8/7/2019 jf00121a031

1/5

J. Agric. FoodChem. 1984, 32, 123-127 123vivo activity at least when administered in the diet, (2) invivo inhibition correlates well with synergism of methylparaoxon, (3)lack of antiacetylcholinesterase activity bymost compounds make them useful tools for selective in-hibition of certain B-esterases, and (4) the degree of syn-ergism is too low to be of economic value in boll weevilcontrol.

1, 1642-44-0; , 85480-01-9; , 78-48-8; ,egistry No.78788-155; ,3819-72-5;,68598-42-5;,68598-41-4; ,2797-64-0;9,68598-40-3;0,68598-39-0; 1,68598-38-9;2, 8598-37-8;3,85480-02-0; 4, 85480-03-1; 5, 85480-04-2;6, 68598-36-7; 7,85480-05-3; 8, 68598-35-6; 9, 85480-06-4;0, 85480-07-5;1,85480-08-6;2, 26115-85-5; 3, 26115-86-6;4, 24067-01-4;5,85480-09-7;6, 24067-02-5;7, 85480-10-0; 8, 30299-04-8;9,9016-18-6.85480-11-1;0,4081-23-6;ethyl paraoxon, 950-35-6; -esterase,

Supplementary Material Available: A listing of data fromIR and H N MR spectra of esters 1-30 and MS of esters 3,8,18,21, nd 24 (7 ages). Ordering information isgiven on any curren tmasthead page.L ITERATURE CITEDAbbott , W. S. J . Econ. Entomol. 1925,18, 265.Arbuzov, B. A.; Razumov a, N. A. Invest. Akad. Nauk SSSR 1945,Bakunick, E. Pol. Pa t . 91602;Chem. Abstr. 1979, 0.Casida, J.E. Biochem. Parmacol. 1961, , 332.Chambers, H.W., Departmen t of Entomology, M ississippi Sta teUniversity, Mississippi State, MS, unpublished data, 1973a.

167,5-11.

Chambers, H. W. Ann . Entomol. SOC. m. 1973b,66, 63.Cook, J. W.; Blake ,J. R.; Yip, G.; Williams, M. J.Assoc. Off. Agric.Ellman, G. L.; Courtney, K. D.; Andres, V., Jr.; Featherstone, R.Eto, M. Organophosphorus Pesticides: Organic and BiologicalEto, M.; Casida, J. E.; Eto, J. Biochem. Pharmacol. 1962,11,337.Eto, M.; Oshima, Y. Agric. Biol. Chem. 1962, 6, 452.Eto, M.; Oshima, Y.; K itakata , S.;Tanaka, F.;Kojima, K. BochuKagaku 1965, 31, 33.Ford-Mo ore, A. H.; Perry, B. J. Organic Syntheses; Wiley: NewYork, 1963;Collect. Vol. IV, p 956.Frawley, J. P.; uyat, H. N.; Hagan, E. C.; Blake, J. R.; Fitzbuch,A. G. J. Pharmacol. E x p . Ther. 1957, 21 , 96.Kobayashi, K.; Eto, M.; Oshima, Y.; Hirano, T.; Hosoi, T.;Wakamori, S. Bochu Kagaku 1969, 34, 165.Lowry,0. .; Rosebroug h, N. J.;Farr, A. L.; Randall, R. J. J .Biol.Chem. 1951,193, 65.Murphy, S. D.; Anderson, R. L.; Dubois, K. P. Proc. SOC. x p .Biol. Med . 1959,100, 483.Schrader, G. Die En twicklung Neuer Insektizider, Phosp honau re,Ester; Verlay Chemie: Weinheim, West Germ any, 1963; 281.Steube , C.; Lesuer, W. M.; Norma, G. R. J. m. Chem. SOC.955,77, 3526.Sun, Y. P.;ohnson, E. R.; Ward , L. F. J.Econ. Entomol. 1967,60, 28.Wilkinson, C. F. Adv. Environ. Sci. Technol. 1976, , 95.

