September 1947
INDUSTRIAL AND ENGINEERING CHEMISTRY
low humidities and the beginning of multilayer and capillary adsorption. The isotherms of Figurr 2 suggest that the'adsorption data should follow the theory of Brunauer, Emmett, and Teller ( 1 ) . Calculations based on their equation were made but are not presented here, because more rsperimental points would be desirable for proper handling of these data. However, it was noted that, ) approximately in the region of relativo pressure ( P , ' P o betncxen 0 . 2 and 0.5, the data approsimatecl a straight line in accordance with their thc~ory. 4CIiNOW LEDGMENT
The' :assistance. of Dale Eoivlin in making sonw of thc vapor piwmeawi'enients is gratefully acknowledged. This work was supported hy the Kansas Industrial Developmc~ntCoininision.
1162
LITERATURE CITED (1) Brunauer, S..Enimett, P. H., and Teller, E., J. Am. C'hem. Soc., 60, 309 (1938). (2) Emmett, P. H., "Advances in Colloid Science," p . 1 , New Tork. Interscience Publishers, I n c . , 1942. (3) Makower, B., IXD.ESG.CHEY.,37, 1018 (1945). (4) Makower, B., and Dehority, B. L., Ibid., 35, 193 (1948). ( 5 ) Makower, B., and Myers, S . , Proe. I n s t . Food Tech., 4th C o n f . . 1943, p. 156. (6) Othmer, D., and S a w e r , F. G . . IKD. E s u . CHEY.,35, 1269 (1943). (71 Silker, R. E., Schrenk, W.G.. and King, H. H., I b i d . , 39, 1160 (1947). (8) Stamin, .-i. J., and Millett, AI. A , , J . P h y s . C'heni., 45, 43 (1941). (9) Wilson, R. E., and Fuwa. T., IXD. ESG. (:HEM., 14, 913 (1922).
SUI'('
Contribution 320 f r o m t h e Department of C h e ~ i i ~ ~ tl irnyn ., i i r Y t a t e C'ollege.
Industrial Aspects of Browning RGaction H. Ar. BARNES AND C.
w.
K.IUFMAN
Geriercrl Foods Corporntiori. Hobolien,
T h e browning reaction in foodstuffs ib attributed to a reaction between reducing sugars and proteins or other amino bodies. This reaction has steadily achiered increased recognition as a factor in the forniation of color and flaror in foodstuffs since 3Iaillard published his paper on the reaction of amino acids with glucose in 1912. The flavor and color effect of the browning reaction niaj be either desirable, as in the drying of malt and i n bahing. or undesirable, as in the storage of dried fruits and iepetables. The experience in the General Foods laboratories with dcsiccated coconut and tomato is summarized. The stud, of the formation of flaiors from the reaction of pure amino acids with reducing sugars rebulted in the discorery that aminobutSric acid and others of similar molecular weight giie rise to a flaior similar to that of maple.
M
AILLARD presented his paper in 1912 on the action of amirio acids on sugars to produce what, he called inelanoidins, dark colored nitrogen compounds of high carbon content ( 2 4 ) . Lafar (19) followed t,his 4.ith his study on foam fermentation in the sugar industry, and Ruckdeschel ( 2 9 ) with a consideration of this reaction in kiln malt, lvhereupon t h P ;\Iaillard r ~ e tion took on industrial importance for the first tinie. Ppasmodically, during thc early 1920's and in increasing tempo through the 1!)3O's, the information on the Xaillard reaction grew, t o reach a climax during the recent riar years. Today it is the subject of estensive programs in many industrial laboratories and a major effort of a coordinated attack by university, government, industrial, and private laboratories under the sponsorship of the United States Office of the Quartermaster General. Until the war ?-cars the AIaillard reaction had been just slightly more than a laboratory curiosit,y. Then operations in the tropics developed food spoilage problems attributable t o the browning reaction and necessitated immediate study to develop a means of control. =\ctually. the problrm was not ney, but it now became inipoi,tant. Earlier investigators had recognized the reaction in the sugtr industry (9, 19, 53). the beer industry (11. 201, malting (21, 22, 23, 2.91, and milk (28). 1Iore recent work has demonstrated like drt,eriorative effects on dried eggs (3,4,20, SB), molasses ( 7 ) , dehyclrated vegetables (8, 271, dried fruit (1.5, 30.
