Determination of Chlorogenic Acid in Coffee R. G. SIOORES, DOROTHY L. &'lcDEK\lOTT, I ~ T. D K. WOOD1 Central Laboratories, General Foods Corporation. Hoboken, N.J .
A new- method for the rapid and precise deterriiination of chlorogenic acid in green and roasted coffee is based upon the absorption of chlorogenic acid a t 324 millimicrons. The chlorogenic acid is extracted from faked coffee samples with water, and for green coffee the iiltra,iolet absorption is nieasiired directly on the extract. For roasted coffee, the absorption is nieasured on the extract before and after precipitation of the chlorogenic arid with basic lead acetate. The Slotta and Neisser iodometric oxidation method For rhlnropenic acid has been investigated and improted.
C
HLOROGENIC acid, which occurs in green coffee to the extent of 5 to S%, is one of the major water-soluble constituents of the bean. I t s importance in coffee chemistry has long been recognized, and considerable attention has been devoted to it in the literature.
325 milliniicroiia o i a misture of phenolic compounds, including chlorogenic acid, which they found in sweet potatoes. PURIFIC.ATIOS lIVD PROPERTIES OF CHLOROGEYIC ACID
Three saiiiples of chlorogenic acid were isolated from green coffee by two different met,hods and purified by repeated crystallization to serve as analytical standards. Sample 1 and a further recrystallized fraction, KO.2, were isolated through the potassium caffeine romplex, while sample 3 was separated by solvent pnrtition. .Ischlorogenic acid is not available commercially, an est,raction and purification procedure is included in this paper. The isolation of chlorogenic acid from coffee through the potassium caffeine complex is a modification, by Kremers ( I O ) of this laboratory. of previously reported methods (6, 7 ) .
The compound was first isolated by Gorter ( 7 ) , who found It widely distributed in the leaves and seeds of numerous plants. Freudenberg (6) and Fischer (6) described the structure of the compound as the quinic acid ester of caffeic acid. Charaux ( 4 ) was one of the first workers to attempt the measurement of chlorogenic acid in plant materials. His method nas based upon the separation of the acid by lead precipitation, regeneration of the free acid, saponification, and extraction of the resulting caffeic acid with ether. The caffeic acid mas then weighed directly. This method can have only preparative value because of losses during saponification and ether extraction. Hoepfner (8) in 1932 developed a method based upon the red color formed when chlorogenic acid is treated with nitrous acid. Plucker and Keilholz (12) questioned the reliabilitl- of this method for roasted coffee, because catechol and protocatechulc acid, likely t o be present in the roasted product, give red colols under the conditions of the reaction. These authors devised an indirect method based upon the hydrolysis, of the chlorogenic acid, extraction of the resulting caffeic acid with ether, and a final estimation bv the nitrite color reaction. Efforts to carry out these steps quantitatively in this laboratory were unsuccessful. Jurany (9) proposed a method based upon the separation of chlorogenic acid as its lead salt, regeneration of the free acid, and a final estimation of the chlorogenic acid by optical rotation or titration with alkali. These methods cannot be considered valjd in vierv of the presence in coffee of isomers of chlorogen!~ acid which have differentrotatory powers and different neutrahzation equivalents, as shown bv recent work in this laboratory ( 3 ) . Slotta and Xeisser (16) recognized the unreliability of the results obtained by earlier methods and developed a new technique for measuring chlorogenic acid. This method consists of a Soxhlet water extraction of a defatted sample, lead precipitation, regeneration with hydrogen sulfide, and estimation of the liberated chlorogenic acid by its reaction with alkaline h? poioditc.
