Estimation of Antioxidants in Lard and Shortening J. H. MAHON AND R. A . CIIAP3I.iN Food a n d Drugs Laboratory, Department of RTational Health and Welfare, O t t a w a , Canada Because mixtures of antioxidants are now widely used in the stabilization of lard and shortening, methods for the determination of combinations of antioxidants are required. Procedures have been developed for the extraction and colorimetric determination of propyl gallate, butylated hydroxyanisole, nordihydroguaiaretic acid, and tocopherol in lard and shortening. The procedure permits the determination of all combinations of these four antioxidants except when propyl gallate and nordihydroguaiaretic acid are present in the same sample. The lower limits of the determinations, with a high de-
P
ROPYL gallate (PG), butylated hydroxyanisole (BHA),nordihydroguaiaretic acid (NDGA), and tocopherol are considered to be the most important antioxidants used in food products in Canada and the United States. Methods arr described in the literature for the determination of t’hese individual antioxidant’s,but there are no procedures for t,he quantitative determination of mixtures. As the present trend is to employ combinations of antioxidants in lard and shortening, there is a definite need for a procedure by means of which mixtures of antioxidants may be determined in these products. The following methods are capable of determining mixtures of antioxidants present in lard and Shortening in the concentrations usually employed. Mitchell ( 7 ) reported that gallic acid and other trihydroxyphenols react with ferrous tartrate to form an intense purple color which he believed to be specific for the pyrogallic grouping. Later, Glasstone ( 2 ) investigated the influence of p H on this reaction with gallic acid and recommended the use of an ammonium acetate buffer of p H 7.0. Lundberg and Halvorson (6)stated that ferrous tartrate gives a colored reaction product with most; if not all, polyphenols but that the product,s are insoluble in many cases and, therefore, cannot, be estimated colorimetrically. Mattil and Filer ( 6 ) ,using Glasstone’s modification, found that the soluble complex formed with gallic acid exhibited a maximum absorption a t 540 mp and obeyed Beer’s law over the range of 0.2 to 1.0 mg. per 100 ml. of solution. Propyl gallate has been found to react with ferrous tartrate in an analogous manner. The 1,l’-bipyridine reagent has been employed by Hill ( 3 ) for the colorimetric estimation of iron in biological material. Enimerie and Engel ( 1 )later used ferric chloride plus 1,l’-bipyridine for the colorimetric estimation of tocopherol. The literature contains many applications of this reaction to the determination of reducing wbstances, including propyl gallatr, nordihydroguaiaretic acid, and butylatrd hydroxyanisole (4).
gree of accuracy, are propyl gallate 0.004%, but)lated hydroxyanisole 0.005%, nordihydroguaiaretic acid O.O05F/O, and tocopherol 0.0157& However, satisfactory results have been obtained at much lower levels. Recoveries ranged from 96.4% for nordihydroguaiaretic acid in combination with butylated hydroxyanisole, to 98.4% for nordihydroguaiaretic acid alone. These procedures permit accurate determination of combinations of antioxidants in lard and shortening, as well as study of the relati\e rates of destruction of individual antioxidants in mixtures in lard and shortening.
Lundberg d n2. ( 6 ) have reported that t h t cpiintitative extraction of a number of antioxidants from fats, tvith the r s c q t i o n of tocopherol, could be achieved using 80% ethyl alcohol. In this invest,igation it was found that repeated extraction of a petroleum ether solution of the fat with 80% ethyl alcohol quantitat,ively removed propyl gallate, butylated hydroxyanisole, and nordihydroguaiaretic acid and also a portion of the tocopherol. However, by reducing the ethyl alcohol content to 72% it \vas possible to extract the propyl gallate, butylated hydroxyanisole, and nordihydroguaiaretic acid and leave 98.5% of the tocopherol in the petroleum ether solution of the fat. When alcohol concentrations lower than 72% were employed, the extraction of butylated hydroxyanisole \vas incomplete and therefore 72% ethyl alcohol was chosen as the extracting medium. The effect of ethyl alcohol concentration on the extraction of butj lated hydroxyanisole and tocopherol from a f a t solution is shown in Figure 1. Owing to the similar solubilities of propyl gallate and nordihydroguaiaretic acid, these antioxidants cannot be separated completely when both are present in the same saniple. However, as combinations of propyl gallate and nordihydroguaiaretic acid are not employed as antioxidants in foods in
P I
c
Preliminary experiments indicated that the ferrous tartrate reagent for propyl gallate and the ferric chloride plus 1,l’-hipyrine reagent for butylated hgtiroxyanisole, nordihydroguaiaretic acid, and tocopherol were satisfactory; therefore, these reagents were studied more intensively.