Chem. 1958,41, 99.M. iochem. Pharmacol. 1961, ,88.Chemistry; CRC Press: Cleveland, OH , 1974; 134.

Received for review July 2, 1982. Accepted March 11, 1983.

Quantitation of Free and Hydrolyzable Phenolic Acids in Seeds by CapillaryGas-Liquid Chromatography

Kazimierz J. Dabrowski and Frank W. Sosulski*

Free phenolic acids were initially extracted with tetrahydrofuran, followed by extraction of soluble phenolicesters with methanol-acetone-water. Alkaline hydrolysis was employed to release aglycons from solublephenolic esters (2 N NaOH) and residue (4 N NaOH). The MeaSi derivatives of the released phenolicacids were quantitated by capillary gas-liquid chromatography within an elution time of 20 min. Greaterprecision was achieved by using methyl heptadecanoate instead of n-tetracosane as the internal standard.Although operations were conducted in the dark under nitrogen, recoveries of p-hydroxybenzoic, vanillic,and syringic acids exceeded 90%, but those of p-coumaric, ferulic, and sinapic acids were 87, 82, and78%, respectively, and that of caffeic acid was only 17%.

Phenolic compounds occur widely as microconstituentsin plant foods and there is increasing interest in theireffects on food quality. Among the various forms ofphenolic compounds, the free phenolic acids, esters, andglycosides that contain an acrylic acid group conjugatedwith the aromatic ring are of particular concern. Thesemonocyclic phenolics are readily oxidized, leading to theformation of quinones which further react to form poly-mers or bind to proteins and carbohydrates (Sabir et al.,1974; Van Sumere et al., 1975). The o-dihydroxyphenolssuch as caffeic acid are particularly reactive and, thus, are

DeDartment of CroD Science and Plant Ecolom. Univ-ersity of Saskatchewk, Saskatoon, Saskatchew&,-CanadaS7N OWO.

found in only low concentrations in seeds (Sosulski, 1979).However, esters of these phenolic acids, such as chlorogenicacid (3-caffeoylquinic acid), occur widely in plant tissuesand represent 3-4% of the defatted flour from sunflower.Sabir et al. (1974) demonstrated that, at neutral pH,one-third of the chlorogenic acid in slurries of sunflowerflour was covalently bonded to low molecular weightproteins and polypeptides. Under alkaline conditions,oxidation and bonding of chlorogenic acid with sunflowerproteins may be extensive, developing dark green colorsthat oxidize further to a dark brown on prolonged storage.Simple phenolic acids have been characterizedashavingsour, bitter, astringent, and phenol-like flavors (Arai et al.,1966; Maga and Lorenz, 1973). However, it is the cholineester of sinapic acid, sinapine, which occurs in concen-trations of 0.5-1.5% in defatted crambe (Austin and Wolff,

0021-856118411432-0123$01.50/0 0 1984 American Chemical Society

8/7/2019 jf00121a031

2/5

124 J. Agrlc. Food Chem., VoI. 32,No. 1, 1984 Dabrowski an d Sosulski

F igure 1. Procedure for the extraction and separation of free,esterified, and residue phenolic compoundsand their hydrolysist o phenolic acids for analysis by capillary GLC (DE/EA/THF= diethyl ether-ethyl acetate-tetrahydrofuran, 1:l:l).1968) and rapeseed (Kozlowska et al., 1975). Glycosidesof sinapic acid (3,5-dimethoxy-4hydroxycinnamiccid) arealso present in Brassica meals (Fenton et al., 1980). Theseesters of sinapic acid are responsible for the strong bit-terness and astringency of rapeseed flours and proteinconcentrates. The relative order of bitterness of equimolarconcentrations of the ester and ita components were si-napine > sinapic acid > choline (Ismail et al., 1981).Simple phenolic compounds usually do not bind as ex-tensively with proteins as the polymeric tannins, and thereare few reports of antinutrit ional effects on protein uti-lization. There are literature reports of phenolic acidsbinding with enzymes (Loomis and Battaile, 1966).Pierpoint (1969) has elucidated the mechanism of chlo-rogenic and caffeic acid binding to the -amino group oflysine or thiol group of cysteine. In leaf protein concen-trates, the reductions in available lysine during extractionwere attributed to reactions with reducing sugars or qui-nones (Allison et al., 1973).

