d asa huge ,potentia l sourceo fhighl y nutritive vegetable protein forhuma n consumption.

3.2.2.5 Germination.

During germination many biochemical changes take place inth e seed. VON OHLEN (1931)ha sgive na review inwhic h hestate s thatmos tworker sar eo fth eopinio ntha tth eoi li noil yseed s break down into glycerinean dfatt y acids during germination,an dtha t carbohydratesar efinall y formed.Starc happear so rincrease s ina large numbero foil y seedsdurin g germination. Reducing sugarap ­ pears inth eembry oan di susuall y quiteabundan t inth edevelopin g seedlings. Proteinschang erapidl yt oa solubl e formari dasparagin e appearsi nth eseedlin gb yth efourt ho rthir dday . Forth estud yo fth efirs t stageso fgerminatio nManch u soybeans were soaked foron ehou ri ndistille d wateran dwer e then germina­ ted betweenwe tpape rtowels.Th eolde rseedling swer egrow n inpot s filled withquart zsan dan dkep ti nth edark .Vo nOhle nstudie dth e translocation andtransformatio no fth ereserv e foodso fmatur e soybeansdurin g germinationan dgrowt ho fth eseedling si nth edark . These,reserv e foodsconsis to fa largeamoun to fprotein ,oil ,som e non-reducing sugar,an da smal lamoun to fstarch . The firstchange s foundwer eth eappearanc eo freducin gsugar,an d an increase instarch .Th eamoun to fstarc h inth ecotyledon sin ­ creased untilth efift hda yan dthe nremaine d nearlyconstan t until the ninth day, afterwhic h therewa sa rapi d decrease.Th estarc h disappeared fromth edifferen t partsi nth efollowin gorder ; root, hypocotyl, epicotyl,an dcotyledons .Reducin g sugar disappeared from theseedlin g inth esam e order buttw oday s later. A marked decrease inth eamoun to foi li nth ecotyledon s occurred by thefourt h day; depletion begana tth ebas eo fth ecotyledon s and progressed towardsth eopposit e end;th epalisad ewa sdeplete d last. Asparagine appeared inth ehypocoty l onth ethir d day. Itgra ­ dually increased duringth edevelopmen to fth eseedling,unti li twa s abundant inth ehypocoty lo nth etwenty-nint hday . The organically bound phosphorusan dmagnesiu m inth ecotyledon s graduallychang et oth einorgani c forms,a considerabl e portiono f

27 which remainsi nth ecotyledons, ,O nth eothe r hand mosto fth e po­ tassium disappears fromth ecotyledons . Germination also effects thevitami n content ofth esoybea na s has been shown by. WA I ,BISHOP , MACK, andCOTTO N (1946), They comparedth evitami n contento fmature ,unsproute d soybeanso f the Bansei varietywit htha to fth esam e variety ofsoybean s after differentperiod so fsproutin g under controlled conditions. Since theusua lpractic e inorienta l countries ist ous eth ehypocotyl s andt odiscar dth ecotyledons,th erelativ edistributio n ofvitamin s wasascertaine d separately inthes e portionsafte r5 4hour so fger ­ mination, theoptimu m period fromth epoin t ofvie wo f Quantity coupled with organoleptic properties, Assaysfo rcarotene , thiamin,riboflavin , niacin, reducedan dde - hydro ascorbic acid showed increase except forthiami n through5 4 hourso fgermination . Thiamin showed alternate increases andde ­ creases throughout thegerminatio n period investigated. Thecoty ­ ledons contained notably greater amountso fal lvitamin s forwhic h the testswer emade .Th eproces so fsproutin g beans, therefore,in ­ creasesth enutrien t valueo fth eproduct . Theentir e bean should be eaten, sinceth ecotyledons ,whic h frequentlyar eno tretained , area riche r sourceo fvitamin s thanar eth ehypocotyls , Asma yb esee ngerminatio no fth esoybean s brings about important changes inth ecompositio no fth ebeans .N odoub t theyiel d figures forth eextractio nwil lb eeffecte d toa larg e extentb ythos e chan­ ges,s otha t facts found for theextractio n ofth enon-germinate d beansma yno tb eo fdirec t applicationt oth egerminate d beans.

3.3 EXPERIMENTAL PART

3.3.1 Scope of work.

Prom literature itappear s that incountrie s where soymilki s produced themethod s applied areno tth esame . Prom theforegoin g it iseas y topoin tou tth eexistin g differences. However, iti s alsopossibl et ostat etha t important points ofresemblanc eexist. , Duet othi s resemblance theproces s ofmanufactur e canb edistin ­ guished infou rstages . The firststage , the pretreatment, servest omak eth era wmateri ­ ala ssuite da spossibl efo rth eextraction .Fo rthi spurpos e appa­ ratusesfo rth ereductio no fth eparticl e sizear eused , This ope­ rationmay b ecombine d with thefollowin ga sa napparatu s suited forth epretreatmen t mayals ob euse d asa nextractor . The second stageconsist so f the extraction proper, inwhic h con­ stituentso fth esoybea nhav et ob ebrough t intosolution ,emulsio n

28 orsuspensio na swel la spossible.Effor tshoul d bemad et o increase the output of the extraction asmuc h as possible. No doubt this yield isinfluence d byth edegre e inwhic h the bean particles are reduced insize . The third stage consistso f the separation of the milk from the residue. One should aim atwithdrawin g as much milk as possible from the residue. The laststag e consists ofsevera loperation s aiming at increa­ sing the quality of the product e.g. applicationo fhea tt oincrea ­ se thenutritiona lquality , increasing thepreservabilit yetc .

Prom literature itma yals ob esee ntha tfe wgood , and systematic studies have been carried out concerning the manufacture of the milk. Especially regarding theyiel d thatca n bereache d the data are veryscareean d -ifavailable -no t very reliable.A sth eyiel d of theextractio nha sa n important influenceo nth ecos tpric eo fth e milk iti sdesirabl e tofocu sou r study inth e firstplac e onthi s point. Inthi swor kw ear e concernedwit hth e firstan d second stageo f themanufacture ; thepretreatmen tan d the extractionrespectively . The separationoperatio nha sno t beenstudied, ,a s intentativ eex ­ perimentsn o important problemspresente d themselves (atleas twhe n using our method of manufacture)* The separation has only:t o be carried out insuc ha wa y that ifth emil k stands fora long time nosedimentatio noccurs .Whe na centrifug e isuse d thisai m isrea ­ ched without difficulties.Th e quality of themil k isno t studied in thischapter . Thestartin gmateria lo fal lou rexperiment swa sa commercia lva ­ riety of soybeans, obtained from an oil-factory. Presumably the country oforigi nwa sAmerica .Th ewate r contentwa sabou t 10%,an d therefore thebean swer eprobabl yno tdrie dartificially.Pro m these soybeanssoyflou rwa sobtaine d bygrindin gth ebean s includingsee d coats. Thecompositio no fth esoybean suse dwa sa sfollow s (%o fdry ;matter ) crude fibre , 11.60% crude protein 40.50% as« 5.64% Concerning thepretreatment .separat e studieswer e madewit hth e whole, unground beans and thesoyflour , while the most important extraction factors -th e extraction ratio,pH , temperature, time and method of mixing -wer e studied. Forth e experiments onth epretreatmen t acertai nextractio nme ­ thodwa sadopted,whil eth eexperiment so nth eextractio nstag ewer e alwaysprecede d bya certai npretreatment .

29 3.3.2 Design, ofexperiments.

Fromth epretreatmen tstag eth einfluenc eo fth edegre eo fgrin ­ dingo nth eyiel d ofth eextractio nwa sstudied . Tothi sen dth e whole beanswer e first soaked for1 6hour san dafte r grindingth e extractionwa scarrie dou taccordin gt oa certai nmethod-I na firs t experimentcoars e grindingwa sperforme d withadesintegrato r (1.1), whilei na nex t experimentth eparticl esiz ewa sfurthe rreduce db y meanso fa pebbl emil l (1.2). Witha Braunmixe ra highe r degreeo f particle size reductionwa sobtaine d (1.3), while inth elas tex ­ periment (1.4)o fthi s seriesth esam e apparatuswa suse d witha higher numbero frotation s perminute , The standard extraction method isth eextractio n carriedou ta ta 1C^ratio ,a p H6.5 ,a temperatur eo f20°C ,fo r1 0minutes ,usin g a Vibromixera sa nextractor . Inexperiment s withth eBraunmixe rth e useo fth eVibromixe r isomitted,a sth eBraunmixe rha sa sufficien t extractingaction ,Fo rth eseparatio na centrifug e isused, , Thesecon d seriesconsiste do fsi xexperiments .Th estartin g ma­ terialwa sobtaine d bygrindin gth edr ybeans .Excep tfo rexperi ­ ment2. 2th eflou robtaine dwa salway ssoake d for1 6hours .Th eex ­ tractionwa scarrie dou ta sdescribe d above except that inexperi ­ ment2. 5an d2. 6th eBraunmixe ran dth eIk aUltraturra x were used as extractors instead ofth eVibromixer . Ina firstexperimen tth ebean swer e ground througha 1-m msiev e (2.1), afterwhic h flouro fa fine r particlesiz ewa sobtaine db y usinga siev ewit h 0.5-mmopening s (2.2). Inth efollowin gexperi ­ mentth esam e typeo fflou rwa ssoake d before extraction (2.3)while attempts were madet oobtai na furthe r reduction inth eparticl e sizeb ymean so fwe tgrindin gi na pebbl emil lwit h porcelain balls (2.4). Inth elas ttw oexperiment sth eVibromixe rwa sreplace db y theBraunmixe r andIk aUltraturra x respectively.

The study of the extraction proper comprisedth einfluenc eo ffiv e factors-,.th eextractio n ratio,pH ,temperature , time,an dmetho do f mixing.Themanufacture.wa scarrie dou tapplyin ga certai npretreat ­ mentan da certai nextractio nmethod.I neac h seriesth eonl yvaria ­ ble factorwa sth eon et ob einvestigated . The pretreatment consisted ingrindin g ofdr ybean s (seedcoats included)b ymean so fa hammermil lprovide d with asiev eo f0.5-m m openings.Th eflou r thusobtaine dwa ssoake d for1 6hours . Thethir dserie sconsiste do fexperiment so nth eextractio n ratio. Waterwa sadde d toth eflou rt osuc ha nexten t that theweigh t of thedr yflou rwa ssuccessivel y5 ,10 ,15 , and 20%o fth emixtur eo f flouran dwater .Th erati owa s then called5 ,Id , 15,an d2 0res -

30 pectively. Distilled waterwa suse d forth e extraction at 20°C for 10minutes . The fourthseries,experiment s4. 1 to4.4 . concerned thepH.Enoug h alcaliwa sadde d forth ep Ho fth emixtur e to increase to7.5 , 8.5, and 9.5. Inth e experiment 4.1 no alcaliwa s added. The ratio was 10%, using a temperature of 20°Can d an extraction time ofte nmi ­ nutes. The fifthseries ,experiment s 5.1 to5.4 ,wer e meant tostud yth e effect oftemperature.B ymean so fa water-bat h the flour-watermix ­ ture was brought to 20°,30° ,4O 0, and 50°C. The extraction was carried out at a 10% ratio, a pH 6.5. for 10minutes . The sixthseries ,experiment s 6.1 to6.4 , dealt withth e influen­ ce of the extraction time, which was 2% 10,30 ,an d 60 minutes respectively. The extraction was carried out at a 10%ratio , a pH 6.5, and a temperature of 20°C. The seventh series,experiment s 7.1 to 7.4, concerned the method of mixing. TheVibromixe r isreplace d By the Ika Ultraturrax. The numbero frevolution swa s1800 ,2100 ,2600 ,an d300 0rp msuccessively . As regards the extent of this investigation it should be noted that thiswa s limited onpurpose , as the research programmawoul d be tooextensiv e ifal lpossibl e combinations 6$ the factorst ob e investigated had to be considered. In general itma y be said that the investigation wascarrie d out insuc ha wa y that acertai n fac­ torwa svariabl e while theothe r factorswer ekep tconstant . For the experiments only the undamaged, whole beans were used, while possible other kinds of seeds were discarded. So there had been somekin d ofselection . Ifnecessary , the beanswer egroun d by a hammermill. Each experiment wasreplicate d once inorde r to check the relia­ bility of the experimental results.Beside s the determinations of drymatter,an dcrud e proteinwer e done induplicate ; thesam e holds forth eacidity , pH,viscosity , and specific weight.Thu s thevalue s represented in the tables of this chapter are the averages of at least four determinations.

The following isa summar y ofth eexperiment s carriedout :

Series 1Pretreatmen t :desintegrator,ditt o+ pebblemil l whole beans Braun (8000rpm) , Braun (12000rpm )

Series2 Pretreatmen t :l-mm, 0.5-mm,0.5-m msiev e beansgroun d 0.5-mmsiev e+ pebbl e mill ditto + Braunmixer ditto + IkaUltraturr a

31 Series3 Extractio n ratio : 5, 10, 15, 20% Series4 Extractio np H : 6,5, 7.5, 8.5, 9,5P H Series5 Extractio n temperature : 20, 30, 40, 50C Series6 Extractio n time : 2%, 10, 30, 60min . Series7 Metho do fmixin g :1800 , 2100. 2600, 3000rp m

3,3,3 Technique of experiments,

A shortdescriptio no fth eapparatuse s usedi sgiven ,afte rwhic h theexperiment s aredescribed , 1, Desintegrator, CONDUX typeLVl5m ,P S0,75 ,n 2900 (photo2) , Inth edesintegrato rth ebean sar egroun d betweentw odiscs,bot h provided with metal teeth.Th estationnar y oneform sth ecove r ofdesintegrator,whil eth eothe ron ei sdrive nb ya nelectromotor , 2, Pebble mill, material porcelain, shape cylindrical, innerdia ­ meter 12,5cm , inner length1 0cm , Thebow li splace do na rotatio n installationan dkep tmovin ga t +3 0rpm ,

3o Braunmixer, 250Wat t (photo3 an d4) , Thecu p hasa loos e bottom provided witha naxi swit ha numbe r ofknives ,Th eaxi si sdrive nb ya nelectromoto ran dca nb ead ­ justeda ttw ospeed s (8000an d 12000rpm. ) 4> Vibromixer, type El,4 0Wat t (photo3) , Thisvibratin g mixerconsist so famotor, aro dan da meta ldisc , 6c mi ndiameter ,provide d withperforation s placedi ntw ocon ­ centric circles.Th eamplitud eca nb eadjusted , buti ti sno t indicated,Th enumbe ro fvibration s is5 0pe r second. 5. Hammermi11, Peppink,,typ e 261D 20/2 ,P K2 ,n i s2850 . Thismil li sprovide d with6 arms ,9. 5c mi nlength ,an dsevera l sievesdifferin gi nsiev eopening . 6. Ika Ultraturrax, typeTM V45 ,n 8000 ,V 220 ,A 1. 8(phot o3 an d 5). Thehea do fth emixe rconsist so ffour'concentri cring spro ­ videdwit h teeth.Th einne ran doute rrin gar e drivenb yth ero ­ tor.Th enumbe ro frevolution s perminut e isadjustabl eu pt oa maximumo f8000 . 7. Centrifuge, Bairdan dTatlock, .typ eV 808 . This centrifugeha sfou r plastic tubeso f3 c mdiameter , which canb efille dwit h5 0m lo fliquid .Th edistanc e from the axis ofth e centrifuget oth eti po fth etube si nhorizonta l position isabou t 14cm .Th eseparatio n was carriedou tb ycentrifugin g for five minutesa t300 0rpm .

32 The experimentswer ecarrie dou ta sfollows .

1.1 Desintegrator. 60go fbean swer esoake dan dthe nthe ywer efe d toth edesintegrator .Th ebea nparticle s left behind between the teethwer ewashe d awaywit hwate ran dth ewas hwate rto ­ getherwit hth eground,swolle nbean swa sbrough tint oa nename l can.Wate rwa sadde dunti lth erati owa s10% .A ssoo na sa nex ­ traction temperatureo f20° Cwa sreache d thero do fth eVibro - mixerwa splace di nth eca n axiallyan dth eextractio ncarrie d outfo rte nminutes.The nth eseparatio no fth emil kb ymean so f the centrifugewa seffecte da t300 0rp mfo rfiv eminutes:. "Th e clarified suspensionwa sweighe da swel la sth ecentrifug etu ­ bescontainin g theresidue . 1.2 Desintegrator and pebble mill, the groundbean sfro mth edesin ­ tegratoran dth ewas hwate rwer epu tint oth epebbl emil lan d thebea n particleswer e furthergroun dfo rthre ehours .The n thismas swa sbrough ti nth eextractio nca ntogethe rwit hth e washwate rfro mcleanin gth emil lan dth epebbles .Afte raddin g morewate rt oobtai nth edesire dratio ,th eextractio nan d se­ parationwer ecarrie dou ta sdescribe dabove . 1.3 Braunmixer. thesoake dbean swer ebrough ti nth ecu ptogethe r withth erequire d amounto fwater ,an dthe ngroun d forte nmi ­ nutesa t800 0rpm .A sth eBraunmixe rha sals oa nextractin gae - tion,n oothe rextracto rwa sused .The nth eseparatio nwa sper ­ formed. 1.4 Braunmixer. See1.3 ;th enumbe ro frp mwa s12.000 . 2.1 Hammermill. Toth e6 0g o fsoyflou rsoake di nth e extraction candistille dwate rwa sadde dunti lth e 10%rati owa sreached .B y meanso fa waterbat h thetemperatur ewa sbrough tt o20° C± 1 . Thenth eextractio nan dseparatio nwer ecarrie dou ta sdescri ­ bedbefore .Th eexperiment s2. 2an d2. 3wer ecarrie dou ti nth e sameway ,excep ttha ti nexperimen t2. 2th eflou rwa sno tsoa ­ ked. 2.4 Hammermill and pebble mill. Aftersoakin gth e flourwa spu tint o thebow ltogethe rwit hno tmor etha n 125m lo fwate ran d then groundfo r threehour sb ymean so fporcelai npebbles . Thecontent swer ethe npu t intoth eextractio nca ntogethe rwit h thewashwate rfo rcleanin gth emil lan dpebbles ,an dwate rwa s addedunti lth edesire drati owa sreached ,Extractio nan dsepa ­ rationwer ecarrie dou ta sbefore . 2.5 Braunmixer. Thesoake d flouran dth erequire d amounto fwate r werepu ti nth ecu po fth emixe ran dth eextractio nwa scarrie d

33 outa susual .Befor ecentrifugin gth emixture ,whic hha dattai ­ ned atemperatur eo fabou t35°C „ wascooled . The numbero fro ­ tationso fth eknive swa s8.00O . 2.6Jka Ultraturrax. The soaked flourwa s extracted with there ­ quired amounto fwate ra t3.00 0rp m for 10minutes .Th etempe ­ rature increased by2 5t o30°C .whic hmad e itnecessar y tocoo l themixtur ebefor ecentrifuging . 3 Extraction ratio. Theseexperiments 'wer ecarrie d outi nth esam e way asexperimen t 2.1. The different ratios were obtained by addingth erequire d amountso fwate r toth e6 0g o fsoyflour . 4 Extraction pH.A si nexperimen t2.1 .Th erequire d pHwa sobtai ­ ned bypipettin g 11.25N ofalcal i to themixtur e (which was kept inconstan tmovemen t bya mixer) , while at thesam e time thep Hwa s indicated bya pH-meter . 5 Extraction temperature._As inexperimen t2.1 .Th erequire d tem­ peraturewa sobtaine d byplacin g theextractio n can inawaters - bath. 6 Extraction time..A s inexperimen t 2,1. 7 Method of mixing. As inexperimen t2.1 . Inorde r topreven tth e large increase intemperatur e the extractionca nwa splace d in a largetan kwit hwater.Th eextractio nwa scarrie d outa t1800 , 2100, 2600,an d300 0rp mrespectively .

Allth esuspension s and sediments (residues )wer e analysed for drymatte ran dcrud eprotei ncontent . Inaddition , theacidity ,pH , viscosity, and specific gravity of themilk ,wer edetermined . All analysesan ddetermination swer edon e induplicate ,an d carriedou t asfollows .

Dry matter. From aweighin gbottl emil kwa spoure do na porcelai n disc. Ina nove nth emil k was dried at 100°C for two hours,afte r which thedis cwa scoole d ina nexsiccato r containing silicagel. Then the discwa sweighed . About2. 5 go fth esedimen ti nglas sweighin gbottle swa sdrie da t 100°Cfo rtw ohours .Afte rcoolin g ina nexsiccato r the bottlesan d theircontent swer eweighed .

Crude protein. About five gramswer epu t intoa Kjeldahl-flask , adding0. 7 gHgO , 10g Kg S0 4an d2 5m lH 2S0496%. The destruction wascarrie dou tunti lth econtent swer ecolourless ,afte rwhic hth e heatingwa scontinue dfo rtw ohours .Afte rcooling ,5 0m l8 %Na„s 20o- solutionan d9 0m lNaOH-solutio nwer eadded .

34 Thenth econtent so fth eflas kwer epu tint oa Parnas-Wagne rap ­ paratus.B ymean so fstea mth eammonia kwa sdistille d andreceive d in5 0m lbori cacid .Afte raddin g2 %Tashiro-indicato rt oth ebori c acidi twa stitrate d witho. 1n o fbase .

Acidity. 10m lo fmil kwer e titratedwit h0. 1n NaOH,usin g0. 5m l phenolphtaleine asa nindicator,Th etitratio nwa scarrie dou tunti l thecolou rwa sth esam ea stha to f1 0m lo fth esam emil kt owhic h 0.5m lo f0.0005 %fuchsine-solutio n hadbee n added.

pH-determinationswer ecarrie dou tb ymean so fa nElectrofac tpH - meter.

Specific gravity. Thiswa sdetermine d bycomparin gth eweigh to f soymilk in5 m lpycnometer s with thato fdistille d water ofth e same temperature (20°C).

Viscosity., A"Hopple rviscosit y meterwa suse dfo rth edetermina ­ tiono fth eviscosity . Inprincipl ei tconsist so fa nincline d glass tube throughwhic ha glassspher e fallsdown . The timeneede db y thesphere.t opas sfro mon epoin tt oanothe ro fth etub e ismeasu ­ red. Around thetub ewate r ofth edesire d temperature (20°C)i s circulated.

3.3.4 Results, discussion, and conclusions.

Themetho do fcalculatin gth eyiel do fdr ymatter ,i sa sfollows . Afterdeterminin gth eamount so fdr ymatte ri nth emil kan dth ere ­ sidue,th epercentag eo fth eorigina ldr ymatte ri nth emil kan di n theresidu ewa scalculated . Itshoul db enote d thati ngenera lth e sumo fth eyield s thuscalculate dwil lno tb e 100%,du et oanalyti ­ calerror san dlos so fmaterial . Promth eamount so fcrud eprotei n found inth emil kan dth eresi ­ dueth epercentage so fth esu mo fth eamount so fcrud eprotei ni n suspension andresidu ewer e calculated. Evidently thesu mo fth e yields calculated inthi swa yi salway s100% .

Asa nexample ;o fth ereproducibilit y ofth eanalyse sof.dr ymat ­ ter,crud eprotein ,an dth edetermination so fpH ,.acidity, - viscor sityan dspecific-gravity ,th eorigina lresult san dthe.average so f theexperiment s3.2 ,4. 1an d6. 2ar ementioned .

35 MILK RESIDUE dry crude pH aci­ visco-• specific dry crude matter protein dity sity gravity matter protein % from (ml) (20°C) (20°C) * fromn dry dry matter matter % % 3 k2 Extraction 7.59 47.83 6.59 1.4 8 1.936 1.0192 19.19 34.78 ratio 7.58 "47.46 6.58 1.44 1.924 1.0191 19.12 33.81 19.30 30.15 duplicate 7.68 46.60 6.59 1.44 1.838 1.0192 7.63 46.52 6.57 1.42 1.805 1.0191 19.24 30.12 -average 7.62 47.10 6.58 1.45 1.876 1.0192 19.21 32.22 4 .1 Extraction 7.70 43.19 6.53 1.63 2.095 1.0197 19.24 40.36 PH. 7.69 43.16 6.52 1.55 2.093 1.0196 19.23 duplicate 7.52 4 0.11 6.42 1.67 2.111 1.0192 19.51 28.46 7.51 38.42 6.41 1.66 2.088 1.0190 19.50 25.53 average 7.61 41.22 6.47 1.63 2.097 1.0194 19.37 31.45 6 .2 Extraction 7.68 47.40 6.59 1.898 1.0199 20.10 31.43 - time 7.66 47.16 6.57 1.894 1.0197 20.03 31.91 duplicate 7.81 47.31 6.58 1.45 1.874 1.0195 20.95 28.82 7.79 46.97 6.57 1.40 1.872 1.0195 20. 7S 28.72 average 7.74 47.21 6.58 1.43 1.885 1.0197 20.46 30.22

The reproducibility of the manufactureca nb echecke d fromth e following exampleso fth e yieldso fdr ymatte ran dcrud e protein.

MILK RESIDUE

weight % 'rom % from weight % from % from (8) orig. ori g. (8 1 ori E- ori g. dry crude dry cru de ma :ter prot ein matl er prot 3 .2 Extraction ratio 418 47 58 45 65 75 119 65 42. 21 34 25 duplicate 416 17 58 .67 67 70 121 .83 43 25 32 30 average 417 32 58 56 66 73 120 74 42. 73 33 ?8 4 1 Extraction H P 42 2 40 59 8S 61 07 114 81 40. 66 38 9 3 duplicate 420 50 58 .20 66 J? 118 50 42. 56 33 51 average 421. 45 59 03 63 78 116 66 41 61 36 .21 6 2 Extraction time 417. 24 58 ,93 66 49 119 .50 44. 16 33 51 . duplicate 425. 63 . 61 14 69 92 112 35 43. 13 30. OR average 421, 44 60 .04 68 21 115, .93 43 65 31 80

Itma yb esee n fromthes e figures thatth eanalytica ldetermina ­ tionsar ereliable,an dtha tb yreplicatin g eachexperimen t fromth e beginningth emetho do fmanufactur e provedt ob ereproducable .

Intabl e1 an dfi g3,th eresult so fth e investigationo f the first stage, the pretreatment, aresummarize dan drepresente d graphically. Thewhol e beansan dth egroun d beanswer estudie di nserie s1 an d2 respectively. 36 Theresult s ofth eexperiment so n the second stage ofth emanu ­ facture, the extraction proper, aresummarize d intabl e 2an d re­ presented graphically infi g4 .The y contain theresult so ffiv e serieso fexperiments : 3) extractionratio ; If) extraction pH; 5)ex ­ traction temperature; 6) extraction time,an d 7) method ofmixing .

Itshoul db ementione d thati nstudyin gth emethod so fpretreatmen t the method ofworkin gwit h each apparatuswa sno tvaried , thati s to say,n oattempt swer emad et oobtai nth eoptima lworkin gmethod . Otherwisea systemati c studyo fsuc h factorsa sdegre eo ffilling , numbero frotation swoul d benecessary . Neverthelessa nappreciabl e differencewa sobtained .

Whenconsiderin gth eresult so f the pretreatment of the whole beans (table 1,serie s 1),i tma yb esee ntha tth edr ymatte r con­ tento fth emil k showsa regula r increase inth esequence : coarse grinding, coarse grinding followedb yfurthe r grindingi na pebbl e mill,fine grinding (Braunmixer.8.00 0rpm),an d finegrindin g (Braun­ mixer, 12.000 rpm). Thedesintegrato r proved tob eunsatisfactor y underth econdition s used.Th eresul to ffurthe r grindingwit hth e pebble millwa srathe r disappointingan dmigh t have been betteri f the timeo fgrindin gwa slengthened.Bu ti ntha tcas eth emil kwoul d alreadyb ecoagulate ddu et osouring . Therefore fine grinding with the Braunmixerwa stried .Th eresult s obtained were comparatively favourable; increasingth enumbe ro frotation so fth eknive sappea ­ redt oincreas eth edr ymatte rcontent . Inth etheoretica l case thatth egrindin g iss ofin e thatn ose ­ diment islef to ncentrifuging , thedr ymatte r contento fth emil k wouldb eio% .I fhowever,par to fth ebean si sno tgroun d fine enough thedr ymatte rconten to fth emil kafte rcentrifugin gwil lb elowe r than 10%.Th ecoarse rth eparticle sth egreate rth edecreas e inth e dry matterconten to fth emilk .S oth efigure si nth ecolum nO fth e dry matter content ofth emil k form a measure for the> fineness of grinding. Iti sclea r thati fth edr ymatte r contento fth emil k decreases theyiel do fth edr ymatte rwil l decrease too.Pro mth ecolumn"dr y matter% »an d "percentageo fdr ymatte ri nsuspension "i tma yb eseen , however, thatth eyiel d decreasesmor e thanth edr ymatte rcontent . Itappeare d thatth eyiel d decreased ifth egrindin gbecam ecoarser . Thisma yb eascribed toth efac t that ifth edr ymatte r content decreases duet ocoarse r grinding, the.particle s that are cen- trifugedou tfor ma heavie ran dmor evoluminou s sediment. Thusth einterparticl e volume isincrease d andretain smor e milk- and consequently alsomor edr ymatter-tha nwhe n there isless , more

37 Oi vo VO • . CN vO CN r-. r» o CO VO 00 00 ^r CO ft m m o\ o>

*• CN VO CO o CO ov o vO m men t % o f to CO to CN

»H 1 I U -rt ft i> 00 in o ft o CN m m w TT 1> m CN r-* 00 m 00 00

se d VO *• CN ft vO CO in dr y c* 1 r* m igi n att e % o f men t m f t m co CO CO m t. EC «fr «* 0 •-<

"3 , c c 1-4 m "fr 00 v 00 ^ CO o r* ^ £ 4J -H 41 - co ft 1-1 m vO 00 *• *• ft o 0 -S "° » o- 0 m I-* to VO Ol vO o Ot CO S?M J; £ S » f m

ft • r- i> O CN ft CN CN 0\ 00 ft <*•* . • *• VO *• ft CO vo 00 t w CN CN VO tn

dr y m m o , m ^r sio n igin a a t te r % o f suspe n co *r w m w vO VO vo VO S B-s «

* o CN o\ m © r- © CO ft CO vO vO O*

crud e fO CO to CO CO CN CO CN CO CN matte r

H prot e i n fro m dr y Z c Id u 00 i> to CN N 00 o vO m o t«. rr CO CO tn 0\ •H £ m «fr * c% M t CN CO CO O Cn . o> VO t- t* E ft ft »-t ft CN ft fl ft ft Q s «-« o ft] ' CO 4J vO vO 0\ 00 CO VO r* ft in CN •c ,-, r» o ft *-4 00 CN VO ft ft M M « ft ft CN •* © 0\ ft ft > r* 0) ~' es o.>. 00 I-. CO CO o o ft ft c i CN ft r-t ft ft ft o »-» *"* ^ ^ •H •P O C\

X ft • *H ft ft • 0. M *-" 1-1 H ft ft CD to ' •"* ' •"*

u Ov CN o 00 00 ^« t". re Ov o s m vO CN w VO o o CO o o• " o a? r-* o\ • OY > ft r*. •^r © m•fl- CO CO eg z ft CN CO 5 - S* J! JI CN CN CN CN CN CO ft o» o\ CN m O rf o\ CN CO ^« -p z o. o CO m t tn V C8 0 -H 8 !H u to "0 o ft ft ft ft ft ft ft ft O) a. a en vO Ov o vO *H t«. Ol CN VO V D X W w vO m vo in in in m m E a W vo vO vO VO vO vO vo « vo vo

O± Bc •H -P o U io -O M P 0 CO ^ 43 CO X 0) to tu o -C in- •w IS 00 CO S VO Pi ^ ^N 4 CN to a o C o •i-> •H cu 4-> O MU u. +J a. X o CD £ O « v e •> a •tJ ,^f J5 ft * 0. r-l u u o u •!-U> m to •*H •rf o a o e o a o t- •H +> -P u §g S ° H d a • 19 CN V t- CO fe -H u B '~ Bj ^> a z •s. o o C -P o O ^-^ i> >0 0>) • A>t >41 cd « •H •H •H •H •* )H > 00 GO 00 00 •P Q « tt 6 X! to a a a . a a g 0) a a a a a O e • • • e a io in tn in in CO u g I i-l z -P o •H w "O a a a a i a a 4) h •w O « a •< B 3 B a a a a a u H-, O e o a a o a x a a 3 O Bj a A a a js 0 -i CO CO 0 a H H bl £ Ji •a o A » V •d o .s to J3 J: x o X Z n «J •H 09 •C •n J: JS - A .C 3 JS M < » TJ 41 c -a c c •o c Cd 4J 3 41 3 •d 3 « 3. 3 "0 fi V c c H U •H C e c S S C J! 0 4) V 0 S 41 ae 4J C3 4) a a ae -a J B. J3 * •H 41 A 4) o 41 A 4> A 4) 0 CO BJ 4J m c 03 *J S3 V a oi £§ < £=1 4J « 00 (N vo CM © f in rH CO m rH 00 o cs m o © m in C CO (N t (N W cs o m cs CO o CO CO OV © Ov Ov oo vo a; « 00 CO CO rH tN % o f B CO VO VO CO © O I* CO CO m CO © t«» co CO crud e M CO CO t CO CS cs cs CO CO cs cs CO CO CS CO es CS cs cs protei n i n sedi - origina l

»H I co (H -H tn >p CO m o\ i-< o o> cs cs cs 00 vo m m w cs Q\ CO IN 00 tN C r* VO CN VO o m *• cs cs CO •

Be d aj dr y

att e f-< 1-< igi n (N % o f CN CO CO vfi B oo CO m t*. m o © b. r* *• to in o o u SB , f CO CO TT * CO CO ^T«f *• CO CO CO CO 0 -H ^ ^* ^ ^ ^.