Y. J .
S I ) , dried juicos (161,bwaliiast cwc,al ( I T ) , wcoriut, (f7),soups (171, ancl dried fish and ineats (17). I t is undoubtedly a factor in the crumb color of h e a d s and cakes and is suspected of being responsible for off colors developed in canned goods during process ing. There is some indication of its occurrence in wines. One glance at this list is, the authors believe, sufficient,justification for the importance no signed to the Ifaillard or browning rwvtion, in so far as the food industry is concerned, and its eventual control assumes great importance in our plans for nwareli today. Thv course of the reaction between amino acids arid sugars has bec.11 .studied by carbon dioxide evolution (1, 2 , 6 , 1 4 ) , color change of reaction products (23), decrease in amino nitrogen during the course of the reaction ( 2 , 6, 18), decrease in pH which occurred a s the free amino groups were used u p (6, I S ) , increase in property of reducing methylene blue (6, 12, 1.41, freezing point determin& tion ( 1 2 ) , and the change in optical rotation ( I $ 14, 26). Euler ( 1 2 ) showed t h a t the decrease in free glucose determined by \~illstatt.er-Schudel method coincides with the decrease in amino acid deternlined according t o Van Slyke, and both are i n agreement with the results of the freezing point determina tioris. The velocity of the reaction depends on t,he nitrogen compound and carbohydrate used, conditions of coiicwitration, temperature, p I l and time of heating, and the presenre of accelerating and inhibiting agents. Nore recently, fluoresccmce has come to the fore as a tool in measuring the coursc of the rcac-' tion, ancl Dutton and Edwards ( I O ) report the colored compounds and the fluorescent compounds as identical. The present work confirmed this to the extent of showing that t,hc color and fluorescrnce curves are parallel, but with an interesting exception. I n tho pi'esence of bisulfite the ratio of fluorescence to color was shout five times that of an unprotected samplc,. BASIC RESEARCH WORK
In our own laboratories n-e have attacked this general problem i n two ways. Our basic rcscarch group has proceeded from the premise that a study of the reactions between individual anlino acids and individual carbohydrat,es would ultimately develop a sufficient understanding of the reactions involved, so as t o e m b l e this type of change to be controlled much more rationally than al;
1168
I N D U S T R I A L A N D E N G I N E E R I N G CHEMISTRY
/d THE YELLOWING OF DESICCATED C
ONUT
Vol. 39, No. 9
5 . Tiitsrca appoars to he a high clogwe o t spcseificity i n the kinetics of these reactions. For inbtsric-r, thc rr:tction characteristics of a gireri pair of rcwtnnts in the dry st:ite cannot IN! uwd to p r d i c t with assurance the chararttLrintics of the suine pair reacted in aqucwu -olutiori. Ilipeptidcs also participatc in tlli typi. of rwction, but their rate of reaction may br (,ither greater or lms than that of t l w component amino acids. 6. L I r a s u r d by color developnieIit, sult'ui~ dioside is still the best known inhibitor of these reactions.
e; COCONUT WITH 400 RRM.