Esrrttct 1 iig. of flaked green coffee by stirring viith 6 liters 01' 70% isopropanol for about 3 hours, filter on a Buchner funnel, and estract the residue again for a short time with 6 liters of 70% xsopropanol. Concentrate the total extract by vacuum distillation to 2.4 liters and cool to about 5O.C. for several hours. Remove thP precipitated fats by filtering p i t h a filter aid. Concentrate tht. filtrate under vacuum t o about 300 ml. Dissolve in this concentrate, by heating if necessary, 8 grams of potassium acetate anti 20 grams of caffeine. Add 300 ml. of 95% ethanol and keep at about 5' C. for 2 days for crystallization of the crude complex. Collect the crude complex by filt,ration on a Buchner funnel and wash with a little 50% ethanol. Dissolve the complex in thc minimum volume of hot water and add sufficient 95% ethanol with stirring to initiate precipitation. Keep the solution at 5' C. overnight, collect the reprecipitated complex by filtration, atid dry a t room temperature. About 70 grams of comples are 011tained from 1kg. of coffee. -1dd 100 grams of the complex in small portions to 225 ml. of hot ITater containing 24 grams of tartaric acid, cool for several minutes, and remove the potassium bitartrate bv filtration. Extract the filtrate Trith chloroform in a liquid-liquid extractor for about 30 hours, or until the caffeine is completely removed. Keep the aqueous phase at, 5 " C. for 1 day for crystallization of the chlorogenic acid. Collect the acid by filtration, wash with cold ivater, and dry at 70 O C. in a vacuum oven. The chlorogenir acid as ohtained melts at' 203-206" C.
Early experience with the Slotta and Seisser method revealed certain weaknesses and indicated the need for modifications. Therefore, the initial efforts in the current investigation were directed toward improvement of the method. The manipulative difficulties inherent in the Slotta and Seisser method made it desirable to devise a new analytical procedure for chlorogenic acid. It was found that chlorogenic acid has intense light absorption a t 324 millimicrons. Additional work was undertaken in order t o develop an analytical method for chlorogenic acid based upon ultraviolet absorption measurements. This paper reviews briefly the recommended modifications of the Slotta and Neisser iodometric method and presents a new ultraviolet abaorption method for chlorogenic acid. Subsequent to the development of this ultraviolet method, Rudkin and Selson (IS) reported a mauimum absorption at about 1 Present
Recrystallization from water two or three times gives a pure product, as indicated by Table I. The samples as used were dried for 16 hours in a vacuum oven a t 70" C. and were found to cont,ain 0.5 molecule of water of crystallization, which was measured by drying to constant xeight in vacuum over phosphorus pentoside at 135' to 140" C. -4 weight loss of 2.4 to 2.7% was observed j tthe theoretical moisture content for the hemihydrate is 2 . 4 8 5 , The presence of 0.5 molecule of water of crystallizatioii in the oven-dried samples is further verified by titration data. trhich R ~ O W a neutralization cquivalcnt of 360 to 367; the calculatcd valur for the hemihydrate is 363. A titration curve for snmplt. 1 is shown in Figurc 1. For uaiformity of comparativr data. d l t t i c chlorogenic acid values on coffee samples reported herein a w espressed in terms of the anhydrous compound, C:,€I:& riiq>lecular weight 354.
address, Research Laboratory, Merck and Company, Ralinay,
N.J.
620
'
V O L U M E 2 0 , NO. 7, J U L Y 1 9 4 8
62 1
IODOWETRIC OXIDATION >lI.:THOI) FOR CH LORO(:b::\IC
ACID
Reagents. Petroleum ether, c . P . , boiling point 20-40" C. 1.ead acetate, c . P . , neutral, saturiittd water solution. Iociine solution, 0.1 S,prcpit,ccl :is directed by -i.O.-\.C. ( d > 43.21 J. Soclium thiosulfate solutioii. 0.1 .\., pi,epai.td as directed by .\.O..LC. (2, 43.28). Starch indicator, 1C;. pi'r~patwl:I> tlirec.terl hy h.O..\.C. (e,
Tahle 1.
Saiii[,le
'?.
Melting SeurraliPoint Oi,tical zation Corrected, Rotatioil, Eilriirn-
SO,
'C.