Phenolic acid esters can be quantitated directly byHPLC (Ong and Nagel, 1978). Alternatively, Krygier etal. (1982a) developed a procedure for fractionating the totalphenolic constituents of flours into free, soluble ester, andhydrolyzable-residue forms and, after hydrolysis, deter-mining the relative proportions of the released phenolicacids by GLC. Rapeseed flours contained 6-98 mg/ 100g free phenolic acids, 768-1196 mg/100 g phenolic acidsfrom hydrolyzed esters, and no phenolic acids from theresidues (Krygier et al., 1982b). The total phenolic acidcontents of wheat, rice, and oat flours ranged from 71 to87 ppm while corn flour contained 309 ppm and potatoflour 410 ppm (Sosulski et al., 1982). The principal phe-nolic compounds were esterified sinapic acid in rapeseed,bound ferulic acid in the cereal flours, and free chlorogenicacid in potato flour.The objectives of the present investigation were to im-prove the procedures for quantitating the free phenolicacids in seed materials, avoiding hydrolysis of esterifiedacids during the extraction process. The effectiveness ofalternate solvent and purification systems was evaluatedby thin-layer chromatography (TLC), gel filtration (GF),capillary gas-liquid chromatography (GLC), and massspectroscopy (MS). Recoveries of the common phenolicacids under the proposed extraction and hydrolysis pro-cedures were determined. Reference phenolic acids and,because of their wide distribution of phenolic acids andesters, rapeseed flours were utilized in the various deter-minations. The modified procedure was then applied tothe analysis of defatted flours from 10 oilseed species

0 10TIME (MIN)

20

Figure 2. GLC chromatogram of Mes% derivatives of phenolicacids released on hydrolysis of the soluble ester fraction in ra-peseed flour using a WCOT capillary column of fused silica (0.2mm i.d. X 25 m) coated with OV-1. 1 = p-hydroxybenzoic,2 =vanillic, 3 = protocatechuic, 4 = syringic, 5 = trans-p-coumaric,6 = cis-sinapic, 7 = trans-ferulic,8 = trans-caffeic, and 9 =trans-sinapic acids. A = methyl heptadecanoate as the internalstandard.(Dabrowski and Sosulski, 1984) and the flours and hullsof 10 legume species (Sosulski and Dabrowski, 1984).EXPERIMENTAL SECTION

Sample Preparation. Reference phenolic acids fromICN Pharmaceuticals, Inc., and Sigma Chemical Co.(salicylic, trans-cinnamic, p-hydroxybenzoic, gentisic,protocatechuic, vanillic, syringic, trans-o-coumaric,trans-m-coumaric, trans-p-coumaric, trans-isoferulic,trans-ferulic, trans-caffeic, and trans-sinapic acids) weredissolved in tetrahydrofuran in known concentrations.Prior to silylation, aliquots were evaporated under nitrogenin vials to which the internal standard, n-tetracosane ormethyl heptadecanoate, was added. Phenolic acids weresilylated by slight warming with N,O-bis(trimethylsily1)-acetamide Formula D (Pierce Chemical Co.).Rapeseed flours used in the study were from the lowglucosinolate cultivars of Brassica napus (Tower) andBrassica campestris (Candle). The seeds were flaked anddehulled by hand before oil extraction with hexane. Si-napine was isolated as he bisulfate salt from a portion ofthe Candle meal by the method of Clandinin (1961).Extraction Conditions. To avoid light-mediatedisomerization and oxidation of phenolic compounds theextraction, hydrolysis, and concentration steps were con-ducted in the dark under nitrogen atmosphere.Free phenolic acids were extracted from the rapeseedflours with tetrahydrofuran, tetrahydrofuran-acetone (l:l),and acetone-methanol (1:l)aswell as the methanol-ace-tonewater (7:7:6) solvent for free phenolic acids and esters.In each case, 1g of flour was extracted 6 times with 20 mLof solvent by using a Polytron mixer (Brinkmann Instru-ments, Inc.). The combined extracts were centrifuged, andthe supernatant was evaporated to dryness under vacuumat 30 "C and stored for GLC analysis. In parallel exper-iments, the supernatants were concentrated to 10 mLunder nitrogen a t room temperature for fractionation byTLC and GF.Procedure for Flours. Initially, the scheme of Krygieret al. (1982a) for extraction and separation of free, ester-ified, and residue phenolic acids was followed. The finalprocedure adopted involved the separate extraction of free