MONTHS AT R O O M TEMPERATURE
Figure 1 l i i the coutw of t i i t , tia>ii*r trc.h ~ ~ l ' l ~ g i ' ~liiall)~ l 1 1 1 Ilrulclrc~db ot tests were made, c w h test g rally involving only oiie aniino acid and onr sugai,. In this \\.a!- moat of tho simple biriitry rnisturcs n-hich at'r pos.cit)li. h e t w e t i the kno\vri amino acids anti thc knoxn simpler cwtiohydrates \\-ere reaoteti undrr :it Icast. one .st>t df conditions. The field is still far from c*ovcwd,S O thnt :iny ~ I I clusions reported here mu!st be rrgarticd as purely t c a n t utivc,. lrsing dekt,rose and glycine mixtures in a 50% aqucwus solution :ic the simplest fomi of the reaction, a technique \vas tit:vc,lopcd f i i t the compitrison of the rates of color drvcxlopment Iiy iiic~arisof t l i c half transmission time a t 400 millimicrou-: in a cc~llof 1.3 centirheters thickness. The factors influcnring this d(~ve1opnicnthavc, been studied. p H has been found to he important. thc rea.rtioii proceeding more rapidly at higher pH valuc~s. Appar(~nt1yt h e h t . e of brovming is doubled at p H 6 to thc, rate obtuincd :it 1111 1 Temperature is an important factor. T h i w or four hours a t 100" appear to be equivalent to onc w r k at 40" C!.: this indica:itc.h n two- or thrrefold incrcasr for each 10' rise in teniperaturv. General conclusions whirh have d(vr~1opcdfmm this pr(iyi':i t i 1 to date arc as folloivs:
1. In many inataiicrs thc. (It lopnic~nt of flavor ib rathei specifically dependent upon the particular aminci arid that part,icipates in thrs reaction. 2. The reactivity of t h e sugars generally patxllt~lat he ri'activity of these rompounds in other respects. T h r aldupentoses, for instance, react more rapidly (or at loiver tempcraturr) than dothe aldohesosrs, and these, in turn, more rapidly than the tli-, t,ri-, and polysaccharides. Carbohydrates having a f r ~ ealdc,hyde or ketone group naturally react more rapidly than thosc t h a t first have t o be hydrolyzed or othcwvise broken down to release these groups. Xylose and arabinose were so reactivt. under the conditions used t h a t color developnient coul satisfactorily measured even in t,he prewnce of sodium Additional work indicated that, sylose is about forty tin active as dextrose, and arabinose about twenty times as as dextrose. 3. The over-all reaction of the amino acids and suyais a p pears t o be a complex series of stepwise reactions about which very little is known, except that the reaction mixture becomes more acid. Carbon dioxide is evolved under some conditions, and an aldehyde corresponding in a t r u c t u r ~t o the amino acid
lIo,t of the present tvork h:ts lii~eri i i i I tic. area of inhibitors. Approximately 100 differtsnt compounds have been tested. IIalf trarisiniouiot~ tinics have been dctermincd for c w l i (if t l i w , (Table I) compared to the uninhibiti~ltl(.strouc.glycine mixture. The ratio of glucoac,, glyciiir,, and rvater used was 4:1:5 by w i g h t . AI1 r'csult> to date> tcnd to confirm thc statt:nicxiit givc-ri SO, that, sulfur tiioside or sulfur-containing conipounds are effective t o a drgwe rornparablc with their ability to liberate sulfur dioside o i i hydrolysin. Hon.evcr, hydrosylamine, hydrazinc.. nnd semicarbnzide yere nioderatcly effcctivt:. In genrral, inhibition produced by sulfur dioside increased with increasing amounts of iri1iit)itur. .\I the hiillit1 time, tht, d f u r dioxide content dccrcascd markedly during tiit. course of the reaction, and it could he tlernonstratcd that such loss \vas duc t o irrevr~rsihli~ hintling mid not t o n\;itlntiori t o slllfatr'.