[e];:
lent
1
207-Y 206-7 207-8
-39 8 -37.7 -38.0
332 337 354
2 3
ii 2) . .-,.
Sulfuric acid, 2 S . Sodium hydroside; 1 .\, 1'iltt.i. paper, Grwn's fliitt,il So. 488'
Properties of Pure Chlorogenic Acid" Iodine EquivaIrnt,
Atoms per 11o1e 10.55 10.55
10,55
Wave Length of Ahximum Absorp13; tion, nip E1 cnl. 324 324 324
532 522 525
18.5 ( - i n . : TT7hatman's SO.I , 5.5-~11i. alld 12.5-~111. Celite filter aid, Johns-llanvillt:, A-(I. 545. -liabcstos, acid-washed nirdium fiber. Procedure. Prepare the sample by grinding thruugh a 1inin. ~ C N T I I on a laboratory Kiley mill and flaking on a 3-roll Lelimann chocolate mill. Determine moisture h!. cliying a 2gram saiiiplt: in a 70" c'. vacuum oven for 16 h0ui.s. DeCat a 4-gram sample of the flaked green coffee by n.ashing v i t h three 25-nil. portions of petroleum ether, allow the residue to italic1 about 5 minutes or until most of the solverit has evapoi,atkd iiiid transicr to a 2-liter volumetric flask. For roasted coffee w i g h 1 grains of the flaked sample directly into the volumetric flask, add 1600 ml. of water, and extract 20 minutes with occasiorial shaking. Dilute t o volume, mix, and filter by suction through N o . 1 Whatman paper on a 12.5-cni. Biichner funnel. . of filtrate, transfer exactl5- 1 liter to a 2id roncetitrate under v a c u u ~ i iat 20" to
Ti~ari~fer t o a 250-nil. Iir;tlwr, keeping the total volui~ieat about 100 ml. .Idd slowly with stirring 2 ml. of neutral lead acetate solution, and allow to stand overnight. .Idd 1 gram of Celite, stir thoroughly, and filter through a thin asbestos bed in a 5.5-cm. Biichner funnel. K a s h three times with 15 ml.of water, or until the volume of filtrate and washings is about 130 ml. Do not allow the bed to dry, as this makes deleading difficult. Transfer the lead precipitate and bed to the same 250-ml. beaker used for the precipitation, keeping the volunie at about SO mi. While stirring continuously with a mechanically driven glass stirrer, add slo~r-ly2 ml. of 2 S sulfuric acid, and continue stirring for 20 minutes or until the chlorogei?ate has decomposed. Filter through a S o . l Khatman paper in a 5.L-cm. Biichner funnel, and wash with four 20-ml. portions of water. Seutralize the filtrate a,nd n-ashings n-ith 2 S sodium hydimide to a pEi of 6 to 7 , t,ransfcr to a 200-nil. volunietric flask, and make to volume. Transfer 30 nil. to a glass-stoppered Erlenmeyer flask and add 25 nil. of 0.1 .\- iodine solution. Inimediatel>., with constant stirring, add 25 ml. of 0.1 S sorliutii hydroside dropwise from a buret over a 3-minute period. Stopper the flask, and allow t o stand in a dark place for 1 hour. During this period the tempemture should he 25' * 2' C. h t the end of the reaction period add 10 nil. of 2 S sulfuric acid, and after .iiiiinutes titrate with standard 0.1 .\. diuin thiosiilfatc solution, using st a i ~ liiir i lica tor. Ituri a blank titration on 25 nil. o t the iodine solution, using 50 ml. of watt'r in place of the ,5O-ml. aliquot of sainple. Calculate the per cent of anhydrous chlorogeriic acid present in the coffee (dry basisj by the follonirig equation:
ML. O . I O O N N I O H
Figure 1.
Titration Curie for Chlorogenic Acid IIetnihylrate
zH
"; chloi.ogr.nic acid = (volume of Sa&O, blank volume of SalSrO:! sample) X S Sa2S& X 3.36 t h y xeight of sample titrated
W
280
240
320 WAVELENGTH
IN
360
2.