/ TOCOPHER
t
-
o
EXTRACTION OF ANTIOXIDANTS FROM LARD AND SHORTENING
~
60
It appeared possible that propyl gallate, butylated hgdroxyanisole, nordihydroguai retic acid, and tocopherol could be extracted individually or in groups on the basis of their differential solubilities in wat,er, rthyl alcohol, and a petroleum ether solution of the fat. Preliminary experiments indicated that propyl gallate and a portion of the nordihydroguaiaretic acid could be removed from a solution of lard or shortening in petroleum ether by aqueous extraction, xvhereas neither butylated hydroxyanisole nor tocopherol JTas removed by this treatment.
65
I
k
70
75
80
% E T H Y L ALCOHOL I N EXTRACTING SOLUTION
Figure 1.
Effect of Alcohol Concentration
on Extraction of Butylated Hydroxyanisole
and Tocopherol from Petroleum Solution of a Fat
Ether
2 mg. of BHA, 10 grams of kat in 40 m l . of petroleum ether 10 m g . of tocopherol, 10 grams of fat i n 40 m l . of petroleum ether Four 3-minute extractions with 25-ml. aliquots of alcoholic solutions
BtIA.
Tocopherol.
1116
V O L U M E 23, NO. 8, A U G U S T 1 9 5 1 either Canada or the United States, it is possible, by applying qualitative tests, outlined below, to ascertain the presence or absence of propyl gallate. If propyl gallate is present, it can be extracted into the aqueous phase and the absence of nordihydroguaiaretic acid assumed. If propyl gallate is absent, the aqueous extraction is omitted and nordihydroguaiaretic acid and/or hutylated hydroxyanisole is extracted with 72% alcohol. The aqueous and alcoholic exti action procedures can be used consecutively for the extraction of propyl gallate and butylated hpdroxj-anisole from a petroleum ether solution of a fat. I n all cases the tocopherol remains in the fat solution. DEVELOPMENT O F R E i G E N T S
1117 hydroxyanisole preparations and nordihydroguaiaretic acid on reaction with the ferric chloride plus I, 1'-bipyridinr reagent. For all subsequent work commercial butylated hydroxyanisole preparation No. 1 was used as a reference standard and 30 niinutes were allowed for color development.
