Direct Determination of Niacinamide in Multivitamin Preparations By 0. PELLETIER and 1. A. CAMPBELL A simple method for the direct determination of niacinamide in pharmaceutical preparations is described. T h e method involves the reaction of niacinamide in potassium dihydrogen phosphate with cyanogen bromide and barbituric acid. T h e resulting purple color has a maximum absorbance at 5 5 0 mp. Niacin does not interfere in the reaction unless present at about three times the concentration of that of the amide. T h e method is accurate and reproducible. Results o n multivitamin preparations compare favorably with those by the A.O.A.C. procedure and recoveries of added niacinamide are quantitative. volume required for each batch of assays by adding 100 ml. of 37, KH2P04to each 2 Gm. of reagent grade barbituric acid; let stand for one hour with occasional shaking; filter before using. SoluNiacinamide Standard Solutions.-Stock tion.-Weigh 50 mg. U. S. P. Niciriamide Reference Standard, dissolve in 60% ethyl alcohol, and make up to 200 ml. Store in refrigerator. Working Standard Solution.-Pipet 2 nil. of stock solution into a 100-ml. volumetric flask and make to volume with 0.3% KH2P04. This solution contains 5 mcg./ml. Assay Solution.-Place a measured amount of sample in a flask and add a volume of 0.3% KHzPOl equal in ml. to a t least twice the amount of niacinamide in mg. Tablets should be ground first. For capsules, add about 2 ml. of ethylene dichloride to aid in dispersion. If sample is not readily soluble, shake and heat so that it may be evenly dispersed. Dilute t o about 5 mcg. per ml. with 0.3% KH2P04. Filter if necessary. Reaction.-To 1 ml. of dilute standard or assay solution in colorimeter tube, add 0.5 ml. cyanogen bromide, mix, stopper, and let stand twenty-five to thirty minutes. Add 10 ml. barbituric acid and mix by swirling. (If barbituric acid cannot be added a t this time, transfer tubes to a bath of crushed ice to stabilize cyanogen bromide reaction.) Set spectrophotometer a t zero absorbance a t 550 mp with a standard blank in which cyanogen bromide is replaced by water and measure maximum absorbance of reaction product in two to four minutes after addition of barbituric acid. Prepare sample blanks by replacing cyanogen bromide with water. METHOD Calculations.-Calculate on basis of aliquots taken as follows Apparatus.-Suitable photoelectric colorimeter or spectrophotometer. mg. per unit sample = Au - X __ 5 x As 1,000 Reagents.-Cyanogen bromide solution, 10yo, dilution factor prepare and use under hood; do not let cyanogen bromide or solution come in contact with skin; where A u = corrected absorbance of unknown and dissolve a weighed amount of cyanogen bromide in A s = absorbance of standard. warm water a t about 40’; allow to come to room temperature and dilute to 10% solution; store in EXPERIMENTAL RESULTS refrigerator and bring to room temperature before using. Potassium dihydrogen phosphate, 3% (KH2Selection of Extracting Medium.-Hydrolysis of POa); dissolve 30 Gm. in water to make 1,000 ml. niacinamide occurred in most extracting solutions Potassium dihydrogen phosphate, 0.3yo; prepared by tested, including very dilute solutions of acid and diluting 374, KH2P041 to 10 with water. Barbituric solutions of dibasic and tribasic phosphate. Howacid bufe‘ered solution, saturated, (2y6) ; prepare ever, KHzPOl did not cause hydrolysis and therefore appeared to be the most suitable extractant. Received February 15, 1961, from the Food and Drug Laboratories, Department of National Health and Welfare, It was found that maximum absorbance was proOttawa, Ontario, Canada. duced with a concentration of about 0.3% KH2P04. Accepted for publication March B, 1961.
rather than niacin is generally used in multivitamin preparations because of the possible undesirable side effects of the latter. Although the chemical estimation of total niacin-niacinamide after hydrolysis is relatively simple (1, a),the methods are not specific for the amide. Chromatographic methods ( 3 , 4) for the differentiation of the two compounds are timeconsuming and other methods (5-9) are too complex for routine use. There is obviously a need for a specific method for the direct estimation of niacinamide. An approach t o this problem was made b y Lisboa (10) who demonstrated t h a t niacin and niacinamide in Sorensen’s buffer at pH ’7.2 reacted differently with barbituric acid and cyanogen bromide. He determined niacinamide directly in the presence of niacin by measuring t h e resulting absorbance at 513 mp. Preliminary tests of Lisboa’s method indicated that i t lacked sensitivity and did not follow Beer’s law a t higher concentrations. This paper describes factors affecting the reaction of niacinamide with cyanogen bromide and barbituric acid and the applicability of this reaction to the direct estimation of niacinamide in complex multivitamin preparations.
