April, 1945
ANALYTICAL EDITION
more reproducible by the cerate procedure and are obtainable with less manipulation and in a shorter time. ACKNOWLEDGMENT
The writer gratefully acknowledges the collaboration of members of the Inspection Board Staff. This paper is published with the permission of the Inspection Board of the United Kingdom and Canada.
217 LITERATURE CITED
(1) Moran, R. F., IND.ENG). CHEM.,ANAL.ED.,15, 361 (1943). (2) Smith, G . F., “Cerate Oxidimetry”, Columbus, Ohio, G. Frederick Smith Chemical Co., 1942. (3) Smith, G . F., and Richter, Frederic, “Phenanthroline and Sub.
stituted Phenanthroline Indicators, Their Preparation, Properties, and Applications to Analysis”, Columbus, Ohio, G. Frederick Smith Chemical Co.. 1944. (4) Sommer and Pincas, Ber., 48, 1963 (1915).
Polarographic Determination OF Vitamin in Fruits and Vegetables W. S. GILLAM,
Purdue University, Lafayette, Ind.
A polarographic method for determining ascorbic acid in fruits and vegetables affords a reliable means of determining vitamin C in quantities ranging from 4 to 85 micrograms per ml. of solution with an accuracy of 3.3 to 4.39& depending upon the supporting electrolyte used. Variable results were obtained on some extractse.g., dehydrated beets. The half-wave potential and the diffusion current, respectively, were found to b e independent of, and proportional to, the ascorbic acid concentration in the four supporting electrolytes studied. Diff usion current constants, half-wave potentials, and diffusion coefficients were determined in four supporting electrolytes. Results of vitamin C analyses on fruits and vegetables b y four different methods-polarographic, visual titration, photometric, and Roe’s-were reported and discussed. The polarographic method showed good agreement with the visual titration and photometric methods on certain fruit extracts. By making use of the diffusion current constant polarographic determination of vitamin C in certain m.%ials i s possible without the necessity of calibrating the instrument.
M
C
OST of the methods that have been proposed for the estimation of vitamin C have been based upon the oxidation of ascorbic acid to dehydroascorbic acid by a specific oxidant such as 2,6-dichlorophenolindophenol. Roe (11) has developed a method which is based on an entirely different reaction of ascorbic acid-namely, its reaction with 2,+dinitrophenylhydrazine. The use of a dye as a specific oxidant has certain limitations. The indophenol dye is not ideal because of its relatively high oxidation potential; materials other than ascorbic acid may be osidized. Its reaction with ascorbic acid is not instantaneous and therefore the time of reading the end point and the rate of addition of the dye are factors that may introduce errors. In the absence of the vitamin the dye will fade, and cause a drift in the galvanometer if used in connection with a photoelectric colorinieter. For highly colored extracts the visual titration method cannot be used, and some modification of the method must be resorted to, such as the one suggested by Bessey (2). Potentiometric methods have been published for estimating vitamin C (1, 6) but a continuous drift in potential throughout the whole course of the titration with the dye makes it very difficult to locate the end point with any degree of accuracy. Harris et al. (6) using a platinum-mercury electrode greatly improved the potentiometric method. Their electrode, however, was found to be slightly sluggish a t the end point, and when working with materials that were high in content of reducing substances they used the less desirable bright platinum electrode. Lewis (S), in this laboratory, found the platinum mercury electrode acceptable on extracts from fresh materials but unsatisfactory for dehydrated products. The ’polarographic method, described in this paper, differs
from most methods in that a specific compound is not used as the oxidant. Instead, ascorbic acid is oxidized at the dropping mercury electrode. Essentially, it is based on current voltage curves and since ascorbic acid possesses striking reducing properties this method should be applicable to vitamin C determination. The determination is specific, is sensitive to small concentrations of ascorbic acid, and colored extracts do not necessarily interfere. Furthermore, an inspection of the anodic wave will definitely indicate whether or not interfering materials are Dresent in the solution being analyzed. Kodicek and Wenig I ’ ‘ 1 ( ? ) stated that dehydroascorbic acid was not reducible a t the dropping mercury cathode, but a characteristic wave was obtained when the dropping mercury electrode was polarized &s an anode. They suggested using a 0.066 N phosphate buffer (pH 7.0) in all polarographic analyses for vitamin C. Their method of analysis was I I ,I , suitable for orange juice +0.3 +0.2 t O . 1 0 but they were not sucVOLTS cessful in applying it to (vs. s.c. E.) extracts of plant or Figure 1. Shift of Half-Wave animal tissues. Other Potential with Change of pH constituents apparently 1, 28 micrograms of ascorbic acid per ml. at hindered the electrode PH3.41. 2, 28 micrograms of ascorbic acid per ml. at p H 2.17. 3, residual current, reaction. supporting electrolyte II (phosphate bufferCozzi (3) reported the 1.5% HPOa). Broken line, residual current after aeration use of the polarograph in the analysis of fruit juices and concluded that it gave high results. The present paper presents a polarographic method for determining vitamin C in plant materials vhich eliminates the necessity of frequently standardizing solutions, is not affected by colored extracts, and is sensitive to very low concentrations of the vitamin. Diffusion current constants and the half-wave potentials of ascorbic acid in four different supporting electrolytes are given. The results of analyses run in this laboratory by four different methods-polarographic, visual titration, photoelectric, and Roe’s dinitropnenylhydrazine method (f2)-are also reported.