00 CO CO t CO © r* VO CTt CS vo o co m o C CS vO vo in t^ K Ov i-. Ov *" t^ O Ov rH rH 0

»H I-l VO rH ta VO rO• vo CO C\ o.>• •CO cs *t* 00 rH OV cs es VO vo

crud e tv tN tN c % o f k>. VO vo in vo t^ k> VO VO vo vo IN- VO t*. t*. l^ r*

a protei n i n suspen - origina l

CO CO CO , cs to CO t- •<*• t*. C CO VO CS cs rH rH CO *4 _f >, s 0. 0 o OY o\ o in o^ 00 vo o m CO •H fl GO m m o o © Ov CS o Ot rH CO CO CO M O 8 to 01 o es •*-> 0 O m m Tf m VP vo vo vo m vo vo in vo m vo vo vo vO VO S O L. CO M CO CN in CO m VO ov « Ov vo o cs O vo cs CJ» vo 00 *+-, CN CO o o CO r* 1^ CS O) CO 00 CO cs rH CO tn o CO rH CO ^*K \ ^ *^i CO M m CO »H vo rH rH * t CO CO rH o Ov Ov Ov t-* ^f VO

C crud e CO CO CO CO CO cs CS CS CO CO CO CO CO CO cs cs CS cs cs CS

'!•* H matte r fro m dr y protei n 4> Z •*-> u, u o cu 1% 1. 4J s CO +4 o> t cs cs Oi X Ov Ov vo o

Q dr y (% ) t^ c^ TT O) t-. r* VO o o 00 t> Ov o Ov OV vo vo VO m «*-, 1-t * rH r-l rH rH «N M ^H CS CS cs rH rH rH rH W rH matte r » o T3 Cd CU *"* *"* *"* i—1 en +J o ^r v m vo in 00 © o o vo CO m CO ^ Xfi o m 00 CO c i—1 >£> t>. CO 00 CO rH o X ~ Oi VO CO vo t* *• cs o> o. © K Ov VO t* •f-S •iH bO U> •M •H N-* i-H rH -P o TT t VO •» cs 00 t CO ov t-* cs m m v o CO CJ CO

X Oi cs © m V o *r CS vo m rH CO CO t". m oo r- m vo ts. CD •aH 00 o> o y-< Ov o o rH OV Ov o Ov o\ Ov CO OV Ov Ot Ov Ov rH, r>! rH O rH CO cs cs CS cs rH rH rH rH rH rH rH rH

3 gravit y o spec i fi e c J3 CO rH o i * tf) cs 00 N T o OV 00 vo *r vo m m o •fl" VO rH vo CO o t- 00 k>. Oi vo m CO t» rH rH m 00 CO -d- r-* VO V CO -~*N r-i o >»o o O 4J0 <—• CO 00 00 t-. o vo r-* Ov CO CO o\ Ov 00 CO 00 rH vo t* CO 00 bo 01 -H O CU C u -H CO es O 1-t 1-) cs vo N cs cs CS es cs es CS •*J o > >-• m 'r«O 1 S T) 1 >, ^x CO m o> r*. CO cs tN in VO *** rH © rH CO CO rH CO m CO ^ •H +J rH VO t CN n es CO Ov 00 in c 3 * e c>u 00 CO v4 vo m rH CS rH m Ov vo vO rH m •H CO crud e V> ^t V *• •v in m in ««• -* v

matte r •a* m v CO + %> fro m dr y * ^ * * * s £S •*>+A E ; r protei n i-t N in vo i-4 rH cs V CO rH VO m 1 m vo rH •* m o> m Ov lO CO rH m \o o o vo a o> m m 00 VO cs f^^ VO GO rH CO CS

V dr y att e I. CO t* M vo t>. t>. tN CO tN t^ r* tN t- r-. r^ t* CO CO CO CO o**m* a e r-l •w *** 0 rH •P o CN CO o\ m t>. m \0 cs m 00 m rH ^j- rH ^T 00 00 o o 00 CO CO CO CO *f in t*. k. r-l o ^c t* ov m cs N ^r r* *• o V •H * — ^ E 00 t>. cs in* fH CO m Ov rH• '0•0 Ov Ov r- rH cs m 00 CO m in 1-t in CS rH CO cs rH CS rH rH rH rH SH « ^ m m cs cs cs cs s. ^r •w 5 * ^* ^* ^ *• ^« *" <»• ^ ^ ^ as 35 J3 M H M a. 03 0 (8 a 8 0 h U u u t- U rl U .C £ u u t« u u u u u u u u u rl u u u o* V V Oi V 9) ti 9) « 0) V V V V V n p •i-» •P •3 3 s •w •H *> A ji A J3 £« X, A A r3 J5 A A A A A js 4-> +J +J J= y 0 u o u o 0 o O 0 u O O o 0 O to a a **a a •u w (0 (0 0 to « 03 DO V) 09 (0 (0 « u u u b "4 •H •H •H •H •H Tt +J L. * •H » •r«t —1 « •H •H -H •H •H ^ +J r> +J "O C «— • 0 0 0 0 o 0 0 0 0 0 0 0 0 o- 0 3 3 3 c h o o • s CO 3 u u u h b u u u h u u -1 u r> -t b •P -0 J3 ja J3 A A A A A A A A A A A A 0 a a •H •H •H T| C o K -H ^•H •H •H •H •w «H •H •H -H •H •H -H M j< « M o Cd 01 M M K) M CO .* t-> h > > > > > > > > > > > > > > > > z t> 3 B C c c C c c C d o •H •cH *cH •eH •ew •H •cH •cH •rct •rl "H «eH •rcl •rt •eH •H M •eH •eH •cH •eH o cu H •M e a a a a a' a a a a a a a 8 a a a B B B w .n H © 1 >i h o o o o o © © © © o o *n O o © o O © O M *H iH i-» «H iM rH rH rH rH rH rH rH rn rH o 4> CO VO rH rH «-l s CO O M cs• •roJ Cd +J •a z CU o 8 r-l , k. u 1 !-c s-, a CJ K o 4> < & m m m m in in m in m m in m m m m m m m m \n CD 4J u t4 (O VO CO vo • U cd H vo VO t- o. VO <0 vo VO VO vo VO vo VO IS 0) M « « c X c3o JJ Cd »-j 0 S-, 0) •H

Cu (9 6? tfl 6? 6^ t« *£ £ fe^ *ft ^ ^ & ^ 6? ** t* ^ « s* « St m o in O © o o © o © o O o o o © o o o o u

Ed <-4 M CO t *4 cs CO rH CS CO rH J ^ *t cs fO -* rH CS ro » CO *t m vO < ' t> 100 100 dry matter in suspension, crude protein in suspension, 90 dry matter in sediment. 80 ,».^ "" crude protein 'in sediment

'»••' —• 70 .»' ,*-"'. « --«— -• 60 ..»' 50 Ji-' .« 50 *'' •S-'-'-o. •"•q_ 40 ,-D,, ' -• o. s ' o-„ .-D-- •' -•o --a 30 ~""0. / ^JD— "O 20 V'

10

0 0 1,1 12 1.3 1.4 2.1 2,2 2.3 24 2.5 26 1. Whol e beans. 2. Beans ground.

dry matter in suspension. o dry matter in sediment crude protein in suspension o crude protein in sediment 1. 7. 100 90 80 70 *.. 60 B.. 50 p --•' ,B 40 O—O.-. £>--•-O... ••D—-a D—.-D—-o... 30 o-—«•-.. " o.—a.. 20 10 0 5 10 15 20V. 63 75 8.5 S15 20 30 40 50 *C 23 K) 30 60min 1 2 3 4 . Extr. ratio. 4. Extr. pK 5. Extr temp. 6.Extr time. 7.Metho d of mixing.

42 compactsedimen to fsmalle rparticle swit hles sinterparticl espace . So thefigure sfo rth eyiel d ofdr ymatte rar edetermined.b yth e finenesso fgrindin gan dth estructur ean dquantit yo fth esedimen t formedb yth ebea n particles.I tma yb esai d thatthe yar e a mea­ sure for the quality of the method of manufacture. Asregard sth eyiel d ofprotei n itappeare d that abouthal fth e amounto fprotei nwa sobtaine di nth emilk.A sth egrindin gwa smor e effective,moreprotei nwa sfoun di nth emilk .Al lth eyiel d figures forprotei nwer ehighe rtha nthos efo rdr ymatter .A si ti sou rai m toobtai na proteinric hfood ,i ti sa favourabl e facttha tth e soy- proteinsar ebette rdispersabl ei nwate rtha nth e otherconstituents . Thep Ho fth emil ko fthes e fourexperiment sprove dt odiffe rver y littlefro m6.6 ,whil eth eacidit yincreased'i nth eorde rmentione d earlier.Thi si sevidentl ydu et oth eincreasin gdr ymatte rconten t of themilk , which isals o indicatedb yth eincreasin g valueso f theviscosit yan dth especifi c gravity. Theamount so fresidu edecrease di nth esam eorde ran ds odi dth e drymatte ran dcrud eprotei n content,thoug hno tquit eregularly . Thisdecreas eca nb esee nbette r fromth evalue sfo rth e"percen ­ tageo forigina ldr ymatte ri nsediment "an dth e^percentag eo fori ­ ginalcrud eprotei ni nsediment" .

Consideringth eresult so feac ho fth eapparatuse sw ema ystat e thatth eeffec to f the desintegrator wasno tsatisfactor y asonl y abouton ethir do fth edr ymatte ran dno tye thal fo fth eprotei n -were found inth emilk .Thi swa salread y expectedo nvisua lobser ­ vationa sth eswolle n beanswer eno tgroun dbu ttom e asunder.Evi ­ dently inthi swa yn osufficien t reduction inparticl e sizewa s possible. Thisma yb econsidere dt ob eth ecaus eo fth elo wyield . Although a better resultwa sdbtained'o n further grinding? in a pebble mill yeti twa sles smarke d thanwa sexpected .Fo ra larg e partthi sma yb eascribe dt oth efac tth eprecedin ggrindin gopera ­ tiondi dno tgiv eenoug hreductio ni nparticl esize ,i nothe rwords , we assume thatth epebbl emil luse dwa sto oligh tfo rou rpurpose ; Besidesth epresenc eo ftoug hsee dcoat simped eth edestructio no f the cellstructure . The Brpammixer gavereasonabl eresult sa t8.00 0rpm. Alittl emor e thanhal fo fth edr ymatte ran dabou ttw othird so fth ecrud epro ­ teinwer efoun di nth emilk .Th ereductio ni nparticl esiz eobtaine d wasmuc hbette rtha nwit hth edesintegrator .A smoreove rth emixin g intensity wasappreciabl y higher thanwit h theVibromixe r this higheryiel dwa sno tsurprising . Ifth enumbe ro frotation swa sincrease dt o12.00 0th eyiel do fdr y mattera swel la stha to fcrud eprotei nwa sals ofurthe r increased.

43 When evaluating theresult so fth e pretreatment of the ground beans, i.e. soyflour, aregula r increase inth edr ymatte r content mayb eobserved .Th ehighes tdr ymatte rcontent swer e reached when grindingwa sperforme db ya Braunmixe ro rIk aUltraturrax ,an dals o when further grindingwa scarrie dou twit ha pebbl e mill.Th edr y matter content inth esecon d serieswa sremarkabl y higher thani n the firstwhe n corresponding experimentsar ecompared . Theyiel d ofmil kwa salway shighe r inth esecon d than inth e firstseries .A sa consequenc eo fth ehighe rdr ymatte rconten tan d the higheryiel d ofmil k theyiel d ofdr ymatte r inth e milkwa s also higher inth esecon d series. Theyiel d ofcrude protein increased with theincreas e inth e yield ofdr ymatter . Thep Han dth eacidit ydi dno tsho w important differences,,whil e thespecifi c gravity increased regularly. Ingenera l itma yb esai d thatth eviscosit y increaseda sth edr ymatte rconten to fth emil k increased. Theweigh to fth esediment sdi dno tdiffe r much, except inth e first experimenti nwhic ha larger sieve openingwa suse d andi n thesecon d inwhic hth eflou rwa sno tsoaked . Generally thedr ymat ­ terconten to fth esediment swa shighe r than inth efirs tseries . As, however,-th eweight s ofth esediment s inth esecon d series were lower than inth efirst ,th epercentag e ofdr ymatte r inth e sedimentsi nth esecon d serieswa sa sa rul e also loweri nth e se­ cond series.Th esam e istru eo fth eyiel do fcrud eprotei n inth e sediments.

In contrast toth efirs t series the pebble mill hasgive nth e best results.Evidentl y this isdu et oth efac t that,th epretreat ­ mentcause da sufficien t reduction inmateria l size,thu smakin gi t possiblefo rth epebbl emil lt oeffec ta furthe r reduction inpar ­ ticlesize . Oncomparin gth efigure sfo rth eyiel do fdr ymatte ro fth e Braun­ mixer and.th e Ika Ultraturrax viith those ofth e Vibromixer.it ap­ pearstha tth efigure sfo rth eVibromixe rar elowe rtha n thosefo r the firsttw oextractors . Inth ediscussio no fth epretreatmen to f thewhol e beansw eassume d that thesedimen t retains lessmil ka s the bean particles become smaller duet obette r grinding ands o the milk yield increases. Ifw econside r themil k yield ofthes e threeapparatuse sw ema yse etha tth eyield spracticall yd ono tdif ­ fer,.Thi smigh tb ea nindicatio n thatth efavourabl eeffec to fth e Braunmixeran dth eIk aUltraturra x isdu et oa muc h more intensive method ofmixin g rather than toa further reduction inparticl e size.

44 Itals oappear s that the sieve opening ofth emil l playsa nim ­ portant role.I fth ediamete rwa sreduce d from1.0-m mt o0.5-m mth e yield ofth edr ymatte ra swel la stha to fth ecrud eprotei n inth e milk increased byabou t 10%.Beside sth eweigh to fth emil k increa­ sed,whic h isi naccordanc ewit hou rassumptio n thatthi s isa mat ­ tero fth estructur e andth equantit yo fth esediment . Aftercentrifugin g anunstabl e layerbetwee nth emil kan dth ere ­ sidue appeared inth ecentrifug e tubes ifth eflou rwa sno tsoake d beforehand. Inthi s layerth emil k andth esedimen tar eno twel l separated, sotha t thenno tonl y themil k yield butals oth edr y matter content ofth emil kdecreases .

NAGELet al.(1938 )obtaine da yiel do f64.5 ,69.7 ,72.2 , 81.0. 79.1,an d92.6 %fo rth e firstextractio n ofdefatte d flakesusin ga ratioo fabou t 2Wo (1:40)fo rth epretreatmen t witha Wile ymill , hammermill with2 ,1 ,0.5-m m sieve openings, porcelain ball mill and subsequent useo fa 100an d20 0mes h sieve respectively,an d wet grinding for4 %hour s ina ballmill . Withth e0.5-m msiev ethe yobtaine d 1;8%mor eprotei ntha nw edid . Thiscoul d havebee nexpecte d sincethei r ratiowa smuc hlowe rtha n ours (10%). Infact ,thei ryiel d should have beenhigher . Thiswa s not thecase ,however ,probabl ydu et othei r mixingno tbein gsuf ­ ficiently intensive.Anothe rpoin to fdifferenc ewa sth econditio n ofth £startin gmaterial:whil ethei r flakeswer edefatted,ou r beans werenot . Enlarging thesiev e opening from0.5-m mt o1-m mdecrease d their yieldwit h 2.5%only,wher ea si nou rexperimen t thisdecreas e amoun­ ted to7.5% .Th edifferen tworkin gmetho d ofprocessin guse d byu s may account forthi sdiscrepancy . * With thebal l mill theseworker s obtained 92.6%fro mth e first extraction,whil e inou rcas e thiswa sonl y 79.4%. Thisma yals ob e ascribed toa differenc e inth epretreatmen t andth e extraction proper.

The results of the extraction proper arediscusse d hereafter.

The extraction ratio provedt ohav ea decisiv e influenceupo nth e composition ofth emilk , while also theyiel d waseffecte d ina largedegree . Itappeare d thatth edr ymatte rconten to fth emil k increased asth erati o increased.I nou rexperiment sth edr ymatte r content varied from3. 5t o16.0% . Althougha sa matte ro ffac ti t isquit e possible tolowe rth edr ymatte rconten t still furtherw e expecttha t increasingth edr ymatte rconten tabov e16-20 %wil lno t be easy asth eseparatio n ofth emil k from theresidu e will then

45 becomemor edifficult .I tshoul d benote dtha t thedr ymatte r con­ tent tends to increase faster than the extraction ratio. This is duet oth e facttha tth eseparatio no fth esedimen t bycentrifugin g is effected less efficiently as theviscosit y of the flour-water mixture increases.A s therati o increased themil kyiel d aswei ra s theyiel d ofdr ymatte rdecreased . Whileth epercentag eo fprotei n from.thedr ymatte rremaine dnear ­ lyconstant,th eyiel d ofprotei n inth emil kdecrease d asth erati o increased. Thep H showed a decrease as the dry matter content of the milk increased. This is in accord with the pronounced increase in the acidity. Evidently theviscosit yan d thespecifi c gravity alsoin ­ creased asth edr ymatte rconten t increased. Theamoun to f residue tended todecrease . While thedr y matter contento f theresidu e increasedwit h theratio ,th epercentag e of crudeprotei nfro mth edr ymatte r remainednearl yconstant . Therefore asth eextractio nrati o increased, more drymatte ran d crudeprotei nwer e found inth eresidue .

The extraction pH alsoha d an important influence onth eyield . Thedr ymatte rconten tappeare dt oincreas e ifth ep Hwa sincreased . Thisincrease,however ,wa sno tver yrapid ,an dtherefor e theyield s ofdr ymatte r- th emil kyiel d beingabou tconstan t -increase don ­ lyver y slowly. Incontras tt othi sbehaviour ,th epercentag eo fprotei nfro mth e drymatter ,showe d arapi d increase.Consequently , andals odu et o the increasing drymatte r content of themilk , theyield s ofpro ­ tein inth emil k increased appreciablywit h higherpH . Theincreas e inp Hwa saccompanie d bya decreas ei nacidity,whil e theviscosit y and thespecifi c gravity increased invaiue .Probabl y the increase inviscosit y was notonl y caused by the increase in drymatter . Withth e sediments the decreasing drymatte r content and thede ­ creasingpercentag eo fprotei no f thedr y matter contributed toa decrease inth eyield so fdr ymatte r andcrud eprotei n inth esedi ­ ment.

Inth enex t serieso fexperiment s on the extraction temperature nopronounce ddifference sbetwee nth eresult swer e found.A thighe r temperatures therewa sonl y a slightly higheryiel d ofproteins .

With the exception of the first experiment the extraction time proved tohav e practically no influence on thecompositio n of the milkan dth eyield.A nextractio ntim eo f2f tminutes ,however ,prove d

46 to be too short asbot h themil kyiel d and thedr ymatte r content of themil kwer eHighe rwit h longer extractiontimes .

Also inth eserie so fexperiment so f the method of mixing nosig ­ nificant differences inresult s were found either. However, when using the IkaUltraturra x thedr ymatte r content of themil k was appreciably higher thanwhe n using the Vibromixer.A s inthi sse ­ ries the protein contentwa sa t the same level aswhe n using the Vibromixer, theyiel d ofprotei nwa scomparativel y higher.Evident ­ ly the acidity, specific gravity, andviscosit y were thenhigher . Now we shall compare the results of our studies on the second stage, the extraction, with those of otherworkers .

NAGEL et al. (1938)studie d the amount ofnitroge n extracted fromdefatte d soyflourwit h different amounts of water, theyobtai ­ ned their ultimate results by triple extraction. For the sake of comparison, however,w e areonl y interested inth e results of the first extractions. They used amounts of water corresponding with about 10,5 , 2'/2,114 ,an d 0.6%flake san dobtaine d 59.4,73.6 ,81.0 , 80.8,an d 82.7%o fnitroge n for.thefirs textractions . ..Theiryiel d with 5%flakes ,whic h isabou t thesam e as5 0m l of water,wa s 73.6%.Notwithstandin g thedifferenc e intechniqu e used, this resultclosel yagree swit htha tobtaine d byu s (71.7%). Although they used anextractio n timemuc h longer than ours, it is probable that themechanica l shaker used by theseworker s had not such amixin g intensity asou rVibromixer . Thisma yaccoun t forth e facttha t ina concentrate d mixture theiryiel ddecrease d more than ours. As far back as 1934CASTAGN0 L investigated the influence of the acidity onth eyield so fproteins .Fo rp H8 , 7,6 , and5 yield s of 17.8,18.4 ,18.0 ,an d 16.4%respectivel y were obtained.Hi scon ­ clusionwa stha t inthi srang e thep Hha slittl e influence uponth e extraction. However, invie w of the fact that theseprotei n yield figures are very lowhi sextractio n process doesno t seem to have beencarrie dou tunde rgoo dconditions . Accordingt oou r figuresi n thep Hrang e6i 5t p8. 5 there isalread ya differenc e ofabou t 10%. We shall therefore consider the work of SMITH and C IRCLE (1938)wh o investigated the extraction ofnitrogenou s material by acidsan d baseswit han dwithou tsalts .Thes eworker sobtaine dmuc h betterresult stha nCastagnol . Itwa s found that the extractionwit h water gaveonl y from 84.8 to85.5 %o f the totalnitrogen .Wit h Ca(0H)2 theamoun to fdisper ­ sednitroge nis89.2 %a tp H9. 8an d forNaO H thisis91.5 %a tp H8.3 . These figures aresubstantiall y higher than ours. Itshould , how-

47 ever,b ebotn ei nmin d that thedifferenc e inyiel d isdu eno tonl y to adifferenc e in extraction method, but also toa difference in themetho d of calculation. Theeffec to f the temperature onth e amounto fnitroge n dispersed wasals o investigated byNAGEL , BECKEL, andMILNER (1938). At 1.5° and45° Cthe,percentag eo fnitroge n extracted was 71.4an d 83.1%. They found that in the range 15° to 35° the variation is about 0.25% of nitrogen per degree centigrade. As has been men­ tioned before, inou rwor k no clear differences could be found in the range 20°t o50°C .

Thesam eworker shav eals ostudie d the:effec to fvariatio n in ex­ traction time on the amount ofnitroge n extracted. For the first extractions 1,7.5 , 15,30 ,60 ,120 , 180,an d24 0minute sgav eyield s of71.9 ,75.6 ,78.0 ,81.1 ,80.7 ,82.3 ,83.6 ,an d 83.9%o f totalni ­ trogen. In general this isi naccor dwit h ourresults .

After thisdiscussio n the question ariseswhethe r itmigh t still bepossibl e toobtai nhighe ryield sb y othermean s thanthos e in­ vestigated.A sha sbee npointe d out earlier inthi s discussion the figures for thedr ymatte rconten t of themil k were comparatively highertha nthos e for theyield so fdr ymatter .W e have argued that thiswa s caused by the retention of the milk, and consequently of drymatter , inth eresidue -I f itwer epossibl e toobtai n thisdr y matter inth e milk,th eyiel d would increase appreciably.The nth e yield would increaset oa theoretica l level thatcoul db ecalculate d fromth eextractio n ratioan d thedr ymatte r content ofth emil k by meanso fth eratio .

drymatte r found (%) . 100%= (theoretical)tota lyield . extraction ratio(% )

Thatthi s istru ew e shalldemonstrat e bymean s oftripl eextrac ­ tion. Prom soaked soyflourmil k isprepare d with the Braunmixer (ratio 10%,20°C , pH6.5 , 10min. )The n the residue is again ex­ tracted, butthi stim eb ymean so fth eVibromixe rt oavoi da possi ­ ble grinding effect.Th esecon d residue isals oextracte d with& L water.Pro m thedr ymatte robtaine d ineac h of the threemilk s the individualpercentag e ofdr ymatte r iscalculate d and after adding up these threeyield s the totalyiel d becomes 78%o f the original drymatter . The facttha tth etheoretica lvalu e (8.16: 10).100 % = 81,6%i sno treached ,ca n beascribe d to losso fmaterial .

48 Triple extraction of soyflow Triple ex traction of soybeans dry- yield yield of dry yield yield of matter of dry crude matter of dry crude (%) matter protein (%) matter protein 1st extr. 8.16 59.7 71.5 7.82 53.6 64.7 2nd extr, 1,50 15.6 16.6 1.63 15.9 17.9 3rd extr. 0.29 2.6 3.6 0.39 3.6 4.7 77.9 % 91.7 % .73. IS 87.3%

Ifth esam e isdon ewit hth esoaked , whole beansw ese etha tth e theoretical valuei sno treache deither .Th edifferenc ebetwee nth e theoretical valuean dth eyiel dobtaine d isabou t5% . Neverthelessw ebeliev etha t the figures for the dry matter con­ tent of the milk from the first extraction may be considered as a measure for the total yield of dry matter. Thisimplief cthatb y washingth eresidu etwic ei ti spossibl et oincreas eth eyiel dt oa muchhighe rlevel . Promth e tablei tappears.tha tth eprotei nyiel d figuresar ehighe r than those forth edr ymatter . Thisi smainl ydu et oa differenc e inth efirs textraction .Whe nconsiderin gth efigure sfo rtotal'pro- - teinyield ,w ese etha t theseyield s amountt oabou t 90%.Fo rth e flourth eyiel d ishighe r thanfo rth ewhol e beans.Estimatin gth e lossesdurin gth eextractio no fth eflou ra tabou t3% ,th etheore ­ tical totalyiel do fth eprotein si sabou t95% . Thismean stha tth e proteinca nb eeasil yextracte dnearl yquantitatively .A si ti sou r aimt oobtai na proteinric hfoo db yselectiv e extraction, grinding shouldb ecarrie dbu tt osuc ha degree ,tha tnearl yal lo fth eprotei n is dispersed. Afterhavin gdiscusse d theresult so fou rexperiments ,w earriv e atth efollowin gconclusions .

1.Th eextractio nexperiment so nth emanufactur eo fsoymil kar ere ­ producible.Extractio no fsoake d flour under certain standard conditions gavemil k containing 7.6-7.9%dr ymatter , abouthal f ofwhic hwa scrud e protein.Th eyield so fdr ymatte ran dcrud e proteini nth emil kwer e about59. 0an d68.Of trespectively . Thus theprotein sprove dt ob ebette rdisperse d thanth eres tof-th e drymatter .

2. Extractiono fsoake d flouryielde d moremil k thanextractio no f beans.Wit hth eflou ra swel la swit hth ebean sbette r grinding yielded moremilk ,an d increasedth edr ymatte ran dcrude ,protei n contento fth emilk .Thu sth eyield,o fdr ymatte r andcrud epro -

49 tein increasedo nbette r grinding.

3 Theinfluenc eo fth eextractio n ratioi simportant .I fth erati o increased, thedr ymatte ran dcrud eprotei nconten to fth emil k also increased. Theyield so fbot h decreased, however.

4. Theextractio np Hals oha da nimportan t influence.However ,i f thep Hwa s increased thedr ymatte r contento fth e milkdi dno t increasea tth esam e ratea sth ecrud e protein content.A tp H 9.5th eyiel do fcrud e proteinwa s

5.The extractiontemperatur eha dlittl eeffec ti nth erang e 20°t o 50°C

6. Aftera certai nminimu m (inou r case about 2%minutes )th eex ­ traction timeha dlittl e influenceo nth eyield .

7.Th erotatio nvelocit yo fth e IkaUltraturra xha dpracticall yn o effecto nth e extraction. However,th econtent san dyield so f drymatte ran dcrud eprotei nwer ehighe rtha nwhe n using theVi - bromixer.

8.B ywashin gth eresidu e twiceth etota l yieldso fdr ymatte ran d crudeprotei n increased considerablyan dapproache da theoreti ­ calvalu etha t-fo rth edr ymatter -ca nb ederive d fromth eper ­ centageo fdr ymatte ro fth emil kan dth eextractio n ratio.Th e totalyiel do fcrud eprotei nobtaine dwa sabou t 90%; theestima ­ tedtheoretica lvalu ei sabou t 95%.

Therecommendation s resulting.fromthi sstud yo nth e manufacture ofsoymil kp na technica lscal ear e giveni nth e lastchapter ,th e general discussion.

