Solvent Extraction and Anion Exchange Chromatography for the Spectrophotometric Determination of Calcium Pantothenate in Pharmaceutical Products THAVIL PANALAKS and J. A. CAMPBELL Food and Drug Laboratories, Department of National Health and Welfare, Ottawa, Canada
b A method was developed to separate calcium pantothenate from interfering substances in pharmaceutical multivitamin products. The solvent extraction procedure for d-panthenol was modified, following which the calcium pantothenate was chromatographed and hydrolyzed, and the determination was based on the reaction of the hydrolysis product with 1,2-naphthoquinone-4-sulfonote. The method was evaluated by recovery tests of calcium pantothenate added to complex multivitamin-mineral preparations, and b y comparisons with a microbiological method. Mean recovery of 101% with a relative standard deviation of * 5 % was found, and the results were in good agreement with those obtained b y the microbiological method. The method was specific, as judged b y the lack of appreciable interference from biological extracts, free amino acids, and other substances present in multivitamin preparations.
A
the only method generally accepted for the determination of calcium pantothenate in pharmaceutical preparations is the microbiological procedure ( 5 ) . Although a number of chemical procedures (6-9) based on a color reaction of one of the products of hydrolysis of calcium pantothenate have been reported, they are limited to certain types of pharmaceutical products. Attempts were made in this laboratory to remove the interfering substances by using Dowes 50-X4(H) and Florisil as recommended by Schmall and Wollish ( 7 ) , and by using direct chromatography of calcium pantothenate on a column of activated alumina as reported by Szalkoa-ski, Mader, and Frediani (9) and Crokaert (1). Xone of the methods yielded quantitative results when applied to pharmaceutical products, especially those containing high concentrations of biological extracts. A method suitable for the purification of samples containing highly complex mixtures of interfering substances has been developed. It utilizes a modificaT PRESENT,
64
ANALYTICAL CHEMISTRY
tion of the solvent extraction system used for d-panthenol [2,Cdhydroxy-N(3 - hydroxypropyl) - 3, 3 -dimethylbutyramide] (4), and a new application of anionic exchange chromatography for the final purification. The color reaction is based on Frame, Russell, and W-ilhelmi’s modification (3) of Folin’s method (2) for amino nitrogen with a product of hydrolysis of the pantothenate, @alanine. The reaction may readily be applied to routine analysis. EXPERIMENTAL
Reagents. Type I.
Amberlite
CG-400,
BORATEBUFFER.Dissolve 25 grams of boric acid in about 900 ml. of 0.3.Y sodium hydroxide. Adjust the p H to 11.0 with 2N sodium hydroxide, and make up to 1 liter. Prepare fresh weekly. FORMALDEHYDE REAGENT. Mix, by volume, 3 parts of 6N hydrochloric acid, 4 parts of glacial acetic acid, and 4 parts of 0.6M formaldehyde. NAPHTHOQUINOKE REAGENT. Dissolve 250 mg. of the sodium salt of 1,2-naphthoquinone-4-sulfonicacid in distilled water, and make up to 50 ml. Prepare fresh daily. PANTOTHENATE STANDARD. Dissolve 150 mg. of calcium d-pantothenate, U.S.P. reference standard, in distilled water, and make up to 500 ml. Prepare fresh daily. Apparatus. Chromatographic glass tubes, approximately 10 mm. in diameter and 30 em. in length, fitted with a stopcock. Procedure. Plug the bottom of the chromatographic tube with a small pledget of glass wool, add the Amberlite, which has been mixed &-ith distilled water, and allow i t to settle to a height of 5 em. Cover with a second pledget of glass wool. Wash the column with approximately 35 ml. of 10% sodium chloride solution, and with approximately 100 ml. of distilled water. Before use, activate the column by passing through it approximately 45 ml. of 5y0 boric acid solution. [The Amberlite column may be re-activated after use by successive washings each with approximately 35 ml. of 0.2M phosphate (iVaaP04 12H20) solution, 10% sodium chloride solution, followed by distilled water and 5y0 boric acid solution as indicated above.]