-0.1
PRELIMINARY INVESTIGATION
In the oxidation of organic compounds hydrogen ions are usualiy involved in the electrode reaction-.g., RH,,==
INDUSTRIAL A N D ENGINEERING CHEMISTRY
218
Vol. 17, No. 4
of 0.25% oxalic acid. Electrolyte I1 consisted of one volume of the phosphate buffer (pH 8.0) and one volume of 1.5% metaphosphoric acid. Both solutions had excellent buffering capacities, and pure solutions of ascorbic acid or extracts from natural products were stable in them for several hours at room temperature. Electrolyte I11 consisted of a 1 to 1 solution of biphthalate buffer and a solution consisting of 0.25 gram of oxalic acid dissolved in 100 ml. of 50% alcohol while electrolyte IV contained biphthalate buffer and 3% metaphosphate in a 1 to 1 ratio. METHOD OF ANALYSIS
4SCORBIC
Figure 2.
+
CONCENTRATloN
PER ML
Calibration Curve for Ascorbic A c i d In Supporting Electrolyte li Phosphate bulirrl.5W HPO8
+
R nH+ ne. Since the hydrogen ion is a component of the electrode reaction, its concentration at the electrode surfme is currentdependent. It is therefore evident that for the Polarographic oxidation of ascorbic acid a strongly buffered solution is required t o keep the concentration of the hydrogen ions a t the and equal to .the concentration of the electrode fairly hydrogen ions in the whole solution. The number of buffer solutions suitable.for use with the polarograph is limited by the fact that they must contain no substance that will interfere with the ascorbic acid anodic wave. Of approximately 24 buffers investigated Only Six Yielded a satisfactory residual current curve. After a critical study of p g buffering capacity, and stability of ascorbic acid in the supporting elect,rolyte solutions, two buffers finally selected: potassium biphthrtlatete-sodium hydroxide buffer (PH 6.2) and a phosphate buffer (pH 8.0). Many extracting media were investigated but all but two, metaphosphoric and oxalic acids, were discarded because they interfered with the anodic wave of ascorbic acid or the vitamin w&s not sufficiently stable in them. Of these two acids oxalic is probably preferable since it can be stored indefinitely and the data indicated that it had a greater stabilizing effect on ascorbic acid and at higher pH's than those solutions containing metaphosphate. Since the half-wave potential of ascorbic acid shifts to more positive values with increasing acidity, the pH of the supporting electrolytes must be kept within certain limits. If the pH values are too low the anodic wave of ascorbic acid is shifted toward such high potentials that the limiting current canriot develop (or is of such short duration that it is difficult to determine) because of the interference of the residual current (curve 2, Figure 1). Thus for supporting electrolytes I and I1 the p H must not be much below 4.6 and 3.4, respectively. On the other hand, if the pH values are much greater than this the vitamin may become unstable. Electrolytes I and I1 were used in practically all the analyses performed in this laboratory. Electrolytes 111 and I V were used only occasionally. I n order to obtain p H values of 4.6 and 3.4, respectively, for supporting electrolytes I and I1 0.25% oxalic acid and 1.570 metaphosphoric acid were used. The sup orting electrolytes consisted of a 1 to 1 mixture of either acid wit% the corres onding buffer. For example .sup rtin electrolyte I consistecf)of one volume of potassium kipht&ate%utfer (pH 6.2) and one volume