3.4 SUMMARY

Although long before the last war articles on soymilk appeared in medical journals* little-work had been done on the technological side. In particular the data on the yield of the extraction were scarce antiif available not very reliable. Therefore a new inves­ tigation of the yield of manufacturing soymilk was started. The many methods of manufacture have four distinctly different stages in common: 1- the pretreatment, . 2. the extraction, 3. the separation, 4,the improvement of the quality. As the way in which the first stage, the pretreatment, is carried

50 outdetermine st oa larg eexten tth ecompositio no fth emilk ,thi s stagewa sstudie do na laborator yscal ei norde rt oge ta ninsigh t inth enatur eo fth eapparatu stha tshoul db eapplie do na techni ­ calscale .Fo rthi spurpos esevera ltype so fapparatuse swer etes ­ ted,usin gon eextractio nmetho dan don eseparatio nmethod ,Whol e beansanijals owhol eflou rwer euse da sstartin gmaterial . Thesecon dstage ,th eextraction ,wa sstudie dt ofin dou twha t factorsmigh tb eadapte dt oge tbette ryield swit ha give nappara ­ tus.Th einfluenc eo ffiv efactors ,th eextractio nratio ,pH jtem ­ perature,time ,an dmetho do fmixin gwer e investigated.Soyflou r obtainedwit hon ean dth esam emetho do fpretreatmen twa sused .

4 PRESERVATION OP SOYMILK BY HEAT APPLICATION (STERILIZATION)

4.i Introduction. t

Soymilki smainl ycompose do fproteins,fats ,an dsugars,Therefor e thisproduc tform sa nidea lsubstrat efo rth egrowt ho fmicro-orga ­ nisms.Whenra wsoymil ki slef ta troo mtemperatur ei tca nb ereadil y observedtha tafte ra shorte ro rlonge rtim esymptom so fdeca ywil l appear.Thi si sespeciall yeviden ta thighe rtemperatures . Inth emanufactur ean ddistributio no fsoymil ksuc hsymptom sca n notb eallowe dt odevelo pa swel la si nan yothe rfiel do fth efoo d industry.Forthi sreaso nw ehav estudie dth epreservatio nbehaviou r duringstorag eo fra wsoymil kafte rapplicatio no fheat .

4#2 Literature.

CHIU (1927)state stha tth egenera ltype so forganism stha tde ­ velopi nmil kar eaci dformers ,ga sproducers ,an dputrefiers .I n soymilkther ear eorganism sactin go nsuga rproducin galcoho lan d organicacids, ,thos eactin go nprotein sproducin gcurdlin go fmil k andevi lsmellin gcompounds ,an dthos eactin go nth efat .Th eaci ­ dityo fth emil kwa sfoun dt oincreas eeac hday.unti lth eeight h daywhe nth eamoun to faci dproduce dchecke dth egrowt ho fth eaci d formers.Carbo ndioxid ewa sfoun ddu et oth ega sproducer sactin g onth esugar .Ammoni awa sals ofoun ddu et oth eactio no fth eor ­ ganismson'th eprotein .Eac horganis mha sa noptimu mtemperature . Thesymptom so fdeca yar edescribe db yMONNIE R (1935). Accordingt othi sworke rafte rsom ehour sth era wsoymil kseparate s intotw odistinc tparts :th elowe rpar ti swhitis han dresemble s curd,whil eth euppe rpar ti sa transparan tgreenish-yello wliquid .

51 By boiling forhal fa nhou r thecolou r becomes alittl e yellower, while thebeanis h odour isreplace d bya nagreeabl e flavour.Th e boiled milk ismuc h stabler thanth era wmil kan dma yb epreserve d for 15t o2 0hour swithou t "turning". After this period, however, the milk losesit shomogeneit yan ddecays .Contrar yt oth era wmil k this decay resultsi ncur d particles floating inth ewhey . The preservationb ymean so fchemical swa sstudied .Sodiu m bicar­ bonate (63g pe rliter )ha sgive nth ebes t results;th eflavou ro f the milkwa simprove d byth eadditio n ofthi s compound; themil k was preserved formor etha n5 4hours . Benzoic acid, salicilyc acid and boric acidar eimpracticabl e asthe ycaus eth eprotein st opre ­ cipitatea tth eantisepti c dose. As described inth echapte r onth emanufactur e ofsoymil k CAS- TAGNOL(1934 )ha sals o givena descriptio no fth efermentatio no f the milk.

4.3 EXPERIMENTAL PART

U.3.1 Scope of work.

Promth einformatio n available itappear s that effortshav e been made toimprov eth ekeepin g qualitieso fth emilk . Thiswa smainl y done eitherb yth eadditio no fchemical so rth eapplicatio no fheat . As, however, inou ropinio n theadditio no fchemical s shouldb e avoided because.inpractic e awron g application mayeasil y cause injuryt ohuma nhealth,w econfine d ourselvest oa stud yo fth e pre­ servationo fsoymil kb yheat .

4.3.2 Desiffi of experiments..

Before carryingou tth esterilizatio n experiments preliminaryex ­ perimentso npasteurizatio ni ncombinatio nwit h coolingwer ese t up. Raw soymilk intes t tubeswa spasteurize d at60° Cfo r3 0minutes . Some ofth e tubeswer e storeda t20°C ,whil eother swer e storeda t 25°C.A numbe ro ftube swer e immediately cooled afterpasteurizati ­ on.Th ep Hwa sdetermine d anda plat e countwa smade . Forth esterilizatio n experiments ofou rstandar d milko fabou t 8%dr ymatte r threeserie so fhea t treatmentswer e designed, using 101°C, 110°C,an d120° Ca ssterilizatio n temperaturesan d5 ,10 ,an d 15minute sa sholdin g timesfo reac hsterilizatio n temperature,Fo r eachhea ttreatmen tnin ecan swer esterilize dan dthe n storeda t30°C . After one,six ,an d twelvemonth so fstorag e each time three cans were examined visuallya sregard sth econditio n ofth eca nan dth e

52 milk,thep Han dacidit ywer edetermined ,an da plat ecoun twa smad e ofeac hmilk , A lastgrou po fsterilization swa scarrie dou twit hmil ko fabou t 12, 16an d24 %dr ymatter .Thes emilk swer efirs tpreheate dwit hth e continuous sterilisator at80° C for4 5se c and then evaporated witha Luw a filmevaporato ra sdescribe d inth efollowin g chapter on evaporated milk.Th esterilizatio ni ncan swa scarrie dou twit h a numbero ftemperature-tim e combinations. Immediately aftersteri ­ lizationth ecan swer eopene dt ose ewhethe rth emil kha dwithstoo d thetemperature-tim ecombinatio napplied .A numbe ro fcan swer e left atroo mtemperatur e (20°C)t ose ewhethe r theywoul dspoil .

4.3.3 Technique of;experiments.

The milkwa sprepare db yextractin g soaked beanswit ha nUltra - turraxi na 1 : 1 0ratio .Thi smetho dwa sfoun dt ob eth emos tcon ­ veniento napilotplan tscale.Afte rextractio nth eliqui dwa scoole d toabou t5° Cb ymean so facoolin gspira lthroug hwhic hcoole d brine of somedegree sbelo wzer ocirculated . Thenth ecoole d liquidwa s filteredwit ha baske tcentrifuge . Themil kwa ssterilize di ncan so f0.3 5liter .Fo reac hsterili ­ zationnin ecan swer epu ti nth esterilizato ran dals oon eca npro ­ vided witha nopenin gwhic hwa sclose db ya rubbe r stop. Through thissto pa thermocoupl ewa splace d insuc ha wa ytha tth een dwa s inth ecentr eo fth ecan .Th etemperatur ea tthi splac ewa sregis ­ tered bya Brow npotentiometer . Thestea mwa slea d intoth esterilizato ri nsuc hawa y thati ntw o minutesth erequire dtemperatur ewa sreached,afte rwhic hth etempe ­ raturewa skep t constantfo rth edesire d holding time.The nth e steam supplywas .cu tof fan dimmediatel ywate rwa sfe dint oth este ­ rilizator.Durin gth esterilizatio nproces sth etemperatur ean dtim e wereobserved ,fo rwhic hth eaverag ecurve sar eshow ni nfig .5 . After sterilizationth ecan swer e labelledan dstore da t30°C . One,si xan dtwelv emonth slate rfro meac hhea ttreatmen tthre ecan s wereexamine d forth evisua lcondition ,pH ,acidity ,an dplat ecount . The tinsma yb ei ngoo d condition (+),a littl e blownu p(-) ,o r strongly blownu p(x) .Fo rth emil kth esame.symbol swer e usedt o designatetha ti twa sstil li na goo dconditio nvisuall y (+),floc ­ culated butstil l milklike (-),o rtha ta curd -an dwhe y forming had takenplac e (x). ThepHan d theacidit ywer edetermine da sdescribe di nth eforme r chapter, whileth eplat ecoun twa scarrie dou ta sfollows .On em l of milk isdilute d with9 m lsterilized ,distille d watert o0.1 ; alsoadilutio no f0.0 1an d0.00 1ar emade.Fro meac ho fthes edilute d

53 milks 1m li smixe d with bouillonagar substrate inPetri-dishes , afterwhic hth eplate sar estore da t37° Cfo rsi xdays .Th eplate s were prepared induplicate. Thenth enumbe r ofcolonie s oneac h platewa scounted . Garewa stake nt opreven t infection.Therefor eth ecan san dals o theopene rwer e first burnedwit halcoho l before openingan dthe n the milkwa spoure d outi na sterilize d bottle. Prom this bottle the dilutions were madewit h sterilized pipetsan dplates .

4.3.4 Results, discussion, and conclusions.

Before carryingou tth esterilization s anumbe r ofpreliminar y pasteurization testswer e performed. Rawsoymil kwa spasteurize d intes ttube sa t60° Cfo r3 0minute s andthe nstore da t20 °an d25°C .Du et oth epasteurizatio nth eini ­ tialnumbe ro fbacteri adecrease d from25.000/m lt obelo w1.000/ml . During storageth ebacteri a increased markedly, thep Hdecrease d firstan dthe n increasedafte rabou ta week . Ingenera lthes eobser ­ vationsar ei naccordanc ewit hthos eo fCastagno l (1934).Ourresult s areshow ni nth e tablebelow .

Plate-countan dp Ho fstored ,pasteurize d soymilk (60°C 30rain).

Platecoun t PH

Numbero f 20°C 25°C Numbero f 20°C 25°C days days

afterpasteu ­ 935 1 6.59 6.49 rization 770 2 5.45 5.12 1 246.000 2.305.000 3 4.96 4.86 220.000 2.430.000 4 4.80 4.53 3 76.000.000 31.600.000 5 4.22 4.53 147.000.000 80.000.000 7 3.88 4.20 4 103.000.000 .318.000.00 0 8 4.05 4.25 137.000.000 1.297.000.000 9 4.25 4.22 5 557.600.000 600.000.000 10 4.35 4.21 740.000.000 1.395.500.000 11 4.42 4.40 12 4.43 4.35

If immediately after pasteurization themil kwa scooled , after threean dals oafte rsi xday sth ep Han dth ephysica lappearanc eo f themil kwer e still quite normal.However , determinations ofth e plate-count showed that-notwithstanding thelo wstorag etemperatu ­ res-a rapi ddevelopmen to fth ebacteri aha dtake nplac e (seetabl e below). 54. . Plate-count and pH of pasteurized soymilk (60°C 30 min) in cold storage.

PH plate-count 3 days 6 days 3 days 6 days

10.5°C . 6.50 6.75 14.500 5.300.000 6 °C 6.45 6.45 3.600 1.550.000 3 °C 6.50 6.55 1.985 34.000

The results ofsterilizin g 8%soymil k areshow n intabl e3 an d fig5 .A smentione d beforeth econditio no fth emil k isdesignate d as l)good (+),2)flocculate d butstil lmilklik e (-),o r3)pronounce d formationo fwhe yan d curd (x).Th econditio no fth emil k fromth e nine canso fa certai n heat treatment isrepresente d byhorizonta l lines infi g5 an di sobtaine d from table3 ,colum n "Conditiono f milk". Above those linesth etemperatur e course inth ecentr eo f theca ni sdepicted . A platewa sconsidere d sterile ifles s than5 colonie swer e pre­ sent. Then itwa sassume d thatth eplate swer e infected. Thisma y alsob esee n fromth efac ttha t thenth ecolonie swer emostl y loca­ ted inth eoute rpar to fth eplate . Belowar egive nsom eexample so fsteril eplates .

Numbero fcolonie so nplate s after6 day so fincubatio n at 37°C '^.dilution 1/10 1/1000 exp.noT^^^ 1/100 5.7 - 1 1

5.8 - •• 1 - - - 1

5.9 • - 1 2

-••• 1 - 6.4 - - - •--..• 1 - 6.5 - ,-

6.6 ' "• - '• •• - •' - •

- '• ••- -

55 TABLE 3 Condition of sterilizedsoymil kafte r 1,6 ,an d1 2 monthsstorag ea t30°C .

Exp. Heattreatmen tStorag e Platecoun t Acidity PH Condition no of can milk

1.1 101°C 5mi n 1mnt h «. 5.18 _ = 2 - - X X 3 - 4.8 X X 4 6 - - 5 X X 6 X X 7 12 X X 8 X X 9 X X

2.1 101°C1 0mi n 1mnt h - 5.31 - - ••

2 -• 5.45 - X 3 - X X 4 6 sterile 4.80 5.27 +. - 5 sterile 4.93 5.39 6 sterile 5.09 5.21 + 7 12 3.54 4.71 +. X 8 4.92 5.34 + X 9 ..... 4.76 X X 3.1 101°C1 5mi n 1mnt h - 5.31 - - 2 - 5.20 -. X 3 - 5.18 ._ - 4 6 sterile 5.26 5.39 + 5 sterile 2.78 5.37 +": X 6 sterile 2.49 5.12 X 7 12 sterile 4.07 5.65 . X 8 sterile 1.88 5.55 - X 9 sterile 4.53 5.37 -* X c 4.1 110° 5mi n 1mnt h sterile 1.65 6.61 + +. 2 sterile 4.05 6.08 + a. 3 ' sterile -. 5.01 - X 4 6 sterile- 3.12 5.65 +.: . 5 sterile 3.69 5.85 + .- 6 sterile 4.75 X X 7 12 sterile. 1.53 6.53 +. + 8 sterile 4.25 5.34 + .. 9 sterile .' •• 1.80 5.79 + X

56 Exp. Heattreatmen tStorag e Platecoun t Acidity pH Condition no of canmil k

5.1 110°C1 0mi n lmnt h sterile 1.51 6.60 + •+. 2 sterile 3.79 6.31 +. 3 • sterile 1.64 6.65 + + 4 6 sterile 1.77 6.43 + + 5 sterile 2.38 5.79 + X 6 sterile 2.06 5.79 + X 7 12 sterile 4.46 5.50 + 8 sterile 4.19 5.58 +• 9 sterile 4.00 5.51 +. 6.1 110°C1 5mi n lmnt h sterile 2.19 6.51 + +. 2 sterile 1.96 6.32 +. +. 3 sterile 1.73 6.20 + + 4 6 sterile 2.64 5.79 +. 5 sterile 2.64 6.51 + + 6 sterile 2.64 6.50 + + 7 12 sterile 1.72 5.84 +. X 8 sterile 1.42 6.56 + + 9 sterile 1.64 5.53 + X 7.1 120°c 5mi n 1mnt h sterile 1.51 6.49 + + 2 sterile 1.52 6.49 +. + 3 sterile 1.61 6.47 + +. 4 6 sterile 1.51 6.56 + + 5 sterile 1.49 6.54 + + 6 sterile 1.28 6.55 +. + 7 12 sterile 1.52 6.58 + + 8 sterile 1.44 6.5.9 + +. 9 sterile 1.52 6.58 + + 8.1 120°C1 0mi n 1mnt h sterile 1.75 6.49 •+ .' . + 2 sterile 1,56 6.69 + + 3 sterile 1.56 6.67 +: + 4 6: sterile 1.41 6.55 + + 5 sterile 1.49 .6.57. + + 6 sterile 1.49 6.56 + • + 7 12 sterile 1.51 6.59 "+. + 8 ;;steril e 1.46 6.58 + >.'"+• 9 sterile >.44 6.58 +." ' +

• -. • •' •••

57 Exp. Heattreatmen t Storage Platecoun t Acidity pH Condition no of can milk

9.1 120°C1 5mi n 1mnt h sterile 1.55 6.67 + + 2 sterile 1.63 6.63 + + 3 sterile 1.64 6.63 + + 4 6 sterile 1.46 6.55 + + 5 * sterile 1.40 6.55 + + 6 sterile 1.34 6.54 + + 7 12 sterile 1.41 6.57 + + 8 sterile 1.46 6.56 + + 9 sterile 1.45 6.56 + .+.

Milk:i ngoo dconditio n + Can:i ngoo dconditio n + Milk:flocculated ,sti>l lmilklik e- Can:blow nu p Milk:curd -an dwhe yformatio n x Can: strongly blown up x

variable coming-uptime holding cooling-time coming-uptime holding -cooling-time coming-uptime holding cooling-time time

1 2 min

i2o*cnumbe r of cans Pig 5» Sterilization of soymilk. Course of temperature in centre of cans. Condition of the milk. Dfi ngoo dconditio n ^flocculated.,stil lmilklik e J^curd-an dwhe yformatio n Itwil l be seentha tth ehea ttreatmen ta t100° Cha dn osterili ­ zingeffect ,a s using aholdin gtim eo f 51 10,or,1 5minute sonl y resultedi nspoile dmilk.I ti snoteworth ytha tafte r6 an d1 2month s

58 nomicro-organism swer e found inth espoile d milk, though there is a pronounced decrease ofth ep Ht o5. 3 and acorrespondin g increase inacidit y to3.0• -4. 5ml . Thesterilizatio na t 110°Cappear st ob ea transitio n temperature, because the number of cans of.goo d milk increases as the holding time increases.I tca nb eclearl ysee ntha tth ewel lpreserve d milk hasa p Ho f6. 5 -6.7 ,wherea sth e pHo fth espoile d milk isbelo w 6. Thusth econditio no fth emil k isexpresse d interm so facidit y aswell.Als owit hthi sserie sn obacteri awer e found inth espoile d milk. ;A11o fth e three lasthea t treatments at 120°C gave favourable results. The values of thep Han d acidity are in accordance with those of the unspoiled milk from the former threeseries . Theresult so fsterilizin gsoymil kwit ha highe rdr ymatte rcon ­ tent areshow nbelow .

Drymatte r Sterilization Physical Bacteriological :, conditions condition conditiono fmil k ofmil kafte r afteron eda y (20°C) sterilization 12.4% 120°C 5mi n normally flowing good 12.4% 120°C1 5mi n normallyflowin g good 16.4% 120°C1 5mi n gelatinized^ 16.4 % 120°C 5mi n gelatinized1 partly 16.4 % 110°C 5mi n gelatinizedf spoiled 16.4% 100°C 5mi n gelatinizedJ 24.3 % 120°C 5mi n gelatinized spoiled

From this table itwil l be seen that 12.4% milk can be safely sterilized at120° C 15min ;th econditio n ofth emil k then remains good, while the heating isals o bacteriologically effective. The 16.4 %milk , however,di d notwithstan d thesterilizatio n neither at 120°C1 5mi nno ra t 120°C5 min.Eve nsterilizatio na t 100°C5 mi n made a gel from themilk . This isno t inaccordanc e withth e heat stability determination made on thismilk . However, itmigh t be possible thatth e useo f arotatin g sterilizatorwoul d have given more favourableresults . Iti simportant.t onot etha tfo rthi smil k sterilized at 120°C 5 min and even 15mi n was not sufficient to preventspoilage .Thi smean stha ta mor esever eheat'treatmen tmus t be applied when the drymatte r content of the milk israised . In view ofou r experience with the 16.4% milk it isno t surprising thatth e 24.3 %mil kdi d notwithstan d thesterilizatio nan d that spoilage occurred atroo m temperature.

59 The followingconclusion sca nb edrawn : 1.Pasteurizatio no fra wsoymil ka t60° C3 0mi nconsiderabl y reduces theinitia lnumbe ro fbacteria ,bu tth emil kwil lno tkee peve non e dayi flef ta troo mtemperatur e (20-25°C). Pasteurization incombi ­ nationwit hcol d storagea t3-6° Cma y preserve themil k forabou t 3 days. 2. Sterilization of8 % soymilk at120° C5 mi n issufficien tt o preserve themil k fora tleas ton eyear , 3. Ifth edr ymatte rconten to fth emil k israise d above 12% th e temperature-time combinationregarde da ssufficien tfo ra n8 % mil k willno tsuffic ea sa sterilizatio ntreatment .

• • '" / 4.4 SUMMARY

As soymilk is mainly composed of proteins, fats, and sugars it forms a good substrate for micro-organisms. In literature some data can be found indicating that the keeping qualities can be improved by application of heat. It is also known that attempts have been made to preserve the milk by chemical means. As in our opinion such methods might endanger human health,we only tried to obtain better keeping qualities by means of pasteurization and sterilization. To this end a number of experiments were carried out. After pre­ liminary pasteurization experiments -some of which combined with cold storage- nine temperature-time combinations for sterilization purposes were studied as to their influence on the keeping quali­ ties of the.soymilk after one,six,and twelve months storage at 30°C. A number of tentative sterilizations of milks with higher dry matter contents, obtained by evaporation, were also carried out.

5 EVAPORATED SOYMILK

5.1 Introduction.

Inchapte r3 w ehav e seentha t-i fth esam emetho do fmanufactur e isapplied -a produc to fconstan t compositionca nb eobtained .W e haveals odemonstrate d thatth edr ymatte r contento fth emil kca n bevarie dwithi nrathe rwid e limitsb ychangin gth eextractio ncon ­ ditions,especiall yth eextractio nratio . Ifi ti sdesire d tomanufactur e soymilk witha hig hdr ymatte r content therear ei nprincipl etw opossibilities.I nth efirs tplac e fresh milk witha lo wdr ymatte r contentca nb e concentratedb y evaporation,an dsecondl yon eca ntr yt omak edirectl ya mil kwit h a highdr ymatte r content byusin g lesswate r forth eextractio n

60 process. However,w ehav e showntha twit ha on estag e processth e output stronglydecrease sa sth edr ymatte rconten to fth emil kin ­ creases.Th eoutpu t becomes particularly lowwhe non eaime sa ta drymatte rconten to f20 %o rmore .Beside s otherdifficultie sarise . Thusi fmil ko fmor e than20 %i st ob emanufactured , concentration by evaporation isinevitable . Thenth einitia ldr ymatte r contentt ob echose ndepend s uponth e total costprice ofth eevaporate d milk. Starting witha ver ylo w dry matter content theoutpu t ishigh , butmuc h waterha st ob e evaporated. Ifmil kwit ha highe rdr ymatte rconten tha st oh econ ­ centrated lesswate r needst ob eevaporated , butth eoutpu to fth e extractiondiminishes .Th eoptimal ,tota l costprice mustli esome ­ where between thesetw opossibilities ;thi soptimum^ca nonl yb ecal ­ culated inconcret e cases.W eestimat e thati tlie sa t about8 %dr y matter. Thereforew estarte dth emanufactur e ofevaporate d soymilk

fromra wmil ko fabou t 8?dr0 ymatte rcontent . Concentrating soymilkb ymean so fevaporator s isno ta nunknow n process. However, some difficultiesma yaris e duringth eproces s which require closer investigation. Therear eindication s inlite ­ rature thatth eviscosit yo fth eevaporate d productbecome ss ohig h at arelativel y lowdr ymatte r content that the milk cannot flow any more.Thi sphenomeno n seemst ob eespeciall y evidentwhe nth e concentrated milk iscooled ; themil k then seems more orles s to solidify. Iti sclea rtha tthi sma y leadt oseriou s troubles during evaporationan di nfurthe rprocessing . Another point ofconsideratio n concernsth efurthe rus eo fth e evaporated milk.Evidentl y thisshoul db epreserve d insom ewa yt o obtain reasonablekeepin gqualities . Ifpreservatio nb ysteriliza ­ tioni sapplied.th emil k shouldb ei nsuc ha conditio n thati tca n withstand thesterilizatio n conditions. Ifpreservatio n ist ob e obtainedb yadditio no fsuga rth eviscosit yo fth econcentrate dmil k shouldno tb es ohig htha tthi sadditio n isno tpossible .I fdryin g to milkpowderi schose n forpreservatio na no tto ohig h viscosity isals o necessary. From theforegoin g itappear s thato nconcentratin g soymilkw e shouldb eintereste d particularly inth eviscosit yan dth eheatsta - bilitya sa functio no fmanufacturin g conditions.I ti sals o impor* tantt oinvestigat et owhic hdr ymatte rconten t soymilkca nb eeva ­ porated while retainingacceptabl e properties.Th ewor k concerning these factorsi sdescribe d inthi schapter .

5,2 Literature.

The literature onevaporate d soymilk islimited; .Probabl yth e

61 firstreferenc et othi ssubjec tcome s fromK A TA YA MA (1906)wh o observed.that ondirec t evaporation of soymilk after addition of sugar theprotei nseparate d intosmal l floccules. Thereforedipo - tassiumphosphatewas -adde d topreven t this separation.Th eauthor ' states thatth e solutioncoul d beconcentrate d invacu ot oa ver y thick liquid. Unfortunately thedr ymatte rconten t reachedwa sno t mentioned. Shortlyafte rthi scommunicatio nastudygrou pwa s founded inPranc e (1910)wit hth eai mt oproduc esoymilk , evaporated soymilkandsoy - milkpowder, but these effortswer e abandoned for unknownreasons . On theothe rhan d iti sknow n that inFrankfur tth e factorySoyam a succeeded tomanufactur e freshan d concentratedsoymil k (Inst.Int. d*Agric , 1936). Also inRussi aattentio nwa spai d toth econcentratin go fsoymilk . HOROWITZ- WLASSOWA, 0B ER HAR D, an dG U TE RM AN N (1931) inRussi aconcentrate dsoymilk ,afte radditio no f 19%sugar ,t oon e fourth ofth e originalvolume .Th e composition of the evaporated soymilkwa s 11.76%protein ,7.64 %fat ,42.12 %sugar ,2.32 %ash ,an d 37.16%water . TheCENTRA L TECHNICAL INSTITUTE of TNO Netherlands (1955)carrie d outresearc ho nth eviscosit yo fsoymil kdurin gcon ­ centration, the choice of the best type of evaporator, and the highest concentration attainable in the evaporator. Themil kwa sprepare d byextractio no f 6k go fsoybean s and 1.5 kg of peanuts with 63 kg ofwater .Thu s 74%o fth e original dry matter, 79%o fth eprotein ,an d 75%o fth eoi lwer eextracte d from thebeans . Inth e course ofth e preparation of soymilk the following heat treatmentswer eapplied .Afte rsoakin g the beanswer ekep t in70° C water for 12minutes .Th e extractionwa s carried out atth e same temperature for 12minute safte rwhic h themixtur ewa sboile d for 20minutes .Th e following filtering of the milkwa s done at 70°C during 12minutes ,afte rwhic hth emil kwa scoole d to5°C . Thenth emil kwa sconcentrate d ina Lurg ilaborator y flashevapo ­ ratorwit h forcedcirculation .Afte r 70minute s20 %dr y matterwa s attained, but itwa sno tpossibl et o increaseth epercentag eo fdr y matter further. The milk then had become so thick that the pump could notcirculat e itthroug hth eevaporato r anymore . With a KurtHerber t evaporator themil k could easily beconcen ­ trated to 20%dr ymatter .Thi styp eo fevaporato r hasa natura lcir ­ culation systemo f the liquid. However, ifth e concentrating was continued to24 %dr ymatter ,th emil k became age lan dclogge dth e pipeso fth eevaporator . Witha Luw aevaporato r itwa sals ofoun d thatth emaxima lconcen -

62 trationpossibl ewa sonl y20%.I nthi styp e ofevaporato r theliqui d flows down avertica l pipe. A rotor forces the liquid to form a thin film against the inside oftri epipe .Th eholdin g time isver y short incompariso nwit hth etw oothe r typeso fevaporators . Measurements ofth eviscosit yo fth eorigina lmil kshowe d thataf ­ terfiv eday sa t5° Cth eviscosit ywa spracticall y thesame .Concen ­ tration to16 %dr ymatte rdi dno tsho wa nimportan t increase invis ­ cosity, and at 18%th emil kstil l flewver ywell .A t20% ,however , the milkwa sver y viscous, but itstil l flew ifth e milkwa s kept inconstan tmovement.I fno tkep tmovin gthe nmil ko f 18%solidifie d inon ehou ran dmil ko f20 %i nhal f-^ nhour .

Asregard sth equestio nwhethe ri ti spossibl et oconcentrat e soy- milkt oa reasonabl e percentage ofdr ymatte r -e.g.30% -w ema yse e fromliteratur e thatver y little information isavailable .However , even in the few publications on this matter the conclusions are very contradictory. Horowitz-Wlassowa showed the possibility of concentratingsoymiIk,wherea sTN O foundthat20 %dr ymatte r should be considered asa limitt oevaporatio n duet o thehig h viscosity and the incrustation. Inth e literature, nodat a can be found on the second factor, the heat stability. Asver y littleo rnothin gi sknow nabou t these two characteristics of soymilk and besides little can be said about the theoretical background ofth eviscosit y andhea tstability , itwa sfoun ddesi ­ rable to examine the literature on animal milk. Although we are quite aware ofth e fact that the animal and vegetablemil k differ in composition iti sprobabl e that some linesma y be found along which a research scheme may bedeveloped .

The literatureo nthi ssubjec t isver yextensiv ean ddate sbac kt o 1919.Fromthi slarg enumbe ro fpublication sw ewil lmentio nonl ya few . Therelationships ,o fconcentratio nan d time toth etemperatur eo f coagulationwer estudie db yHOLM , DEYSHER. andEVAN S (1923). They found adecreas e ofabou t 1.50°Cfo rever y 1percen t increase inconcentratio n ofth emilk .Betwee nte nan d sixtyminute sth e re­ lation betweenth etemperatur et oth etim eo fcoagulatio nprove d to approximate very closely a logarithmic relation with respect to time. Milk of inferiorqualit y had a decidedly lower coagulating temperature butth erelationship so fconcentratio nan d temperature seemed tob eth esam ea sfo rmil ko fgoo dquality. " Inth epape ro fDEYSHER , WEBB-, andHOL M (1929)i ti sshow n thatth emetho d ofpreheatin gth era wmil kprio rt oevaporatio nin ­ fluences thehea tstabilit y ofth eevaporate d milk.Temperature su p to 70°C, applied for tenminutes ,decreas e the stability,whil e

63 higher temperatures markedly increase it.Homogenizatio n of the evaporated milka tpressure su p to400 0pound s per square inchaf ­ fects itshea t stabilitybu t slightly. Theseworker spoin t outtha t body of the evaporated milk should beo fa cream yconsistency . Theirexperiment s indicate thatth econ ­ sistencyattainabl ebefor ecoagulatio nbegin san dth ehea t stability are of a reciprocal nature. Those products of greatest stability had the lowest relative consistency atth e time ofcoagulation . So milkso fexceedingl y highhea tstabilit y areno tth emos t desirable fromth e standpointo fproductio no f body. ',Du et oth e observation that injection ofhig hvelocit y steam ina stream ofmil k caused stabilization to the heat of sterilization WEBB andBEL L (1942)designe d experimentst ostud y the influence of rapid heating onth e heat stability of evaporated milk of 18% solidsno t fat.Fo rth e greaterpar tth emil kwa sheate d toth ede ­ sired temperature in5 seconds,hel d2 5second san dcoole d toa tem ­ perature of lesstha n38° Ci n4 seconds.Th e control samples were forewarmed at95° C1 0mi ni na steamjackete dkettle ,whil e thetes t sampleswer eH T ST heated overa rang eo ftemperature s from 101°C to165°C . Itwa s found that thehea t stability ofevaporate d whole milk of 26%tota l solids was generally twice and occasionally six times greater thantha to f thecontro l samples.Th e relationship between thehig hforewarmin g temperature and thehea tstabilit y ofth eeva ­ porated productdiffer swit heach "milk . Thehig hforewarnin gtemperatur erequire d toproduc e anevapora ­ ted milk ofth e desired viscosity may bewithi n the limits of 2°C for one milk orwithi n limits aswid e as 60°C for another milk. Milks forewarmed toproduc e excessive high stability will be too thin,whil e thosewit hto olo wstabilit ywil l beroug h aftersteri - 1ization. As the increase of the heat stability of concentrated milk by HTST forewarming maymak epossibl eth e preparationan d steriliza­ tionwithou t coagulation ofmilk s ofsolid s concentrations greater than 26%WEBB, . BELL, DEYSHER, and HOLM (1943)wer e thus able to investigate the effect of various degrees of forewarming uponth ehea tstabilit y of milkso fdifferen t solidsconcentrations . Inth e range of solids concentrations investigated, 26t oabou t 33%, thestabilit y increasesa sth eholdin g time is lengthened for temperaturesabov e 100°Ct o 150°C. Onlywhe n theHTS T forewarming temperaturewa s lowo r excessively!hig ho rwhe n the holdingperio d wasto o short, lessstabilit ywa s observed inth e test than inth e control samples. These control samples were heated in a steam- jacketed vessel at 95°C for 10minutes .