Pipet 10 ml. of the aqueous extract of a sample containing approximately 3 mg. of calcium pantothenate into a 40-ml. glass-stoppered centrifuge tube containing 13 grams of sodium dihydrogen phosphate (NaH2P04.HzO). Shake on a mechanical shaker for 15 minutes. S d d 20 ml. of benzyl alcohol, stopper, and shake for another 15 minutes. Centrifuge to break the emulsion. Pipet a 15-ml. aliquot of the upper benzyl alcohol layer into another centrifuge tube containing 10 ml. of toluene, and add 12 ml. of 5y0boric acid solution. Shake for 15 minutes, and centrifuge. Chromatograph 10 ml. of the lower boric acid layer on an activated Amberlite column. Allow the solution to pass through the column by gravity. Wash the column twice with 15 mi. of 5% boric acid solution, and once with 15 ml. each of 7oy0(by volume) acetone, 0.01S sodium hydroxide, and distilled water. Elute the chromatographed pantothenate from the column with 15 ml. of borate buffer, and collect the effluent in a 25-ml. volumetric flask. Autoclave the flask containing the effluent a t 15 pounds for 30 minutes. Cool to room temperature, and adjust the p H of the solution to 8.0 by adding a predetermined volume of 3N hydrochloric acid. Add 1 ml. of naphthoquinone reagent. Steam the flask in an autoclave a t 100’ C. for 10 minutes. Cool to room temperature, and add 1 ml. of formaldehyde reagent and 1 ml. of 0.1N sodium thiosulfate solution. Let stand for 10 to 30 minutes, and make to a volume of 25 ml. with distilled water. Read the absorbance at 465 mM in a spectrophotometer, setting the instrument a t zero with a reagent blank, prepared by chromatographing 10 ml. of 5% boric acid solution and followed by subsequent treatments in the same manner as that of the sample. For each series of determinations treat a standard solution in the same r a g as the sample solution. RESULTS AND DISCUSSION
The solvent extraction system of Panalaks and Campbell (4) for dpanthenol was unsatisfactory when applied to calcium pantothenate in samples containing high concentrations of interfering substances, such as yeast and rice bran extracts. Direct application of this procedure in combina-
tion with the anion exchange resin Amberlite CG-400 lowered the sensitivity of the naphthoquinone method and also gave too high apparent recovery of added calcium pantothenate. When distilled water was replaced by 5% boric acid solution, the apparent recovery was somewhat improved. It was necessary to replace ammonium sulfate with sodium dihydrogen phosphate for saturation of the aqueous solution before benzyl alcohol extraction, and also to use boric acid solution in place of distilled water. Sodium dihydrogen phosphate was chosen as a salting-out agent among various salts shown in Table I because of the high partition coefficient for calcium pantothenate alone, and also for pantothenate in the presence of a complex multivitamin preparation containing yeast and rice bran extracts. The partition coefficient was cdculated from the ratio of the concentrations in the organic and in the first aqueous phases in the first extraction. The pantothenate content in the organic phase was estimated by the general procedure. For the purposes of comparison it was assumed that stripping of the pantothenate by the boric acid solution, elution from the Amberlite column, and hydrolysis of the vitamin were complete. The concentration in the aqueous phase was obtained by difference from the total amount of 5 mg. of calcium pantothenate originally added. Boric acid solution was used in the second partition because it was not only an effective stripping agent for the slightly basic calcium pantothenate, but vias also the medium in which the pantothenate was later t o be chromatographed on the borated Am berlite column. Experiments were conducted to determine the effectiveness of the AmberIite column for removal of interfering substances present in multivitamin preparations. Two solutions were pre-
Table I. Effect of Various Salting-Out Agents on Partition Coefficient of Calcium d-Pantothenate in Pure Solution and in Solution of Complex Multivitamin Preparation Containing Yeast and Rice Bran Extracts
Salting-Out Agent SaH2P04.10 H20
KHsPO4 KzHPOd
Partition Coefficient Preparation" Ca CaPan. Pan. 1.25 1.28 0.35 0.43 0.19 0.41 0.01 0.02 0.09 0.16
+
NaaPOa.12 HzO (",)zSO, NaCl 0 04 0.20 Sample 1 given in Table 111.
Table II. Relative Reactivity of Calcium d-Pantothenate and Amino Acids Determined b y Proposed Method
pn
OF
ELUTING BORATE
BUFF'ER
Figure 1. Absorbance of reaction products o f calcium d-pantothenate and sample coniaining yeast and rice bran extracts Determined after hydrolysis in HCI A. Ca Pan., 2 mg. 6. Sample, 2 ml. Determined after hydrolysis in eluting buffer C. Ca Pan., 2 mg. D. Sample, 2 ml.