REAQENT.Potassium biphthalate buffer (pH 6.2), 500 ml. of 0.1 M potassium acid phthalate plus 470 ml. of 0.1 N sodium hydroxide, made up to 1 liter. Phosphate buffer (pH 8.0), 50 ml. of 0.066. M potassium dihydrogen phosphate plus 950 ml. of 0.066 M disodium hydrogen phosphate. Oxalic acid solution, 0.25%; metaphosphoric acid solution 1.5%; and 0.25 gram of oxalic acid dissolved in 100 ml. of 60% alcohol. APPARATUS. The current-voltage curves were obtained by means of a Leeds & Northrup Electro-Chemo raph. All the polarograms were obtained by using an externaf saturated calomel electrode as cathode, connected to the electrolysis cell through an a ar bridge saturated with potassium nitrate. Afl experiments were carried out in a water bath at 25.0° i 0.2' C. The amount of mercury flowing from the dropping electrode per second, designated by m, w&B determined droppin in air (10). The n and t values of the dropping capillary use8 were respectively 2.33 mg. of mercury per second, and 3.0 seconds when the capihary tip dipped into the supporting electrolyte and no potential was applied to the mercury drop. The drop time, of course, varies with the nature of the supporting electrolyte and the potential of the dropping electrode but it W&B found much more expedient, a t the start of an anafysis, to adjust the pressure on the mercury so that 1 = 3.0 seconds in all the SUPPOrtinK electrolytes used. i n all cases the drop time was also measured at the approximate Potential a t which the diffusion current was messured. Thus for supportin electrolyte I (biphthalate-oxalic acid) t = 2.85 seconds, for efectrolyte I1 ( hosphate-l.5% HPOJ t = 2.67 seconds, for electrolyte II? (biphthalate-oxalic acid-alcohol) t = 2.81 seconds, and for electrolyte I V (biphthalate-3% HPOJ t = 2.68 seconds, respectively, when a potential of 0.19, 0.26, 0.14, and 0.30 volt was applied to the mercury drop. The value of mala t l / b for the capillary used was 2.08. PROCEDURE. A given weight of the material to be analyzed is whi ped in a Waring Blendor with a known K h I n e of metam phospKoric or oxalic acid for 2 to 5 minutes ( 4 ) . The extract is then filtered through a dry fluted filter paper. A 2- or 5-ml. aliquot of the filtrate is withdrawn, an equal volume of the buffer
+0.4
m
-
0 -
W
K
~ 0 . 3 t0.2
+O.I
0
-0.1
V 0 L T S (VS S.C.E.) Figure 3. Current-Volta e Curves of Ascorbic A c i d in Supporting Efectrolyte I I at 25' C. 1 23 microgrirnr of ncorblc acid per al P 11 7 mlsrogramr ol &corbie scld per ml. 3, reslduil eurredt brok;a h e , r d d r a l current i f f i r aeration. Range of recorder, 10 nlcrwmpau
ANALYTICAL EDITION
April, 1945
solution is added, and the well-mixed solution is polarographed between -0.1 and +0.3 volt. The diffusion current is then measured on the polarogram, corrected for the residual current of the supporting electrolyte, and the corresponding concentration of vitamin C is read off the calibration graph, Figure 2. If the concentration of ascorbic acid in the aliquot used is too reat to give a satisfactory limiting current, the sample may be %luted with the supporting electrolyte being used or a lower sensitivity of the instrument may be employed. I n analyzing solutions one merely adds a 1- to 5-ml. aliquot to an equal volume of the supporting electrolyte, and polarographs the solution. ANALYTICAL RESULTS
219
ml. aliquot was polarographed. The anodic waves obtained from the extracts do not always coincide with the residual current curve in the range -0.1 to 0 volt, the former usually lying below the latter. Thus a correction must be applied to obtain the true diffusion current. Under such conditions, or when the limiting current is not well defined, i t has been found advantageous to determine the true diffusion current by taking the difference in current a t two predetermined potential points. By this method it is only necessary to select two reference potential points to be used in calibration and in subsequent analyses of solutions. With this particular electrolyte (I) the potentials used were 0 and 0.2 volt. The corrected diffusion current is then taken as the difference in current a t the two potentials, minus the difference in current a t these two potentials when the ascorbic acid concentration is known to be zero. Many of these materials were analyzed by four different methods-polarographic, visual titration, photometric, and Roe's methods (Tables 111and IV).