1
rIi! -0-0
Prop$ Gallate. The ferrous tartrate reagent as modifird l ~ y Glasstonc ( 2 ) we used for the colorimetric analysis of proriyl < 0.2 i gallatr. This reagent was found to be specific for propyl gallate among the four antioxidants under consideration. .4lthough :i high concentration of nordihydroguaiaretic acid does produce a hlue prcxcipitate with ferrous tartrate, this precipitate is insoluhlr and can t w removcid b y rrntrifuging, thus produchg no txi'ror it] the colorimetric determination of propyl gallate. Blthough thc. 0.0 ferrous tartrat,? reagent ir I c w sensitivr than the frrrir chloi~iti~~ plus I,l'-hipyridinv reagent, the former \\-a~adopted t r c a u r e of 10 20 0 0.50 I .o it3 specifiritl-, rapitfit!- of color formation, and stability of the TIME I N M I N U T E S color. The ferrous tartrate-prop?-l gallate colored complex exhi1)its :t maximum ab~orptiona t 540 nip, and obeys Rrer'a la\\- ov('r Figure 2. Rate of Reaction of Propyl Gallate with Ferrous Tartrate Reagent the rangr of 60 to 360 microgrimis of propyl gallatr per 25 mi. ( i f solution. The rat,? of color dwrlopmrnt is shox-n in Figure 2. Butylated Hydroxyanisole and Nordihydroguaiaretic Acid. Tocopherol. For the determination of tocopherol, 2 nil. of hlasinium color formation was ohtained when 2 ml. of 0.2% frriir 0.2y0 ferric. chloride plus 2 ml. of O . ~ j 7l~l'-~)i~)>.i.iiliii(, ~ \ri~rrfound chloridr (FeC1,.6HzOj reugent and 2 nil. of 0.2% l,l'-l)ipyi~iditie t o give the most satisfactory results unclrr the conditions emreagent \ \ w e added to not more than 50 micrograms of the foris~ ) l o y r d .The exact procedure is given in the wction outlining the going antii~xidantcvontninrd in 8 ml. of 82% eth>-lalcohol (hut>.method for the determination of tocopherol. Iatrcl hytiroxyanisol~extrart in 5 nil. of 72y0 ethyl alcohol plus 3 nil. of 10070 rthyi iilcoholi. With this reagent conii-)ination i ~ f REZCEZTS fective nxixinia of coloi formation were obtained after 30 minutes Petroleum Ether. 1Iix 1 voluinc of 30" to 60' C. pc'troicum for coninic.rcial buty1:itt.d h>-droxyanisolepreparations and i i f t c s r c~thtt~ with 3 voluiiics of 60" to 1003 C. prtroltwm c8tht:t. (SkeI1,~3 m i n u t e for nordihydroguaiarPtic acid. Commerrial preparasolve 11j :ind shaltc t h e niixtui~:with one tdantli its ~ - o l u ~ nof c ' colitions of 1)utyIiited hydroxyanisole available for thi ccntratcd ,sulfui.ic acid for 5 iniiiutcs. Itun off t h f a y~~Ilo\\-ish mixtures of the isomer$. 2-t~,.t-but~l-i-hj-droxyanisoleand 3-trrtacid layer :rnd wash the pctrolrum ethor rvj)eatc:dly \\.it Ii \\':iter and theti with 1% p o t : t ~ ; ~ i uIiytli~ositl(~ m s i ~ l u t i ~ until ~ti i ' r r ~of I)utyl-~-hydrosyanisolr. It \vas found that the 3-tert-l,utyl-l-:icid. I1i;Rtill the prtrolt~umether, using :iti alI-gl:i~+; t p p : i i ~ ; i ius. hydrox>-;triisole react.? more rapidly with ferric chloride plus 01,01'ohol. T o :it)solutc alcohol add :t~)pr~)xitii:rt~~l~~ 0.1$; hipyridiiie than 2-t~rl-butyl-4-hydroxyanisole, but thitt 2-tevtn hydroxide, and 0.1 c; of pot iuiii ~,c~t~i~inti~:inate. I)utyl-i-hydrosyanisole produces more color per unit \\-eight -glass apparxtus. T h r distillatr 1s u*cd for, the, 100TO ~ilc~oliol aticl is diluted \\-it11 witer to contain 86 :ttld 7 2 r , :Llcohol than 3-tof-butyl-4-hydros~-~~iii~oIe. Results indiratr that maxib y vollllllc~. nium color formation is obtained in 15 minutes with 3-tert-butylFerrous Tartrate, 0. I cC hydrated fwroui. wlE:itr 4-liydros~~niiisoIe,in 45 miiiutes with 2-tert-butyl-i-h?-dr(~sy- and O..icC of Rochelle salt i u distillrd w a t r r , f i , c 4 i l ~ . :inisole, : t i i d in 30 minutes n-ith three commerc-ial hutj-1att.d Ammonium Acetate, 1.23, 1.67, and 10.Ori w l u t tiori is :ilso rcquircd containing 1 . 