8/7/2019 jf00121a031

3/5

Quantitationof Phenolic Acids J. Agric. Food Chem.,Vol. 32,No. 1, 1984Table I. Precision of Gas Chromatographic Determination of Me, Si Derivatives of Reference Phenolic Acids Using TwoInternal Standards (Means of Eight Determinations)

125

~_ ~ ~ ~ ~ ~ ~ ~~ n-tetracosane methyl heptadecanoatequantity quantity SD, quantity SD,applied, recovered, recovery, SD, % of recovered, recovery, SD, % ofu g meanhenolic acid RRTa f i g ug % f i g mean !Jug %- - -

salicylic 0.33 113.9 113.0 99.2 11.4 10.0 110.3 96.8 9.2 8.3trans-cinnamicb 0.35 112.0 114.8 102.5 12.4 10.8 111.9 99.9 5.1 4.5p-hydroxy benzoicvanillicgentisictrans-o-coumaricprotocatechuictrans-m-coumaricsyringictrans-p-coumarictrans-isoferulictrans-ferulictrans-caffeictrans-sinapic

a Relative retention time.

0.47 134.00.66 188.00.70 186.00.71 129.10.75 129.10.80 170.90.84 193.70.89 193.71.07 208.91.09 199.41.16 237.31.28 364.6

Not a phenolic

136.3190.6184.9132.0129.2174.2197.5198.2213.5201.5238.7366.5

acid.

101.7101.399.4102.2100.0101.9101.9102.3102.2101.0100.6100.5

phenolic acids with tetrahydrofuran (Figure 1). Thissolvent was also included with diethyl ether-ethyl acetate(1:l:l)o transfer the hydrolyzed phenolic acids from theaqueous phase in the analysis of phenolic acid esters andresidue phenolics.One gram of defatted flour was extracted 6 times for 5min each with 20 mL of tetrahydrofuran in the Polytronmixer, and the combined supernatants were tested for freephenolic acids as above. The residue was then extracted6 times with 20 mL of methanol-acetone-water (7:7:6) atroom temperature. After centrifugation, the combinedsupernatants were evaporated under vacuum at 40 "C toobtain a water suspension containing the soluble phenolicesters and glycosides. The aqueous suspension was com-bined with an equal volume of 4 N NaOH for hydrolysisat 2 N NaOH for 4 h under nitrogen at room temperature.The hydrolysate was adjusted to pH 2 with 6 N HCl andextracted in a separatory funnel 5 times with hexane ata hexane to water phase ratio of 1:l to remove free fattyacids and other lipid contaminants. The released phenolicacids were then extracted 6 times with diethyl ether-ethylacetatetetrahydrofuran (1:l : l) at a solvent to water phaseratio of 1:l. The ether-ethyl acetate-tetrahydrofuranextracts were dried with anhydrous sodium sulfate, filtered,and evaporated to dryness under vacuum at 30 "C. Methylheptadecanoate (250 pg),dissolved in tetrahydrofuran, wasadded to the dry residue and the solvent was evaporatedunder nitrogen at room temperature.The moist residue from methanol-acetone-water ex-traction was hydrolyzed directly with 10 mL of 4 N NaOHunder the same conditions as above. Further purificationsteps for the released acids were the same as for the abovehydrolysates.Gas-Liquid Chromatography. The Me3Si derivativesof the phenolic acids were separated on a Hewlett-PackardModel 5880A gas chromatograph equipped with flameionization detector and connected to a Spectra-PhysicsSP4100 computing integrator. A WCOT capillary column offused silica (0.2 mm i.d. X 25 m), coated with OV-1, wasused. Initial temperature of 112 "C was held for 3 min,after which the temperature was programmed to 260 "Cat 4. 5 "C/min. The injector temperature was 25 0 "C anddetector temperature 300 "C. Phenolic acids were iden-tified on the basis of relative retention times of Me3Siderivatives in the sample and Me3Si derivatives of refer-ence acids. The results were confirmed by GLC-MStechniques as described by Krygier et al. (1982a) .