64 Several experimentswer eals o conducted todetermin e whether the stabilizing effectwa sderive d fromth eheatin gtemperatur e or from the rapid heating. Itwa s found that rapid heating increased the stability ofmos tmilk sabov etha timparte dt osample sheate d slowly. They alsoobserve d that the generalhea tstabilit y relationships ofmilk ssubjecte d todifferen t forewarming treatmentswer e notaf ­ fected by thedevelopmen t of acidity inth e fresh milk or by the additiono fstabilizin g saltst oconcentrate d milks;th e difference instabilit y caused byaci do rsalt swa son eo fdegree . Thehea ttreatmen t canals ob eapplie d afterconcentration . WEBB and BELL (1943)heate d the milk-before and after concentration. HTST heatingo fth econcentrate d milk increased itsheat,stabili ­ tyt oa maximu man d thendecrease d it.Withn oholdin gtim eth esta ­ bility increased witha risei ntemperatur e upt oth epoin t of ini­ tialcoagulation . Factorswhic hcontrolle d theattainmen t ofmaxi ­ mum heatstabilit ywer eth etim ean d temperature of forewarming the rawan d heatingth econcentrate d milkan d thesolid sconten t ofth e concentrate. The greatest heat stability wasobtaine d when the raw milkwa s heated at120^ Cfo r4 minute san d theconcentrat e to 150°C withn oholdin gperiod . The authors point outtha tth e most desirable heat treatment is not always the one that givesmaximu m stability, since the effect on the body and the colour must be considered. Since thehea t stabilitiesan dconcentration s ofevaporate d milks could be greatly increased by the use of high temperatures, this provided a meanst oprepar esample swit ha wid e range ofconcentra ­ tionsan d stabilities. Inthi swa y itwa spossibl e for DEYSHER, WEBB, and HOLM (1944)t ostud yth eviscosit ybehaviou r ofeva ­ porated,milk s during sterilization and storage as influenced by HTST heatingbefor e evaporation.Thei rresult sindicat eho w rapidly the viscosity ofevaporate d milks increased just before and after coagulation. Besides the body of themil kwa s shown to be greatly influenced byth ehea tstabilit y ofth e milk.Areciproca l relation­ ship between thehea tstabilit y and theviscosit y aftersteriliza ­ tionwa s noted. The highest viscosities were produced inmilk s of highconcentratio n and low stability. From their data it appears that the method of forewarming also influencesth e viscosity behaviour ofth e concentratedmiT Rdurin g storage. The viscosities from evaporated milks from 26t o 36 per­ cent solids concentrations followed awell-define d pattern during processingan d storage.Thickenin g occurred during thesteriliza ­ tion process. This was followed bya losso f body and,a thinning early in the storage period. The low storage viscosity was main­ tained for variousan d unpredictable periods oftim e duringwhic h

65 fatseparatio noccurred . Late inth estorag e period some evaporated milks, especially those receiving light heat treatments and'those witha highconcentratio n ofsolids,bega nt osho w increases invis ­ cositieseve nt oth epoin to fgelation . Another exampleo fth e influence ofhea t onth eviscosit y ofeva ­ porated animal milk can be found inth e publication of BELL, CURRAN, and EVANS (1944)wh o HTST heated the concentrated- milk prior tocanning . Ifth emil kwa s forewarmed,HTS' T sterilized after concentration andcanne d aseptically,th eevaporate d milkswer e of fairlysatis ­ factoryqualit y foronl yapproximatel y 4month swhe nstore d at30°C . However,brie d additional heating incan s at 115°C increased the storage life ofthes e high-short sterilized milks.

The publications referred toar e only part ofth e existinglite ­ ratureou to fwhic hw ehav echose nthos earticle swhic hmigh tb eo f importance toou rwork . Promth e informationo nevaporate d animalmil k itma yb esee ntha t it ispossibl e to influence theviscosit y of the concentrated pro­ ductb yapplicatio no fhea tbefor e orafte revaporation . DEYSHER, WEBB, and HOLM (1944)demonstrate d that the viscosity of the concentrated productafte rsterilizatio nca n be influenced bypre ­ heating.Th eviscosit y behaviourdurin g storageals odepend so nth e method of forewarming the milks as they have shown. Thewor k of BELL, CURRAN, andEVAN S (1944)i sanothe r example ofth ein ­ fluence of heat on theviscosit y behaviour duringstorage .Beside s it isa n illustrationo fth e fact that the influenceo fhea ttreat ­ ments hasno t only effect before evaporation but alsoafte revapo ­ ration, infac t this influenceapplie gt oth esterilization-proces s as well. Briefly we may say that the application of heat in the process of manufacturing evaporated animal milk results inchange s ofth eviscosit yafte rth esterilizatio n processan ddurin gstorage . Returning toou r ownproblem ,th epresenc eo f aviscosit y barrier onconcentratin gsoymilk ,w e mayexpec t that application ofhea ti n somestag e ofth e processwil l influence theviscosit y behaviour of theconcentrate d product. Evidently it isno t possible to predict what the results of such experiments will be,bu t considering the favourable resultsobtaine d with animalmil k wema yhop e that this influence willb esuc h ast omak e itpossibl e toevaporat esoymil k toa t least 30%solid scontent .

Inconnectio nwit h ourow nwor k theresult so fDEYSHTER , WEBB, andHOL M (1929)a sregard s thehea t stability ofanima lmil kar e important. They havedefinitel y shown,tha t it ispossibl e to in-

66 fluenceth estabilit yo fth eevaporate d productb yvariatio no fth e heat treatment priort oevaporation . WEBB andBEL L (1942)hav e demonstrated that ifrapi d heating intubula r heaters isapplied , thestabilit yo fevaporate dwhol e milko f26 %tota lsolid si sgene ­ rally increasedt otwic etha to fth econtro lsamples.I nthi swa yi t was possible forV TEBB , DEYSHER, BELL, andH O LM (1943)t o preparea mil ko f32 %tha tcoul dwithstan d sterilization. They also demonstrated thati ti sth evelocityof.heating , inothe rword sth e coming-up time,fro mwhic hth estabilizin g effect isderived . Also whenth econcentrate d milk itselfi sheate d WEBB andBEL L (1943) found thatth ehea tstabilit yi sinfluenced . Promthi s informationi ti sclea rtha thea ttreatment s havegive n favourable resultsa sregard sth ehea tstabilit yo fth econcentrate d animal milk.Therefor ei tseem sjustifie dt ostud y thepossibl eef ­ fecto fth e influenceo fhea to nth estabilit yo fevaporate d soymilk.

5.3 EXPERIMENTAL PART

5,3.1 Scope of work.

Promth eforegoin g discussion ofth eliteratur e iti seviden t thato nconcentratin g soymilk difficultiesma yb eexpected ,especi ­ allya sregard s theviscosit y ofth eproduct . These difficulties seemt oaris e inparticula rwhe na dr ymatte rconten to f20 %o rhighe r isreached . Thereforea stud yo fth einfluenc eo fth econcentratio n and other manufacturing conditionso nth eviscosit y isdesirable . Withcowmil kprio rt oevaporatio na hea t treatment isapplied .I t isknow ntha tthi spreheatin go rforewarmin g influencesth evisco ­ sity. Possiblywith'soymil k preheating alsoha ssom e influence on. the viscosity. Prom preliminary experimentsw ealread y knew that fresh soymilk hasa comparativel ylowhea t stability.Ifthi shea tstabilit y should not become bettero nconcentrating ,o reve nbecom eworse ,the n ste­ rilizationo fth econcentrate d milkwoul d bea difficul t problem. Promcowmil ki ti sknow ntha tth ehea tstabilit y isimprove do nap ­ plicationo fheat.O nthes e groundsastud yo fpreheatin gi nconnec ­ tionwit hhea tstabilit ywas desirable . Alsofo r'othe r reasons heatingth era wmil kprio rt oevaporatio n is important; inth efirs t placepreheatin g lowersth einitia lnum ­ bero fmicro-organisms ;thi si so fimportanc ewhe na nevaporato ri s choseni nwhic hth eproduc tremain sa comparativel y longtim e under conditions favourable forgrowt ho fmicro-organisms ,an dals o when the concentrated milki suse dfo rth emanufactur eo fsweetened,con ­ densed milko rmilkpowder .A lo wplat ecoun ti sals o favourablefo r

67 the latersterilizatio n process/Besides preheating is important because enzymesar einactivate d whichmigh thar mth equalit yo fth e product lateron .Amon gthes eenzyme son ei sundesirabl e froma nu ­ tritionalpoin to fview . Thisenzyme ,th etrypti c inhibitor, must bedestroye d anyhow. Lastly, preheating isrecommende d asa metho d ofdesodorisation . Wemigh tas ki fth eevaporatio nan dth ehea t treatment couldno t becombined . Inprincipl ei ti squit epossibl et oevaporat ea tcom ­ paratively hightemperature san dt oadap tth eholdin g time ofth e productt ocondition smos t favourablefo rth eviscosity , heatsta ­ bility,an dothe r factorsmentioned . However,fo ra hea t sensitive producta ssoymil ka nevaporato r ispreferre d witha shor t holding timean da lo wevaporatio n temperature. Insuc ha typ eo fevapora ­ torth emil khardl yundergoe sa hea ttreatment . Ifsuc ha nevapora ­ tori sapplie d itwil lb enecessar ytha ta separat e heatingpreceed s theevaporation .The nth eadvantag e istha tth ehea t treatmentca n becontrolle d more accurately. Summarizingi twa snecessar yt opa yspecia l attentiont oth epro ­ cesso fpreheating ,an dt oit sinfluenc eo nth eviscosit y andhea t stabilityo fth eevaporate dmilk . Forthi spurpos esoymil kwa sprepare d according toa unifor mme ­ thod recommended inth econclusion so fth echapte ro nmanufactur e ofsoymilk .Afte rpreparatio nth emil kwa sheate d invariou s ways; thisprehea ttreatmen t isdescribe d inth efollowin gchapter . Then themil kwa sevaporate d underconstan t conditions;th eevaporatio n Wascarrie dou ta t40° C (about5 0m mmercury )b ymean s of vacuum steam of70°C . Sampleswer e takeno nwhic h determinations ofdr y matter, viscosity, heat stability, acidity, andp Hwer e madeb y threeperson ssimultaneousl ys oa st ohav eth emeasurement smad ea s soona spossibl e aftersampling .

5.3.2 Design-of experiments.

Inactua l practice themos t common method ofheatin g isbatc h heatingi na vesse lwit ha stea mjacket . Sucha vesse lwa suse di n ourwork .I nou rfirs theatin gserie sth elowes ttemperatur e applied was 60°Cwit ha holdin g time of3 0minutes . Itwa sno tconsidere d tous elowe rtemperature sa sthi swoul dmak eth ehea t treatmentto o much questionablea sregard sth edestroyin go fmicro-organisms . Inth ediscussio no fth eliteratur e itwa spointe dou ttha t WEBB and others (1943)hav e observed adefinit e effect ofth eholdin g timeo nth estability . To-investigatewhethe r this also holds for" soymilka secon d serieswa sdesigne d inwhic hth esam e temperatures asi nth efirs t serieswer eapplie dbu tusin ga holdin g timeredu -

68 cedt ozer ominute stheoretically .I npractic ei ttoo kabou tthre e minutesbefor eal lth emil kha dbee ntransferre dfro mth eheatin g intoth ecoolin gvessel . Asha sbee nshow nb yWEB B andother s(1943 )rapi dheatin gin ­ creasesth ehea tstabilit yabov etha timparte dt omil ksample sheate d slowly.Sucharapi dheatin gi sonl ypossibl ewhe nth emil ki sheate d continuously,tha ti st osay ,i na flowing:condition .Fo rthi spur ­ posea speciall ydesigne dcontinuous ,flas hsterili^ato rwa scon ­ structedb yth emachin efactor yo fStor k (Holland).I tconsist so f threeconcentri cspiral so fdoubl epipes ,th espiral ssuccessivel y representingth eheating ,th eholding ,an dth ecoolin gtube . Theorigina lai mwa st ohea tth emil ka ttemperature sbetwee n 120°an d150 PCwit hthi sflas hsterili^ator .However ,whe n120° C wasapplie dth eformatio no fincrustatio nwa ss ostron gtha ti twa s necessaryt oincreas eth estea mpressur efro m2 t omor etha n6 at m tokee pth emil kheate da t120°C .Beside safte rbein guse da secon d timeth eapparatu sprove dt ob eobstructed .I nou ropinio nth erapi d incrustationma yb eascribe dt oth egrea ttemperatur edifferenc e andth erapi dheating . Onfurthe rexperimentatio nwit hth esterilizato ri tappeare dtha t temperatureslowe rtha n100° Ccoul db eapplie dwithou tto oman y difficulties. Inth ethir dserie so fhea ttreatments ,th elowes ttemperatur e appliedwa s70° Cwit haholdin gtim eo f4 5seconds,whil eth ehighes t temperaturewa s120° C4 5seconds .A si twa sals odesire dt oinves ­ tigate:th einfluenc eo fth eholdin gtim ewhe nth emil ki sheate d rapidlyi nth efourt hserie sthi sholdin gtim ei sreduce dt ozer o secondstheoretically .Actually ,accordin gt oth econstructio n data,th etim eneede db yth emil kt opas sfro mth eheatin gt oth e coolingtube-i sabou ton etent ho fa second .Als oi nthi scas ei t wasno tcon sidere dt ous elowe rtemperature stha n70°C ,a sth emea ­ ningo fsuc ha hea ttreatmen tfo rth edestroyin go fmicro-organism s wouldbecom ever yquestionable . Summarizingw estudie dth epropertie so fth eevaporate dsoymil k asinfluence db yth efollowin gheatin gmethods .

Series1 Series2 Series3 Series4 Batch Batch Continuous Continuous :- - 120°C4 5se c -. ._ - 100°Co se c 90°C3 0mi n 90°C0 mi n 90°C4 5se c 90°C0 se c 80°C3 0mi n 80°C0 mi n 80°C4 5se c 80°C0 se c 70°C3 0mi n 70°C0 mi n 70°C4 5se c 70°C9 se c 60°C3 0mi n 60°C0 mi n -— •'

69 5.3:3 Technique of experiments.

In processing the milk to evaporated milk the first step was the preparation of the raw soymilk. This milk was heated by a steam- jacketed vessel or by the flash sterilizator, after which the milk was cooled to about 5°C.Then it was sent through a film evaporator, where it was evaporated at 40°C by means of 70°C vacuum steam. Sam­ ples of increasing dry matter content were taken andeac h sample was cooled. Determinations'weremade of the dry matter content, visco­ sity, heat stability, pH, and acidity.

In preparing the mi Ik about 10 kg of dry beans were soaked over­ night and extracted with water for half an hour in a 1 : 10 ratio by means of an Ultraturrax provided with knives. The mixture was cooled to about 5°C, after which it was sent through a centrifugal filter. As it was important for the milk to be completely free of small solid particles a second filtration was carried out. During the whole filtering process the milk was kept cool in order to pre­ vent the growth of micro-organisms. The prepared raw milk was kept in the refrigerator at about 5°C Then the milk was heated either by a vessel described above or by the continuous flash sterilizator. If the vessel was used (photo 6), the milk was kept in constant movement by a stirrer. This improves the heat transmission and avoided incrustation. After the required period of heating the milk flew into a container in which a cooling spiral and a mixer were placed. For heating 2 atm. steam was used, about 10 minutes was required to reach the desired temperature. It took about 30 minutes for the cooling operation to reach 20°C.

-•In the continuous sterilizator the coming-up time is 3-4 seconds. Also the cooling operation is very rapid, as it takes only 12 se­ conds to reach 20°C, The sterilizator consists of three concentric spirals of double pipes (photo 7). In all three sections the milk flows through the inner pipe. In the-inner spiral, the heating tube, the milk is hea­ ted by steam;: in the outer spiral, the holding tube; the milk can be kept for o, 15, 30, or 45 seconds at the temperature obtained; in the middle spiral, the cooling tube, the milk is cooled with water (fig 6)» The temperature is measured with thermocouples which are present in four places. The first is placed at the inlet of the soymilk, the second immediately after the heating tube,the third just before the cooling tube, and the last at the milk outlet (fig 7).

70 Thedat ao nth econstructio no fth eapparatu sa t10 0L/hou rcapa ­ city-a sgive nb yth emachine-factory -ar esummatize di nth efollo ­ wingtable :

Plash-sterilizator:dat afo r10 0L/hou rcapacity .

length inner velocity Re K (m) diameter (m/sec) (Kcal/m2.hr.°C) (mm) heatingtub e 5 5 1.4 12.000 2.500 holdingtub e 3.6+7.2 12 1.4 coolingtub e 12 12 ,- 0.28 7.000 1.000

Afterth ehea ttreatmen tth emil kwa scoole dwit hcol dbrin et o about5°C ,an dthe nth emil kwa ssen tthroug hth efil mevaporator .

Theevaporatio ninstallatio nconsist so fa film evaporator,a con- densor, a vacuum system, anda vesse lfo rth eproductio no f vacuum steam (figg$ .

Thelowe rlonge rpar to fth eevaporato r (fig9 )consist so fth e evaporatorbod ywhic h issurrounde db ya steamjacket .Th euppe r shorterpar tserve sa sa vapou rliqui dseparator .Betwee nthes etw o partsth eliqui d isintroduce dtangentiall y intocth eevaporato r body,wher ei ti scaugh tb ya four-blad eroto rextendin gfro mth e topo fth euppe rpar tt oth ebotto mo fth elowe rpart .Du et oth e movemento fth erotor ,th eliqui di sthrow nagains tth einne rwal l andflow sdow na sa thi nfilm .Th eroto rca nb eadjuste da tthre e speeds.Th eadvantag eo fthi skin do fevaporato ri stha tth eliqui d comesonl yfo ra ver yshor ttim eint ocontac twit hth eheatin gsur ­ face. Theevaporato rbod yha sth efollowin gdimen sions : innerdiamete r 60 mm outerdiamete r 64 mm length 90 mm thicknesso fwal l 2 mm heatingsurfac e 0.0954m 2 Thediamete ro fth eroto rvarie sfro m5.6 7- 5.7 7cm.S oth edis ­ tancefro mth eroto rt oth eheatin gsurfac evarie sfro m1.0-1; 5mm . Thethre espeed so fth eroto rar e 1) 1010rotation spe rminut e 2) 1470 3) 2120

71 Fig6 .Stor kdoubl epipe ,continuous ,flas hsterilizator .

steam

soymilk f

NT thermocouple . direct tocoolin gtub e cooling t 15se c holding time water ) 30 „ <5„ Pig7 .Plas hsterilizator ,positio no fthermocouples . FLOWSHEET FILMEVAPORATO R

product.

receiving tanks, vacuum line. Pig'8. Evaporation installation.

FILMEVAPORATOR

Pig 9. LUWA film evaporator.

73 Promth evapou r liquid separator the vapourcome sthroug ha con ­ nection pipe into the glass condensor, where it is condensed by three glasscoolin g spirals.Throug h these spiralswate ro rcoole d brine canb ecirculated .A t itsbas eth econdenso r isprovide dwit h twocontainer s for the condensed vapour.Th econdenso r isdirectl y connected withth emai nvacuu m systemb y themai nvacuu mpipe . Thevacuu m inth e evaporator isobtaine d by means of one water andtw ostea mejectors.Th eworkin gpressure swer e3 atman d 2.8at m respectively. Thevacuu m ismeasure d by amercur y manometer Man d amountst oabou t 73c mo fmercury .Th evacuu m iskep tconstan t by a leakage ofvalv eA ,o nth eto po fth e leftcontainer .Unde rworkin g conditionsth evacuu m isadjuste dt oa constan tvapou r temperature which ismeasure d bya mercur y thermometer T inserted inth econ ­ nectionpip ebetwee nvapou rseparato ran dcondensor .T o insureth e vacuumtightnes sa vacuu m (V)an dawaterloc k (W)ar eapplie d around theuppe ran d loweren do fth erotatin gaxle .

Theevaporato rbod y isheate d by vacuum steam which isgenerate d by avesse lwit ha now nsecondar yvacuu m system. Thevacuu m system is notconnecte dwit h themai nvacuu m system. Due toa n automatic heating regulation thevacuu m steam had aconstan t temperature of 70°C. The secondary vacuumwa s about 53 cm of mercury.

The actual operation wasa sfollows.Th emai nvacuu msystem ,wate r andvacuu m lockswer eswitche d on.afterwhic hvacuu msteamwa sgene ­ rated.The nth eroto ran d condensorwer eswitche don . Thecol dmilk ,abou t5°C ,wa ssen t into*th eevaporato rbod y after havingpasse da flowrator.Th e milkwa sno tsen t intoth e evaporator at boilingtemperatur e becausefro mpreliminar y experiments itwa s knowntha tthi swoul dcaus etroubl edu et oth e foamingo fth emilk . However, if the temperature ofth emil k was lowered and the speed of the rotor increased the foam was completely suppressed. This flowratorwa sadjuste d insuc hawa y thatth e inletamoun twa sabou t 24L/hour .Th emilk , flowingdow na sa thin ,film, 'was -Heated 'b yth e steamjacketan d thevapou rescape da taboilin gtemperatur e of40°C . •Th emil kha sa strongtendenc y tofoa man d ifth espee d of the rotor isno t high enough the foamwil l not be completely destroyed and passesint oth econdensor.Unde rth echose ncircumstance s thesecon d speed of the rotor, 1470r .p.m. ,wa s high enough topreven t this. Theconcentrate d milkwa sreceive db y oneo fth etw oglas scontai ­ nersa tth e lowerpar to fth eevaporator . When thiscontaine rC wa s filled,themilkflo wwas >switche dt oth eothe rcontaine rb ymean so f tapK . Then thecontaine rC wa s shut off fromth e vacuumsyste m by means ofta pC lf.andairwa sadmitte db yta p C2. By:Openingta pC 3

74 themil k flowsou to fth econtainer .Afte remptyin g the tapsC 3 and C2wer eclose d and asta pC iwa sgraduall yopene dth epressur eo fth e mainvacuu msyste mwa slowere dunti lth erequire d vapour temperature of40° Cwa sreache d again,Inthi swa yth e glasscontainer swer e fil­ ledan d emptied. The milk,whic hcam eou t of thetw o glasscontai ­ ners, was immediately cooled with acoolin g spiral through which cooled brine wascirculated . Inthi swa y themil kwa s senta fewtime sthroug h the evaporator at the rate ofabou t 24L/hour , and during each pass a number of sampleswer etaken.Durin gth efirs tpas ssom ethre et o fivesample s withdifferen t drymatte rconten twer e takenb y loweringth e supply of milk to the evaporator. The greater part was evaporated under constant conditionsan d thusa firstconcentrat ewa sobtaine d with a drymatte r contento fabou t10%.Afte rth e firstpas sth emil kwa s sent through theevaporato r for a second time; the second concen­ trateha dabou t 14%dr ymatter .Als odurin gth esecon dpas ssample s of increasingdr ymatte rconten twer etaken.I fth eviscosit ybarrie r wasno tye t reacheda thir dan d fourthpas swa seffecte d and samples were obtained.Altogethe rth emil kwa sevaporate d twot ofou r times until ithardl y flowsowin gt oth e increased viscosity. Thesample s obtained inthi swa ywer epu t inbottles ,whic hwer e cooled inwate r ofabou t 0°C.Th eviscosit y then increased due to the decrease inth e temperature ofth e evaporated milk. The milks of lowviscosit ycoul db epoure d outeasily,bu twit hthos eo fhighe r viscosity -thoug hth emil kwa sstil l flowing -th ecircula r upper surface ofth emil ktende d toretai n itsshap ewhe nth e cylindrical bottlewa skep tsloping,Whe nthi shappene d suchamil kwa sdesignate d a gel,thoughth econsistenc yo fsuc ha gelwa sno tnecessaril y very high. This gellingoccurre d above 100cP . The gels of the other sampleswit h successively increasing dry matter content gradually became firmer until at last,keepin g the milk bottle upsidedown , the contentsdi d not flowou to fth ebottl e but remained together like a very firmpudding .

The viscosity determinationswer ecarrie d outwit ha Haak eVisko - waage,(fig 10),whichi sa modificatio no fth eHopple rviscosimeter . Itconsist s of a standard with a fall tube and waterjacket. The standard carriesa yok ewhic hha sa weight , rod,an d metal ball on the lefthand ,a lon gneedl e inth emiddle ,an d ascal e forweight s on the righthand . When theglas stub ewa sfille dwit hth e evaporated milk,th ebal l wasplace d in itslowes tpositio n inth e tube.Byputtin gweights'o n thescal eo nth erigh t thebal lwa spulle dupwar dthroug hth e liquid and theneedl e movedove ra marke d scale.Th etim ewhic hth eneedl e

75 needst opas sbetwee ntw opoint so nth escal ewa sobserved .Th ede ­ terminationswer ecarrie dou ta t20° C Thewate raroun dth eglas s tubei scirculate db ya Hopple rthermostate ,whic hkeep sth ewate r temperaturea t20° C± 0.05° . Theviscosit ywa scalculate dwit hth eformula . n= F G T n= viscosit y(cP ) P= ro dfacto r (cm-2) G= weigh t(g ) T= tim e(sec )

fig.1 0Viskowaag e

The heat stability determinationswer ecarrie dou ti npyre xglas s tubesope na ton eside. "Thi ssid eca nb eclose db ymean so fa nut . Thetub eha sth efollowin gaverag esizes :lengt h20. 5cm ,oute rdia ­ meter1. 2cm ;abou t80 %p 8ml )o fth evolum ewa sfille dwit h> eva­ poratedmilk .Th etube swer efastene dt oa movabl ero db ymean so f twoclamp san dafte rimmersio ni nth e120° Cparaffin ebat hkep ti n constantvertica lmovemen ta sindicate di nth efollowin gfigure . Thistemperatur eo fth eparaffin ewa schose nbecaus eth esteriliza ­ tiono fevaporate dsoymil ki ncan swoul db ecarrie dou ta t120°C . Thetub eend swer emove dup -an ddownward sabou t2 5t o3 0times. a

76 minute. A stirrer kept theparaffin e inconstan t movementi n thetan k whicha tth efron twa sprovide d witha glas swindo w through which observationscoul db emade .Thes eobservation swer emad ei nduplicate .

«mii..ij.«*.iujii...... 1,1. i..•-,...,;...... I.l..i.l-I.M.JIII.'.-.W

• ^^^|^•JJJ.•^.^l•u^w•^•J•JH.J•^^•l^^^•g|J•^^J•J•.••,^••..••^•.^A^^•l^J•l>•.^•.•_•^..JJ^.^w^^JJ^w•^lJ••JW'^l•7^«

fig. 11 Pyrex glasstiibe

I>\//»////T7-77 „t/t/i///WJ//l>/l>/bl/IFTr. >»//»»////>>!/////////J//. 7 .A ' i ,xr i i,-

i i i i

.---T ^,K

bs/W/W»JW/WWWWSSW»W/777777n; 1 fig.1 2Paraffin etan k

As soona sth etube swer eimmerse dth etim ewa smarke dwit ha stop * watchan dth emil kwa skep t flowingi nth etubes .Durin gth emilk - flowsmal lspeckle sappeare do nth einne rsurfac eo fth etubes i; n the milk itselfn osoli d particlescoul db eobserve da syet .The nth e milk begant oflo wjerkingl y (theflowlin eo fth emil ki sno tsmoot h anymore )an dwhe nthi spoin twa sreache dth ewatc hwa sstopped.Th e time between immersiono fth etube san dthi smomen twa sconsidere d asth ehea tstability .O nfurthe rheatin gsoli dparticle sappeare d throughoutth emil kan dth emil ksolidifies .

77 5.4 Results, discussion and conclusions.

Before enteringth ediscussio n ofth eexperimenta l results iti s necessaryt odwel lupo nth ereliabilit yo fth edetermination so fth e viscosityan dth ehea tstability .T othi sen dw ecit e some figures ofviscosit y determinationswhic hwer edetermine d intriplicate ,:.;• .

Heat treatment 60°C 30 sec.

concentration viscosity concentration viscosity 11.7% 3.02c P 21.3% 17.88c P 3.18 17.80 3.02 17.65 16.9 % 5.82 27.5% 123.62 5.72 119.72 5.62 120.— 19.6 % 10.11 33.5% -287:60 10.18 271.60 10.29 264.80

Roughly estimated upt o1 0c Pth evariatio n is.0. 2cP ,a t10 0c P iti s2 cP ,an da thighe rviscositie si tbecomes/abou t1 0cP .A thig h viscositiesth evariatio n rangema yb eeve n larger, but'its•meanin g forth ereliabilit y ofth eviscosit y curvesi sestimate d tob esmal l asth elas tpar to fth ealread yver y steep curvewil lno tb einflu ­ encedmuc hb yeve ncomparativel y largevariations . Thehea t stability determinations were carried outi nduplicate , someo fwhic har egive nbelow .