6 7 7 of animoni hydroxyaiiisoli. prep:irations. Over the range of 10 to 50 niicro5'; aqiieous cthvl alcohol. grams of Iiutylated hydros>-anisolein 12 ml. of solution, the caolor Ferric Chloride, 0 . 2 5 solution of hvdt,atcd f t > i , r i c . 1.11Ioritie forinrd obeyrd Hrer'v h \ v , The color formed with nordih>-dro(Fe('L6H20) iri purifird :il)solute (Jthyl'aIc.olro1. Tlli; ~,e:igcsnt guaiaretir ac4d d s o olx~yrdBrei's law over the same concentration must tie fresh to avoid highly c*oloi,c~l I)lnnks. 1,l'-Bipyridine, 0 . 2 5 and 0..jri solution.: iii purifictl :il)i.oliite rxnge. Ivhcn the a t ) w l ) a n r y 1v:is nic:isuretl niter 3 minutes. ethyl alcoliol. T:O)Ir I s h o \ ~ t.he s ;r)).wi+)anrypc'r rnicrogr:tni of various butylated Borax, 2.05; aqucous solution of 11or:tx (X;t?B4O:.101120),
I
Q U .4 LIT 4T I V E TESTS
l a h l c I . -4bsorbancy per L-nit Weight of 1-arious Butylated Hydroxyanisole Preparations and Norclihyl r o guaiaretic Acid on Reaction w-ith Ferric Chloride plus l,l'-Bipyridine Keagent hntioxidant
Pure 3-BH.4
Ileactiun Tirne, lliniite-
.4beorbancj- '> at 515 t i l +
30
0.0154
so. 1
30
0 01z:,
5-0. 7
.? 0
0.013fi
:ill
0 0136 0 0170 0,015.5 0 0161
Cotiiiiiercial Blt.4
x
50. Piire 2-BII.4
XDGA 1-11(id
30
x
30
intl i P:I t ('5 the presc.n CP of t)ut ?-lat i d I1yd ro~yitrl i.*oh~. J i : i n cesrivi, :iniount of 2,6-dichlor~~quirioii~~~~hI~ii~iiiiiil~ is : I I ~ ~ I I Y I tlic, . color formed may he greenish t)lucJ. This te-droguaiaretic acid. However, when both conipounds are present it is necessary to set up a duplicate series of tuhes and to determine the absorbancy of one series of tubes after 1 minute and the absorbancy of the other series after 30 minutes. The absorbancy measured after 1 minute is due to the ieaction of CO% of the nordihydroguaiaretic acid and 10% of the but>lated h\-droxyanisole. Similarly, the absorbancy measured after 30 minutes is due to the reaction of 100% of the nordihydroguaiaretic acid and 100% of the butylated hydroxyanisole. Therefoie, it was possible to derive the following siniultaneous equation.:
L a t 1 minute L a t 30 minutes
+ 0.9 S = 1.0 B + 1 . O S
=
0.1 B
several times with petroleum ether. Combine the washings arid dilute to volume with petroleum ether. If qualitative tests have shown that the fat does not contain propyl gallate, butylated hydroxyanisole, or nordihydroguaiaretic acid, a 4% solution of the fat in petroleum ether can be employed without any previous extraction. Place three duplicate aliquots of the fat solution, varied from 1 to 8 ml., in duplicate series of 125-ml. low actinic, glass-stoppered separatory funnels. T o one series add 2 ml. of a petroleum ether solution containing 15 microgranis of d,a-tocopherol per ml. Dilute the fat solution in all separatory funnels to 10 ml. with petroleum ether. .4dd 2 ml. of O.2yO ferric chloride reagent and 2 ml. of 0.5yo 1,l'-bipyridine reagent. These reagents should be blown from pipets into the fat solution to ensure satisfactory mixing. Six minutes after addition of the ferric chloride, add 14 ml. of 86% ethyl alcohol (specific gravity 0.843 a t 2Oo/2O0 c.)from a rapid delivery pipet. ( A 15-ml. pipet with the tip cut off was found satisfactory for this purpose.) After 9 minutes run off the alcoholic layer into 40-mm. absorption cells and measure the absorbancy a t exactly 10 niinUtes after the initial addition of the ferric chloride reagent, employing a Coleman Universal spectrophotometer set a t a nave length of 515 mp. Measure all absorbancies relative to the blank prepared by using 10 ml. of petroleum ether in place of the fat solution. The tocopherol content of the aliquot employed for the d(3trrmination can then be determined by the following formula:
("-) ST - S
X 30 = micrograms of tocopherol per aliquot
where S is absorbancy due to tocopherol in the fat, and S T is absorbancy due to tocopherol in the fat solution 30 niicrograms of d,a-tocopherol.