12.5 9.213.2 6.911.3 6.18.5 6.46.9 5.39.2 5.39.8 5.08.9 4.57.6 3.66.7 3.34.4 1.89.7 2.6

133.0190.9181.6127.9129.0172.3197.0196.4208.7203.4236.4363.1

99.2101.597.699.199.9100.8101.7101.499.9102.099.699.0

6.1 4.65.1 2.74.6 2.52.5 1.92.5 1.93.6 2.12.0 1.04.4 2.22.3 1.13.2 1.63.9 1.69.3 2.6

Thin-LayerChromatography. The TLC plates werecoated with silica gel IBZF containing fluorescent indicatorUV-254 (J.T. Baker Chemical Co.). A solvent system ofbutanol-acetic acid-water (407:32) was applied for sepa-ration of rapeseed extracts and reference compounds.Phenolic compounds were visualized on the TLC platesas blue fluorescent spots under long-wave (366-nm) U Vlight.Gel Filtration. Gel chromatography of tetrahydrofuranextracts from rapeseed and reference sinapic acid wasconducted according to Lattanzio and Marchesini (1981)using Sephadex LH-20 resin (Pharmacia Fine ChemicalsAB, Sweden). The column (bed volume 19X 2.5 cm) wasequilibrated by using an aqueous ethanolic (50%)solutionacidified to pH 2 with concentrated H3P04. A 5-mL ali-quot of rapeseed extract or 1 mL (2.0 mg) of referencesinapic acid was applied to the column. The elution so-lution was the same as above with a flow rate of 39 mL/h.The fractions (3 mL) were collected and the absorbanceread at 332 nm.RESULTS AN D DISCUSSION

Capillary GLC. The WCOT capillary column of fusedsilica coated with OV-1 used in this study gave an excellentseparation of Me3Si derivatives of phenolic acids (Figure2) . Total chromatographic analysis of phenolic acids,including sinapic acid, could be performed in 20 min.Reference Standard. Poor repeatability of referencephenolic acid determinations was observed when n-tetra-cosane was used as an internal standard. Among threeother internal standards tested, methyl heptadecanoateproved to be most suitable for phenolic acid analysis.Table I shows that higher precision was obtained withmethyl heptadecanoate than with n-tetracosane. Also, therelative standard deviations of GLC determinations ofreference phenolic acids were decreased by approximatelyhalf when methyl heptadecanoate was used instead ofn-tetracosane. This could be due to the better solubilityof the former in the silylation reagent.Extraction Conditions. The methanol-acetone-waterused to extract the soluble phenolic acids and esters byKrygier et al. (1982a) was effective in the solubilizationof 1 6 standards. Only trans-sinapic acid required slightwarming and shaking for the solubilization of 10 mg of acidin 0.20 mLof solvent. However, the subsequent separationof the free acids from the esters gave variable results. The

8/7/2019 jf00121a031

4/5

126 J. Agric. Food Chem.,Vol. 32,No. 1, 1984Table 11.Flour Determined Using Fo ur Solvent Systems forInitial Extraction

Apparent Free Sinapic Acid Conten t in CandleDabrowskiand Sosulski

alternatesolvent systemacetone-methanol-waterte t rahydrofuran-waterte t rahydrofuran

wa te r (7 :7 :6 )