1
MILLIMICRONS
Figure 2. Ultraviolet ihsorption of Green Coffee Extract 1.
4
The factor 3.36 = 100 X milliequivalent weight of chlorogenic acid iodine equivalent
Original water extrnct Filtrate from lead preoipitate of water extract, deleaded
Conditions Affecting the Reaction between Chlorogenic Acid and Iodine. The Slotta and Seisser iodometric
622
A N A L Y T I C A L C H E MIS TRY
Table 11.
Effect of Chlorogenic Acid Level upon Iodine Titration"
No. of Titrationb
Excess Iodine
Chlorogenic Acid Taken
7c
Mg.
90 106 145 168 282 425 810
47.8 41.0 34.2 27.3 20.5 13.7
Chlorogenic Acid Found Q
.
41.1 40.5 34.1 28.2 21.6 13 9 9.2
70 88.0 98.7 99.7 103.4 1051 116.1 133.3
6.8 Values calculated using ratio of 10.55 atoriis of iodine per niole of chlorogenic acid. a
Table 111. Effect of pH upon Reaction of Chlorogenic .4cid and Iodine p H before Adding Iodine
pH during Iodine Reaction
3.3 7.2 8.1 9.7 11.6
10.9 11.0 11.0 11.0 11.1
Table IV.
Chlorogenic Acid Found (34.1 Mg. Added) Jlg.
rc
34.1 34.3 33.8 31.6 24.0
100.0 100.6 99.1 92.7 70.4
precipitation and recovery of chlorogenic acid are: ( a ) concentration of chlorogenic acid and of excess lead, ( b ) pH of the solution, (c) time and temperature of precipitation, ( d ) method of collecting the precipitate, and ( e ) method of regenerating the lead salt. Early work indicated that several hours were required for complete precipitation of chlorogenic acid by lead under the conditions of the Slotta and Keisser method. In the later work on the ultraviolet absorption method, it was found that by suitable manipulations essentially complete lead precipitation from dilute solutions could be made in about 1 hour. This was achieved by the use of basic lead acetate in the presence of potassium acetate, and by heating the lead-treated solution briefly, then cooling to about 0" C., with constant stirring, for the remaining period. These conditions were not applied to the iodometric method. Regeneration of the chlorogenic acid from the lead salt has been one of the most difficult steps in the iodometric procedure, The substitution of sulfuric acid for hydrogen sulfide as the reagent for decomposing the lead chlorogenate improves both the precision and convenience of this step. The sulfuric acid should be added in a limited excess (25 to loo%), inasmuch as a large excess results in an apparent loss of chlorogenic acid. High temperatures during deleading in acid solution also result in low recovery.