+
here L is observed absorbancy, R is absorbancy due to butvlated hvdroiyanisole, and S is ab-orhancy due to nordihydroguaiaretic acid. By nieans of the foiegoing equatioiii it is possible to calculate the respective absorbancies due to nordih? droguaiaretic arid and 1)utylated hydroxyanisole. The value of S divided bv a constant of 0 0161 (see Table I ) and the value of B divided by 0.0155 nere found to give the concentrations of nordihydroguaiaretic acid and butylated hydroxyani-ole, respertively, in microgranis per aliquot. Tocopherol. T h e fat solution in petroleum ether, from vihich the propyl gallate, butylated hydroxyanisole, and/or nordihydroguaiaretic acid has been extracted, is used for the determination of tocopherol. Owing t o the depressing effect of fat on the color produced per unit weight of tocopherol in the presence of ferric chloride plus l,l'-bip>ridine, it i- necessary to use an internal tocopherol standard. I\
The most accurate range was found to be from 20 to 120 niicrograms of tocopherol per aliquot of fat solution. Table IV.
Recoveries of Antioxidants Added to Fat Freed of Oxidizing Rlaterials Determinations s o . of
.4ntioxidant PG XDG.4 BHA
BH.4
after extraction of P G
6
Added. Amount%
Recovery %
0.01 0.01 0.02
98.0 98.4 97.4
0.01 0.02
96.4 97.0
0.02
96.9
REPRODUCIBILITY OF RESULTS
Transfer the fat solution after extraction of propyl gallate, hutylated hydroxyanisole, and/or nordihydroguaiaretic acid to a 250-ml. volumetric flask and nash the separatory funnel
Table 111. Precision of 3Iethods for Determination of Prop>-I Gallate, Butylated H>-droxyanisole, and Sordihydroguaiaretic .4cid, and Tocopherol in Lard and Shortening Substance
Mean of 6 Detna.,
yo
Standard Deviation
997; Confidence Liniitc for Single Detn. (4.032 X Std. Deviation)
I'( ; 0.01036 o.onno47 -0 I'OG 0.00677 o.ooon46 +n RH.4 0.02143 n 000247 -0 BHA (after PG estraction) 0.01917 o.oooo28 -0 s D(;A o 01029 o.000134 *n NDGA (rvith B H A ) 0 01059 0.000063 =n BH.4 (with S D G A ) 0.01942 0.000556 in Toropherol 0.1303 0.00291 -0 Tocopherol (after BHA extraction) 0.1233 0.0033i 10 .4ddition of ortanol t o f a t Qolritionfollowed by estraction 1.67'3 amnioniiini acetate in 5 9 aqueoii,- ethyl alcohol.
00019 00019
onin0
om11
nnn.i4 00023 om24
RECOVERY OF .4NTIOXIDANTS
0117
013.5
of
P(; x i t h
I f
Standard deviation =
12
'I'
(ri
- 'I2
- %2
nhcre 7 = mean of 6 values, column 2 4.032 = value of student's t for I' = 0.01 and degrees of freedom = j
I n order to determine the precision of the foregoing analytical procedures, six identical fat samples were analyzed for their respective antioxidant or antioxidants. I n all cases the extract from a single sample was analyzed a t three concentration levels, and the i.esults were averaged and reported as a single figure. The results of six such determinations of each antioxidant, or group of antioxidants, were statistically analyzed and are shown in Table 111. Yo attempt wa'i made to employ the same amount of antioxidant in each test, although in all cases t,he concentration was in the range normally expected in foods. I t appears t,hat adequate prerision can be obtained by rrporting a single value which is actually the average of three simultaneous determinations at different concentrationP.