% ofmg/10 0 the t o t a lg of flour in floura75.3 6 .580 .5 6 .974.2 6 .4trace 0.0

a Based o n th e tota l s inapic ac id in f lour (Ta ble 111).difficulty was overcome by using tetrahydrofuran as thesolvent for the initial extraction of the free acids (Figure1).Most of the phenolic acids @-hydroxybenzoic, alicylic,trans-cinnamic, vanillic, gentisic, o-coumaric, proto-catechuic, m-coumaric, p-coumaric) were immediatelysoluble in tetrahydrofuran; the remainder (syringic,trans-isoferulic, tram-ferulic, trans-caffeic, trans-sinapic)dissolved readily with shaking. On the other hand, chlo-rogenic acid and sinapine were completely insoluble in thissolvent. As a further test of the solvent, tetrahydrofuranused in six 1-min extractions with the Polytron was ef-fective in extracting 95.3% of sinapic acid added to ra-peseed flour, but no sinapine or added chlorogenic acid wasrecovered. The solubility of free phenolic acids was notimproved by combining tetrahydrofuran with acetone.The rapeseed flour samples were extracted with freshtetrahydrofuran 6 times for 5 min each in the Polytronmixer or in a single 24-h extraction using a magnetic stirrerand beaker. Essentially no free phenolic acids were de-tected in the rapeseed flours, except for occasional tracesof sinapic acid. These results were confirmed by TLC andGF of the extracts and GLC-MS of the peaks in the GLCchromatograms. The low concentrations of free phenolicacids found in the investigations of Krygier et al. (1982b)and others (Fenton et al., 1980; Kozlowska et al., 1975) mayhave been artifacts of the extraction and purificationprocedure.Stepwise analysis of t he fractionation scheme revealedthat the initial steps, including extraction, concentration,and acidification of the aqueous extract to pH 2 (Figurel) , had caused partial hydrolysis of esters from which si-napic acid was liberated. This phenomena occurred inseveral solvent systems in which water was a component(Table 11). In these conditions 74-80 mg of sinapicacid/100 g of flour (constituting about 7% of total sinapicacid present in flour in form of esters) was released.Similarly, chlorogenic acid was solubilized in methanol-acetone-water, and after evaporation of the organic sol-vents and adjustment to pH 2, quinic and caffeic acidswere found in the GLC chromatograms (Figure 3) .

' ICG

0 10 20 3aTIME IN MINUTES

Figure 3. GLC chromatogram of MeaS i derivatives of compoundsextracted in the free phenolic acid fraction after subjectingchlorogenic acid to acidic (pH 2) conditions of the Krygier e t al.(1982a) procedure. QU = quinic acid, IS = internal standard,CA = caffeic acid, and CG = chlorogenic acid.Isomerization and oxidation of chlorogenic acids (Son-dheimer, 1964) would account for some of the small peaksthat eluted near the chlorogenic acid peak in Figure 3.As a result of these observations, the Krygier et al.(1982a) procedure has been modified to include an initialextraction of the defatted flour or other plant material withtetrahydrofuran to detect and quantitate the free phenolicacids only (Figure 1).The residue would then be extractedfor soluble esters and glycosides with methanol-acetone-water, with the organic solvents being removed by evap-oration in order to conduct the hydrolysis under aqueousconditions. It was also found that the transfer of thehydrolyzed phenolic acids from the aqueous phase wasimproved by the addition of tetrahydrofuran to the diethylether-ethyl acetate extracting solution. The recoveries ofsome minor phenolic acids were increased severalfold,permitting quantitation of most peaks (Figure 2; Table III).Also, the isomerization of trans-phenolic acids to the cisform was reduced substantially.Recovery of Phenolic Acids. As mentioned previ-ously, no free phenolic acids were found in the Candle orTower flours nor were any acids recovered from the hy-drolysates of the residues. However, nine phenolic acidswere released on hydrolysis of the soluble ester fractionin both rapeseed flours (Figure 2; Table 111). Sinapic acidcontaining compounds predominated in the ester fraction;the small quanti ty of cis-sinapic likely arose via isomeri-zation of the natural trans form. trans-Ferulic and p -hydroxybenzoic acids appeared in much greater concen-trations than obtained by Krygier et al. (1982b).Previously, Krygier et al. (1982a) reported that the al-kaline hydrolysis conditions (2 N NaOH) employed in theprocedure would destroy about two-thirds of the caffeic