-
'Yhi' tiahic research groups at this laboratory have (1 IIY c:oiiclusion that a t the present time sulfur dioside, propc,rly :tpplird and in limited but, adequate quantities, represents the I
txat i i i ~ a ~of i sprewnting the deteriorative rfferts of the, brouxing
rivictioii. HYOLOGI(:&L WORK
' r h v w i . i i t i i 1 approach to the general problcin \vas Iiy t h c s rcc4hiiology group who endeavorrd to find solutions to specific product problems as they arose, fitilizing all of the informatioli available from the littlrature, or from thc haeic research of this rompany. It appearid that drih?.dratrd vcgctubles, coconut, s o u p , etc.. :ill o f irittlrest to the a p p l i d group, f(3ll inta this reaction class. Ai least they k t i t v that tach of these foods had a characteristi? dark c*olor,off-flavor, and off-odor. In desiccated coconut the authors had a probk~nir q ~ w e n t i n gtheir first acquaintance n i t h this type of reaction. Here the reaction takes the form of producing saffron yellow coconut-an unsalable item. DESICCA4TED < ' O C C I S ~ T . T h e applied work on drsiccatfsd coc'onut, has just been completed. It is summarized in the following way :
INDUSTRIAL AND ENGINEERING CHEMISTRY
September 1947
1 , .1 rcductioii i n moisture levi.1 retards but does riot stop thi; bi,owiiiig artion. This statement should be qualified by a practical limitatiori-tiamely. the inability or inlpracticability of drying belon- 2'; nioirture for general purpose use. 2. -1 shift in p f i to\r-ard the acid sidi, retards but does not stop t h e rcaction. 3 . .in irlcrcme in either invert sugar, protein, or ariiino rlitrogeri acwlerates the react ion. I t has been established that invert. sugar and protelin decrease on ripening of the coconut. A correlation tjctn-cvii analytical fa,ctors, iiinturity, and rate oi browning has h v i i establishid. 4. Partial reinoval by learili,ng. oxidation, ferment,atioIi. etr., of c'ithcr aniirio nitrogen or mvert sugar retards the reartion, but hy all techniques used to date has so altered this fl:tvoi, of t h protluc't as t o iiialte appliration impossible. 3 , Aniong :ill rheinicd trcBatnients ti,icd, sulfur dioxide ill c o i i c i ~ n t ~ ~ : ~ tofi i ~levs ~ i sthan 300 parts per niillion p r e v e n t d t h i s reactiori From occurring. This ic tbffectively demonstrati4 in Figurc 1 , \vliic.h s h o w ( a ) a coloi. vhaiigP curve actually t lislicd by plotting w l o r change occurring in norxiial ( i o o F.) storapt' ovi'i' a long p i ~ i ~ i oofd timc, and i b ) the cclor change, i l l thi. prcw~tii-c~ i i f ~ ~ u l i ' dioxidc ui~ 0x1 tlie s i n i t ' c o l o ~ srirlcx. ~
Figul,c, 2 a h o w ( ( L ) an accclc~ratedchange cul'vc nladt! 111. 173" F., ( b ) ail accelerated chaiige curve showing the effect of adding pulfur dioxide ab an inhibitor, (c) the increased rate of color tioii in this pi~cscrircof $odium acetate, pH 8, ( d ) the de rat0 iii til13 p i ' i w n w o f calciuin chloride, pH 3, and ( e ) ti i'atc' rausrd bj- leachiiig out the 1r.atc~r-solut11ematerial. DI:FXY~~K ~ T E I )TOM &TOM. :Is anot hi,r example of even more coniples naturi~,dehydrated tonlatoes ~ w r e, ~ k ~ t for e d study. It, was recognized that two rhemic~ilctiariges \ r e r i ~irivolvtd i n the storaae effects iiotcd OII t h k product. Iri addition to t,hc bitter! tlark cuiiipli~xin the swum! this developmt~ntis p1.t.t w l d or paralleled by the appearanrcl of the hay flavor aiid odor characteristic of soiiic of thil highlh- pigniented dehydrated vegetables. Oiie i,caction is apparent,ly catalyzrd by tiit: othcBi,, and as a result dehydrated t o m a t o r s have an t~srrcmr4y short shelf life. Bilcaurtl o f ita il~~licatc~ naturc the product is also tistwnicly hcat strnsitivc~.and owrhcating dui,ing drying, with c~onsequcnt vhemical changes, att~ the ifi-flavor appears t o a i * i ~ I ~ ~ rinarki~illy c*hangesiu .-torag(:. w Although niuch 01 t h i h work is still going mi, it ninv tw wmriiai,izc*clNI: follon-.; to datc,: ~