Heat treatment 70°C4 5 sec.

concentration heat stability concentration heat stability 10.8 % 61mi n 6se c 13.3% 16mi n4 4se c 61mi nl ose c 16mi n4 0se c 12.9 % 34mi n2 9se c 16.1% 8mi n ose c 34mi n 9se c 7mi n4 0se c 13.9% 18mi n 6se c 19.2 % 3mi n5 0se c 18mi n 4se c 3mi n4 5se c

Theresult so fou rexperiment s arerepresente d intabl e4 ,5 , and 6,an dfi g13-25 . The first four graphs,fi g13-16 ,dea lwit h the heatingexperi -

78 TABLE 4 Influenceo fpreheatin go nth eviscosit yan dhea tstabi ­ lityo fevaporate dsoymilk .

S E R IE S I

Heat treatment dry matter viscosity heat stability PH acidity (ml)

90°C 30 min 10.9 %• 9.85 cP 60 m 15 sec 6.63 2.16 11.4 10.21 46 m 45 6.67 2.49 12.1 11.14 41 m 17 6.64 2.36 13.1 21.03 33 m 52 6.62 - 14.2 44.16 18 m 33 6.62 - 14.4 43.60 18 m 39 6.67 -

«0°C 30 min 14.4 5.38 23 m 17 6.56 15.5 6.28 8 m 18 6.56 16.3 7.35 7 m 53 6.56 3.34 19.1 23.03 1 m 6.57 4.23 19.5 26.09 2 m 48 6.54 4.13 22.2 82.87 1 m 6.54 - S3124.7." 176.93 1 m 6.50 - ^125.8 220.13 1 m 6.52 -

70°C 30 min 12.8 3.52 23 m 14 6.53 2.36 14.0 4.48 8 m 58 6.53 2.45 17.1 7.41 2 m 25 6.51 3.43 19.2 11.56 1 m 6.49 3.47 20.3 14.34 1 m 9 6.50 4.11 20.8 22.28 1 m 7 6.50 3.91 23.7 44.74 - 6.52 - 25.3 80.81 - 6.50 - 05125.4 102.57 - 6.52 - L • 60°C 30 min 11.7 3.07 36 m 42 6.41 j 3.06 13.1 4.02 37 m 38 6.43 j 2.72 13.4 3.71 22 m 34 6.43 | 2.97 13.9 4.65 11 m4 6 6.45 j 2.97 15.9 6.01 5 m 38 6.45 | 3.51 16.9 5.72 2 m5 4 6.45 | 3.87 18.2 8.92 2 m 42 6.44 j 3.95 19.6 10.19 2 m 6.44 | 3.55 21.3 17.78 - 6.45 ! 4.60 23.3 26.45 1 m 54 6.45 j 4.61 27.4 131.37 0 m 25 6.43 | - 33.5 824.10 - 6.40 • - 79 SERIES II

Heat treatment dry matter viscosity heat stability pH acidity (ml)

90°C 0 min 12.8 % 6.55 cP 31 m 30 sec 6.60 2.50 12.9 6.37 33 m 46 6.60 2.55 13.1 7.55 35 m 13 6.62 2.45 13.2 7.67 32 m 19 6.58 2.83 gel 16.5" 64.68 4 m 38 6.62 - gel 17.2 99.87 2 m 56 6.58 -

gel 18.5- 286.98 -• 6.60 *"

80°C 0 min 13.7 4.21 48 m 53 6.59 2.50 14.9 5.19 23 m 34 6.55 2.71 16.5 6.77 7 m 29 6.53 2.93 18.9 12.02 3 © 17 6.50 3.82 22.6 36.32 1 m 30 6.50 4.43 clumps 24.5 85.06 1 m 10 6.49 4.67 clumps 26.1 110.24 1 m 4 6.50 - gel 29.11 234.93 1 m 10 6.46 -

70°C 0 min 20.0 15.80 1 m 27 6.42 3.80 20.8 23.27 1 m 22 6.46 .4.17 23.1 29.17 1 m 4 6.44 4.69 25.4 78.26 0 m 47 6.41 4.97 26.2 64.99 1 m 4 6.42 - gel 28.5 174.00 0 m 32 6.40 - gel 30.0 260.70 0 m 41 6.41 -

gel 32.9 563.85 -• 6.32 - 15.2 8 m 38 15.8 6 m 47 17.6 3 m 16

60°C 0 min 20.2 9.97 2 m 24 6.50 3.82 20.5 9.97 2 m 24 6.49 4.08 21.3 11.31. 1 m 58 6.50 4.51 21.7 11.85 2 m 30 6.50 4.19 22.7 14.74 2 m 6 6.50 4.44 26.4 31.06 - 6.49 4.95 30.4 66.65 am 30 6.42 - 32.9 208.37 0 m4 0 6.40 clumps 33,6 157.45 0 m3 0 6.43 .-

80 SERIES III

Heat treatment• dry matter .viscosit y heatstabilit y pH acidity (ml)

90°C4 5se c 12.4% 3.71 cP 87m 5 6se c 6.59 2.78 15.0 4.74 17m 3 3 6.53 3.51 15.5 5.43 16m 4 9 6.53 3,;7 3 17.4 7.95 9m 3 6 6.49 4.17 22.3 40,13 2m 5 9 6,56 - 22.3 41.6 4 2m 5 7 6.51 •- 22.9 64.40 2m 5 8 6,42 5.63 gel26. 3 298,41 6.49 - -• - gel28. 8 368.80 '- 6.50 ° 80°C4 5se c 12.7 3.32 144m 3 0 6.56 - 13.5 » 100m 2 5 - 14.4 4.49 83m 6.58 17.9 15m 2 3 - 19.8 9.63 7m 2 7 6.57 21.0 13.26 7m 7 6.56 23.6 20.27 4m 6.54 25.2 - 5m 1 0 6.52 25.3 33.84 3 m2 0 6.52 26.2 49.54 1 m2 9 6.52 gel30. 2 253.93 6.55

70°C4 5sec . 10.8 3.25 61m 8 6.43 2.83 11.4 3.31 54m 5 9 6.44 3.20 12.9 3.51 34m 1 9 6.49 2.61 13.3 3.62 16m 4 2 6.44 3.70 13.9 3.93 18m 5 6.45 3.17 14.2 4.03 23m 1 4 6.48 3.13 16.1 5.10 7m 5 0 6.45 3.20 19.2 •8.34 3 ml8 6.47 4.24 21.3 11.20 2m 1 8 6.43 4.69 26.8 30.16 0m 55 ' " 6.44 7.70

27.7 45.16 0m 3 9 " 6.41 _ 28.1 46.54 - 6.40 - ( clumps32. 8 153.60 - 6.37 - gel37. 3 622.80 - 6.33 - gel39. 0 730.90 - 6.34 -

120°C4 5se c 18.3 17.22 16m 1 0 6.71 18.9 22.45 14m 5 9 6.72 19.-6 26.18 14m 5 9 6.69 20.1 37.86 12m 1 6 6.73 22.7: 162.49 1 6m 5 3 6.69

81 SERIES IV

Heat treatment dry matter viscosity heat stability PH acidity (ml)

90°C 0 sec 13.7% 3.78 cP . 6.61 2.62 15.0 5.08 46 m 19 sec 6.64 2.72 15.6 5.11 39 m 12 6.64 3.06 15.8 5.61 31 m 51 6.62 2.64 17.7 6.94 19 m 31 6.64 3,49 21.5 22.94 4 m 24 6.65 4.30 24^2 44.33 3 m 39 6.64 24.5 48.05 3.m 43 6.64 24.2 29.85 4 m 30 6.64 gel 31.0 411.60 6.61 - gel 33.2 545.74 - 6.59 - -

80°C 0 sec 12.0 3.32 94 m 16 6.59 2.29 13.8 4.61 63 m 11 6.65 2.76 14.3 4.18 46 m3 6 6.60 2.84 16.7 6.43 16 m 8 6.59 3.37 17.8 6.80 13 m 36 6.62 3.31 18.8 8.37 5 m 58 6.63 3.82 18.9 8.71 6 m 49 6.63 3.96 19.9 13.99 3 m 25 6.63 3.95 21.4 13.47 2 m 53 6.63 4.03 24.7 26.64 1 m 6 6.56 2.53 27.0: 40.- 0 m 57 6.58 5.14 27.3 51.64 - 6.58 5.84 28.4 101.04 0 m 52 6.58 5.84 gel 33.7 189.44 6.56 - gel 35.3 246.91 - 6.58 gel 35.2 338.14 6.55 -

70°C 0 sec 11.5 3.09 62 m 6.61 2.28 13.7 4.04 31 m 49 6;6 1 2;74 14.5 4.25 22 m 6.62 2.81 14.9 3.80 5 m 25 6.61 3.21 14.3 4.31 21 m 4 6.61 2.79 17.4, 6.08 1 m 38 6.61 3.71 20.3 8.14 M 1 6.59 4.32 22.0 9.54 6,58 3.37 23.0 10.90 0 m 51 6.58 5.06 32.3 39^17 0 m 22 6.56 6.21 36.1 67.29 . - 6.54 - • • . gel 36.7 115.69 - 6.52 - gel 37.2 186.54 - 6.52 gel 41.2 320.21 6.51 - gel 44.2 458.70 6.49 -

82 Heat treatment dry matter viscosity heat stability PH Acidity (ml)

10CPC 0 sec 15.4% 5.90 cP 26 m4 4 6.52 16,3 9.63 20 m 6 6.53 17.5 9.73 17 m 15.5 6.54 17.6 9.80 16 m 6.5 6.57 18.0 11.63 12 m 43.5 6.54 19.4 17.55 7 m 47.5 6.52 20.4 49.50 5 m 30.5 6.50 20.5 64.17 6 m 6 6.54 gel 20.7 50.62 7 m 33 6.53 gel 21.5 162.30 3 m 56 6.50 gel 21.8 131.64 - 6.51 gel 23.5 236.25 - 6.51

: Heat Dry matter Viscosity Heat Dry matter Viscosity treatment treatment Not heated 11.4 -3.22 70° 0 sec 14.1 4.30 13.1 3.57 (duplicate) 21.9 9.75 19.2 7.03 28.9 30.83 19.3 7.28 31.4 46.53 21.0 9.20 36.2 153.13 21.0 15.50 40.2 362.04 21.9 10.40 25.8 15.50 70° 45 sec 14.3 4.08 28.9 31.00 (duplicate) 18.9 7.15 30.5 35.98 23.1 16.07 30.7 53.72 25.7 25.62 33.0 57.65 27.9 40.78 36.6 205.15 30.1 69.75 36.9 187.59 33.5 158.65

Heat stability Not heated 20.4 7.57 Not heated 13.7 29 m 8 22.3 10.53 13.7 39 m 45 26.9 25.73 15.6 8 m 41 28.0 34.66 16.5 1 m 56 29.5 56.80 17.5 2 m 54 31.8 96.66 17.8 2 m 13 32.3 92.13 19.2 1 m 40 20.T 1 m 36 21.3 1 m 7 • '

83 cP

.250

h.stab. vise. 90'C 30 mm a • eo'c o 0 70"C a a 60*C V •'

VESSEL.

% dry matter afttr cone. Fig 13. Series.1. Preheating with vessel,holdin g time3 0minutes . Effecto ftemperatur e levelo nth eviscosit yan d heat stabilityo fevaporate d soymilk.

.min. cP 100 J 250 I y 190*5- 0 min

\ * 90*C. Emit.. I - VT '"- , /aotOmin,, ' \\ \80*C-0min. | it \ \\ \ • I n .1, / /TlfcOmin. J A J 7 70*C . v &tr^~ -'L 60'C . . / •"-•=••5..

VESSEL. «!»-»—»

% dry matter after cone. Pig 14.Serie s 2. Preheatingwit hvessel , holding time ominutes . Effecto ftemperatur e levelo nth eviscosit yan d heat stability ofevaporate d soymilk.

84 9fl*C A 30 min A 0 mtn »0"C o . . . 70*C o . • . eot i . i .

VESSEL.

FM ff 1 R '% dry matter after cone. Preheating with vessel. . Effect of holding time on the viscosity of evaporated soymilk. ...

VESSEL

% dry natter after cenc.,

Pig 16. \<- -••" ";'" •-•••'• • -';'\ " - Preheating with vessel. - : Effect, of?holding time on the;heat 'stability of evaporated soymilk. > 85 90*c.t5«ic:| „ V

\ 4—/ \ 120*C.t5sec. , j I . \% .4i : !

80t<5s«c. \ ^/ / / 7tfc>5s«c \ h.stab vise. 30 C15sec & I 80°C . o . o 70eC „ • a 12(fC „ . a ' *

STORK.

_.__—. _. • - % dry matt«r afUr cone. Fig 17. Series 3. Preheating with flash- sterilizatorr holding time 45 seconds. Effect of temperature level on the viscosity and heat stability of evaporated soymiIk.

min. 90 C One 100 80*C . 250 70'C , WCfC .

\V TUtftO 90'C.0li \ " \ . 8Q'C.0«»cj ,/'

,9QtQ«.e \ 7 /'• 70'C0a«c /

MSHMK.\ \|K^.. / / . 7

STORK. sttrilisator v—-"#-• 20 30 <0 . _ . *Udr ymatU r aftw cone. Pig 18. Series 4. Preheating with flash sterilizator, holding time o seconds. Effect of. temperature level on the viscosity and heat stability of evaporated soymi-lk....' cP 250 90t.45s«c. a Osac. & \ k t i 801: . o . . ™t . o . • t2dt . 100*C . w

! J/ .' /70*C.t5iic , / /I / / o/ ' C 45860. -/; 1I /1 //' / / /*' / r ' -h-j j-—i j I /BO*C<5«C- / "ft"™- U—/-—J ! /ao'c«..B / ' / / •' / / /* / / / / / /»O*C-0MC7^ . / ^MV / / / // / 100*COs. c */7 '//

STORK. %6£^--^ steritisator *£££&••m^»

f, dry matUr after cone. Pig19 . Preheatingwit h flashsterilizator. . Effecto fholdin gtim eo nth eviscosit yo feva ­ porated soymilk.

:\ V\\\° 90*C.<5«»c.

90*C.«5>tc A OStC.A 80*C. . o • •',\ \ 'v.*\ 8Q'C<5». H 7o"C. . D • • t20"C. . a

IOO'C • - 9Q-C.0«^\\\\J^\\v^%\. la-c.i5,K 80*CL0».cA ^ ,v i^V„\ \ \ X ".>, -e 120*.»«.c.

• \ V «NSS. • 70'C.OMC. \ \ „ \. *VJ-J,

\ \ „ V ^ \ . \ \. \ 'v.... \ >.

STORK. sttrilisator

It dry matttr aftw cone.

Pig20 . Preheatingwit h flashsterilizator . Effecto fholdin gtim eo nth ehea t stabilityo feva ­ porated soymilk.

87 Omirt. Oaec 90"C. A A 80"C. 0 ' I'U ! i 250 70"C, • B 60*C. V ill ' Wtffc. • 1 i' 90*C.Osec./ J j r I 9DC0mint

fcomin. ' // / /" ~~1i\ iI l*C.O»ec,/ I / / i 7/7 '•' J 70'CQrtiin././ / /- / 70C0sec. ; ./ ~~/// / / / / // 7 ,60*C0mln . I / Hi I i ••-/ //A 7 A / 1 •//•// / VESSEL-STORK J*^a*s&2i&Z=^ steriUsator

% dry matter after cone. Pig 21. Comparison of preheating with vessel and flash sterilizatoiv Effect of temperature level onth eviscosit y of evaporated soymilk.

eo c . '\ \ \ 90'C0».e. 7o'c • 60*C T

v\ \\ \

\\ \ " '"•••., 60C°m'"-

2°Ma^> \N.. "•-... • \ ?>^/~~*-o—o

i VESSEL-STORK sterilisator

% dry matter after cone. Pig 22. Comparisono fpreheatin g with vesselan dflas h sterilizator. Effecto ftemperatur e levelo nth ehea t stability ofevaporate d soymilk.

88 VESSEL-P"™"L Omin.' »" «°

% drymatte r after cane Pig 23. Effecto fth emetho do fheatin go nth eviscosit y ofevaporate d soymilk.

90'C. «0-C VESSEL-[30min . Omin. »5s«c. STORK \_ Osac. \ \\ 80*C30mln.

% dry mattir afttr cone.

Pig 24. Effecto fth emetho do fheatin go nth ehea tstabi ­ lityo fevaporate d soymilk.

89 mm 100L

dry matttr afttr cone Pig25 .Duplicatio no fexperiments .

TABLE 5 Drymatte rcontent so fevaporate dsoymil kcorrespondin g witha viscosit yo f10 0cP . Influenceo fpreheating .

Series1 Series2 Series3 Series4 Batch Batch Continuous Continuous

^s. Holding ^v time Heating- ^v 30mi n 0mi n 45se c 0se c temperature ^v

120°C - • _ 22.7% _ 100°C - - 21.7% 90°C 17 .,% 17.8% 24.8% 27.3% 80°C 23.2..% 25.8% 28.4% 30 % 70°C 26.2,.% 26.7% 31.4% 36 % 60°C 28.8% 31.5% - - unheated 32.0% • - . • 33.8% • • ' '-

00 mentscarrie d outwit h the vessel. The first graph, fig 13 shows theviscosit y and heatstabilit y curveswhe n themil k is preheated with 3CTminutesholdin g time, and the second one, fig 14,wit h 0 minutes holding time. The third graph, fig 15,compare s thevisco ­ sity curveso f3 0minute sholdin gtim ewith/o f 0minutes . The same isdon e infi g1 6wit hth ehea tstabilit ycurves . Thenex t four graphs dealwit h thewor k done onpreheatin g with the Stork sterilizator. The fifth graph, fig 17,show s the curves forviscosit y andhea tstabilit ywit ha holdin g timeo f4 5seconds , and thesixth , fig 18,fo ra holdin gtim eo f 0seconds .Th e seventh graph, fig 19,compare sth eviscosit y curves of4 5 secondsholding , timewit hthos eo f 0seconds ,whil eth esam e comparison ismad e for thehea tstabilit y curves infi g 20. A comparison ofth e viscosity curves of preheating zero minutes with thevesse l and 0.secondswit h theStor k sterilizator ismad e infi g21 ,whil e infi g2 2th esam ecompariso n ismad e forth ehea t stability curves.Th e influence ofth emetho d ofpreheatin g onth e viscosity curveso fevaporate d milki sshow n infi g23 .Pa r thehea t stability curvesthi s isdon e infi g 24.

Indiscussin g theresult sw ewil lpa yattentio n toth edr y matter content-reached after acertai n heattreatment -correspondin g with a viscosity of 100cP . This viscosity ischose n asw e may safely assume that thenn ogellin g occurredyet .Th e heatstabilit y will bediscusse db yconsiderin g thedr ymatte rconten t reachedb ydrawin g a horizontal linea tth e 10minutee-leval-Thi svalu e ischose na s it should beconsidere d aminimu m ofsterilization .

When we take fig 15 (Preheating with vessel,effec t of holding time and temperature levelo n theviscosity )an d fig 19 (Prehea­ tingwit h flashsterilizator , effect ofholdin g time and tempera­ ture levelo nth eviscosity) ,w eca ndra w an imaginarylin ea ta ,vis ­ cosity level of 100cP .A t the pointsWhere'thi slin e crosses'the viscosity curvesth edr ymatte rcontent sca nb erea db y comparison with thehorizonta l axis.Th evalue sobtaine d inthi swa yar e shown intabl e5 . Ifw e compare thedr y matter contents ofserie s 1reache d after heatinga tdifferen t temperatureswit hth esam eholdin gtim ew e can readily see thata sth e heating temperature increases thedr y mat­ ter content decreases. Thisdoe sno t only apply to the points of the curvesa tth e10 0c P levelbu tals ot omos t otherpoint s ofth e curvesperhap swit hth eexceptio n ofthos ea tth enearl yhorizonta l part ofth ecurves. r: ... Consideringth efigure so fserie s2 w ema ysa ytha tthe .sam e holds

" 91' for this series.A sth etemperatur e increasesth edr ymatte rcon ­ tentsdiminishes .Thi smean s thati fw eprehea tth emil kwit ha ves ­ sel^th eviscosit y barrier arises later if lower temperaturesar e used. " ' ''.". : .'.'.'".•','• . ]'"'..• ' Comparisono fth eresult so fSerie s1 an dserie s2 show stha tfo r eacho fth eheatin g temperatures'Usedth edr ymatte r contentso f series %ar e higher.""This-mean stha tshortenin g the holding time causesth eviscosit y barrier toarise'later , itma yb e"see ntha t the influenceo fthi s factor isno ts opronounce d astha t ofth e temperature level. '"'•''' Theresults'fro mserie s3 sho wth esam eregularit ya si nth efor - mertw oseries ;Thi si sals oth ecas ewit hserie s4 .I fw eno wcom ­ pareth eresult so fserie s3 wit hth ecorrespondin g oneso fserie s 4w ese etha ta highe rdr ymatte rconten ti sreache di fth eholdin g timei sshortened.Thoug hth eholdin gtim eha sa definit einfluence , it isno tver y important.Wha tcount si nth efirs tplac e isth e in­ fluenceo fth etemperatur e level. ' • Weca nmeasur eth edifferenc e ineffec to fheatin g withth eves ­ selan dth econtinuou s sterilrjjatorb ycomparin g series2 an d4 , whichtheoreticall y haveth esam eholdin g time.Thenw ese e(fi g21 ) thatwit hth econtinuou ssterilizato rappreciabl yhighe rdr ymatte r contentsca nb ereached.Thi si sespeciall yth ecas ewit h preheating at 70°Can dals oa t90°C .Evidentl ythi sma yb eascribe dt oth emuc h more rapid heating bfth econtinuou s sterilizator andpresumabl y alsot oth erapi d codling.

Thusw eca nconfir m thatther ea^tov-y iexist ssuc h athin ga sa viscositybarrier .Whe nusin gth eter m "barrier"w eespeciall yhav e inmin dth estee ppar to fth eviscosity"curve , whichi nth eprope r senseo fth ewor d represents aCarrie r thatcanno tb eovercom eb y evaporating. I? : .•;••-„• The.ver y regular sequenceo fth ecurve si nfi g1 5an dfi g 19ar e proveo fth esystemati c influenceo ftemperatur e levelan dholdin g timeo fpreheatin go nth eviscosit yo fth eevaporate d product. Ifw ejudg eth eresult sb yquantitativ e standardsw ese etha ta t

about2 Of0 orafte r 20%,a smentione d inth eliterature , therei sa viscosity barrier. Howeverw ehav e found that alsobelo w 20%"dry matter this barrier canaris e ifW e choose the'right preheating conditions"(vessel;90°C). We have also-show n that theviscosit y curvesan d-thu sth eviscosit y barrierca nb emove d tohighe rdr y mattercontent sb ylowerin g theheatin gtemperatur eand .reducin g therholdingtime .I nthi srespec tth econtinuou s flashsterilizato r ismor e effectivetha nth evessel .Wit hth eunheatet fmil kw eeve n reacheda dr ymatter "conten to f32-34% . It"is:difficult-t oexplai n

92 that flashheatin ga t70° Cwit hn oholdin gtim egive samil ko feve n a higherdr ymatte rconten t thanth eunheated , evaporated milka t 100cP ,

Whenw etak e fig1 6 (Preheatingwit hvessel ,effec t ofholdin g timean dtemperatur e levelo nth ehea tstability )an d fig2 0(Pre ­ heatingwit h flashsterilisator ,effec to fholdin gtim ean dtempe ­ rature levelo nth ehea t stability)w eca ndra wa nimaginar y line atth estabilit y levelo f1 0minutes .A tth epoint swher ethi slin e crossesth estabilit ycurve sth edr ymatte rcontent sca nb erea db y comparison withth eX-axis .Th evalue sobtaine d inthi swa yar e shown intabl e6 .

TABLE 6 Drymatte rcontent so fevaporate d soymilkcorrespondin g witha hea tstabilit yo f1 0min . Influenceo fpreheating .

Series 1 Series 2 Series 3 Series 4 Batch Batch Continuous Continuous

>v Holding >v time 30 min 0 min 45 sec 0 sec Heating-S. temperatureN.

120°C ». . 20.6 % - 100°C - - .- . . 18.8% 90°C 15.0 % 15.3 % 17.0% 19.1 % 80°C 15.9 % 16.7% 19.6 % 17.4 % 70°C 14.0% 14.8 % 15.8 % 14.8 % 60°C 14.9 % - -

unheated 15.5 % ' - - • -

Onexaminin gth efigure so fserie s1 w ese etha theatin ga t8 0C givesth ehighes tdr ymatte rconten tan dthu sw ema ysa ytha tthi s heat treatment givesth e higheststability .Fo rth ethre ehighes t temperaturesth esequenc eo fincreasin gstabilit y impartedt oth e evaporated milk is70° .90° ,an d80°C . Alsoi nth esecon d seriesheatin ga t8 0C impartsth ehighes tsta -

93 bility to the evaporated milk. It is interesting to note that also with this series the sequence in which the heat treatments increase the stability is the same as in the former series for the three highest temperatures. Comparison of the figures of series 1 and 2 shows that for each of the heating temperatures the dry matter contents of series 2 are higher. This means that-decreasing the holding time increases the stability. Though the results of series 3 show the same order-as in the for­ mer two series it should be noted that the determination of the heat stability of the heating experiment 90°C 45 sec was carried out about, one day later as usual. Normally we may expect the order to be 70°„ 80° and 90°C as in the following series. This experiment could not be repeated due to the fact that the original stock of soybeans was exhausted. It will be seen that heating at 120°C im­ parts an even higher stability. In series 4.th e order of heat treatments increasing the stability is 70°, 80°, and 90°C, if only the three lowest temperatures are considered. Clearly,heating at 90°C, has given the best results. If we now compare the results of series 3 and 4 we see that in general the values of series 3 are higher. This means that -contra­ ry to batch heating- lengthening of the holding time increases the stability, except for the heating experiment at 90°C, for 45 sec. As has been explained we may expect that this value would be higher if the stability determinations had been carried out in time. A comparison of the stabilizing effect of batch (series 2) and continuous heating (series 4) will show (fig 22) that continuous heating imparts a higher stability to the evaporated product. The difference in stability .seem s to become greater with an increase in temperature.

Thus it appears that preheating has a systematic influence on the heat stability of the evaporated product. In comparison with the unheated milk the evaporated milks preheated with the vessel do not much improve their stability. The regular influence of preheating, however, is undeniable. The continuous sterilizator imparts a much more pronounced stabilizing effect to the milks.

The overall effect of preheating on the viscosity is depicted in fig 23. In order not to overcrowd the picture only the curves of 70°C and 90°C are drawn. As may be seen from this figure as "ell as from table 5 the curves.shift to the right in the order series 1 - series 2 - series 3 - series 4. The same order applies to the 80°C- curves.

94 The overall effecto fpreheatin go nth ehea tstabilit y isshow n in fig24 ,wher e onlyth ecurve s of90 °an d80° Car edrawn .Th e orderi nwhic hth ehea ttreatment s increaseth estabilit yi sserie s 1 -serie s2 - serie s4 - serie s3 .A sha sbee npointe dou tbefor e the ordero fth elas ttw ocurve so f90° Ci sreversed .

Theexperiment s which have been duplicated, 70°C4 5sec ,70° C 0 secan dth eunheate d milk,ar eshow ni nfi g25 . In accordance withth eforegoin gdiscussio nw ewil ljudg e theirvariatio n range atth e10 0c Plevel .Fo rth e70° C4 5se cduplicate sth edifferenc e is lesstha n 1%,whil efo rth e70° C0 se ccurv e this variationi s about1.5 %o fdr ymatter .Th eunheate d milk givesa largervaria ­ tionrange ,abou t2% .Pro mthes eduplicate son eget sth eimpressio n thatth emor eth emil ki sheate dth eles si sth evariation . Judging fromthes e resultsth eviscosit y curvesma yb eregarde da srepro ­ ducible.

The foregoingdiscussio n leadsu st oth efollowin g conclusions: 1.Ra wsoymil kca nb econcentrate dt oa tleas t30 %withou tdiffi ­ culties. However,thi sproduc tca nno tb esterilized .

2. Heatingth era wsoymil k prior toevaporatio n gives riset oa viscosity barrier,whic hmake si timpossibl et oconcentrat eth e milk furthera sth eevaporate d milkcanno t flowi nth eevapora ­ tor anymore .Th edr ymatte r contenta twhic h this viscosity barrierarise sdepend supo nth emetho do fheating .

3. When a steamjacketed vessel isuse dt oforewar m thera wmilk , lowering theheatin g temperature from 90°Ct o60° Ca swel la s reducingth eholdin gtim e from3 0t o0 minute smove sth evisco ­ sity barriert ohighe rdr ymatte rcontents . 4.Whe n the continuous, flash sterilizator is usedt oforewar mth e milk loweringth eheatin gtemperatur e from 120°t o70° Ca swel l as reducingth eholdin gtim efro m4 5t o0 second smove sth evis ­ cosity barriert ohighe rdr ymatte rcontents .

5. In comparison withth eviscosit y curveso fth evessel ,th ecur ­ veso fth econtinuou s sterilizatorfo rth ecorrespondin g heat treatmentsa t70° ,80° , and90° Car elocate da thighe rdr ymat ­ tercontent s thanthos eo fth evessel .

6.Usin ga vessel ,fo ra holdin g timeo f3 0minute sa swel la s0 minutes,th ehea tstabilitie so fth eevaporate dmil kar eincrea ­ sed inth eorde r 70°,90° ,an d80°C .Reducin gth eholdin gtim e

95 from 30 to 0 minutes increases the stability. 7. Using a continuous sterilizator, for a holding time of 45 seconds as well as 0 seconds, the heat stabilities of the evaporated milks tend to be higher in the following order of heat treat­ ments: 70°, 80°. 90°C. Increasing the holding time from 0 to 45 seconds tends to increase the heat stability. 8. In comparison with the heat stability.curves of the vessel those of the continuous sterilizator for the corresponding heat treat­ ments at 70°» 80°, and 90°C are situated at higher stabilities than those of the vessel. 9. In the range 70°, 80°, and 90°C the overall effect of preheating is the shifting of the viscosity curves to the right, i.e. to higher dry matter contents, in the sequende* vessel 30 min-vessel 0 min-flash 45 sec-flash 0 sec. After these conclusions we should consider again our starting point, where we assumed that -notwithstanding differences in struc­ ture between soyan d animal milk- principles of preheating valid for animal milk might prove useful in application to soymilk. As presumed, heating raw soymilk prior to evaporation does effect the viscosity of its evaporated product. With cowmilk, preheating may increase or decrease the viscosity of the concentrated product, whereas with soymilk the only effect of preheating is an increase in viscosity. Thus the influence is an unfavourable one. Also the heat stability appears to be influenced by preheating. It may be said that in general preheating has a favourable effect on the stability. The opposite effects of preheating on these two properties of the concentrated soymilk constitute a difficult problem that will be discussed in the last chapter of this work.

Addition of salts to the evaporated•soymilk.i

Prom the literature on animal milk it is known that the heat sta­ bility can be influenced by salts.With the purpose of increasing the heat stability of evaporated soymilk some experiments concer­ ning this point were carried out to get an insight in the eventual role of salts in the sterilization process. . The experiments were carried out as follows: 0.75, 0.50, and 0,25 ml of 0.5n solutions of sodium citrate and potassiumphosphate were pipetted in separate flasks, adding 0.25, 0.50. and 0.75 ml water respectively. The same was done with 0.05n solutions. To,each of

96 theseflasks ,containin gon em lo fsal tsolutions ,5 0m lo fcol d evaporatedsoymil kwa sadde dan dthoroughl ymixed .O nthes emilk s heatstabilit ydetermination swer emade . Inth efollowin gtabl esom edat ao nth esal tinfluenc ear eshown . Inthre ecase sth esalt sha dn oinfluence ,whil ei nth eothe rthre e adefinit eeffec tcoul db eobserved .Thi srati oi sno trepresenta ­ tivefo rth etota lnumbe ro fsample sinvestigate da sthes edat awer e choseno npurpos es oa st oillustrat eth edifference si nsal tin ­ fluencefound .I twil lb eeasil yunderstoo dtha tfro mthes elimite d datan ogenera lconclusion sca nb edrawn .The yonl yindicat etha t researchalon gthes eline smigh tb euseful .