Table 111. Phenolic Acids Released o n Hydrolysis of th e Soluble Ester Frac t ion f rom Candle Flour and Recoveries ofAdmixtures of Authent ic Phenol ic Acids wi th Candle Flour af te r Subjec t ion to Alkal ine Hydrolysis Condi t ionstotal recovery ofpresent in flour, admixed , recovered, added phenolicphenolic acid pg/g of flour pg/g of flour pg/g of flour acids,a %___-p - h ydrox ybenzoic 56 i 5a 184 223 91 .0vanillic 7 i 2 1 6 1 1 5 6 9 1 .4protocatechuic t racesyringic 1 1 5 2 1 6 4 1 6 0 9 0 .4trans-p-coumar c 5 t 2 23 5 209 86.7trans-ferulic 150t 26 280 380 82 .3cis-sinap c 330 i 46 304trans-sinapic 111 0 7 t 1 0 3 3 11000 19 716 78 .3t rans -caf fe ic 5 i 2 281 52 16.7

a Based o n four de terminat ions.

8/7/2019 jf00121a031

5/5

Quantitation of Phenolic Acids J. Agric. Food Chem., Vol. 32,No. 1, 1984 127acid and one-third of the sinapic acids in rapeseed flour.Appropriate quantities of the authentic phenolic acidspresent in rapeseed were added to the flour (Table 111).The tetrahydrofuran extraction step was bypassed in orderto subject the phenolic compounds to the alkaline hy-drolysis conditions. Recoveries of p-hydroxybenzoic,vanillic, and syringic acids exceeded 90%. Increasing losseswere evident for p-coumaric, trans-ferulic, and trans-si-napic acids, and correction factors would be justified inreporting data on their composition. Nearly all of theadded caffeic acid was lost under the conditions of thehydrolysis, and no estimate of the composition of caffeicacid esters could be made by the present procedure, in-cluding that of chlorogenic acid. HPLC methods forquantitation of these phenolic acid esters would be rec-ommended (Ong and Nagel, 1978).The low recovery of caffeic acid (3,4-dihydroxycinnamicacid) can be attributed to the instability of the o-di-hydroxycinnamyl moiety in the presence of an alkalinereagent (Harborne, 1964; Sutherland, 1958). The rapidoxidation rate of caffeic acid reflected its low oxidationpotential of +0.35 V in 1 M acetate at pH 4.7 (Felice etal., 1976). The oxidation potentials of other phenolic acidswere as follows: for p-hydroxybenzoic, +0.99; for vanillic,+0.72; for syringic,+0.55; for p-coumaric, +0.73; for ferulic,+0.57. The oxidation of syringic acid (Table 111)wa s lessthan predicted from the oxidation potential because twoof the three hydroxyl groups were methylated as in the caseof sinapic acid.Salomonsson et al. (1978) have recommended the useof enzymes rather than alkali for hydrolysis of alkali-labilephenolic esters. The addition of reducing agents such assodium tetraborohydrate prior to the alkaline treatmentresulted in high recoveries of compounds containing the3,4-dihydroxycinnamyl moiety (Schroeder, 1967). Herr-mann and co-workers (Schmidtlein and Herrmann, 1975;Schulz and Herrmann, 1980a,b) have used this reducingagent to suppress the formation of o-quinones in TLC andGLC analysis of caffeic and sinapic acids; their recoverieswere in the range of 87-88%. The present authors wereunable to improve the recoveries of caffeic or sinapic acidsthrough the use of sodium tetraborohydrate but this ap-proach is still being investigated.