pared, one containing the standard calcium d-pantothenate, the other containing B vitamins (including d-panthenol but no calcium pantothenate), and yeast and rice bran extracts. The absorbance of the products of reaction is shown in Figure 1. Curves A and B represent the absorbance of reaction products of standard and sample solution which have been chromatographed, washed, eluted with the borate buffer at p H 4.4 t o 10.0 hydrolyzed in HC1 (autoclaved at 15 pounds for 30 minutes), and reacted with the naphthoquinone. The products of hydrolysis of the standard and the sample could not be differentiated on this basis. On the other hand, elution with the borate buffer a t pH 9.0 to 11.0 and hydrolysis of the effluent without further pH adjustment resulted in no appreciable absorbance of the reaction product of the interfering substances (curve D). The color reaction of the hydrolysis product of the pantothenate (curve C) was a direct function of the p H of the eluting buffer. When elution and hydrolysis were carried out a t p H higher than 11.0, however, it was no longer possible t o differentiate between the standard and the sample. On the basis of these results, it was decided to use a borate buffer of pH 11.0 as a n eluting agent as well as a hydrolyzing medium for the pantothenate. Although a maximum degree of hydrolysis of the vitamin was not reached at this pH, the effect of the interfering factors, which gave a positive naphthoquinone reaction, mas minimized. A study mas made to determine the optimum pH for the naphthoquinone reaction with the hydrolytic product of calcium panthenate. Under the experimental conditions, this was found to be p H 8.0, in contrast to p H 8.8 to 9.5 for d-panthenol ( d ) , and p H 9.3, reported by Schmall and Wollish (7). Since the naphthoquinone reagent was originally intended t o be applied t o the determination of amino acids (d), i t
Relative Compound Reactivity" Calcium d-pantothenate 10,000 Neutral amino acids DL-Valine 0 L-Leucine 0 DL-Isoleucine 1.3 DrcMethionine 0.52 DL-Phenylalanine 0.52 DbThreonine 0.52 Basic amino acids ~ l l r g i n i n emonohydrochloride 0 L-Lysine monohydrochloride 0 Acidic amino acids ~ ~ B s p a r tacid ic 8.9 DL-Glutamic acid 17 0 Based on determinations of IO-ml. solutions containing 3.0 mg. of calcium d-pantothenate and 300 mg. of amino acids. Relative reactivity expressed as ratio of absorbance per mg. of compound. Table 111. Recovery of Calcium dPantothenate Added to Complex Vitamin Preparations
Re-
covery,
cSample" /U B vitamins,b clpanthenol yeast, and rice bran extracts 102 2. Sirup, B vitamins,b d-panthenol, liver extract 99 3. Coated tablets, B vitam i q b B,C,D, d-pan106 thenol, minerals 4. Gelatin capsules, B vitamins,b A,C,D, d-panthenol, minerals 101 5 . Sirup, B vitamins,b A,C,D, d-panthenol 96 6. Liquid, B vitamins,b dpanthenol, lipotropic factors 99 7. Liquid, B vitamins,* glycerophosphates, minerals, 106 liver extract 8. Gel?tin capsules, B vitam i q b C, citrus bioflavonoids, d - panthenol, 96 desiccated liver 9. Coated tablets, B vitamins,b C,D 105 10. Gelatin capsules, B vitamins,b C, intrinsic factor 94 concentrate 11. Gelatin capsules, B vitamins,* A,C,D, d-panthenol, minerals 110 Mean 101 a Composition given by label declaration; 80 mg. of pantothenate added to 100 ml. of liquid samples or aqueous extracts of 10 tablets or capsules. * Including three or more of vitamins: biotin, riboflavin, folic acid, thiamine hydrochloride or mononitrate, niacin or niacinamide, pyridoxine hydrochloride, and vitamin B12.
1. Sirup,
VOL. 34, NO. 1, JANUARY 1962
65
Table IV.
of interfering substances, which did not appreciably affect the results obtained by the microbiological assays and were removed in the proposed method. The results of analysis by the proposed method were generally slightly lower than those obtained by the microbiological procedure. This difference may be due to deterioration of the vitamin, since the determinations by the present method were conducted about 6 months after the microbiological assays were completed.
Calcium d-Pantothenate Content of Multivitamin Preparations
Coated tablets, B vitamins Liquid injectable, B vitamins Coated tablets, B vitamins, liver fraction, minerals Coated tablets, B vitamins, C,D, minerals, yeast, and liver extracts Coated tablets, B vitamins, C,E, minerals, whole desiccated liver Coated tablets, B vitamins, minerals, liver extract a Applied about 6 months after other methods.