TABLE 7 INFLUENCE OP THE ADDITION OP SALTS ON THE HEAT STABILITY OFEVAPORATE D SO YM IL K

o Heat treatment 120°C 60 C 60 C 90°C 90°C 45 sec 30 min 0 mln 30 min 0 Bin

Dry matter 18 . 9% 18. 1% 20.8% 11.4% 12.8% 15.5% Blones 18 B 2.4: 1 B 42 2 B 25 20 B 14 28 B 55 6 B 57 19 B 18 1 m 39 2 « 24 19 B 55 29 B 5 7 m 5

Na-citrate 0.75 B1 D.5 •» 20 B 50 1 » 25 2 a 34 48 a 0 48 24 34 50 0.50 ml 22 B 50 1 B 46 2 a 36 38 a 0 21 37 37 43 a 30

0.25 B1 21 B 34 1 B 33 2 a 24 32 a 30 19 B 40 36 25 33 a 0 0.75 nl 0.05 " 20 m 28 1 m 26 2 a 26 22 a 30 18 a 50 10 28 19 a 53 0.50 al 20 B 38 1 B 39 2 a 26 17 a 30 19 B 58 33 28 15 a 30

0.25 B1 18 B 37 1 a 39 2 a 26 21 m 18 19 B 58 33 26 50

K-phosphate 0.75 Bl 0.5 n 24 B 0 1 « 41 2 B 14 60 a 16 B 0 22 B 4 36 22 15 a 49 0.50 ml 21 B 58 1 a 23 2 a 33 60 a 12 a 54 24 B 0 36 30 44

0.25 B1 23 B 2 1 a 48 2 a 24 60 a 11 « 41 22 B 5 44 26 55 0.75 nl 0.05 " 23 B 2 1 B 48 " 13 a 16 21 0.50 ml 20 B 6 1 a 48 • 10 a 2 22 B 10 47 10

0.25 B1 21 B 58 1 B 42 • 10 B 3 24 B 43 40 17 1no effect 1 no effect 1 no effect 1 effect 1 effect effect

97 5,4 Summary. • •

The existing contradictory datao nth eevaporatio n ofsoymil k promptedth ewor k done on.thi ssubject.Som eauthor s state thatsoy ­ milkca nb eevaporate d without difficulty,whil eothe rsource scon ­ clusively proveth eexistenc e ofa viscosit y barriera tabou t20 % drymatte r content. Therefore thebehaviou r ofth eviscosit yan d also thehea t stability were studied. Asi ti sknow n thatth eviscosit y andhea t stabilityo fevapora ­ ted animalmil kar esensitiv et oheat,a nattemp twa smad et oinflu ­ ence these twopropertie s ofevaporate d soymilk byapplicatio n of heatprio rt oevaporation .Ra wsoymil kwa ssubjecte d tofou rserie s ofhea t treatments.Th eheatin g experiments werecarrie dou twit ha steamjacketed vessel,an da speciall y designed continuous, flash sterili*ator.Th eexperiment s were designed insuc ha wa ytha ta possible effecto fth ecoming-u ptime ,temperatur e level,an dhol ­ ding timewoul d manifest itself. Inou rcas eth eheatin g provedt ob eth efacto r that determines thepositio no fth eviscosit ycurve san dthu sals o thato fth evis ­ cosity barrier.I ngenera li tma yb esai dtha tth eles sth eheatin g thehighe rth edr ymatte r content thatca nb ereache d byevapora ­ tion.Thi shold sfo rth etemperatur e levela swel la sfo rth ecoming - uptim ean dth eholdin gtime . Iti sinterestin gt onot e that ifth e rawsoymil k isevaporate d without having beenpreheate d even milks ofmor e than40 %dr ymatte rd ono tsho wan ysig no fgelling . Itseem s thatw eno wca nbridg eth etw ostandpoint s concerning theevaporatio no fsoymilk ,whe n taking intoaccoun tth einfluenc e ofheatin go nth eviscosity .Th era wsoymil kca nb econcentrate dt o at least30 %dr ymatte r content. If,however ,th emil k isheate d priort oevaporatio n this heating will give rise toth eexistenc e ofa viscosit y barrier,th epositio no fwhic hdepend s uponth e me­ thodo fheating . Alsoth ehea tstabilit yo fevaporate d soymilk provest ob einflu ­ enced bypreheatin ga sexpected . This influence depends uponth e coming-up time, temperature level,,an dholdin g time ofth ehea t treatment.

6 SOYCURD

6.1 Introduction.

From soymilk aproduc t called "toufu"i sobtained . Iti sals o called tofu (Japan), taohu (Java), anddauph u (Indochine). This

98 product isobtaine d by precipitating the proteins of soymilk and pressing theprecipitate .Du et oth e analogousmetho d ofmanufactu ­ re ofchees e fromanima lmil k it isofte ndesignate d as soycurd or beancurd. Inth e followingw eshal lus eth e termsoycur d orshorte r "curd".A sha sbee npointe d out inth e introduction of the chapter onth emanufactur e ofsoymil kth eprocessin g ofsoybean s tosoycur d wasalread ymentione d atth e beginningo fth eChristia nera . Soycurd may bedescribe d as a product obtained from soymilk by precipitating themilkprotein s followedb ypressin g theprecipitate . Dependingupo n thecompositio no fth emil kan d themetho d ofmanu ­ factureth edr ymatte rconten to fth ecur d isabou t 15t o25% .Som e analyseso fth ecompositio no fth ecur dar e givenbelow .

Java Indochina Indochina Japan Philippines Prinsen Bloch Monnier Inouye Gibbs Geerligs 1906 1935 1894 1912 1896 frozen baked osmosis

Water 76.15% 83.85% 81.40% 15.32% 73.0% 55.76% Proteins 13.15 11.25 29.28 41.42 13.88 14.56 Fat 7.09 4.33 5.66 23.65 10.78 7.12 N-freeextrac t ; 1.40 0.64 15.05 2.22 2.13 Ash 2.21 0.57 0.67 3.08 1.2

The figures from Inouye refert o frozen tofu, those of Gibbs to baked tofuan d tofukep t ina sal t solution. Prom thesoycur d a number ofproduct s isderived , which incom ­ parisonwit hth eorigina lcur dcontai nles swater .Thi sdecreas e in watereonten t isobtaine d invariou sways:greate rpressure,osmosis , drying, freezing,an d baking.Als oproduct s areknow n inwhic hmi ­ cro-organismspla y arol e inth e ripeningprocess . When soycurd is to be produced on a factory scale a number of technological prerequisites must be fulfilled. First the starting material, soymilk, must havea reasonablyconstan t composition. In the experimental chapter on soymilk it hasbee n shown that this willcaus en odifficulties .Second , theprecipitatin g agent should give uniformyield swhe n theprecipitatio n conditionsar ekep tcon ­ stant.Thi smean stha tsom ekin d ofstandardizatio n ofth eprecipi ­ tating agentmus t beestablished . Third, theyiel d ofprecipitate d proteins must be ashig h aspossible . Asth e firstpoin twa salread ystudied ,no wattentio nwa sdirecte d towardsa standardizatio n ofth eprecipitatin gagen tan d toa stud y ofth eyiel d ofprecipitation .

99 6.2 Literature.

6.2.1 Methods of manufacture.

The literature on tofu is much more limited than that on soymilk. In the following some methods of manufacture are described. In Java according to PRINSEN-GEERLIGS (1896) after the soaked beans are ground, the mass is boiled and filtered through a large cloth. The coagulation is effected by means of sour soymilk. and two hours later the half solid mass is pressed between two boards. Then it is cut into small"'cake's. In Japan the proteins are precipitated with 2% of a mother solu­ tion, obtained from the salt manufacture out of seawater. The whey is filtered through a cloth and the precipitate is pressed slowly and then cut into pieces (INOUYE, 1894-97). BLOCH (1906) described that in Indochina the milk is boiled and after 10-30 minutes it is transferred into another container to whichthe coagulating agent is added. The curd is placed between two sieves coveredwith cloth, and dependent upon the desired pro­ duct more or. less pressure is exerted.. Then the cake is cut into pieces.

PIPER and MORSE (1923) using a 1% salt-solution of MgCl2- added one part of salt solution to four parts of milk. Whole soy- flour and defatted presscakes ga,ve as good results as the soaked beans. Prom several varieties of -soybeans the yield of soycurd was determined; the dry matter and crude protein content,of this curd, however, were not mentioned." INOUYE (1894-97) states that from the salt solution used the calcium- and magnesiumsalts are active but the sodiumchloride is not; on addition of calciumnitrate or magnesiumsulfate a precipita­ te is created,which is hot the case with even large amounts of so­ diumchloride or sodiumsulfate. An excess of magnesiumsulfate dis­ solves the precipitate again. Analysis of the salt solution showed the presence of 27.9% magnesiumchloride and 7.0% sodiumchloride. According to his calculations only about one fourth of the original amount of proteins in the beans was obtained in the soycurd. The manufacturers, according to BLOCH (1906), used two solu­ tions: an original solution and a solution for immediate use. This last one is a dilution of the original solution with four or five times the amount of water. The original solution is the residue of the salt manufacture from seawater. According to this worker 100 ml contain: MgCl2 29.21 % MgSO 1.12 Na^t^ 6.24

100 Noneo fbot hsolution scontaine d animalo rvegetabl e rennet. Ifth e original solutioni suse dth efloccule sar eto olarg ean dto othic k resulting ina cur d different ofappearanc e andconsistenc y from thenormal . Furthermore BL0CH (1907)state s thatth echloride sand'nitrate s ofcalcium , barium, strontium, magnesiuma swel la smagnesiu m sul­ fate have coagulating properties. This property isstronges t with magnesiumsulfate: "LesChinois ,e nemployan t lechlorur ed emagne ­ siuma l acoagulatio n du laitd esoj abouillant ,son tempiriquemen t tombe'su rl ecoagulan tdeselectio nd ec eproduit " Ifth eprecipitatio n iscarrie d outa ta lo wtemperatur e before cooking only with magnesiumchloride aprecipitat e is originated quickly and that only iflarg eamount sare'used ,~If'the .precipita ­ tion iscarrie d outa tlo wtemperature s after cooking twiceth e amounto fmagnesiumchlorid e isnecessary . InTonki n unpurified salti suse d forprecipitatio n (MONNIER, 1935).Thesolutio ncontain s2 0g o fsal tpe rliter.Thi sworke rto o states that magnesium ison eo fth estronges t precipitating salts. Itwa sfoun d thatprecipitatio n inth evicinit yo fth eboilin gpoin t proceeds fasteran dca nb ecarrie d outwit hsmalle ramount so fsalts . Ifth epreparatio no fdauph ui srepeate d every day, thesaltrso - lutionca nb ereplace db yth e"petit-lai taigre "o r"eau-mere "fro m thepreparatio no fth eforme rday . This liquidha sa p H2. 9an dha s the followingcompositio n (gpe rliter ) drymatte r 12.77g chlorides 0.24g nitrogen (total) 1.92 ash 2.62 oil 0.51 acidity 5.36 glucides traces The:liquid isstrongl y acidan dthu si tprecipitate sth eprotein s ofth evegetabl e milk.Th efollowin gacid swer e identified: lacticaci d present formic acid traces acetic acid traces

MATAGRIN writes thataccordin gt oBeltze r (1911)aceti c acid givesa powder y curdonl yafte r longboiling ; diluted sulfuric acid givesa clea rprecipitat eafte r short boiling,whic hha sa yellow ­ like colour;th eactio no fhydrochlori c acid liesbetwee n thato f both foregoingacids .

6.2.2 Ripening ofsoy cheese.

The literatureo nthi s interesting subject isver y limited.

101 Aneffor t toinitiat e ripeningi sreporte d by INOUYE (1894-97), whomixe d Swisschees e with 10%sodiumchloride . The results,how ­ ever,wer epoor . KA TA YA MA (1906)obtaine d symptomso fbacteriologica l ripening by mixing 450 gtofu , 60 gcaseine , 60 gsodiumchlorid e and 20g Swisscheese .Afte r storage at 15°C for fivemonth s thechees eap, - peared tohav e acompac t consistency and^formatio no f acrus t had takenplace.Th etast e isdescribe d asbein gagreeable .Notwithstan ­ ding:th e'additio no f 2 gmilksuga r nohole s -duet o gasformation - appeared. Increasing the amounts ofmilksuga r did not giveholes . Thecoars e surface,however ,becam esmooth ,whil eth e taste ofsoy ­ beanha dcompletel ydisappeared . • InRussi aBELENKI I andPOPO VA (1933)fermente dwar msoym iI k with milk Streptococci and cheese bacilli habituated to soymilk. After fermentationth eproduc twa sallowe d to-stand at33° Can d the curd obtained waspressed . According to the InstitUt International d8Agriculture (1936) the factoryCaseo-sojaine ,situate d inth evicinit y ofParis ,suc ­ ceeded inproducin g severalsort s -soft,hard ,ripened,not-ripened - of soycheese„ InGerman y experiments on giving soycheese the same taste ascertain-Europea ncheese swer e successful.'Thi swa s accom­ plished bymean so fselecte d cultures ofmicro-organism s used for thepreparatio no fEuropea ncheeses . SMITH (1948)describe ssom e sortso fcheese si nChin aa sfollows . Chee-fan isa brownis hsof tcheese . Itha sa nagreeabl e smellan d taste. The curd cubesar e inoculatedwit hmold , salted, and placed ina n appropriate storage'hous e forabou t seven'day sfo rdevelop ­ ment ofmold.iThe :mold -(Mucor ) isgrow n onwhea t flour. It exists inChin a inmol d ofwin e and iswhit e in colour. Also Aspergillus glaucus, blue incolour , apparently takespar t inth e cheesedeve ­ lopment. Next, the cubes are placed in an earthen crock orwide - mouthed bottle,art dyello w wine and mold ofwin e are added. It is allowed toag e inth ewin e forabou ton eyear . Tsue-fan ("drunkencheese" )i sanothe r typeo fcheese .Th ecur di s boiled inwater,cooled,an d partly dried.It ismolde d and placed in yellowwin e (ricewine )wit hwin emol dad'de dan dage d forsi xmonths . Hon-fan isa red cheese made inth e same manner as tsue-fan ex­ cept that soy sauce rather than wine isuse d in aging thecheese . Significant variations inth eprocesses ,beside s themicro-orga ­ nisms, areth eproportio no fsal tan dth etyp eo fsolutio n inwhic h the cheese isaged . The cheese appears tovar y somewhat with the locality inwhic h it isproduced , avariatio n due probably toth e influence ofclimati c conditions onth eactivit y ofth e fermenting organisms.

102 6.3 Experimental part.

6.3.1 Scope of work.

Asma yb esee n fromth ediscussio no fth eliteratur e thedetail s ofmakin gtof udiffe r locally. Ingeneral ,however , iti spossibl e todistinguis h fourdifferen t stagesi nth epreparatio no ftof uan d derivatives. Inth efirs tstage ,th eprotei n isprecipitate d fromth emil kb y theus eo fa precipitatin g agent. Iti sdesirabl e that this agent hasa constan t composition. Besidesth eoutpu to fth eprecipitatio n shouldb ea shig ha spossible . Inth esecon d stage,th ecur di sseparate d fromth ewhey . Inprac ­ tice this isdon e byfilterin g througha clot ho ra sieve , after whichth ewhe y isdiscarded .Th ecur dha sstil la ver y loosestruc ­ ture. Inth ethir d stage,th edesire d amount ofwate ri sremove d from the curdb ypressin go rotherwise . Inth elas t stageth ecur d isinoculate d with micro-organisms, thus initiatingth eripenin gprocess . Experimental worko nth efirs t stage,precipitatio no fproteins , isknow n from literature.Thi swor ki smostl y qualitative innatur e and only littleattentio nwa spai dt oth equantitativ easpec to fth e precipitation process.Therefor ea stud ywa sundertake no fth eyiel d of protein precipitation, usingvariou sagent sfo rcoagulation . No researchworkwa sdon eo nth eripenin gprocess ;onl yth epreparatio n ofth efres hcur dwa sinvestigated .

6.3.2 Desigi of experiments.

Inactua l practice onlya limite d numbero fagent sar euse dt o precipitate theprotein s fromsoymilk . These agentsar eth emothe r liquid from seawater, sour milk,an dspontaneou s souring. Itha s been shown thatth eprincipa l agents inth emothe r liquid causing precipitationar eMg ,Ca ,CI ,an dS 0 4 -ions. Thereforew euse d threesalt s MgCl,*,MgS0 4, CaCLj,an dlacti caci d asprecipitatin g agents. Different concentrationsan dtemperature s were used. The influenceo fth edr ymatte r contento fth emil kwa s also studied. Theprecipitatio nwa scarrie dou taccordin gt o«n e uniformmethod , while alsoth eseparatio no fth eprecipitat ewa salway sdon e inth e samewa yb ymean so fa centrifuge . Determinations ofdr ymatter ,crud eprotein ,pH ,an dacidit y were carriedou to nth eorigina lmil kan do nth ewhey .Th eamoun to fdr y

103 matteran dcrud e protein inth ecur d isobtaine d bysubtraction . Allexperiment sa swel la sth edetermination swer edon ei nduplica ­ te.Thu sth efigure so ftabl e5 ar eth eaverage so ffou rdetermina ­ tions. Intota ltwelv eprecipitatio n serieswer edesigned .

Inth efirs tserie s MgCl2wa suse da sa precipitatin g agent.Pro m a IN solution 1,2,4,an d8 m lwer e added tp10 0g o fsoymil ko f 80°C. Roughly thesal t concentrations were 0.01,0,02 ,0.04 , and 0.08N . Inth esecon d seriesth etemperatur evarie d from40 °t o90° C usingtw om lo fth esam esal tsolution .

Thesam ekin do fprecipitatio nwa scarrie dou twit h MgS04 (series 3 and4) , andCaCl 2 (series5 an d 6). Toapproac hth eeffec to fth emothe r liquid equal amountso f1 N solutionso fth eforme rsalt swer emixe dan dvaryin g amountso fth e resulting liquiduse d forprecipitatio na t80° C (series 7). Intryin gt orais eth eyiel dth esam e amountsan dtemperatur e were used,bu tth ep Hwa slowere dt o4. 5 (series8) . Inserie s 9th einfluenc e ofth ep Ha t20° Cwa s investigated, while inth enex t seriesth etemperatur ewa svarie d ata constan t pH4.5 . Inth elas ttw oserie s anotherdr ymatte r content ofth emilk , about 10an d4% ,wa suse dfo rprecipitatio nwit hMgSO ,a t80°C .

6.3.3 Technique of experiments.

Theprecipitatin g agentwa sadde dt o10 0g o fsoymilk ,place d ina 325m lcentrifug e tube.The-milk'an d thesolutio n were well mixed witha glas s stirrer. Ifhig h temperatures were used thetub ewa s placed ina water-bath ;whe n adjustingth ep Ho fth emil kth eelec ­ trodeso fth ep Hmete rwer e placed inth emil k while thesolutio n was beingadded . Immediately after additiono fth eprecipitatin g agentth econten to fth ecentrifug etub ewa scoole dt o20° Cb ymean s ofic ewater .The nthe .tub ewa spu tint oa Marti nChris tcentrifug e andcentrifuged for5 minute s atabou t 3000r.p.m . (diametero f centrifuge4 0cm) . Theweigh to fth ewhe yan dcur dwa sdetermine d andth edry .matter, - crudeprotei ncontent ,pH ,an dacidit yo fth ewhe ywer e determined as described forth emanufactur e ofth emilk .Th eindicato ruse d forth etitratio no fth ewhe ywa sphenolphtaleine . Thedr ymatte r andcrud e protein content ofth eorigina l milk wereals odetermined.B ysubtractin gth edr ymatte ran dcrud eprotei n inth ewhe y fromthos e inth eorigina l milkth epercentage.o fth e drymatte ran dcrud eprotei ni nth ecur d couldbe'calculated .

104 6%3t4 Results, discussion, end conclusions.

Before starting the discussion some figures are given which may serve to illustrate the degree of accuracy of the determinations and also of the reproducibility of the experiments. Below are given the complete figures of series 2.

whey cu rd

grams % dry crude PH aci­ % of orig. % of % of % of matter protein dity dry mat­ orig. orig. orig. ter in crude dry crude whey prot. matter prot. in whey in curd in curd

Whey 2.1,1 71.50 5.04 1.77 6.05 1.39 44.23 37,35 55.77 62.65 40°C „» - - 5?02 .* .., 6.02 1.35 .. li 2 71.55 4.19 1.57 6.03 1.37 38*19 31.37 61.81 68.63 4t,17 1.56 6.02 1.34 71.53 4.61 1.63 6.03 1.36 41.21 34.36 58.79 65. 64

Whey 2.2. 1 75.85 3.26 0.94 6.03 1.16 30.34 20.88 69.66 79„12 60 C 3.25 0.94 6.01 1.15 •2.2 69.10 3.04 0.92 6.03 1. 11 26.69 17»65 73.31 82.35 3.02 0.90 6.02 0.96 72.48 3.14 0.93 6.02 1.10 28.52 19.27 71.49 80.74

Whey 2.3.1 66.50 2.79 0.53 6.08 1.03 22.85 10.29 77.15 89.71 80°C 2.79 0.53 6.08 1.0* •3.2 67.20 2.72 0.55 6.03 lo04 22.37 10.36 76.63 89.64 2.71 0.54 6.02 1.03 66.85 2.75 0o54 6.05 1.03 23.11 10.33 76.89 89.68

Whey 2.4.1 66.15 2.69 0.50 6.09 0.92 21.87 9.71 78.13 90.29 90 C 2.68 0.50 6.08 0„89 •4.2 65.10 2.75 0.55 6.03 0.99 22.86 10*08 77.14 89.92 2.75 0.55 6.02 0.97 65.63 2.72 0.53 6.06 0.94 22.37 9.90 77.64 90.11

Starting milk dr.m. 8.15-8.13. cr .prot. 3.42*3.38' acidity 1.75-174. P» 6-36-6.33 dr.m. 7.84*7.81. <=r .prot. 3.64*3.50. acidity 1.73*1.69. -'•" PW„6-40-6.39

The experimentalresult sar esummarize d intabl e8 an di nfi g2 6 and27 .

First,th einfluenc eo f salts (andtemperature )wil lb ediscusse d nextth e pH (andth etemperature )an dthe n the combined action of salts and pH.Last ,th eeffec to f the dry matter content isdiscus ­ sed. Onconsiderin gth eresult sobtaine di tappear stha tth ecurve sfo r the salts MgCl„, MgSO, CaC l, an dth ecombinatio no fthes ethre e Os OS Os f* 00 Ol Ol b c o -H

wt^rft^ a a a ^ o CM « o N* H « ota n HI •* o> •"• m a 3 « w Mui t-» v oo >o oo >o ot in w«.n ^ w r*. oi oo i-t m t*4 f» u >> »-• co i-i t-» oo ton io n r* CN r> c-> t> I'D -rt

ct) o i. a t» to K co o v r*. co w oo io «H ** U)Q NU} o •> v t> o in (ON(0o> W Ol I> T* OhOO Qto ^ IO

-4 *» U -P >, ** •«• t*. oi *-» N 1-4 t* OSNO ffl f* Ol w »-c w oi in MOW o a V H tNO ci in »4 co CI CO © io Ol «N o i-4 van B £ * co « m . OlNtNpi *C4 MM t- CI CM D HdNfl SHMM »• C nciii sco -!

>> * J t J Jo;* t--4CN0 oc«is.io oo^

>» 2225 mcjinio o»oi» ^imt^* eN-.»t> e e Mn c >a « X e *» eoinm loioioe •£ n ooo* uimnm »»'»>! « c o ft •0 o e l< • -4 s h •0 » .-» S2Di SS.tS ^""^ COOJMOI Oinooa on*^ o « 3 *» JR l« 0 *-* .*** "oiinio Huiinn menoin •# S v 5- m en m in O b co o o o <-iooo mo o o ^ooo . • u ?222 S22N -W-wr, ««»„ ««>io o*ioo> *> "0 l> CO 00 O "INN n 00 COO -ICHOOOO n»»« r»o>»oo o •a a w <~NMnr >. 3 u SS.#2 ^?SS §£2? S§22 ssss ss§„ 0 >• 3 • Of -4 M a » w SSSS SSS2 -SK» «^»« »»oi «««^^ a -e u * oi«'i>i> NI>IOIO coe N t. i-t-0,0 i.eNN IOI-IOVO ou •e ft• * "H •? o o -< o 0 0 0 0 B o 0* o 0 B 0 O 0 o i. « oooo oooo oooo • « *» H Mt CO * , « B ft t. • CO •o +» a e « a. o CD u •» 3 o a T! * c ft -< e •H u 3 m o *> « O » «H h **4 C ft Q M

106 CO CO •* 00 a o in M N H N 00 in oo ci Hin HOI oo oo o m H IN * rt P500 N«N nn o oi c> w M • •• * OlOiOO O W CI O V OtOl *H CO t- 00 Oi Oi Oi'Cft •> 00 oo *o 00 00 Ol Oi Ct 00 00 00 00 00 00

0> OOt «3 O *H O (O IH o w) W ro Oi W CO *f m n ^ o» v>

C^ 00 « •* ci o •* *• ot oi m

oo »H *H »H 01 H H H

o vo o m

Oi mo N m co © m oo t- m oo ©'•••is •»! CO Oi Ct co 00 O Ol 00 r* c» oi oo o *-i co *-« CO CI CI oi in tn m

CM CI CI d

c- io oi oo cowoot

H P)0 O) * VO ffl N o^-cit«» OlO^-CO ^- 00 00 CO * *r *• "*• o m o in to m ^ «t ^i> CO CO CI CI

>a . o

c» ID m »o Oi^MO Ol *0 *" •«• OMNM o i> m o ^r io m oo C> Ol **• CO Ol O CO •"* O• O»l * o •00 *** © Ol o 00 H t* Oi o m -or « COCOMCO oo ^r co co

O O 00 in uf in o CO CI Oi 00 «-i 00 o m CO co »n tn 00 O fO, o CO o N N« CO ct CI Oi m ** *-t 10 CO W *-* e- CO tN t* C| 00 tttw« o IOOH T-t in CI CO «o Oi to m M ^ •«t CO *• o CI Ol Oi 00 oo •^ ^fr­ 00 to Oi c* 00 00 oo «o c* c* s t- et* 00 00 00 c* r-* « oo «n *>• o

»-t o o o o s 8 in m w o t0 0 C > 0 in +» o oo o 9 *-l CI 00 0 HNt 00 «-i CI •* 00 tn• •* CO fO *• *o«o Oi * o •-< « ^r *0 e ,i o« o *o -i © c ~x> *0 u, U h •o •o o o u 3 3 •H -ft h 00 U 00 a +» - +J o o m o K K a a ^s *> G ••>* c —I •w m^_^ in +» ^ 4J *-* a o 0 0 oc e e •H -4 B _««• -H ^ •P o * a o •* O 4J >>co 01 10 oo o X X u m 4J Q O C3 a o a a tn co CO %^» J CI J Q 3 a s a

107 a dry matter in whey. • dry matter in curd. o crude protein in whey. • crude protein in curd.

•/. 100

90 9 f-*—• ^--» •-—.«—_-. —* / «-—•-—« ,-*" 60 X »—-• .-*" —• 70 '•••? jf—-- «' ,j»;' ^ i * • /' 60 iii,' *' / i, 11 •* W; * \' /1 >*>\,; / 50 'ii 1 40 'A q. I 30 // \\ -a '•a. l <\ t "~Ck. li - -a „ %va. 10 • i2, .—.o—-- o '•o-,-0 6.—--o-— o "'O-T --o 0 12 4 8ml 40 60 80 90C 1 2 4 8ml 40 60 80 SCCC 12 4 8ml 40 60 80 90*C

MgClx( 1n,80°C) 2.Temp.(MgCI,1n.2ml) 3.MgS04( 1n,80t) 4.Temp.(MgSO,1nl2ml)5.CaCll0n8C(t) 6,Temp.(CaClt1n,2ml) Pig 26. Precipitationo fprotein s from soymilk. Distributiono fdr ymatte ran dproteins ^i ncur dan dwhey , Influenceo fsalts , concentration, andtemperature .

• dry matter in curd. dry matter in whey. • crude protein in curd. o crude protein in whey. : •1 1 . •.- . •' 1 00

»—•—•—.. .*---•• •'1 • 90 ,9 • 80 :r"\ B r" AT" 70 .-•:-"'•--»,.-.. / •, • - V ft'*-" • t ,V--»- .''

50 ir/ ! • X". I•••-. ' 'Ml 40 A n% V--°' li 't'i a 30 \ ,p o~--a... • . •J 'A-—-D--' a...^-^;-° '-0—.^ 1%.-*-° 20 /L-— t \ " \ 'w' k. ! 6.. 10 • * r &-—-O-—-O ''-o—--0 O....-0--~o — -o "'••o—-o 0 . . . i . 12 4 8ml. 1 2 4 8ml 30 35 45 55ph. 40 60 80 9CTC 1 2 4 8ml. 0,5 1 2 4ml. 1 MgCI,«MgSC^Caa B.MgCI^MgSCVCoCljft Lactie acid. 10. Loctie acid. 11.Mil k 10'fcdrm. 12.Mil k 4 7. dr.m. 80"C 80'C ph.4 5 20*C ph.4. 5 Pig 27. Precipitationo fprotein s from soymilk. Distributiono fdr ymatte ran dprotein si ncur dan dwhey . Influenceo fsal tmixture ,pH »concentration , temperature, anddr ymatte rconten to fmilk . 10S salts (seriesl ,3 ,5 ,an d7 )sho wa marke d resemblance..These four serieshav ei ncommo n thatwhil e additiono f1 m lo fsal t solution precipitates very little,7 5t o80 %o fthe.dr ymatte ro fth emil k isobtaine d inth ecur db ydoublin g thisamount .Th eyiel do fdr y matter inth ecurd ,however , doesno tris e ifth equantit yo fth e saltsolutio ni sincreased . Ingeneral ,th esam ei stru eo fth eprotein .An ,importan tdiffe ­ rence, however, istha tth eprotei nyiel d figuresar econsistentl y higher than thosefo rth edr ymatter . Evidentlyth eprotein sar e better precipitated, asmigh thav e been expected.. Aris e inthe amounto fsal tsolutio ndoe sno tincreas eth eyield .Al lfou rcur ­ ves show about thesam e maximum yield which averages 90%'ofth e original crudeprotei n inth emilk .O nusin g2 m lo fsal t solution thecur di ssoli d (thoughsoft )andth e wheyi sligh tyello w trans­ parent, whereasth ecur dha sth eappearanc eo fa viscou sliqui dan d thewhe yi smilklik ei f1 m lo fsalt/solutio ni sused ._ — Consequentlyth eprecipitatin g effect of:each salt aswel la s thato fth ecombinatio n isquantitatively , practically thesame . This implies thatt oobtai na hig hyiel d - it is not necessary to use such a complicated solution as that applied by the population. Aso nextractio nth eprotein sar edisperse d better thanth eothe r constituents of the:dr ymatte ro fth ebean ,an dthe-protein sar e precipitated bettertha nth eothe rconstituent so fth emilk ,w ema y statetha t the isolation of soyproteins, by extraction followed by precipitation can give a high yield. ••'" ' When consideringth einfluenc eo ftemperatur e onth eactio no f .theindividua l salts (series2 ,4 ,an d6 )w efin d again amarke d resemblance.Th ecurve sfo rth eyiel do fdr ymatte r riseregularl y from about60 %t o80% . whileth eprotei nyiel d figuresar ehighe r than thosefo rdr ymatter .Raisin gth etemperatur ecause sth edif ­ ferences't oincreas efro mabou t7 %t o 13%.