Effect of Temperature upon Iodine Reaction
Determinations
Temperature
3 3 3
18-20 25-27 35-40
Chlorogenic Acid Taken Found 34.1 34.1 34.1
30.5 34.3 35.6
89.3 100.7 104.3
method (f5)is based upon the oxidation of chlorogenic acid v i t h iodine in alkaline solution. The solution to be analyzed is mixed with a measured amount of standardized iodine, then hypoiodite is generated by adding sodium hydroxide. Following a given reaction period, the solution is acidified and the liberated excesy is back-titrated Tvith standard thiosulfate solution. The theoretical aspects of this reaction and its application in the measurement of catechol and its isomers, 1,2-naphthoquinone-4-sulfonic acid, adrenaline, tyrosine, and thiamine have been presented by Slotta and Xeisser (14, 16). Their work indicated that under specified conditions o-diphenols such as chlorogenic acid react with about 10 atoms of iodine and showed that the hypoiodite oxidation involves the aromatic ring only without affecting the side chain. Current observations have shown that exact control of conditions during the iodine reaction are necessary for precise measurements. Reliable results are obtained when: (a)the amount of chlorogenic acid in the aliquot treated Tvith iodine is 30 to 40 mg. (see Table 11), ( b ) about 150% excess iodine is present (see Table 11), (c) the pH of the solution prior to the addition of the iodine is between 3 and 8 (see Table 111), (d) the sodium hydroxide (25 ml., 0.1 S )is added over a period of 3 minutes, and the final pH is between 10.9 and 11.1, ( e ) the temperature is 25" * 2 " C. (see Table IV), and (f)the reaction time is 1 hour. Under these arbitrary conditions, 1 molecule of anhydrous chlorogenic acid reacts with 10.55 atoms of iodine. This figure is an average of 20 titrations on a pure chlorogenic acid sample. If one or more of the factors listed in the preceding paragraph is varied, the amount of iodine consumed per mole of chlorogenic acid will also vary. Conditions Affecting Formation and Decomposition of Lead Chlorogenate. The precipitation of chlorogenic acid by means of lead acetate has been used by many workers as a means of separating the acid from other soluble constituents of coffee. In using this step in an analytical procedure, i t is essential to employ the conditions that will give quantitative precipitation and recovery of the chlorogenic acid and a t the same time provide optimum specificity. The conditions known to affect the lead
ULTRAVIOLET ABSORPTION METHOD FOR CHLOROGEVIC ACID
Reagents. Petroleum ether, c.P., boiling point 20" to 40" C. Potassium acetate, c.P., crystalline (CH,COOK.3H20). Lead acetate, C.P. basic dry powder, saturated water solution. Filter paper, Whatman S o . 1,ll-em. Green's Fluted To. 4881 18.5-em. Procedure. Prepare the sample for analysis by grinding and flaking. Determine its moisture content by drying 2 grams t o constant weight in a 70' C. vacuum oven. Defat a 2-gram sample of thr flaked green coffee bv washing with three 25-ml. portions of petroleum ether, boiling point 20" to 40" C. Allo~vthe residue t o stand about 5 minutes, or until most of the solvent has evaporated. Transfer to a 1-liter volumetric flask. For roasted coffee Tveigh 2 grams of the flaked sample directly into a 1-liter volumetric flask. d d d about 800 ml. of distilled water and shake occasionallv for 20 minutes. Dilute to volume, shake thoroughly, and filter through a Buchner funnel, using a No. 1 IThatman filter paper. Discard the first 25 to 50 ml. of filtrate and collect about 150 ml. Dilute the filtrate of both the green and roasted coffee 1 to 10 with water and measuie the optical density at 324 millimicrons. Calculate the per cent chlorogenic acid in green coffee using an E:?,, value of 526 for anhydrous chlorogenic acid. Analysis of roasted coffee requires a lead precipitation step. Place 100 ml. of the roaqted coffee extract in a 200-inl. volumetric flask. Add 1 gram of potassium acetate and 2 ml. of saturated basic lead acetate solution Tvith swirling. Place the flask in a steam or boiling Lvater bath for 5 minutes, then transfer to an ice bath a t about 0 " C., and stir mechanicallv for 1 hour mith a glass stirrer. I\lake to volume, shake thoroughly, and filter through a fluted filter paper. Discard the first 25 to 50 ml. of filtrate and collect about 25 ml. Take the density reading of this filtrate without dilution a t 324 millimicrons. Subtract one fifth of the densitv reading of the lead filtrate from the densitv reading of the 1 to 10 diluted original extract. Calculate the per cent chlorogenic acid using this corrected density value. The absorption measurements reported herein were made with a Beckman quartz spectrophotometer, Model DC, using 1-cm. quartz cells and a slit width of 0.3 to 0.5 mm. /9,