The instrument was calibrated by polarographing solutions containing known amounts of crystalline ascorbic acid, shown to have a purity of 98% by a n iodine titration. The residual current was determined for each supporting electrolyte by polarographing the solution between -0.1 and $0.3 volt at each rmge of the recorder--e.g., 40,28, 10, 5, 4, 3, and 2 microamperes. This corresponds to a range of l/m to full sensitivity of the galvanometer. The corrected diffusion currents were then obDISCUSSION tained by measuring the distance between the limiting current and The calibration curves show that ascorbic acid can be deterthe residual current. mined with an accuracy of 3.3% in supporting electrolyte I when All the residual current curves were similar to the one plotted the ascorbic acid concentration lies between IS and 85 microin Figures 1 and 8, in that they showed a pronounced cathodic grams per ml. I n the concentration range 5 to 13 micrograms current between -0.1 and +0.1 volt. Although cupric ions per ml. the accuracy is 4.0%. I n electrolyte I1 the accuracy is will give a cathodic current within this voltage range, the wave 4.3% over a range in concentration of 3.5 to 85 micrograms per noted was due t o oxygen. Consequently it can readily be removed by aerating the solutions with either nitrogen or carbon dioxide. The residual eurrent c w e , after aeration, is indicated Table I. Reproducibility of Results by the broken line in Figurea 1 and 3. (Micrograms of ascorbic acid per ml. of solution) Oxygen was not removed from the solutions analyzed because Material Material it d i d p o t interfere with the ascorbic acid determinations and by 61.6 Spinach 60.0 Tomato eliminating the aeration procedure considerable time mas saved. 60.1 60.0 61.6 62.2 Table I illustrates the reproducibility of the results in a num60.8 58.0 ber of repeated determinations on extracts from natural products 60.0 61.6 Mean and on several pure solutions containing varying amounts of as1.0 Average deviation 0.6 1.6 Coe5cient of variation, % 1.0 corbic acid. The maximum coefficient of variation for the de94.0 Cadiilower 180.0 Tomato terminations listed on natural products is ~ 1 . 6 % . 94.0 182.0 94.0 180.0 The zlf-wave potentials, determined experimentally, the 94.0 180.6 Mean diffusion coefficients, 'and diffusion current constants are listed Average deviation 0.0 0.9 in Table 11. The diffusion current constants for supporting Coe5cient of variation, % 0.0 0.5 electrolytes I, 11, 111, and IV, respectively, are constant t o 137.6 40.0 Pepper Orange 132.0 40.0 t2.2, *2.2, t 2 . 0 , and *3.9oJO. 130.0 40.8 133.2 39.6 Theoretically, the shift of the half-wave potential with chang40.1 133.2 Mean ing p H values should be 0.059 volt per p H unit. This shift was Average deviation 0.4 2.2 experimentally determined in two supporting electrolytes 1.0 1.6 Coe5cient of variation, % (phosphate bufTer-l.5% metaphosphate) and phthalate bufferPure Solution of Aacorbic Acid 0.25% oxalic acid) and was found in each case to be 0.056 volt Contained Found per p H unit, which is in very good agreement with the theoretical 300 300 (Figure 1). 300 300 The value of the half-wave potential was found to be inde321 325 38 38 pendent of the ascorbic acid concentration in all the supporting 30 30 30 30 electrolytes studied (Figure 3). 20 30 Many naturally occ&ing products have been successfully 22 21.8 analyzed in this laboratory by the polarographic method and a 11.7 11.7 few of these results are reported in Tables I11 and IV. Fresh materials such as orange, lemon, grapefruit, tangerine, Table 11. Half-Wave Potentials (Volts vs. Saturated CaIomeI Electrode a tomato, pepper, spinach, lettuce, apple, and 45' C.), Diffusion Coefficients, and Diffusion Current Constants potato, and plant tissues such as tomato, mustard, turnip, Supporting Half-Wave Diffusion Diffusion and beet tops, have been analyzed and found to yield aatisElectrolyte Potential Coe5cient Current Constant factory anodic waves. Dehydrated potatoes and carrots SQ.em. sec.-1 ~ J C i,j/Cm'//'t'/I(g) Biphthalate bufferlikewise have been successfully analyzed, and with special :if5% oxalic acid precautions the vitamin C content of dehydrated beets 0.10 5 . 7 X 106.25 3.00 Phoaphate bufferand onions can be determined. 1.5 0 HPOa (11) 0.17 6 . 2 X 10-0 6.34 3.04 Several polarograms obtained from fruit and Biphtralate huffer0.25% oxalic acid vegetable extracts are shown in Figure 4. From 5 in so% alcohol 2.34 0 06 3 . 6 X 104.88 to 25 grams of the material were extracted with Biphthaletebuffer(111) 2.88 0 . 2 4 5 . 3 x 100.25% oxalic acid, except in the case of H and I, fol3% HPOa (IV) 6.02 lowing the procedure previously described, and a 5-
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
220 Table
111.