was of interest to test the effect of free amino acids as possible interfering substances in the general procedure. Results shown in Table I1 indicated only traces or no reaction of neutral and basic amino acids. The acidic amino acids exhibited only slight reaction by this method, and the responses were vc-ell within the range of experimental error. The results of recovery tests (Table 111) for calcium pantothenate, added t o complex multivitaniin products, indicated a mean recovery of l O l % , rT-ith a relative standard deviation of &5%. The presence of amines and pyridines as represented by the group of B vitamins together with other substances frequently present in vitamin preparations caused no interference in the method. This mean recovery was approximately the same degree of accuracy and precision as that of a similar
Assay Value, Mg. per Tablet or M1. Solvent extraction Schmall and Microand anion biological Wollish exchange“ assay (6) (7) 1.32 1.50 3 0 1.13 1.34 5 0 0 58
0.97
7.5
0.65
0.66
2.6
0.21
0.20
1.o
1.48
1.80
3.1
ACKNOWLEDGMENT
The authors acknowledge the assistance of H. 6. Lakke Gowda, a Colombo Plan Fellow from the Public Health Institute, Bangalore, India, in some of the analyses.
method for d-panthenol ( 4 ) . Precision of the method was further evaluated by making determinations of 10 aliquots of a standard solution and 10 aliquots of a liquid sample containing a complex vitamin mixture, including d-panthenol together with a n addition of calcium pantothenate a t a level of 0.4 mg. per ml. Relative standard deviations obtained n ere 5.0 and 2.2%, respectively. The results of analysis by the proposed method and by the U.S.P. microbiological procedure ( 5 ) , together a-ith those obtained by the method of Schmall and Wollish ( 7 ) , are compared in Table IV. The values obtained by the present method agreed closely with those by the microbiological procedure, but not with those obtained by the method of Schmall and Wollish, \\-hich are much higher. This was presumably due t o the presence in the samples
LITERATURE CITED
(1) Crokaert, R., “Contribution A 1’6tude
de la a-alamine et des ses compos& dans les milieux biologiques,” Acta Medica Belgica, Brussels,-l953. ( 2 ) Folin, 0.. J . Bzol. Cheiii. 51, 377 nnn ,1
\
JIYLLJ.
( 3 ) Frame. K,G.. Russell. J. .\.. Kilhelmi. A. E., Idid., 149, 255 (1943). \ - ,
(4) Panalaks, T.) Campbell, J. -\., -1s.~~. CHEU.33, 1038 (1961). ( 5 ) “Pharmacopeia of the United States of hmerica.” XVI, v. 871, Mack Publishing Co., Easton, P a . , 1960 (6) Schinall, 11.. Wollish, E. G., Bx.4~. CHEX.22, 1033 (1950). ( 7 ) Ibzd., 29, 1509 (1957). (8) Sxalkowki> C. R., Davidson, J. H., Ibid., 2 5 , 1192 (1953).
(9) Szalkowski, C. R., Mader, JY. J., Frediani. H -1..Cereal Chevi. 28. 218 (1951).
RECEIVEDfor review July 21, 1961. Accepted October 16, 1961.
.
I
Use of Gas Chromatography for Rapid Determination of Carbonate at Low Levels FRANK G. CARPENTER National Bureau o f Standards, Washington,
D. C.
b Carbon dioxide was evolved with acid in a closed system, a known fraction expanded into a gas buret, and the amount determined b y gas chromatography. The procedure is about 100 times more sensitive than gravimetric or volumetric methods. A detection limit of 5 pM or 0.2 p.p.m. is readily obtained. The time required is about 10 minutes per determination.
measure. I n a gravimetric method the lowest readily measured weight change is 0.1 mg., which corresponds to 2 pmoles of COZ. I n a volumetric determination the smallest readily measured volume change is about 0.05 ml., which also corresponds t o 2 pnioles. Both methods can be refined to yield greater sensitivity, but only a t the expense of ease and speed of manipulation. On the other hand, gas chromatography readily offers a sensitivity 100 times better without any special refinements.
I
determination of carbonate a t concentrations below about 10 p.p.m. the amount of carbon dioxide that is liberated fromareasonablesample size is so small as to be very difficult t o N THE
66
ANALYTICAL CHEMISTRY
METHOD OF ANALYSIS
The procedure, in brief, consists of evolving the carbon dioxide with acid
in a closed system, expanding a fraction into a gas buret, and then measuring the amount of gas chromatography. The evolution flask (125 ml.) was attached directly to the gas chromatography apparatus as shown in Figure 1. After the sample to be analyzed (solid or liquid) was placed in the flask and attached to the apparatus, any residual carbon dioxide that might have been in the air in the flask was displaced by sweeping with C02-freegas for 5 seconds at a flow rate of about 100 ml. per second. The flask was then evacuated to about 0.5 atm. and 10 ml. of 3N HC1 was added. The contents mere stirred a t a high rate for 5 minutes. I n the meantime the spaces marked