Within the limits used, 5 to 20 mg. per liter, the ultraviolet absorption of chlorogenic acid in water solution follows Beer's law of direct proportionality between absorption and concentration, and molecular extinction values of potassium chlorogenate and potassium caffeine chlorogenate are identical with those of chlorogenic acid. The effects of the presence in coffee of isomeric forms of chlorogenic acid and of free caffeic acid upon the absorption have been neglected, all measurements being made with pure chlorogenic acid as the standard. Preparation of Sample. The importance of proper preparation for analysis of a plant material such as coffee cannot be
623
V O L U M E 20, NO. 7, J U L Y 1 9 4 8
chlorogenic acid, absorbing a t 324 millimicrons. I t is not known whether this represents a modified chlorogenic acid, some entirely different substance formed during roasting, or an interference with the precipitation of lead chlorogenate. These observations indicate that lead precipitation should be included in the procedure for roasted coffee, Factors Affecting Light Absorption of Chlorogenic Acid. The absorption of chlorogenic acid remains constant over the p H ranges 2.4 to 5.9 in solutions bnfW A V E L E N G T H IN M l L L I M I C R O N S fered with citric acid-disodium phosphate mixtures. In a singleexperiment the ab. Figure 3. UltraFiolet Absorption of Roasted Coffee Extrart sorption of the acid in pure solution was 1. Original water extract found to decrease by about 18% a t 324 2. Filtrate from lead precipitate of water extract, deleaded millimicrons when the p H was raised above 8; this change is characteristic of phenolic compounds. Tt is also possible that the decrease is a overemphasized. Extensive investigations in this laboratory secondary effect of air oxidation in the slightly alkaline solution. have shown that accurate analytical values on green coffee can be S o change in the absorption a t 324 millimicrons was observed in obtained only on samples in which most of the cell structure has pure solutions or coffee extracts held at pH 4 to 6 for 1 to 3 days been broken. I t is recommended that, for the analysis of any a t room temperature in diffused daylight. water-soluble constituent of green and roasted coffee, all samples be milled to a particle thickness of 50 to 100 microns. A 3-1-011 Precision and Specificity of Ultraviolet Method. A comparatively high degree of precision is attainable with the ultraviolet chocolate mill, cereal flaking mill, or any mill giving comparable absorption method because of the simplicity of the manipulative grinding can be used for this purpose. procedure and the reproducibility of the final light measurement. Extraction and Precipitation of Chlorogenic Acid. Several The statistical methods of Moran (11) and the A.S.T.M. ( 1 )have different techniques have been used in previous methods for been applied to the calculation of the precision of the ultraviolet extracting the chlorogenic acid from coffee. Snalyses with both the iodometric and ultraviolet methods show that a 500 to 1 method. The valueof 1 2 u has been uscd in calculating the L C g S batch extraction gives higher and more reproducible chlorogenic value (limit of uncertainty for 95 out of 100 determinations). h acid values than a 10 to 1 batch extraction or Soxhlet extraction. correction factor of 0.869 has been applied in these calculations as h slight turbidity in the dilute water extracts of green coffee recommended by the A.S.T.N. (1). On the basis of the detercan be avoided by petroleum ether extraction. This solvent does minations