1169
in general be coiltrolled by the use of sulfur dioxide in quailtitics of less than 1000 parts p t niillion, ~ so that the eventual consumer iJbt,aini: a product superior t o that which he would othcrrvisr obtain. FL4vons FROM A M I ~ O ACIDS
111 tlie eourst of the fuiidainc~ntalstudies, it \VM not4:d that the fiavors and odors developed were apparently c:haractt!ristic 01 the particular amino acid being used. Aisurvey t o t the literature indic:ati:s a similar observation by earlier workers. Lint,ner (2Jj rcpo.rtc:d that tht: aroma, tax, arid color of the reaction products of rhe mnirio arids and sugars arc%different, with c!achamino acid u d . Iadretention in hcer aiid that t1it:y f u r t h ~ iitn, j)i,ovetl tht>tastc: of beer. Tht: prtwnt, authors noiv belirve that, in addition I O twiiig reS~)lJiiSibkfor many deteiiorative changes thal owur iii i'ood produrts, t h e Alaillard reaction may also be the contributirig factor iii the dcvclopnient of many of our characttxri-tic. food flavom. .Although no evidence is as yet available, thercl is rcason t o suspect that the distinctive flavor differences in breakfast foods, the crust of baked bread, r o s t e d coffee, et,c., may be attributed to cheniical votnbinationc:h m i g h t ahout during t,he heat ti~tvttmi:ntopf~ation.
w: THE YELLOWING OF DESICCATED
OCONUT
(11 I% SODIUM ACETATE 12) CONT90C-UNTREATED CCCONUT BIWHOLE COCONLlT L E A C H E D 3 HRS. IN WATER (4)1% CALCIUM CHLORIDE
1. Good htoi'age life depends upoil adequate coiitrol of di,ying to w e that the product is The criticaal tempei,ature of c-uti'enicly Ion.. probably belon150" F. 2. 1)arl;c~iiirig 01 the w r u m is attributable t l J tlie a i i i i i i o acitl-aldehyde reactioii plus a protx~ble t I'UC c~aramelizatiori of this invert sugar by tlic high acid in the dry firial mixture. This c~araiiic~lizittiorl. as ne11 as the Maillard reaction, is t.xtrenic~lyausceptibli~to traces of
moisture. l l e t a l ioiis in tracc, qua ' ' the rat(i froni pig liidiie>-s, Trcatirig the. juice \\-itti di~amiriwc~. had a ,siniilar effect. 3. T h c hay flavur dr.velopiiieiit, paralleling 1o*s of red color,
2
3 HOURS
4 AT
5
6
7
I75'F.
Figure 2
As ti specitic instarice, the applied laboratories of this company have denionstrated that soy flour may be processed with special sugars in a variety of ways to develop distinctly different flavors. The flavor characteristics resulting in the final product appear to be closely correlated with time, tmiperature, rnoisturr,, :rrid pH rc,littioris of the haking opeiitt ion.
originallj- s n s p i ~ c t i d . \ I P L E FI,\VOK
In the course of coiisidering the f'uiidarnciital rcactions of amino is inaiurninciti. acids with reducing sugars, the authors went over again some of is a ncwwity to lic~c~piiig 5 , Pro1ic.r parkaging at lo\v rnr)istui~~s the same ground that was covered by LIaillartl thirty-five years ago quality. and covered by many workers oince. It was objerved that iiiilriy I n wiiniiiig. up all of t h c w p'actical obscrvat,iom, the auth0i.h i)f the amino acids when hwteti with glucose or levulose gi:nc>rated ~r~riclutlc. that thew complcx amino acid-aldchjdc rcvictioiis rail odors which rrerc similar to thix aldehydes having o w less c::t~~l)oii
INDUSTRIAL AND ENGINEERING CHEMISTRY
'1170
atom than the amino acids. The e\aniplc of this which iq most easily and certainly recognized is the benzaldehyde odor which is formed by the reaction of glucose n ith a-aininophenylaeetic acid.