The influenceo fth ep Ho nth eyiel do fprecipitatio n at20° C followsa nunusua l course (series 9). Whileo ndecreasin gth ep H the rise inyiel d ofdr ymatter -wa ssmalle r than expected, it appeared thatth eamoun to fdr ymatte rprecipitate d decreasedi fth e decreasei np Hwa scontinued. 'Th emaximu myiel dwa sabou t55% . In thiscase ,too ,ther ewa sa mor emarke d reactiono fth epro ­ teinst oth eprecipitatin g agents otha ta nappreciabl yhighe ryiel d (85%)wa sreached . Fromth egrap hi tca nb eclearl y seen thati f thep Hdecrease st o4. 5th eprotein sar eprecipitate d ina large r degree thanth eothe rconstituent so fth edr ymatter . Ifa tp H4. 5th etemperatur e israise d from40 °t o90° C (series

109 10). it appears that the yield of the dry matter as well as that of the crude protein increase regularly. The difference in yield re­ mains constant when the temperature is raised. The highest yield of proteins was 94.0%. As mentioned before MONNIER (1935) states that the pH of the coagulating agent is 2.9. As this pH'increases due'to dilution this makes it understandable that an important part of the precipi­ tating effect may be ascribed to the low pH as obtained with' the empirical method used by the population of some East and South East Asiatic countries. :Furthermore''.th e precipitation is often carried out while the milk is boiling. If simultaneously a low pH (4.5). high temperature (E0°C) and salts are applied (series 8) a rather good reproduction of conditi­ ons of the empirical method-as applied,by the population- is ob­ tained. The curves for the yield of dry matter and crude protein remain almost horizontal, the protein yield figures being about 18-20% higher than those for the dry matter. The protein yield fi­ gures do not exceed 92.4%. Hence the empirical method may be consi­ dered very effective.

If milks with a higher or a lower dry matter content are used,10% and 4% respectively, it appears that in principle the course of the curves is the same as that for 8% milk precipitated by a single salt. With-the thin milk 1 ml of salt solution proved to be suffi­ cient to obtain a solid curd, clear whey and a reasonable (85%) yield. As pointed out before,this amount of salt solution is not sufficient with the 8%milk , for which about 2 ml are needed.- Howe­ ver, with the io% milk it will be seen that these 2 ml are already insufficient. Apparently, as the dry matter content of the milk is increased more salt is needed to obtain efficient precipitation.

INOUYE (1894-1897) states that about one fourth of ttietotal amount of proteins in the beans is obtained in the curd. LOSKA and MELNICK (1950) obtained 54% of the original proteins in the curd by adding 5 ml 0.2 n MgCi2 solution to 25 ml of their milk. With another method they."increased' their, yield'toT74%, the curd being fluid, however. According to these workers the maximum yield for precipitation with MgCl2 is between 0.01 and 0.03 n. This is in agreement with our experimental results. It is interesting to note that on extraction of the soyproteins with MgCl^solutions a minimum yield is obtained at 0.025 n, ac­ cording to SMI TH, CIRCL E , and BROTHER (1938); at 0.010 n almost 30% more is extracted, whereas at 0.05 n only about 4%mor e is extracted. In general this is confirmed by our results.

110 Promth eforegoin gth efollowin gconclusion sca nb edrawn .

1.Ther e aren oappreciabl e differences between MgCl9, MgS04,an d CaCl^, asregard sthei r precipitating actiono nth eprotein so f soymilk.Th ecombine d actiono fthes e three salts doesno tsho w any essential difference either. About 90%o fth ecrud e protein canb eprecipitate d byth eus eo fa singl esalt . 2. Until upt oabou tp H4. 5mor e protein isprecipitate d ifmor e lactic acidi sused .A ta lowe rp Hles sprotei n is precipitated. Ifth etemperatur e israise da tleas t 90%o fth eorigina l crude proteino fth emil k isobtaine d inth ecurd . 3.O nlowerin gth edr ymatte rconten t ofth emil ka les sconcentra ­ ted salt solution suffices,o nraisin gth edr ymatte r contenta stronger salt solutionmus tb eused .

6A SUMMARY

Soycurd,a produc t derived from soymilk,isprepare d inman y ways. All thesemethod so fmanufactur e havea numbe ro fstage si ncommon : 1.Th eprecipitatio no fth eproteins ,2 .th eseparatio no fth ecurd , 3. thelowerin go fth ewate rcontent ,4 .th eripenin gb ymean so f micro-organisms. Inthi s chapterth efirs t stage,precipitatio no f theproteins ,wa sstudied . Theprecipitatin g effecto fa numbe ro fsalts , MgCL,, MgSO., CaCl,, lactic acid, ando fthei r combinations wasstudie d atdifferen t concentrationsan dtemperatures .Th eeffect so feac ho fthes e salts proved tob epracticall y identical, while lactic acid produced a maximum yielda tabou tp H4. 5 (20°C). Whenth edr ymatte r contento f the milk waschanged , thecours eo fth ecurve s remained thesam e and proved onlyt ob eshifte di na horizonta ldirection ,

7 SOY- YOGHURT

7.1 Introduction,

Preserving animal milk bymean so fmicrobiologica l souringha s probably been practised ifro'mancien t timesi nman y areaso fAsi a where nomadic tribeswer e living.A smor e thanon esou r milk pro­ ducti sknow ni nth e followingw eshall-confrneourselves't osourin g bymean so fbacteri a foundjUheso-calle dyoghurt. . Thoughth eproduc tma yhav ebee n longknown ,accordin gt o LITT- MANN (1934)i twa sno tbefor eth e1 7 century thatstatement so n itsus ea shuma nfoo d becameavailable .B ywa yo fTurke yth eprepa ­ rationo fyoghur t seemedt ohav esprea dt oth eBalca ncountries .I n

111 WesternEurop eyoghur t becameknow nno tearlie r thani nth ebegin ­ ningo fthi scentury . Asha sbee nshow ni nth echapte ro nhea tpreservation ,soymil ki n a rawconditio ni sver ysubjec tt odecay . Thereforei ti susefu lt o investigatewhethe ri ti spossibl et opreserv esoymil kb ymean so f bacteriologicaltransformations .Needless't osay'that'th ebacteri a constitutingyoghur twil lhav et oaccep tsoymil ka sa substrate *

7.2 Literature.

Withth eexceptio no fa Frenc h statement anda reference toa Russian investigation, hardlyan yliteratur e isavi^dabl eo nsoy - yoghurt. MATAGRIN (1939)state stha tvegetabl emil ki sprobabl ysubjec ­ ted tothi s kindo fprocessin g inthos e eastern countries where Bhudism prohibited theus eo fco wmilk . Accordingt othi s worker inEurop ea swel la si nAmeric ather ehav ebee nonl yfe winvestiga ­ tionsconcernin gthi spoint .H edescribe s thepreparatio no fsoy - yoghurtan dsoykefir .However,i ti sno tclea rwhethe rthes emethod s ofmanufactur ewer edevelope d afterexperimenta lwor ko fth eautho r orwhethe rthe ywer ecited . Accordingt oMatagri n inth epreparatio n of soy-yo$iurt thesoy ­ milk iscpncentra€e dt oon ehal fo ron ethird , afterwhic h2- 3m l "rnaya"(Bact .bulgaricu mBarthel )o r 15m lo fyoghur t isadde dpe r litreo fmilk .On em lo fa streptococcu s culturei srecommende dt o avoida peptone-taste .Afte rmixin gth etemperatur eo fth emil ki s maintaineda t35-40° Cunti lcoagulatio noccurs .The nth eproduc ti s cooled. MILYUTINA (1948)produce d soykefir.by adding4 %Dispor acau - casicat oth esoymilk . Thebeverag eha da n unpleasant';tast ei f pure soymilk wasused . Separationo fth eprotei nwa sprevente db y theadditio no fabou t 0.05%o fa naqueou sagar 2solutio ni'usttbefor e souring. Asma yb esee nth eliteratur eo nthi ssubjec t isver ylimited .How ­ ever,w ehav ego tth eimpressio n that lactic acid bacteriad oac ­ ceptth esoymil ka sa substrate .

7.3 Experimental part.

7.3.1 Scope of work.

Judging fromth escarc e literature dataavailabl eon emigh tsa y that therei sa reasonabl e chance that lactic acid bacteriawil l

112 grow-an dperhap seve n growwel li nsoymilk . Therefore the first point wehav et oconfir m is,tha t ifw etak e the twolacti caci d.bacteri apresen ti nyoghurt ,th eThermobacteriu m bulgaricuman dth eStreptococcu s thermophilus,an dinoculat e them theni nsoymilk.eithe ro fthe ma swel la sbot h togetherd ogro wo n thisunknow nsubstrate . . ., , The second point -ifth ebacteri a grow-i sto .stud y whether they growsufficientl yt ohav ea preservin g actiono nth emilk .W ecoul d do thisb ycountin gth enumbe ro fthes etw omicro-organisms .Bu ta s wear emor e interested inthei r preserving action- a smeasure db y their capacityo faci d formation-w estudie d theaci d formationb y meanso ftitratio nan dth echang e inp Husin ga nElectrofac tpH - meter.

7.3.2 Desigi of experiments.

Firstth ebehaviou ro feach.o fth etw obacteria , Thermobacterium bulgaricuman dStreptococcu s thermophilus,wa sdetermine d andthe n theircombine d actionwa sstudied ..

Prom tentative experiments itappeare d that45°C (wasa favourabl e temperature forth erodlik eThermobacterium . Therefore the.influ­ enceo fth einoculatio npercentag ewa sstudie d at thistemperature - Separate portionso fsoymil k weresuccessivel yinoculate dwit h 1,2$, 5,and 10%o fa pur ecultur eo fth eThermobacterium.B ya ninoculatio n percentageo f1 %i smean t that1 0m lo fth epur e culture was.adde d to 1000g o fsoymilk . Theinoculated ,mil kwa spoure d into 50m l bottlesprovide d with tightfitting glass stopsi nsuc ha wa ytha t noai rwa slef ti nth ebottles .Th ebottle swer epu ti na waterbath . Then, usinga ninoculatio n temperatureo f 5%,th einfluenc eo fth e incubation temperatures35° ,40° ,45° .50° , and55° Cwa sinvestiga ­ ted. After 1,2 ,4 ,5 ,6 ,an d7 hour san dals o after 24hour sth e pHan dth eacidit y were determined.Fo rthe ,acidit y determination 10m lwer e titrated with0. 1n NaO H (indicatorphenolphtaleine) , The same investigationwa sextende d toth eStreptococcu s thermo­ philus. However,th etemperatur e chosen for,thestudy ,o fth einflu ­ enceo fth einoculatio npercentag ewa s40° C Theeffec to fth etem ­ peraturewa sinvestigate d inth erang e from35° Ct o50°p .. Itappeare d thatfo rth eThermobacteriu ma temperatur e range from about35 °to,50° Cca nb euse d successfully, while for.th eStrepto ­ coccusthi srang ewa sabou t from35 °t o45°C .Henc efo rthe .inoculatio n experimentswit ha mixtur eo fbot h bacteriaa,temperatur erang e from 35°t o45° Cwa schosen. ,I nth emixtur e theratio ,betwee n.th erod s andth ediplococc iwa s L: 1. .Th esterilized ,mil kwas•inoculate da t a 1, 2% and 5%level . .. . . ,; ,.,,. ,

113 7.3.3 Technique of experiments.

Thesoymil kuse d inthes e experimentsha da dr ymatte r contento f 8%.''Itwa ssterilize d at120° Cfo r1 0minute san dthe n stored. The inoculationmateria luse dwa seithe ra mixtur eo fbot hbacte ­ riao ra pur e culture ofth ebacteri a isolated from themixture . The original mixturewa sobtaine d froma ninstitut e where these mixturesar eprepared . Theisolatio no fth ebacteri awa scarrie dou ta sfollows .Cn em l ofa i o dilutiono fth eorigina lmixtur ewa smixe d witha suitabl e substrate, according toBRIGGS , inPetridishes , andincubate da t 37°C Afteron eo rtw oday sth ecolonie s appeared onth eagar .A s the colonieso fth erod san dthos eo fth ecocc i differ inshap ea preliminary separationwa sobtaine d byinoculatin ga numbe ro fth e coloniesseparatel y intube scontainin g sterilizedco wmilk . After incubation at37° Cth emil k coagulates, after which thecontent s were examined microscopically toensur e thatth econtent swa smad e upo feithe r rodso rcocci .The n these pureculture swer e inocula­ ted inlarge r amountso fco wmil k (100ml) . Everyda ytw oflask so f 100m lco wmil kwer e inoculated with 2l/&o fth eThermobacteriu mo r the Streptococcus. Then thefollowin g dayon eflas kwa suse dfo r the inoculation experimentso fth esoymilk , whileth econtent so f the other flask were usedt oinoculat e anothertw oflasks.Th e rods were incubated at45°C,th ecocc ia t40°C ,an dthe n stored overnight ina refrigerator ,s otha tth ep Ha tincubatio nwa sabou t4.5 . The mixturewa sinoculate d inco wmil ka ta 2Woleve lfo r2 &hours , cooled inwater ,an dthe nstore d inth erefrigerator .Th ep Ha t in­ cubationwa sabou t4.5 ,

7.3.4 Results/ discussion, and conclusions.

The resultso fth einoculatio nar eshow n in table9 an drepre ­ sented graphically infi g28 ,29 ,an d30 . In generalw ema ysa ytha t both bacteria accepted thesoymil ka s a substrate. Itma yals ob esai d that -depending uponth eexperi ­ mentalconditions -the y grewwel la swa sindicate d byth eaci d for­ mationan ddecreas ei npH .

Fig2 8show sth einfluenc eo fth einoculatio n percentage whenth e Thermobacterium wasincubate d at45°C .Th efou rcurve sfo racidit y andth efou r forth ep Hfollo wa regula r course. Whena lo wpercen ­ tage (1%)wa suse d theacidit y increased slowlybu ta thighe r per­ centagesth eincreas ewa smuc h morepronounced . Thenal lfou rcur ­ ves tended-t obecom ehorizontal 1an dthe ydi ds osoone ra sth e in o-

114 CO 00 1* 00 ^ O O* N T H mo o tfjwi^n •> oo .*}• ro e«* ci »-< 0«-*CN O w ^ *-* •"• w o co N to IN t» w r» o\ ^•^^••tf ^^.^w •* *• ^ •«• ^ •** v^-^P w ^ ^ •**• ^r^^^u>

ft «-* « ft n o o to Ci t rt eo N i-i >N m •"* 00 O H 00 t-. ft ft ft JN ifi O MO 10 IN to i£ \£ ft m *• *-* <**•*•<«• <• •»*• *r •»*• m «•**•*• w *• -* «o *f *r ^- *• 5f "0* "»r •* ^

Ot co o< m (0 00 ft *? ft ft

co~ '•*' t*. vs m is *- ft oo n not N oo n HO <0 •* Irt

ft •* • O' tNWOi W co t w N « * * CO <0 tMOtOO oftinoi n oo * X « n A « «n«M 0.

tO t-» IN CO *IO»<<0« m H in r4 o •* n m * « M ft W O OtOt N s CH *-4 <* O ft « m <* *« comm• •••o • nnn• • • n -••-ft (ift• *«OI* • .H» •f *• C O• *0 ««nv) Minnwio tO-iO t© t© « «o » * mwt

r»etfttfi mo * ft *- ft oo M e in r* » r* * oo «* •» ft to »* »* o» © •* WW M W M •>-• l> W -• 00 • ••• ••*•• •• • • • • • *o «e * * * * v> o « «o « *6 * m

t o coc o< ne oo fl HriN *4 Ol f4-C0 C o * r>< o toe om r> o IfthNn IOOOIN o IO t» o u u r»«- 4o tt o IO * m in

«o « • in *o to «0 >0 c c oi n in to •4 » « Mttn«oi ao in >> to to oo in O f V) •«• to tO n oo ^ r» • co m co ^r co «n * ^ ininto «inte * •* a -»• u • •-' m to w wi n « t» m * HhOfl rt» tf tl a M « to to m t^ CO 14 to ot oto«o> o«toto -I o >. n w * m B»*' co co co m co co m *•* JS c •* WD T> -4 a -i OtO (OH NO/AO h i<«t>k<-(«^ et t« o n Ot ^ in in to co Kb u •n oi o •*• to oo c oo to co Ol o> •« in • a hvnf) coco.mwoo x a. cico^in ««»*»»< ^ H (0 P) rtP4«H •5 IO -< o o n m M)0 en to Ot Ot in o •> CO m o w 00 •f ° 03 r» K. oo c- Ol CO » to to tn m in oi in to « in T 3 * ft * « ft * o4 r* CO n •0 CO CO CO t^ «* •4 -4 v4 *H »i >^ *-« -4 M m to « • • f-l in fH tO Ot CO to — » to £ «HO o M «o ft t* 3 IO 3 ft u m to r* ot o 0* in in » oo in in c- to 0. • o * in V) O °z *•H* 0 •^ e o v-t *4 r4 ~* t-» *4 X v<4 M• M• • H •4• W• •« H •*• C•O •CO • W o u• ^* « -4 • • a0 U O i • c 3 *k 4« JO u in • O Vi 9 * • » 09 » 3 e e h 0 U u U + ,0 0 in o in o K* 01 •* "fi • in on o in (K * o M «t* "4 ft m ii f^ co in m • «_m «) — «WH m ».»r in u 0 <-* J u 3 u •-( . «-i CO «>» * »4 CO (O H M •*) M f-l co «n >r •* co rt •n •u nnt - A «n a » tt rt X m «s »M CO m m » is. «

115 TntrmofcecUrium kutfaricim

•- mCainHaO M a-*

K2$ ^s*i^s% —&-—-j£=* :~ Q

'I. 9 3 » e 7. ' 24 n 1 ' 2 S ' 5 • 7 24 » I Inoculation parcanteg* 9 tKuMw* timpiritur

Fig 28. • ';- • - , Inoculation withThermobacteriu mbulgaricum . Influenceo finbculatid hpercentag ean dincubatio n temperature upon the acidityan dpH . •.-•'.-:•• -

Straptocetaua IhariwopNIm

N£3

X*

••' s . ' =^^=~-

9 14 3 4 7

Fig 29. ' Inoculation withStreptococcu s thermbphilus. Influenceo finoculatio n percentage andincubatio n temperature upon the acidityan dpH .

116 n ^ 1 ? m \ i ; II

> • Ml i 111; 35 a •a 0c3

->p> 7/ X: •H / >. \ •a 1 •r-l v O 3w Oj O *H ; 1 •H

W *aH> 3 3 o -P o .cS o -SH TTT o C o -P E& a a>. f\ it? c o w •.-I •a -p c OS cS J2 3 \ f E o 3 •cH •oH tn T3 it • 03 C M 03 I—1 3 I ' « ,. v £! bao) 03 § -P In* •H CDa i u o -a>p auj t> OS a .O c O o E •e f-i •p —1 H 3 43 o 4-> o •i-l •cH

s tw c o o •H a> \*vfr •p o - o3 c a r-i a> C*3 3 3 i—i • / 1! S ^ fco Oo

117 culation percentage was increased. The last part of the acidity curves is nearly horizontal. It may be said that there is little difference in acidity between 7 and 24 hours. It should be noted that the final acidity became higher as the inoculation percentage increased. When the acidity increases the pH decreases.Thi s may be seen from the same figure. In this sense the curves for the pH are the rever­ se of those for the acidity. It can also be seen that the filial pH decreased as the inoculation percentage increased. Under the experimental conditions the lowest pH and highest aci­ dity reached were 4,6 and 5.7 respectively. The effect of the temperature is shown in fig 28. Fifty five de­ grees proved to be too high for the Thermobacterium to form acid. On the other hand, 35°C is too low although after seven hours the aci­ dity does not differ much from that obtained at temperatures be­ tween 35° and 55°* At first, however, the curves for 40°, 45°, and 50° clearly rose more steeply. There are hardly any differences be­ tween these three curves. Judging from therat e of acid formation theStreptococcus also grew well in the milk (fig 29). However, it appeared that the acid for­ mation was much slower than with the Thermobacterium. For instance, the acidity curve for 10% is about equal to that for 2Wofo r the Thermobacterium, up to about five hours. It is surprising to find here a further increase after seven hours, as with Thermobacterium neither the acidity nor the pH hardly changed after this period, The acidities thus obtained are appreciably higher than with the Thermobacterium. In contrast to the Thermobacterium the acidity curves for the lower inoculation percentages did not rise until after three to four hours. The influence of the temperature is very pronounced; 40° and 45°C proved to be favourable for the Streptococcus. Although the poor result at 35°C, however, was unexpected, the bad effect of 55°C upon the bacteria was not surprising. In this series, too, the acid production was very slow. After 24 hours, however, it surpas­ sed that of the Thermobacterium, The results of the inoculation of a mixture of both bacteria are shown in fig 30. It is clear that the course of the acidity curves showth e characteristics of both curves for the individual bacteriar The first part of the curves is typical of the Thermobacterium, while the second part is influenced markedly by the Streptococcus. Comparison of the first part of the 45°C acidity curves and those shown in fig 28 will reveal that -if the same inoculation percen­ tages are compared- the acidity caused byth e mixture is higher than

118 that caused by the rod alone,Th e same istru e of the second part ofth e40° Ccurve si fthe y arecompare d with fig29 .Thu s itappear s that, using thesam e inoculationpercentage , incubation temperature and incubation time,a highe racidit y isobtaine d with themixtur e thanwit heithe ro fth etw obacteri aalone , Whena nincubatio n temperature of40° Cwa smaintaine d theacidit y generally decreased, An incubation temperature of 35°C isto o low asma yb esee nfro m theslo wris eo fth eacidit ycurves . Allth ep Hvalue sreache dafte r2 4hour s liebetwee n4. 5 and4.p .

Asma yb esee n fromthes eexperiment sth eeffort s toconver tsoy - milk intoa sou rproduc tb ymean so flacti caci d bacteriahav e given favourable results.W ema y expect that-due toth e lowpH -a t least thepreservabilit y of theproduc ti sincreased .Whethe rthi sproduc t willcom e intous ewill ,however ,depen dupo nothe r factorssuc h as theorganoleptica l properties and thenutritiv e quality.Thes efac ­ tors aswel l asth epreservabilit y ofth eproduc t were not inves­ tigated inthi swork .

Promth eforegoin gth e followingconclusion sca nb edrawn , 1.Th e lactic acid bacteria, Thermobacterium bulgaricum andStrep ­ tococcus thermophilus, did well insoymil k (t 8%dr y matter). 2. Theacidit y caused byth eThermobacteriu m bulgaricum reached a maximum of 5.7 ml 0.1 nNaO H (pH4.7 ) after about five to si* hours (inocul.perc.107o,incub,temp.45°C) . 3.A t first the Streptococcus thermophilus formed less acid than theThermobacterium . However, afterabou t sevenhour s -whenth e acid formation by the Thermobacterium had ceased the acidity increased further. At an inoculation percentage of 10%an d an incubation temperature of 40°C after 24 hours the acidity was 6.4 ml (pH4,3) . 4. A combinationo fbot hbacteri aals o gavea regularaci dformati ­ on.A ta n inoculationpercentag eo f 2Woan d an incubationtempe ­ ratureo f40° Ca nacidit yo f4. 8m l (pH4.6 )wa sreache d after fivehours .

7.4 SUMMARY .

Prom literature iti sknow n thatpreservin g animal milka ssou r milkb ymean so fbacteriologica l changesi sa long-establishe d prac­ tice. Inorde r to investigate whether lactic acid bacteria might prove useful inconvertin g soymilk into sour milk anumbe r of ex­ perimentswer e designed.

119 The lactic acid bacteria, Thermobacterium bulgaricuman dStrep ­ tococcus thermopbilus, were inoculated separately insterilize d soymilk (8%)a tdifferen t inoculationpercentage s and incubation temperatures,Th esam ewa sdon ewit ha combinatio no fboth . Eithero fth ebacteri aa swel la sbot htogethe rprove dt ob eabl e todecreas eth ep Ho fth esou rmil kbelo wp H5 °

8 GENERAL DISCUSSION,

Theai mo fthi s workha sbee nt oobtai n information aboutth e technological aspectso fsoymil kan dsom eproduct sderive d fromit : evaporated soymilk, soycurd,an dsoy-yoghurt .Th epreservatio no f soymilkb yhea twa sals ostudied, ,i nth eprecedin g chaptersth ein ­ vestigationso fthes esubject sar edescribe d separatelyan dth ere ­ sultsar esummarized . Hereth epractica l consequenceso fthes ere ­ sultswil lb ebriefl ydiscussed .

Inth echapte ro nth e manufacture of soymilk wehav e demonstrated thati fth esam emanufacturin gcondition sar emaintained ,a produc t ofconstan t compositionca nb eobtained . Withrespec tt oth emanu ­ facture ona technical scale we should paydu eattentio n toth e requirementso fsimplicit yan dspeed , What extraction ratio canb erecommende d foractua lpractice ?I n our opinion arati oo fabou t 10%shoul d bechosen .B ymean so fa single extractionmil ko fabou t7.5 %dr ymatte r isobtained , There aren oreason sfo rapplyin ga lowe rratio :i nsom e caseshighe rra ­ tiosma yb eused ,bu tthe ni twil lb emor edifficul t toobtai nsa ­ tisfactory-yields. Nowi tmigh tb equestione d whether it is desirable to dehull the beans. Froma technologica l pointo fvie w thisi sno tnecessar yan d even lessdesirabl ei nvie wo fth epossibl eyiel dan dth e extent to whichth esuspensio n canb efiltered.Therefor edehullin gwil lonl yb e applied ifi ti sconsidere d necessary froma nutritiona l pointo f view. Themanufactur eo fsoymil k isa matte ro f considerable reduction in particle sizeo fth era wmateria lan do fdispersin g inwater ; both operationsma yb ecombined .Th enex t questioni swhethe r a wet or dry reduction in- particle size shouldb echosen .Genera lexpe ­ rience aswel l asou row ninvestigation s have shown thata firs t

120 reduction inparticl e size isbes t carried outdry ,an d a further

reduction weta Thereforew ewoul d first recommend togroun d the beans inva simple grinder,e ,g .a hammermill ,t oa particl e sizeo fabou t 0,5 mm, Then thispowde r should besoake d fora tleas t4 t o6 hours ,prefe ­ rably ata lowtemperature,i nwate rtwic et othre etime sth eweigh t ofth ebeans , A further reduction inparticl e sizema v be carried outwit hth e ultimatelydesire d amounto fwate ran d isthu scombine d withth e extractionproper . Wehav e seentha tth etemperatur e haslittl e influence on theex ­ traction,,Thu s the grinding/extraction canb ecarrie d out atroom - temperature. Furthermore thep Hprove st ohav ea n important influ­ ence. If, however, the milk is destined for human consumption (either in a fresh, or concentrated form) it isno t advisable to increase the pHconsiderably . Only if the milk ist o be used for themanufactur e ofsoycur d anincreas e inp Hma yb eapplied . An important question is which type of apparatus isbes t suited for the grinding/extraction, Invie w of our experience we are of the opinion that it isbes t tous e amoder n grinder/mixero f which many typeshav e appeared on themarke t during the last fewyears , For our work weuse d an Ultra-Turraxwit h satisfactory results, A grinder/mixerha sno tonl y ahig h capacity forparticl e sizere ­ duction, but inadditio n itha s a strong dispersing action. In a comparatively diluted suspension it isals o effective, but itma y be advisable toad d littlewate rt oth esoake d flour firstan d then after ashor t timeo fgrindin g add theres to fth ewater . With res­ pect toothe r equipment (ballmill ,runne r mill,colloi d mill) a grinder/mixer hasth eadditiona ladvantag eo fsimplicity , lowprice , high capacity, and rapid action. The time of grinding/mixing is dependent uponth esiz eo f thebatch ,an dth etyp eo fapparatus , and has tob edetermine dempirically . Asstate d before,the separation of the milk from theresidu e does not presentgrea t difficulties, For thisoperatio n we used a bas­ ket centrifuge with filter cloth. No doubt this apparatus is the most simple one.However , on using this apparatus the filter cake should be prevented from becoming too thick, aswit h increasing thickness thedr ymatte rconten t of themil k becomes lower.Th e use of a clarifier is better justified as - by the right choice of diameter, numbero frevolutions , and,holdin gtim e - it ispossibl e to control accurately whichparticle s will remain inth e milk and which inth e residue.Howeve rth edisadvantag e ofa nordinar yclari ­ fier is the small capacity for solid substances. Therefore, if larger quantities ofmil k aret ob eproduced , amor eexpensiv econ ­ tinuousclarifie r isnecessary .

121 SOYBEANS

HAMMER MILL

•1

STOCK SOYFLOUR

SOAKING WATER

f

Pig 31.Flowshee to fth emanufactur eo fsoymilk .

122 Ifi t isdesire dt o increase theyiel d furtherw ehav e seen that iti spossible ,t oappl ya nextractio n insevera l stages.N o doubt a countercurrent extraction will then be used. At the second and third stageagai na "grinder/mixe r mayb eused.s otha ta furthe rre ­ duction inparticl e size iseffected, , or a simple stirrer having only a dispersing action. Pig 31 shows a method of processing usinga nextractio n inthre estages .