Analyses of Tomatoes Extracted with
2 to 4% and in two cases the former values were slightly higher than the visual titration values. I n the case of cauliflower (Table IV) the polarographic value was considerably higher than that obtained by the visual titration method.
3% HPOI
(.\Iilligrams of ascorbic acid per gram of fresh weight) Polarographic Visual Titration Saniple Method Methoda
a
633-5-5-3 1956-1-L 1554-2 633-5-8-2 1956-1-5 633-5-1-2 Analyzed b y J. W. Porter.
0.452 0.398 0.479 0.390 0 380 0 361
0.516 0.411 0.665 0.447 0.407 0.398
ml. With electrolyte 111 the accuracy is 4.07, (concentration range, 5 to 95 micrograms per ml.), and in electrolyte IV the accuracy is 4.3% over a range in concentration of 8 to 85 microgranis of ascorbic acid per ml. The data in Table IV illustrate conclusively that the polarographic method checks very well with the visual and photometric methods on several fruit and vegetable extracts. Theoretically, the polarographic method might be expected to give more accurate results than any method which uses the indophenol dye. For obvious reasons the values obtained by the visual titration method might be expected to be slightly higher than those obtained by the polarographic method. Theoretically the polarographic method is more specific because each compound is characterized by its half-wave potential. When several electrooxidizable substances are present in the solution each one will produce its own characteristic wave if the halfwave potentials are far enough apart. Thus, compounds such as glutathione or cysteine should not interfere in certain acid supporting electrolytes. The data in Table I11 may indicate that the -0.5 polarographic method is more specific than the visual titration method in these particular extracts. The tomatoes listed in this table consisted of the red- and green-fruited species, Lywpersicon parnpinellifolium and Lycopersiwn peruvianum. Several of these small tomatoes were whipped in a Waring Blendor with 3% metaphosphate and filtered, and aliquots of the filtrate were analyzed by the two methods. A few of these extracts were slightly pink in color. Sample 1554-2 gave the most highly colored filtrate and both methods showed it to contain the greatest concentration of ascorbic acid. The variation between the two analyses, however, was greatest for this sample, which would indicate that slightly colored solutions tend to give high values by the visual titration method. I n every case (Table 111) whether the filtrate was colored or not, the visual titration values were from 3 to 18% higher than the polarographic values. This may be due to the fact that (a) a small excess of dye must be added to indicate the completion of the reaction, ( b ) the pink color in some of the extracts would tend to obscure the end point, and (e) the dye may have reacted with reducing materials other than ascorbic acid. All .these would tend to increase the titration value. Another sample of tomatoes (Lywpersicon pimpinellifolium) was extracted with oxalic acid (sample 1, Table IV),. This extract was colorless but again the visual titration value, as well as the photometric value, w a ~considerably higher than the value obtained by the polarograph. The tomato samples, 2 to 9, consisted of oxalic acid extracts of F, and FI selections of crosses of Lywpersiwn escukntum, L. hirsutum, and L. pimpinellifoliuin species. Here, the polarographic values check the visual titration values within
Vol. 17, No. 4
I n all the analyses Roe’s method (1.2) gave rather high results. This was to be expected, since his method determines total vitamin C content (ascorbic plus dehydroascorbic acid), and the difference between these values and those obtained by the polarographic method was attributed to the presence of dehydroascorbic acid. However, several of these values, notably those for fresh potato and lettuce, appeared to be too high. For this reason Roe’s method was included in the analyses of certain citrus fruits, where dehydroascorbic acid, if present a t all, would be in very low concentration. Curiously, Roe’s method, even on these products, gave very high values; in one instance (tangerine) the value was 300% greater than the value obtained by the polarographic or visual titration methods. To check this further these extracts were treated with hydrogen sul6de and subsequently analyzed by the polarographic and visual methods (Table V). Results indicated there was very little, if any, dehydroascorbic acid present in the citrus fruit extracts studied. Additional work dealing with these four methods of analysis will be reported in a later paper.