=
(6-1)
To obtain an accurate indication of the recovery of antioxidants i n lard or shorteningusing the forrgoingprocedures, it was necessary to prepare a fat free of peroxides and other oxidizing materials. Oxidizing substances, if present in the fat, react with the antiosidant or antioxidants and thereby rewlt in a Ion- recovery. F a t samples free of oxidizing materials were prepared by holding the fat containing approximately O.O5Y0 of tocopherol at 8O"C.for l0minutes. If, oncooling, tocopherol was still present, it was concluded that t,here vas an excess of tocopherol over oxidizing materials, I n the case of frwh shortenings there is us,w,lly
ANALYTICAL CHEMISTRY
1120 sufficient natural tocopherol present and, therefore, none need be added. Table I V gives the recovery figures for various antioxidants added to fat previously freed of oxidizing materials. S o recovery figure is given for tocopherol because this antioxidant is determined by means of an internal standard. DISCUSSION
I n the determination of propyl gallate with the ferrous tarrrttte reagent, the color reaction is very sensitive t’o variations in the p H . Opt,imum results xere obtained when a p H of 7 was employed and in most cases the ammonium acetate buffer was sufficient t o hold the p H close to 7. If, however, the solution is too acid. the color formation is incomplete and if the solution is a l h line the purple color has a tendency t,o fade. Gallic acid, tannic3 acid, and gallates other than propyl gallate may give a color reaction with ferrous tartrate at, p H 7 which is identical to that produced a i t h propyl gallate. Hoivever, tannic acid, gallic acid, and gitllates ot,her than propyl gallate arc not usually emploj-ed as ailtioxidants in either Canada or the Unitrd States. Guni guaiacum if present in lard or shortening \vi11 be extracted from a petroleum ether solution of a fat by 727, ethyl alcohol. Reducing substances in the gum react with ferric chloride plus 1,l’-bipyridine to produce a red color, which might be mistaken for butylat,ed hydroxyanisole or nordihydroguaiaretic nciti. Guni guaiacum can be readily detected bj- adding a few millilitc~s of the ferric chloride rexKent to an aliquot of t’he 72y0 ethyl alcohol extrart. The appearance of an intense blue color that rapidly fades indicates the presencbe of this material. This color phoultl not be confused with the purple color produced hy propyl gal1at.c.. From the data in Table I, it can be seen that, 2-tert-Lutyl-4hydroxyanisole produces approximately 10% more color per unit However, in the cwe weight than 3-tert-butyl-.1-hyd1.oxyanisole. of commercial hutylatrd hydroxyanisole preparations I, 2, :rnd 3 , the variation ip approximately 1%, measured a t 30 minutw. Therefore, for investigutiond or control purposes the butylated liydroxyanisole prcxparation r.mplo>-(ddiould be used as the 9t:intlard . The ferric chloritle-l.l’-l)ipyridine re%gcnt is not specific for butylated hydroxyanisole, iiordihydroguaialrtic acid, or t.ocopherol and therefore could react viith other reducing eul)stances present in the fat. Hun-evrr, no such intcrfermc-e was encountered in any of the fat samples txx:tniined in this study. The analytical method for the tletcarminiition of tocopherols is not as precise as the methods for the othrr antioxidants. d;a’-