shown in Table V, the L1-95for chlorogenic acid by direct ultraviolet readings on water extracts of green coffee is not remove any substances which absorb light in the range of 324 millimicrons. Turbidity is not encountered in the water extracts 10.24%. When the determination includes lead precipitation as of roasted coffee. used for roasted coffee,the corresponding LVy5value is ~ O . l O ~ c . Experimental results on pure solutions show that basic lead There is some variation between the chlorogenic acid values acetate is a better precipitating agent than the neutral salt and found by the ultraviolet and iodometric methods. The ultrathat effective precipitation can be obtained by heating the leadviolet absorption figures, as shown in Table V, are 5yGhigher than treated solution for several minutes in a steam bath and then the iodometric values for green coffee and 177, lower for roasted stirring in an ice bath for 1 hour. The addition of potassium coffee. These variations reflect the different types of specificity acetate lowers the solubility of the lead chlorogenate. By taking of the two methods. The ultraviolet absorption procedure is the advantage of these factors i t is possible to precipitate in 1 hour more reliable of the two methods for both green and roasted coffee, 98.5% of the chlorogenic acid in 100 ml. of solution containing as it is based on a highly specific measurement. I t has also been 15 mg. of pure chlorogenic acid. The solubility of lead chloroobserved in this laboratory that pyrolyzed products of sucrose, genate in pure solution would account for essentially all of the material remaining in solution in lead-treated green coffee extracts which absorbs a t 324 millimicrons. Excess lead acetate Table Y. Comparison of Ultraviolet and Jodometric does not interfere with the absorption. Under comparable conl l e t h o d s for Determining Chlorogenic Acid i n Coffee ditions, 0.2 mg. of chlorogenic acid is in solution after the lead preUltraviolet -4bsorption Method By difference before cipitation of pure solutions, and 0.3 mg. of apparent chlorogenic Direct and after lead Iodoinetric 3Iethod b acid is in solution after the precipitation of the green coffee reading precipitationa Per Cent ( D r v Basis)---extracts. The basic lead acetate removes 98.07, of the material 7.41 absorbing a t 324 millimicrons from green coffee extracts. These 7.46 data, shown in Figure 2, indicate that the total ultraviolet ab7.41 7.46 sorption a t 324 millimicrons of aqueous green coffee extract is as 7.46 7 .41 reliable a measurement of chlorogenic acid as a method based A v. 7 80 7 74 7.43 upon the absorption bcfore and after lead precipitation. 4 86 4.58 Roasted blended coffee 5.75 The observations made on roasted coffee extracts show that an 5.77 4 88 4.58 4 83 4.56 5,61 appreciable fraction of the material absorbing light a t 324 milli4 84 4.56 6.09 . . 4 82 .. microns remains in solution after treatment with basic lead 4 73 .. .. acetate. The ultraviolet absorption spectrum of a roasted .. 4.78 .. .. 4 83 coffee extract is shown in Figure 3. The soluble fraction contains .. 4 80 .. 4.78 the equivalent of 0.8 mg. of chlorogenic acid or 8.4%of the total 4V. 4.82 4.57 d . 81 of 9.4 mg. in 100 ml. of solution. After the correction factor for a Correction of 0.2 mg. added for solubility of lead chlorogenate. the solubility of lead chlorogenate (0.2 mg.) is applied, there reb Correction of 3.0 mg. added for solubility of lead chlorogenate. mains in solution about 6% of the total material, calculated as
-----
ANALYTICAL CHEMISTRY
624