& O H
-0.5 -1.0
-1.5
-1.5
-1.5
:Qd;R -0.5
-0.
-1.0
-1.0
-0.2 -0.4
-0.
+O, +O.
tO.l
0 -0.1+0.3 +0.2 +O.I
V O LTS Figure 4.
Typical Polarogram
o
-0.1 +0.3+0.2+O.I
o -at
(VS. S.C.E.) ol Fruit 4nd Vegetable
Extracts
H and I polaro#raph*d in supporting ~labolvtrIll, all others in supporting rl&olyte
I
ANALYTICAL EDITION
April, 1945 Table
IV. Ascorbic A c i d in Fruit and Vegetable Extracts
Material Ascorbic acid, 300 rnicrograms per rnl., in 1.5% HPOa Ascorbic acid, 321 micrograms per ml. in 0.25% oxalic acid Orange Orange Grapefruit Lemon Tangerine Tangerine Tangerine Cauliflower Apple Tomato 1 2 3 4 5 6
7 8 9 Pepper Lettuce Lettuce Fresh potato Dehydrated carrots Dehydrated potatoes 1 2 3 4 5 Dehydrated onions
Ascorbic .4cid. Micrograms per Milliliter of Extract PhotoVisual metriob Zoea titrationa
Polarographic
300
302
306
300
325 40.1 27.0 42.0 51.1 33.0 31.0 41.0 180.0 6.4
315 41.3
312
315
80.0
0.168 0,200 0.138 0.170 0.126 0.207 0.344 0.182 132.5 9.2 4.0 8.4 9.6
.....
40.1 50.4 35.0
.....
.....
161.0 6.0
92.0 0.176 0.204 0.138 0.174 0.123 0.214 0.343 0.178
..... ..... .....
..... .....
...
27.0
... .... .. ... 41.0 ...
... ...
73.0 72.8 108.4 65.0
...
227:2 10.7
95.0
102.0
... .. .. .. ... ... ... ... ...
...
... ... ...
.,. ...... ...
... ...
140.0 18.2
... ...
27.8 10.4
2.4
...
15.0 22.5 ... 27.0 4.0 6.6 0. 3.0 5.0 12.6 6.0 13.0 19.0 30.Oe 25.8 28.OC 20.5 a Analyzed b y J. 13‘. Porter. b Analyzed b y W. R . Lewis. 5 Poor anodic wave indicating a fairly high concentration of interfering material.
..... ..... ..... .....
.....
... ...
...