Asha sbee nshow nth emanufactur eo f raw evaporated milk ofabou t 30% dry matter content isquit e possible. This product, however cannot be sterilized as itshea t stability isto o low. Although preheating appears toimprov eth estability , itstil l remainscom ­ paratively low.Moreove r aresul t ofpreheatin g istha ta viscosi ­ ty barrierarise sa ta dr ymatte rconten tdependin g upon themetho d ofpreheating . This impliestha t itwil lb edifficul tt omanufactur e sterilized evaporated soymilk with more than 20%dr ymatter ,a s intha t case theconcentrate d milk cannot besterilized . Bypreheatin g at 70°C for 45 seconds adr y matter contento fabou t30 %ca n be reached, whileth emicr oorganism sar epartl ydestroyed. ^ condensed milk. As the same concentrated milk can be used for spray drying it willals o bepossibl e tomanufactur e soymilk powder. Ifi nactua lpractic e anevaporato rha st ob echose n itshoul d be remembered that forfoo d products low temparaturesan d short hol­ ding times arepreferable . For thispurpos e filmevaporator s are theonl y appropriate type.Thi styp eo fevaporato r hasals oth ead ­ vantage thatth eholdin g time israthe runiform ,whic h impliestha t themil kparticle sar esubjecte d toa unifor mheating . Ifa n evaporator with a longholdin g time isalread y present as parto fth eequipmen t ofth e factory itshoul d betake nint oaccoun t that theheatin g surface might damage the quality of the milk as regards itsviscosity . Inthi scas e it isadvisabl e to seet o it that the temperature of theheatin g surfacedoe snot exceed60°C . Ina on e effetevaporato r byadaptio no fth evacuu ma boilingtem ­ peratureo f about30° Ccoul db emaintained . Inthi swa ydetrimenta l heating inth e evaporator isavoided . Thismetho d ofmanufacture , however,necessitate sa separat ehea ttreatment . Foaming may bea complicatio n inth e execution ofth eprocess . According to our experience this foaming does not depend on the boiling temperature used,a si tma yoccu rbot hat 'lo wan dhig htem ­ peratures. It seems as ifonl y thedifferenc e between the steam temperature and the boilingpoin t isth edeterminin g factor, be-

123 causewhe n the steam temperature israise d above acertai n level or when thevacuu m isincrease d themil k starts foaming. These twoob ­ servations give the impression that the foamingi scause d byenlar ­ gemento f the foam bubblesdu e to eithera decrease ofth e pressure outside the bubbleso r an increase of the pressure insideth e ,bub« blesbecaus e ofa rais e intemperature .Wit h the film evaporator in­ creasing the feed above a certain level also causes the milk to foam. In this case it seems to be the increase in the amount of foambubble s thatoause sth emil k to foam. Inth e case of the film evaporatorth e foaming canb e suppressed by increasing the speed ofth e rotor,whil e we have the impression that ifth evolum e of the vapour separator is large enough nodif ­ ficulties will occur. Itmigh t also be possible to combat foam by coolingo f thepassag e between vapour separator and condensor, thus causing the foambubble s tocollapse . Further research isnecessar y to investigate the possibility of further increasing thehea tstabilit y and of shifting the viscosity barrier to still higher dry matter contents or entirely dispense with it.W e mention some lines alongwhic h future research may be developed.

1.Dr yo rwe theatin gth ebean s ormeal . 2. Heating theconcentrate d milk after evaporation. Thismetho d is notunknow n inexperimenta l literature. 3.Additio no fsalt s beforeo rafte revaporation . Asha s been shown insom e casesw emanage d to increase the heat stability. Asmore ­ overth e electrical chargeo f theprotein s might bechange d this could influence theviscosity . 4. Germination of the soybeans. As important changes take place during germination the tryptic inhibitor may disappear. Besides the amount ofprotein s diminishes and themil k may then acquire a higher heatstability . 5. Mixing with other vegetablerproteihs..Th e mixing may.caus e the stability tochange .

Furthermore it'shoul d'b enote d'that'.th ework 'don e wascarrie d out with soybeans obtained fromoi l factories, and sopresumabl y belon­ ged to thecommercia l group of soybeans. Varieties of soybeans may bedivide d intothre e groups:commercia l (grain), forage, and vege­ table.Th e commercial varieties are used largely forprocessing"int o oil, meal, flour,an d flakes. Inth eOrien t varieties have been de­ veloped forus ea s greenbeans,bea ncurd , bean sprout, and numerous other food products; the term "'vegetable varieties"i s used to distinguishthe m fromvarietie s grown forothe rpurposes.Thi sgrou p

124 is superior tocommercia l and forage typesfo rvegetabl e milk, soy flour.bean curdetc . (USDA .1952) .I ti spossibl e thatou to fthes e vegetablevarietie swhic hnumbe rove ron ehundred , oneo rmor ewil l give nodifficultie swhe nprocesse dt oevaporate d milk. Inthi scon ­ nection itma y beusefu l toconside rth e influence ofth e stageo f ripeness on processing characteristics. Extensivean dsystemati c research isneede d to investigate thisaspec to fth eproblem .

'.Inth eproduction-distribution-consumptio n chain itwil l be some time before actual consumption of the milk takesplace . Therefore pasteurization or sterilization will have to be applied. As has beenshow npasteurizatio n reducesth enumbe ro f bacteriaappreciabl y and the milk may bepreserve d forsom e days ifpasteurizatio n is combined with coldstorage .Th edrawback , however, istha twhe nth e pasteurized milk acquiresa highe r temperature,bacteri awil ldeve ­ lopa ta hig h rate.Even ifth emil k iskep ta tlo wtemperatures'th e bacterial growth continues. Therefore,especiall y inwarmclimat.es sterilization should onlyb econsidere d forapplicatio n inpractice . Sterilization may be carried out asusua l with equipment already, existing. According to literature the mother liquid from seawater, spon­ taneoussourin go rsou rsoymilk 'ar eth emeans ,use d toeffec tprotei n precipitation insoymilk .Th eobjectio nagains tspontaneou ssourin g is that thedegre e ofsourin g isdifficul t tocontrol ,whil e fur­ thermoreprotei ndecompositio nma y lead toa dangerou sdeca yo fth e milk. Therefore theus eo fth emothe r liquido r itsconstituent s is preferable. Since it has been shown in this work that it is not necessary tous eal lconstituent so fth emothe r liquid the question ariseswhic ho fth econstituent sshoul d beused.A sfa ra sth eyiel d isconcerned , thesalt sa swel la s lacticaci dprov enearl y equally effective. The attractive point inth eus eo f singlesalt s istha t iffro m a nutritional pointo fvie w it isdesire d to increase for instance the calcium content, this can be easily done by using a calcium- salt, e.g. calciumchloride. Prom atechnologica l standpoint,however , the useo f acids (pre­ ferably lactic acid)ha sth eadvantag etha ta one-stag e extraction already yields aver y high percentage of the original protein if the pH is increased.Th eprecipitatio nwit haci dyield s nearly all theprotei npresen t inth emilk .Anothe rpoin t infavou ro f theus e pf acid istha ta lo wp H inhibitsth egrowt ho fdecay-bacteria .

Theattractiv epoin ti nth emanufactur eo f soy-yoghurt istha tth e conversion of the soymilk can be effected in asimpl e way without

125 the useo f complicated equipment. Proper attention should be given toth e incubation of amixtur e of both lactic acid bacteria,Ther - mobacteriumbulgaricu m and Streptococcus thermophilus. Sinceth enutritiona l qualitieso fthi sproduc tar eno tye tknown , thisrequire s further investigation.Technologically , the important pointi stha t theconversio n ofsoymil k intoa sou rproduc tdoe sno t present anydifficulty .

126 SAMENVATTING

Gezien de grote tekorten aan dierlijk eiwit voor de voeding van de wereldbevolking en het feit, dat het in vele landen niet goed mogelijk zal zijn om op korte termijn deze tekorten op te heffen, is het noodzakelijk om'te-.zien naar voedingsmiddelen met een hoog gehalte aan hoogwaardige, plantaardige eiwitten. Het is bekend, dat de eiwitten van de sojaboon van hoogwaardige kwaliteit zijn. In enkele Aziatische landen is het gebruik van de soja als voedingsmiddel een gewoonte, die reeds van vele eeuwen voor het begin van de Christelijke jaartelling dateert, Voedings­ middelen, zoals sojamelk en sojakaas, zijn derhalve geen nieuwe produkten. Technologische gegevens over de fabrikage van deze pro- dukten zijn echter vrijwel niet aanwezig. Het leek ons daarom van belang een onderzoek te verrichten be- treffende de fabrikage van sojamelk en enkele hiervan afgeleide produkten teneinde het rendement en de kwaliteit te verbeteren en derhalve de kostprijs te verlagen. Tot de andere bestudeerde pro­ dukten behoren sojawrongel, sojayoghurt en ongesuikerde, geconden- seerde sojamelk.Ookde verduurzaming van de melk werd in het onder- zoekprogramma opgenomen.

SOJAMELK

Ind everschillend e bereidingsmethodenva nsojamelk , zoalsdi e worden toegepast doord ebevolkin g van enkele Oost-Aziatische lan-- den, kunnenee naanta l duidelijkt eonderscheide n stadia aangewezen wordendi eal svolg taangedui d kunnenworden : 1)d e voorbereiding; 2)d eeigenlijk e extractie; 3)d eafscheidin gva nd emelk ; 4)d everbeterin gva nd ekwalitei tva nd emelk .

Aangeziend ewijz ewaaro p de voorbewerking plaatsvind ti nsterk e mate desamenstellin g vand emel k bepaalt, werden diverse typen maalapparaten onderzocht. Hierbij werd zowelva nd eongepelde , hele bonenal sva nd eto tmee l vermalen bonenuitgegaan . Ookde eigenlijke extractie werd onderzocht teneindeee ninzich t tekrijgen in defactore ndi egevarieer d zouden kunnen wordenwan - neerva nee nbepaal d type apparaatword t uitgegaan. Overd eafscheidin gva nd emel kwer d geenonderzoe kverricht.aan - gezien zich hierbij geen problemen voordoen, althans nietbi jd e

127 dooron sgevolgd e werkwijze.Ee nonderdee lva nhe tlaatst e stadium, dehoudbaarhei d vand emelk , werd inhe tonderzoe kbetrokken . Devolgend e resultaten werden verkregen:

1.D eproeve n overd efabrikag e vand emel k zijn goed reproduceer- baar. Extractieva ngeweek t sojameelbi jee nextractieverhoudin g van 1: 1 0resulteerd e inmel kme tongevee r 7i.7%drog estof ,waar - vanongevee rdehelf tui tru weiwi t bestond.He trendemen tva nd e droge stofe nhe tru weiwi t bedroegenda nongevee r 59.0. respec- tievelijk 68.0%.D e eiwitten bleken derhalve beter dispergeerbaar tezij nda nhe toverig e dee1va nd edrog e stof. 2. Extractieva ngeweek t sojameelga fee nhoger eopbrengs tda nex ­ tractieva nd egeweekt ebonen.Zowe l bijne tmee l alsbi jd ehel e . bonen steegd ehoeveelhei d melk bijbeter e vermaling,evenalshe t gehalte aandrog e stofe nru weiwit v Dooree nbeter e vermaling stegen derhalve ookd erendemente nva nd edrog e stofe nhe t ruw ••eiwit . ;• 3.D eextracti e verhouding heeftee ngrot e invloedo pd esamenstel - lingva nd emelk .Bi jee ntbenemend everhoudin gstijge nd egehal - tesaa ndrog esto fe nruw 'eiwit,terwij lva nbeid ed erendemente n dalen. 4.Oo kd eextracti e pHi sva ngrot e invloed ophe trendement .He t droge stof gehalte steegbi j toenemendep Hechte rnie ti ndi ema ­

teal she tgehalt eaa nru weiwit .Bi jee np Hva n9. 5bedroeg ;he t rendementva nru weiwi t80% .

5.D eextracti e temperatuur bleek inne t gebied van'20°"to t50° C weinig invloed te'hebberi.

6.D eextracti e tijd had, nadat een( zekerminimu m overschre«lenwas , eveneensweini g invloedo phe trendement . 7.D erotati esnelhei dva nee nIk aUltraturra xha d vrijwel geenef ­ fecto pd eextractie . De gehaltes aane" nd e rendementen vand e droge stofe nhe tru weiwi t lagenechterwa thoge rda nbi jextrac ­ tieme tee nVibromixer . 8. Doorhe tresid u tweemaa lt ewasse nstegen'd etotal e rendementen vand edrog e stofe nhe tru weiwi t aanzienlijke nnaderde nzi j ' eengrenswaarde ,di e- althah svoo rd edrog e stof- berekeri dko n worden.He ttotal e rendementva nhe tru weiwi tbedroe g 90%;ge - scliatwordt ,da td egrenswaard e bijd etoegepaste'werkwijz e ind e buurt ligtva n95% . " "'' •..:•••:.' De consequenties voord epractisch euitvoerin gworde n inhe thoofd - stuk GeneralDiscussio n besproken.'

128 HOUDBAARHEID VAN SOJAMELK

Waar de sojamelk in hoofdzaak uit eiwitten, vetten en suikers bestaat vormt deze uiteraard een ideale voedingsbodem voor micro- organismen. In de literatuur zijn enkele aanwijzingen te vinden, die er op wijzen dat toepassing van een warmtebehandeling de houd- baarheid van de melk doet toenemen. Ook is het bekend, dat getracht is met behulp van chemicalien de houdbaarheid van de melk te ver- lengen. Aangezien het gebruik van chemische conserveermiddelen in voedingsmiddelen minder gewenst is, werd slechts getracht door mid- del van pasteurisatie en sterilisatie een betere bewaarbaarheid te verkrijgen. De volgende resultaten werden verkregen: 1. Pasteurisatie van de rauwe sojamelk bij 60°C gedurende 30 min verlaagt het initiele bacterie-aantal aanzienlijk. De gepasteu- riseerde melk blijft echter bij een temperatuur van 20 ~ 25°C nauwelijks een dag goed,Pasteurisatie in combinatie met bewaring bij lage temperatuur verhoogt de houdbaarheid tot ongeveer drie dagen. 2. Sterilisatie van 8% sojamelk bij 120°C gedurende 5 minuten blijkt voldoende te zijn om de melk gedurende tenminste een jaar houdbaar te maken. 3. Indien het droge stof gehaltevande melk boven de 12% stijgt dient de sterilisatie drastischer te worden uitgevoerd dan voor een 8% melk.

ONGESUIKERDE GECONDENSEERDE SOJAMELK

Om meerdere redenen kan het wenselijk zijn om de beschikking te hebben over gecondenseerde melk van tenminste 30% droge stof. Een dergelijke verhoging van het droge stof gehalte zou met behulp van verdampers uitgevoerd kunnen worden. Het was echter bekend^ dat zich bij het concentreren van sojamelk met behulp van verdampers . „„ rn hot knrt kunnen deze moeilijkhe- moeilijkheden kunnen voordoen. In het kort Kunnei vicnn«itoi+ den omschreven wordeA~„n al „is ehe hPt tzee 7per rfatem sterk wtoeneme n van ^d e, viscositei . t j..„. ctnf eehalte van omstreekro o s 20% van de sojamelk wanneer een droge stof genaite ... CTO . ' wordt overschreden, waardoor de hele melk ineen P^ingacht g massa overgaat.Door het optreden van dit verschijnsel is een limiet gesteld aan het concentreren door middel van **«™- Getracht werd na te gaan of een warmtebehandelxng van de r.a.e Koianfrri ike invloed zou kunnen neD- sojamelk voor het indampen en belangrx ke x ^ ben niet alleen op de viscositeit van net me 129 tevenso p de hitte-stabiliteit. Door middel van flash verhitting - metbehul pva nee nspeciaa l ontworpenbuize nwarmtewisselaa r -e n langzameverhittin g inee nkookkete lwer d de invloed vanverschil - lende temperatuurniveau's en verhittingstijden onderzocht. De experimenteleresultate n gavenaanleidin g tot hetopstelle n vand e volgendeconclusies :

1.Rauw esojamel kka n zondermoeilijkhede nto ttenminst e 30%drog e stofingedamp tworden .He taldu sverkrege nproduc t isechte rnie t steriliseerbaar, 2. Bij verhitting van de rauwe sojamelk voor het indampen treedt tijdens het concentreren een viscositeitsbarriere op,di e het nietmogelij kmaak to md emel kverde rt econcentreren .He tdrog e stof gehaltewaarbi jdez eviscositeitsbarrier e optreedthang ta f vand ewijz eva nverhitting . 3.Wannee r een kookketel gebruikt wordt om de melk teverhitten , heefthe tverlage nva nd everhittingstemperatuu rva n90 °to t60 ° C zowelal she tvermindere nva nd everhittingsduu r van3 0to t 0 minuten tot gevolg, dat de viscositeitsbarriere naar een hoger drogesto fgehalt eword tverschoven . 4. Wanneer een continue,flash sterilisator gebruikt wordt om de melk teverhitten ,heef the tverlage nva nd everhittingstempera ­ tuurva n 120°to t 70°Czowe l alshe tvermindere n vand everhit ­ tingsduurva n4 5to t0 seconde nto t gevolg,da td eviscositeits ­ barrierenaa ree nhoge rdrog e stofgehalt eword tverplaatst . 5. Invergelijkin gme td eviscositeitscurve nva nd emel k behandeld in de kookketel zijn de curves voor de flash gesteriliseerde melkvoo rovereenkomstig e temperatuurbehandelingen (700, 80°e n 90°C)o pee nhoge r drogesto fnivea ugelegen . 6. Bij gebruik vand ekookkete l geldt- zowe lvoo r eenverhittings ­ duurva n3 0al sva n 0minute n -da td ehitt e stabiliteit ind e volgorde 70°,90 ° en 80°C toeneemt. Wordt de verhittingsduur van3 0to t0 se cverminderd ,da n stijgtd ehitt estabiliteit . 7. Bij gebruik van de flash sterilisator heeft - zowel voor een verhittingsduur van 45 als van 0 seconden -d e hitte stabili­ teitd eneigin go mto et eneme n ind evolgord e70° ,80 °e n90°C . Wordt de verhittingsduur verhoogt van 0 tot 45 seconden dan vertoontd ehitt estabilitei td e tendenso m toe tenemen . 8. In vergelijking met de hitte stabiliteits curven van de melk uit de kookketel zijn die voor de flash sterilisator, voor overeenkomstige warmte behandelingen op 70°,80 ° en 90°C, op eenhoge rnivea uva nhitt e stabiliteitgelegen .

130 9. In het gebied 70°, 80° en 90°C is het overall effect van de voorverhitting het verschuiven van de viscositeits curven naar rechts, d.w.z, naar een hoger droge stof gehalte, in de volgor- de ketel 30 min- ketel 0 min - flash 45 sec - flash 0 sec, Beide eigenschappen - de viscositeit zowel als de hitte stabili- teit - blijken derhalve sterk door de warmtebehandeling beinvloed te worden, Bovendien blijkt, dat deze invloed aan een bepaalde wetmatigheid gebonden is. In het algemeen kan gezegd worden, dat naarmate de verhitting minder intensief is bij concentreren een hoger droge stof gehalte bereikt kan worden, Dit geldt zowel voor het temperatuursniveau als voor de coming-up time en de verhit- tingsduur.

SOJAWRONGEL

Evenals bij de bereiding van de melk kan ook in het fabrikage- proces van sojakaas een aantal duidelijk verschillende stadia wor­ den aangeduid: 1) precipitatie van eiwitten; 2) afscheiding vand e wrongel; 3) vermindering van het watergehalte; 4) rijping vande kaas, ( t^&*^.. WAV, fr^wc^ Voorwaarde voor een regelmatige produktie op grote schaal is, dat precipitatie onder bepaalde uniforme omstandigheden steedsesncon- stante opbrengst zal geven, Tevens zal men een zo hoog mogelijke opbrengst verlangen. Teneinde constante bpbrengsten te verkrijgen is het nodig het precipiterende agens te standaardiseren. Hiertoe werd van een aan­ tal zouten enva n melkzuur afzonderlijk eni n combinatie de precipi­ terende werking bij verschillende temperaturen en concentraties be- studeerd. Uit de experimented resultaten kunhen de volgende conclusies worden getrokken: 1. Bij het neerslaan van de eiwitten uit sojamelk met behulp van zoutoplossingen bij verschillende temperaturen en concentraties zijn voor MgCl2, MgS04enCaCl2 geen noemenswaardige verschillen aan te wijzen, Ook de werking van een combinatie van deze drie zouten geeft geen essentiele verschillen te zien. Ongeveer 90% van de eiwitten is op deze wijze te precipiteren, 2. Wordt de precipitatie met behulp van melkzuur bewerkstelligd, dan wordt tot pH 4.5 meer eiwit neergeslagen naarmate de pH daalt.Echter blijkt.dat bij lagere pH minder eiwit wordt gepre- cipiteerd. In combinatie met eenverhoging van de temperatuur is

131 hetrenderaen tva nd eprecipitatl e op tevoere nto t tenminste90% . 3. Bijee nverlagin g vanhe t droge stof gehalte van demel kka nme t eer zwakkere zoutoplossing worden volstaan, bij verhoging van het droge stof gehalte dient een sterkere oplossing te worden gebruikt.

SOJAYOGHURT

Sojamelk, inhoofdzaa k bestaande uiteiwitten , vetten ensuikers , vormt een goede voedingsbodem voor diverse micro-organismen. De groei van deze organismen leidt tot een snel bederf van demelk . Nagegaan werd ofhe tmogelij k isd e tweemelkzuurbacterien , Thermo- bacteriumbulgaricu m enStreptococcu s thermophilus, in sojamelk te doen groeien om aldus een zuur melkproduct te verkrijgen. In een dergelijk zuurproduk t zoudeh de rottingsbacterien weinig kansheto - beno m totontwikkelin g tekomen , Hiertoewerde nvoo r elk vand e bacterien afzonderlijk enoo k voor het mengsel van beide bacterien een aantal kweekproeven bij ver- schillende temperaturene nentpercentage s opgezet. Uitd eproefresultate n werd hetvolgend e geconcludeerd:

1.D e melkzuurbacterien, Thermobacterium bulgaricum enStreptococ ­ custhermophilus , groeien in 8ft,sojamel k goed. 2. De zuurtegraad teweeg gebracht door deThermobacteriu m bereikte na ongeveer vijf tot zes uur, een maximum van5. 7m l0. 1n NaOH (pH4.7 ) bijee n inoculatiepercentage van 10%e n eenkweektempe - ratuur van45°C . 3. Inhe t begin vormt deStreptococcu s minder zuur dan deThermo ­ bacterium. Na zevenuu r echter -wannee r dezuurvormin g doord e Thermobacterium reeds isopgehoude n -blijf t dezuurtegraa dtoe - nemen. Na 24uu rwa sd ezuurtegraa d 6.4m l (pH4.3 ) bij een10 % inoculatiepercentage enkweektemperatuu rva n 40°C 4. Een combinatieva n beide bacterien geeft een regelmatige zuur­ vorming. Na vijf uur werd een zuurtegraad van 4.8 mi (pH4„6 ) bereikt bijee n inoculatiepercentage van 2Y&e n eenkweektempe ­ ratuur van 40°C

132 REFERENCES

Allan,K.S., and Circle,S.J. Ind.and engin.chem. 30: 1414-1418 (1938) Belenkii.D.E. and Popova,N.N. Russ.pat.32,908, oct 1933 Beckel,A.C, Bull.W.C., and Hopper.T.H., Ind.and engin.chem.34: 974-978 (1942) Belter,P.A. , andSmith ,A .K . J.Am.OilChem.Soc : 170-174 (1952) Beltzer, Matagrinp .19 6 Bishop,I.R . J,Inst.Brewing35 :316-32 2 (1929) Bloch,A . 3ulLdesScience sPharmacol . 13:138-14 3 (1906) 14: 593-606 (1907) Bogatskii, V.D.,Storozhuk,M. K and Muromtsev,V .A .Schrifte nzentra l biochem. Forschungs inst.Nahr-Genussmitte l ind. (Moscow) 2: 410-30 (1933) Burnett.R.S.,an dFontaine ,T.D , Ind.andengin.chem .36 :2G4-28 8 (1944) Burtis.E.L.,Markle y p.61-111.195 0 Castagnol.E.M . Bull.Econ.de 1*Indochine,sept-oct . 982-994 (1934) Chiu,Yan-tsz Thesis!Cornel l University,pp .38 .192 7 Cartter.L.an dHopper.T.H . U.S.Dep.of Agr.Tec hBul l787 ,6 6PP . (1942) Curran,H.R. ,Bell ,R.W. , andEvans ,F.R .J.Dair ySc i27 :913-91 9 (1944) De.S.S., SoybeanDiges t8 :45-4 6 (1948) Deysher,E.F,,Webb.B.H, ,an dHolm ,G.F .J .Dair ySci .12 :80-8 9 (1929) Deysher.E.F.,Webb,B.H . andHolm ,G.F .J.Dair ySc i 27:345-35 5 (1944)

F.A.O.NutritionalStudie sNo 48 PP.7 8195 2(Brock ,J.F .an dAutret.M. ) Fontaine,T.D. ,an dBurnett ,R.S . Ind.andengin.chem .36 :164-16 7 (1944) Fontaine,T.D., Samuels,C.,an dIrving.G .W . Ind.and engin.chem.36:625"62 7 (1944) Holm, G.E.. Deysher,E.F. ,an dEvans,F.R .J.Dair ySc i6 :556-56 8 (1923) Horowitz-Wlassowa,L.M. ,Oberhard ,J.A. ,an dGutermann,B.J .Schrift.zentra l biochem. Forsch-Inst. Nahr-Gemussmittel,1 : 157-69 (1931) Institut Internationald'Agricultur e lesoj adan s lemonde ,p.158-19 4 (1936) Inouye,M . Bull.Coll .Agr. ,Toky o Imp.Univ .,2 : 209-216 (1894-7) Katayama,T. , Bull.Coll .Agr .Toky o Imp.Univ. ,Japa n7 : 113-115 (1906); 7:117-11 8 (1906) Kondo.N. Ztsch.Untersuch Nahr.W.Genussmittel. 25 (1): 1-56 (1913) Lager Theusefu l soybeanpp.29 5194 5 Leniger,H.A . CentralTechnica l InstituteT.N.O. ,Repor t215 . PP.12. 1955- Notpublished . Li,Y.Y . Matagrinp 17 4 Loska,S .J. , andMelnick , D. CerealChem .27 :127-14 0 (1950) Littmann,0 . Beitragezu rKenntni sde rYoghurtbereitung. ' ThesisHohenheim . pp.46,193 4 • Markley, K.S. Soybeansan dsoybea nproducts ,vo l Ian d II PP.1145 (1950) Matagrin.A., Lesoj ae tle sindustrie sd usoja , pp.390 (1939) X Milyutina,L , MolochnayaProm .9 :34- 7 (1948) Monnier,E. , Bull.Econ.d e 1'Indo-Chine ,janv.-fevr. : 66-87 (1935) Nagel.R.H. ,Becker ,H.C. ,an d Milner,R.T .Cerea lChemistr y 15:463-47 1 (1938) Ohlen,F.W.vo n Am.Jnlo fBotan y 18:30-4 9 (1931) • O-lcott.H.S. , andFontaine ,T.D . J.Am.Chem.Soc. QV 2037"2040(1939 } Piper,C.V. ,an d Morse,W.J . Thesoybea n (1923) Prinsen-Geerligs,H.C . ChemikerZeitung : 67-69 (1896) Rogers,L.A. , Deysher.E.F.. andEvans,P.R . J.Dairy Sci.4 :294-30 9 (1921) Smith,A.K . USDA , AIC -234. June 1949 Smith,A.K . Econ.botany8 :291-31 6 (1954) Smith,A .K. ,an dBeckel.A.C . USDA , AIC -113.Febr.194 6 Smith, A.K.,Circle,S.J. ,an dBrother,G.H. . J.Am.Chem.Soc. 60:1316*132 0(1938 ) Smith,A.K. ,an dJohnson,V.L . CerealChem .25 :399-40 6 (1948) Staker,E.V. , andGortner ,R.A . J.Phys.Chem.35 :1565-160 5 (1931) Visco, A. Voeding 16:301-31 0 (1955) X Voskresenkii,V.M. ,an dDobruinina ,T.K . Proc.Inst.Sci.Res .Foo d Ind.(Leningrad ) 2: 23-31 (1935) Wai, K.N.T.Bishop .J.C. ,Mack.P.B .an dCotton,R.H . Plantphysiolog y22 : 117-126 (1947) Webb, B.H., andBell.R.W . J.DairySci .26 : 1071-1077 (1943) Webb,B.H., Bel1,R.W.,Deysher,E.F.,an dHolm,G.E . J.Dairy Sci.26 :571-57 9 (1943)

x Originalpape rno tread . £ •

O"

4. Knives of Braun mixer /. mVn =s lOu

1. Microphoto of soymilk (720 x)

!Wis. 'C'>^- **>.-&1i/:' #--i "*•<£» J

3. Le/t: Vibromixer - Centre: Braunmixer 2. Desintegrator Right: lka Ultralurrax £;":-1 I-

>-,— i^'

5. Head of UUraturrax S^> V£

L • - 'JlI.^.«»JX—1/

6. Steamjacketed vessel with mixer

-'! r v A'

... j^—.-^..-^ jj . A.25 5j- t-,•c J > i" •:••• • • i••- p.- i * ' r

! 1 I - .,.ni r *^, " i- r •— » \ -—"••

I

.

7. Continuous flash sterilizator 8. Evaporation installation (Luwa film evaporator) page line for read

9 28 In is 13 10 MORSE, (MORCE, 24 16 courses sources 29 14 scare scarce 36 5 4,1Extractio n ratio 4.1 Extraction pH 43 4 beansparticle s beanparticle s 44 20 nog not 44 37 beansparticle s beanparticle s 48 36 is 49 14 implied implies 51 6 an and 63 11 en an 66 6 BULL BELL 66 28 applied applies 67 26 possible possibly :e69,69 , .70,92.92,! e 8.13.20.19,41.6 sterilisator sterilizator 70 35^5 > 1 0 78 18 become becomes 91 5 witho f withthos eo f 92 46\< N, w acually actually 93 16\ o Bath Batch 111 17 The the 21 Call2 CaCl2 33 confire confine 33 outselves ourselves 34 found the found inth e 38 semmed seemed 112 10 avialable available 113 19 as at 120 32 if of

42 Pig3 .Extractio no fprotein sfro mth esoybean . Distribution ofdr ymatte ran dcrud eprotei n inmil kan dresidue . Influenceo fpretreatment .

42 Pig4 .Extractio no fprotein s fromth esoybean . Distributiono fdr ymatte r andcrud eprotei n inmil kan dresidue . Influenceo fextractio nconditions . STELLINGEN

I

De conclusie van Huizenga, dat in 1939-40 een gemiddeld inkomen van tien gulden per maand voor een arbeidersgezin (4-5 leden) op Java als een minimum, redelijk levensniveau dient te worden beschouwd, is niet gerechtvaardigd. Huizenga, L. H., diss. Wag. (pp.296) , juni 1958

II

Vooroorlogse gegevens wijzen op een toestand van ondervoeding bij de bevolking op Java.

Ill

Het is mogelijk plantaardige melk zodanig te fabriceren, dat deze de rol van dierlijke melk als voedingsmiddel kan overnemen.

IV

Op economische gronden dient bij het streven de melkproductie in de tropische gebieden op te voeren meer profijt te worden getrokken van de productie van buffelmelk.

De opvattingen van DOS SANTOS en TROPA betreffende de rol van de veterinair bij de productie en verwerking van dierlijke melk zijn merendeels onjuist. 22 iime Session du Com. de l'Office Int. des Epizooties.XLII: 537-546, 1954

VI

Voor het verhogen van het eiwitgehalte van de sojaboon ware het wenselijk reeds thans bij de cultuur aandacht te besteden aan de hiervoor in aanmerking komende maatregelen.

VII

Het gebruik van gips voor de bereiding van tofu dient te worden afgeraden.

VIII

Zowel voor hoog ontwikkelde als voor onderontwikkelde landen zijn de voor- waarden voor een juiste meningsvorming dezelfde. Deze voorwaarden zijn een objectieve en volledige voorlichting enerzijds en een critisch denkvermogen anderzijds.

Diss. Tan B.H . Wageningen 1958