ACKNOWLEDGMENTand L. Hogge.Technical assistance was provided by H. Braitenbach

Registry No. Salicylic acid, 69-72-7; trans-cinnamic acid,140-10-3;p-hydroxybenzoic acid, 99-96-7;vanillic acid, 121-34-6;

gentisic acid, 490-79-9; trans-o-coum aric acid, 614-60-8;proto-catechuic acid,99-50-3;rans-m-coumaric acid, 1475502-3; syringicacid, 530-57-4; rans-p-coumaric acid, 501-98-4; trans-isoferulicacid, 25522-33-2; rans-ferulic acid, 537-98-4;trans-caffeic acid,501-16-6; rans-sinapic acid, 7362-37-0;cis-sinapic acid, 7361-90-2.LITERATURE CITEDAllison, R. M.; Laird , W. M.; Synge, R. L. M. Br. J. Nut r . 1973,Arai, S.;Suzuki, H.; Fujimaki, M .; Saku rai, Y. Agric. Biol. Chem.Austin, F. L.; Wolff, I. A. J. Agric. Food Chem. 1968, 16, 132.Clandinin, D. R. Poult. Sci. 1961, 40, 484.Dabrowski, K. J.; Sosulski, F. W. J . Agric. Food Chem. 1984,Felice, L. J.; King, W. P.; Kissinger, P. T. J. Agric. Food Chem.Fenton, T. W.; Leung, J.; Clandinin, D. R. J . Food Sci. 1980,45,Harborne, J. B. In Methods in Polyphen oh Chem istry; Pridham,Ismail, F.; Vaisey-G enser, M.; Fyfe, B. J.Food Sci. 1981,473,1241.Kozlowska, H.; Sa bir, M. A.; Sosulski, F.; Coxworth, E. Can. Znst.Krygier, K.; Sosulski, F.; Hogge, L. J . Agric. Food Chem. 1982a,Krygier, K.; Sosulski, F.; Hogge, L. J. Agric. Food Chem. 198213,Lattanzio, V.; Marchesini, A. J. Food. Sci. 1981, 46 , 1907.Loomis, W. D.; Battaile, J. Phytochemistry 1966, 5, 423.Maga, J. A.; Lorenz, K. Cereal Sci. Today 1973 ,18 ,326 .Ong, B. Y.; Nagel, C. W. J. Chromatogr. 1978, 157, 345.Pierpoint, W. S. Biochem. J. 1969, 112,609.Sabir, M. A.; Sosulski,F.W.; Finlayson, A. J.J.Agric. Food. Chem.Salomon ason, A. C.; Th ean der , 0.;Aman, P. J. gric. Food Chem.Schmidtlein, H.; H errmann, K. J. Chromatogr. 1975, 115, 123.Schroeder, H. A. Phytochemistry 1967, 6, 1589.Schulz, J. M.; Herrmann, K. J. Chromatogr. 1980a, 195,85.Schulz, J. M.; Herrmann, K. J . Chromatogr. 1980b, 195, 95.Sondheimer, E. Bot. Reu. 1964, 30, 667.Sosulski, F. J.Am. Oil Chem. SOC. 979, 56, 711.Sosulski, . W.; Dabrow ski, K. J. J.Agric. Food Chem. 1984, thirdSosulski, F.; Krygier, K.; Hogge, L. J. Agric. Food Chem. 1982,Suthe rland, G. K. Arch. Biochem. Biophys. 1958, 75, 412.Van Sum ere, C. F.; Albrecht, J.;Dedonder, A.; de Pooter, H.; P6,I. In T he Chemistry and Biochemistry of Plan t Proteins;Harborne, J. B.; Van Sumere, C. F., Eds.; Academic Press:London, 1975; p 211.

29, 51.1966, 30 , 364.

second paper of three in this issue.1976, 24, 380.1702.J. B., Ed.; Macmillan: New York, 1964; p 25.

Food Sci. Technol. J . 1975, 8, 160.30, 330.30, 334.

1974, 22, 575.1978, 26, 830.

paper of three in this issue.30, 337.

Received for review Jan uar y 24, 1983. Revised manus cript re-ceived May 12, 1983. Accepted August 26, 1983. Th is investi-gation was supp orted by th e Canola Council of Canada an d theNatural Sciences and Engineering Research Council of Canada.