comparable t o those present in ioasted roffee, react with iodine under the conditions used in the iodometric method for chlorogenic acid. 5001 The water-soluble alkaloids, proteins, 400 and sucrose found in water extracts of green coffee apparently do not interfere with the direct ultraviolet absorption .300 method. The absorption at 324 millimicrons of a solution containing pure chloroY genic acid, trigonelline, caffeine, sucrose. and the nondiffusible fraction of green 200 coffee extract, u hich together represent about 907, of the water-solubleportion of flaked green coffee, is almost identical 100 with the absorption of the same amount of vhlorogenic acidin pure solution, as shon n bv the absorption curves in Figure 4. I I I I I Dialysis experiments indicate that the e40 280 320 360 41 WAVELENGTH I N MILLIMICRONS high molecular n-eight components of Figure 1. ihsorption Curves of Chlorngetiic .kcid green coffee extract do not interfere n i t h 1. Pure chlorogenic acid hemihydrate the absorption measuremeiits. The frac2. Mixture of chlorogenic acid hemihydrate, caffeine. trigonelline. eurrose. and nondiffusilrleb tion of the water extract which diffuse. in xreen roffee water extract rhrouah cellolnhane contains essentiallc all of the material which absorbs light at 324 millimicrons. ( 2 , A ~ a o c Official . Agr. (;lieiii., “Oficial aiid Tentative Method? uj 6th ed., Washington, D. C., 1945. Known amounts of pure chlorogenic acid added to the extrac(81 Barnes, H. hl., Feldman, J. R., and White, W.T., “ISOchhJl#>t i m mixture of green coffee and water can also be measured quangenic Acid from Coffee,” presented before Division of 01titatively by direct absorption measurement. For example, th(, . ganic Chemistry at, the 111th Meeting of AM. CHEM.Soc.. recovery of 142.2 mg. of chlorogenic acid added to 2-gram sanipkb Atlantic Cit,y, K.J., April 1947. (41 Charaux, C.. J . Pharm. Chim., 2, 292-8 (1910). of green coffee was 142.5, 142.5, 140.5, and 140.0 nig. for four (5) Fischer, H. 0. L., and Dangschat, G . , B e r . , 65B, 1037 (19328 separate extractions. (6) Freudenberg, K., Ihid.. 53B,232 (1920). These observations indicate that the ultraviolet absorptiuii (7) Gorter, K., Ann., 358, 327 ( 1 9 0 7 ) : 359, 217 ( 1 9 0 8 ) ; 379. method is more precise and specific and a t the same time much 110 (1911). more rapid than any existing procedure for measuring chlorogenic (8) Hoepfner, IT., Chem.-Ztg. 56, 991 (1932). acid in coffee. (9) Jurany, H., 2. a n d Chem., 94, 2 2 5 (1933). (10) Kremers, R.E., personal communication. ACKNOWLEDGMENT (11) Jforan, R. F., IND.ESG.CHEM..A N ~ I ED., . . 15, 361 ( 1 9 4 3 ) . These methods for measuring chlorogenic acid were developed (12) Plucker. W., and Keilholz. IT., Z.Cafersuch. Lebensm., 6 8 , $37 as a part of the program of fundamental researchon the chemistq109 (1934). ( 1 3 ) Rudkin, G. O., and Nelson, J. hl., .J. Am. Chem. Soc., 69,1470 of coffee sponsored by the Maxwell House and Sanka Divisions of (1947). General Foods Corporation. The authors acknom-ledge with (14) dlotta, K. H., and Neisser, K., Ber.. 71B,1611 (1938). gratitude the interest and encouragement of R. E. Kremcrs and (15) Ibid., p. 1616. H. ?*I.Barnes of this laboratory. (16) Ibid., p. 1984.
-
-
-
v
,I’
LITERATURE CITED
(1)
Soc. Test,ing Materials, “Manual on Present,atioii of Data,” t,hird printing. Philadelphia. 1941.
A1n.
RECEIVED September 23. 1947. Presenced before t h e Division of Analytica! and Micro Chemistry a t the 112th Meeting of the ANERICANCHEMICAL SOCIETY, Wea I’ork. N. Y .
Determination of Bromine Addition Numbers A n Elect romet ric Method H. D. DtlBOIS AND D. i. SKOOC,’, California Research Corporation, Richmond, Calif.
B
ROAIIXL addition ib widely used in the petroleum iiidustr\ to measure the olefinic unaaturation of hydrocarbon materials, and a number of methods have been proposed for thiT determination. One of the most common is the Francis method (1 and its modifications by LIulliken and Wakeman ( 4 ) and Le\\ IC and Bradstreet (3). The original Francis method used an m e r \ of a potassium bromide-bromate reagent in acid solution as thr brominating agent, and the excess was determined iodometricallv 1
Present address, Department of Chprni-tri .itanford I-nix wsitl Calif
In the modifications of the method. the e\(*es>of the brominating rpagent and the time allowed for bromination are carefully cuntrolled in order to avoid high results as a consequence of substitution of bromine in the hydrocarbon These procedures have all been found wmewhat time-consuming, and considerable trouble is encountered in determining the proper excess of reagent, particularlv u hen the samples are colored Kaufmann ( 2 ) has proposed a method whichinvolves theuseof a bromine solution in methvl nlcohol saturated v ith sodium bromidr