The polarographic method possesses several advantages over other published methods for vitamin C analysis. After the initial calibration no standardization of solutions is required. If oxalic acid is used as the extracting agent all solutions listed in the procedure can be stored several months. No pretreatment of the extract is required and colored extracts do not necessarily interfere, since the method is concerned with a specific and characteristic property of the reaction. Suspended matter which will affect the light absorption in the photoelectric method ordinarily does not interfere. However, with some materials-e.g., dehydrated potatoes-the colloidal content of the extract may be sufficiently high to interfere with the analysis. I n some instances this can be alleviated by using the alcoholic supporting electrolyte. In any analysig an inspection of the anodic Kave will quickly and definitely indicate whether or not interfering substances are present in the extract, being polarographed. This is not true of the visual, photometric, or Roe’s methods, for nith these one obtains a given value and has no indication of the presence or absence of interfering materials. Another advantage of the polarographic method is that any number of analyses can be run on one and the same aliquot. The method is accurate over a range in ascorbic acid concentration of 4 to 85 micrograms per ml. However, since there is a dilution factor of * / 2 (one volume of extract plus one volume of buffer) the extracts being analyzed should possess 8 to 170 micrograms of ascorbic acid per milliliter of solution. In the analysis of materials having a high content of interfering substances the method becomes less reliable and may fail completely. However, even though interfering substances may prevent the formation of a well-defined limiting current (Figure 4, H), reproducible results are obtained by taking the difference in current a t two predetermined potential points. Since the diffusion current constants for four supporting electrolytes have now been detcrmincd. an asvorbic acid analysis of
22 1
certain materials could be made from a single polarogram. Thus the necessity of preparing standard solutions of ascorbic acid and calibrating the instrument can be eliminated. The diffusion current constant has previously been referred to as i d / C , but according to Lingane (9) a more fundamental constant is id/ Cmdatl/’. Thus, C = i.i/Km2/at1/6,where K is the diffusion current constant expressed as id/Cm2”3tl/6. From this equation it is evident that m and t must be known for the capillary that is used. The determination of these two values is very simple and obviously requires much less time than the calibration of the instrument. Lingane (9) states “that with m values in the usual range from about 1 to 3 mg. per second, and drop time between 2 and 4 seconds, the accuracy of this relation is as good as the accuracy with which one can measure a diffusion current”. This procedure could certainly be used on citrus juices, but for materials such as potatoes, dehydrated products, etc., the slower procedure of recording the current a t two selected potentials must be employed in order to correct for the displacement of the anodic wave above or below the residual current curve due to the presence of interfering substances. The diffusion current constants are also of theoretical interest because they can be used to calculate the corresponding diffusion coefficients. Thus a measure of the apparent size of the ascorbic acid molecule in the different supporting electrolytes can be obtained. Ths diffusion coefficients are calculated by the applicain which tion of the IlkoviC equation, i d = 605 nD1/2Cm2/3t1/6, n is the number of electrons involved, 2 in this case, D the diffusion coefficient, C the concentration of ascorbic acid in millimole3 per liter, m the milligrams of mercury flowing from the capillary tip per second, and t the time in seconds required for the formation of one drop of mercury. Thus D = (K/1210)2when K i3 expressed as id/Cm2/3t1 ‘ 6 (Table 11;. Since the diffusion current constant of ascorbic acid was largeit in the phosphate-l.5% mctaphosphate supporting electrolyte, it follows that the apparent size of the ascorbic acid molecule must be smaller in this medium than in any of the other aqueous supporting electrolytes investigated. Similarly, the apparent size of the ascorbic acid molecule must be larger in electrolyte IV than in either I or 11. In general, the diffusion coefficients are much smaller in alcoholic solutions than in water solutions. I t is, therefore, impossible to make such comparisons, based on diffusion coefficients, between supporting electrolyte I11 and the other three aqueous supporting electrolytes. Table
V.
Comparison of Ascorbic A c i d Contents of Fruit and Vegetable Extracts before and after HIS Treatment (Micrograms per ml. of extract)
Naterial Orange Grapefruit Lemon Tangerine Cauliflower
Before HzS Treatment Polarographic Visual Roe 40.1 41.3 42.0 40.1 73 51.1 50.4 72.8 33 35 108 180 161 227
...
After Hz3 Treatment Polarographic Visual 41 39 45 39.6 55 50.4 40 35 208 17s
LITERATURE CITED
Becker, E., and diGleria, J., 2.Vitaminforsch., 6, 86 (1937). Bessey, 0. A., J. Biol. Chem., 126, 771 (1938). Cozzi, D., Ann. chim. opplieata, 29,434 (1939). Davis, W. B., IND.ENG.CHEM.,34, 217 (1942). Harris, L . J., Mapson, L. W., and Wang, Y . L.. Biochem. J., 36, 183 (1942). ( 6 ) Kirk, M . M.,and Tressler, D. K.,
[email protected]., ASAL. ED., 11, 322 (1939). (7) Kodicek, E., and Wenig, K., Nature, 142, 35 (1938). (8) Lewis, W. R., unpublished data, 1943. (9) Lingane, J. J., IND.ENG.CHEM., A N A L .ED., 15, 583 (1943). (10) ,Muller, 0. H . , J . Chem. Education, 18, 176 (1941). (11) Roe, J. H., and Kuether, C. A , , J . B i d . Chem., 147, 399 (1912). (12) Roe, J. H., and Oesterling. M. J.,Zbid., 152,511 (194.2). J O C R V A L Paper 169, Purdue University Agricultural Experiment Station. (1) (2) (3) (4) (5)