Tocopherol )vas chosen as an arbitrary standard for this work. However, each of the four tocopherol isomers produces a different color intrneity with ferric chloride and 1,l’-bipyridine (9). The procedure of Parker and McFarlane (8) was tried but was not sufficiently sensitive for the determination of the relatively small amounts of tocopherol present in lard and shortening. I t was evident that large aliquots of fat would have to be employed in order to obtain sufficient tocopherol for a reasonably accurate analysis. The foregoing procedure was therefore devised, whereby a fat aliquot up to 320 nig. could be used compared to a maximum aliquot of 20 mg. in the Parker and McFarlane (8) procedure. The use of ferric chloride plus 1,l’-bipyridine in 100% ethyl alcohol e n s u r d complete niisrihility of these reagents with the petrolcum ether solution of the fat, and thus permitted rapid reacti.m \vitli the tocopherol. The subsequent addition of 1.1 ml. of 86Yc ethyl alcohol l o ~ e r c c lthe ovrr-all alcohol content to 90%, ~vhich wilted in :t separation of the alcoholic phase from the petroltwm ethoi, carrying n.ith it the red ferrous bipyridine complex, n.hile niost of the, fat reniained in the petroleum Pther phase. This piucetlure is intended only to give a reasonable estimate of the tocopherol present. On.ing t,o the presence of traces of carotenoitis in some ehorteningp, the results may he high unless the fat solution is treated to remove innrotenoids before determining tocopherol ( 8 ) . ACKNOWLEDGMEXT
The authors wish to express their thanks to C. W. Dunnett fur statistical analJ-sis of the data and to the Universal Oil Products Ch.,the Tennessee Eastman Corp., and the Heyden Chemical Corp. for samples employed in this study. LITERATVRE CITED
(3) (4)
(5) (6) (7)
ninicrie, A,, and Engel, C . , Rec. trctc. chin^., 57, 1331 (1938). lasstone. S.,A n a l y s t , 50, 49 (1925). Hill, R., Proc. R u g . Soc. L o n d o n , B 107, 205 (1930). Iit.ayhi11, H. It.. Dugan, L. R., Beadle. €3. W., Vibrans, F. C.. Swartz, V e x . , and lieaabek, H., J . A m . Oil Chemists Soc., 26, S o . 9, 449 (1949). Luntlherg, IV. 0 . . mid Halvorson, H. O., Pim. I m t . Food Tech/ i d . . 1945, 115. Nattil. K. F.. and Filer, L. J., Jr., IND. ESG. C H E Y . ,- ~ I S A I . . ED., 1 6 , 4 2 i (1944). hIitchell, C. A , , Arin!j/st, 48, 2 (1923).
(8) Parker, 11.. E., and RIcFai,lane, \V. D., Can. J . Resrnrch, 18B, 405 (1940). (9) Stern, XI. H. and Raxter, J. C . , A i s tCHEJI., ~. 19, 902 (1947). RECEITED August 1-1, 1950.
Butylated Hydroxyanisole in Lard and Shortening Control A nalys is J. H. 314IION
4\D
K. 4. CII4PJI.IN
Food and Drug Laboratories, D e p a r t m e n t of \-utional Health und V e l f a r e , Ottawa, Canada
B
UTl-LrlTED hydroxyanimle (J3H.i) has been reported to I)e a very effrctive antioxitlant for animal fats (4,:ind thip information has stimulated interest in nwthods for its qunntitutive determination in l ~ r dand shortening. h niotiificnt ion of the Emmerie antl Engel ( 2 ) fwric chloridr plus 1. L’-liiliyridinc, reagent has Leen employed satisfactorilj- ( 7 ) for thc tic~tc~rmin:ition of this antioxidant, hut this reagent is unstal)le to light nnrl is not, specific for h t y l a t r d hydrox>.anisole. Therefore. this study was initiated for the purpose of developing a rrsgent for the determination of hutj.latet1 hj-tlrosy:inisolr w1iic.h does not exhibit these disadvantages .
Gihlis (3) found that 2,B-dicl~loroquinonec~hlorimide is a senpirive re:igeiit for phenol antl that the iwulting blue indophenol is ?table. Prt,liniinary experiments indicated that this compound iras highly specific for butylated hydroxyanisole among the antioxidants currentlj- added to lard or shortening. Tlierd‘orr, the reagent \vas studicd more intensively with a vieir to ad:iptinp it for thc quantitative tlcsterniination of Iiutyl:ttd lij-drosymisole. EXPERIMERTAL
Previous work in this laboratory ( 7 ) estalilishrd that 727, ethyl alcohol ( b y volume) was a satisfactory solvent for the ex-