SEI, +--C C'OOH H -
d--c=O H
+ dextrose -+=
$; ?
A large nuniber of amino acids fortlied aldehyde odors in this manner, the requisite conditions being that the aldehydes bi> volatile and that they have an appreciable odor intrnsity. Tyrosine, for example, would not be expected to givr a volat,ile aldehyde, and the acetaldehyde derived from alanine \vould riot be expected to have a high odor effect. The amount of aldehyde formed, hovivever, is quite small. The major course of the renct,ion leads elsewhere. Sourness and bitterness are the main flavor products. Illoreover, the aldehyde does not appear to be the onl~, aromatic body formed. Leucine, for esample, gives an odor which undoubtedly contains isovaleraldehgde, but the total odor of thtt reaction is reminiscent of rye bread. Similarly, a-amino-n-butgric acid when heated with glucose yields an odor that is niort' -complex than t,he expected propionaldehyde. The total odor was strongly suggestive of maple. The aminobutyric acid was not the only amino acid w h i h would develop maple flavor on heating with gluc,ose.
Vol. 39, No, 9
snnrIKY The lluillard or bronxiii r(~;wtion,IL. it i- txwmiing 1110re commonly knowi, is respon ihle for niaiiy of the dutcriorative changes t h t take place in our food products, and it is amenable to contrul by the use oi sulfur dioxide in reasonable quantities. At this writing, a less frowned-upon control been found. Slorc recently the browning reac added industrial implication, that of being im of developing flavor characteristics common to many of our food products. This area is only now beginning to unfold. Fundamental research on this aspect may provide a means of chrniically synthesizing some flavors which will nrarly approsiniatt the natural flawit,.. .ICK30WLEUG\IE\IT
Tlir authors are indebted to Clair lietinedy of thc. 0i.gmir Rcsearcxh 1,al)oratory for his ninny contributions t o the fundamental work on browning and to .i. \I-inston ( ~ ftlic Food Teclinology Laboratory for most of the applied work on tii'siwated coc.onut, is n.ell as to a large number of the technical pi.i~sorini4at C 'c-!iti,al I,ahoratories, General Foods Corporation. LITERATURE CITED
Akahori. Shim J . Chem. SOC.Japan, 52, 606-10 (lW..A . K., and Smenson. T. L.. Food Research, 1, 319 (1836). Bate-Smith, E. C., and Hawthorne, J. It., J . S O C .C'hem. I n d . , 64, 297 (1945). Bauminger, Bella, and Lieben, Fritz. Biochem. Z . , 292, 92-7 (1932.
SHr I
Serine
HOC'H-->iI)kto catch the gases evolved, and it was found that 1 mole of tiniinobutyric and 2 moles of glucose formed 0.53 Inole of carbou tiioside, 5 moles of water, a trace of propionaldehyde, and insoluble m a t e rial equal to 147' of the product Twiglit. There was 0.17 molr of amino acid left, judged by formol titration, and 0.30 mole of organic acid formed, judged by straight titration, The pH of the solution of the product was 3.6.
id. Landic.. 42, 737-46
Lichen, E'rita. and Baurtiiuocr, Bella, Biochem. %.. 279, 321-5 \1935!. Ling, -4.I t , , J . Inst. Bracing, 14, 193-515 (1909). Linttie!. C'. J . , Tochschr. Braii., 32, 201 (1915). Liiitner. C .J., Z.ges. Braziu.. 35, 545-8 (1913). 1l:iiIlard. 1,. C., C o m p f .r r m L 3 154, 66-8 (1912). Seuliei.g, Carl, and Iiohei, Maiia, Biochem. %., 174, 1 6 4 - i H
I