N
IACINAMIDE
926
Vol. SO, No. 11, November 1961
927
TABLEI.-PER CENTRECOVERY OF NIACINAMIDE compared with the A.O.A.C. method (1)in assays and FROM NIACIN-NIACINAMIDE MIXTURES in recovery tests on 13 representative pharmaceutical products. The tablets were ground to a Recovery, Yo fine powder and well mixed. Capsules were opened Niacin7 5 mcg./ml.---10 mcg./ml-Test 1 Test 2 amide, % Test 1 Test 2 and their contents uniformly mixed. Duplicate samples were weighed out. In one, niacinamide 95 100.5 99.5 100.0 99.9 75 100.0 99.5 100.0 100.0 was determined directly; to the other, niacinamide 50 101.0 100.5 100.5 100.7 was added in an amount equivalent to label claim 25 103.5 101.0 102.1 102.4 and then assayed. Results are reported in Table I1 5 111.6 107.5 109.2 111.4 in terms of per cent of label claim or per cent recovery of added niacinamide. In the assay of samples, the difference between the two methods was Factors Affecting the Reaction.-Color develop- small except for a liver extract elixir and a syrup ment is dependent on both the reaction of preparation which gave lower values by the A.O.A.C. niacinamide with cyanogen bromide and on the method. These two samples became extremely coupling with barbituric acid. The effect of various dark upon hydrolysis and yielded very high blanks factors on these two reactions was determined by the A.O.A.C. method. Excluding these two using 10 mcg. niacinamide in 1 nil. solution. Con- products the mean difference between the two methcentrations and volumes of reagents and time and ods was 0.27, and in the recovery test a mean value temperature reaction influenced the amount of color of 99.6 f 3.07, was obtained with the A.O.A.C. formed. The optimum amount of cyanogen bro- method, while the proposed method yielded a mean mide was found to be about 0.5 ml. The cyanogen recovery of 100.2 f 1.4%. bromide reaction was complete in about twenty-five Reproducibility of Assays.-To test and compare minutes at toom temperature and remained stable the reproducibility of assays by both methods, an for about five minutes. By cooling in ice water at elixir vitamin B-compound preparation containing that time, the stability could be maintained for yeast extract was assayed on five different days and over one hour. For color development no advantage results are reported in Table I11 in terms of per was gained by using an amouut of more than 10.0 ml. cent of label claim. Values of 101.3 f 1.27, by the or a concentration of more than 2% barbituric acid. proposed method compared favorably with 102.2 The optimum concentration of KHzPOl as a buffer & 2.5% by the A.O.A.C. method. for barbituric acid was 3%. Specificity of Reaction.-To test the specificity of Concentration of Niacinamide.-There was a linear relation between concentration and absorbance up to 10 to 12.5 mcg. niacinamide per ml. depending on the instrument used. The most suit- TABLEIII.-RESULTS OF REPEATEDASSAYS OF IN A VITAMIN B COMPOUND-YEAST able concentration to use appeared to be 5 or 10 NIACINAMIDE BY PROPOSED METHODA N D ELIXIR PREPARATION mcg. A.O.A.C. METHOD Effect of Niacin on the Determination of Niacinamide.-Mixed solutions of niacin and niacinamide Per Cent of Label Claim, containing 5 to 95y0 niacin were assayed a t the 5 1 . 2 5 mg./ml. Proposed Method A.O.A.C. Method and 10-mcg. levels of niacinamide and results are 103.0 104.7 shown in Table I. There was no significant inter102.5 98.2 ference of niacin up to the 507, level. Results were 100.0 99.2 2.3% higher a t the 757, level and about 10% higher 100.6 102.1 a t the %yolevel where the solutions contained al100.5 102.2 most 20 times as much niacin as niacinamide. Mean 101.3 102.2 Assays and Recoveries of Niacinamide in Phars. n. 1.2 2.5 maceutical Products.l-The proposed method was 7
TABLEII.-AssAY
AND
Multivitamin Preparations
Tablets Tablets Capsules Capsules Tablets (minerals) Capsules (minerals) Capsules (minerals) Tablets (minerals) Tablets (minerals) Injectable Elixir (yeast) Elixir (liver) Syrup (minerals) Mean S. D.
RECOVERY OF NIACINAMIDE IN MULTIVITAMIN PREPARATIONS BY PROPOSED METHODAND A.O.A.C. METHOD Label Content mg./Tab., Capsule, ml.
-Label Claim, yoA.O.A.C. Proposed
20 25 100 150 2.5 10 10 20 50 100 1.25 1.25 1.0
100.0 103.1 100.0 103.3 111.9 115.4 105.3 92.0 94.6 99.2 104.7 101.8 91.0
101.o 100.0 102.0 98.9 112.3 114.5 107.6 95.0 95.3 100.0 103.0 114.7 115.0
...
...
...
...
...
...
Difference
+l.O
-3.1 +2.0 -4.4
+l.t
-0.9 -0.7 +3.0 +O.7 +0.8 -1.7 ($12.91" (+24.0) -0.2
...
---Recovery, A.O.A.C.
100.9 94.1 99.7 99.1 101.3 98.7 99.2 99.5 97.0 99.2 107.2 (104.7) (89 .O) 99.6 3.0
%b-
Proposed
101.0 102.0 99.0 99.6 103.7 99.5 97.7 100.0 100.4 100.0
100.0 100.6 99.0 100.2 1.4
~~
Values in parenthesis not included in mean and S. D.
b
Recovery of an amount added equivalent t o labeled claim.
928
Journal of Pharmaceutical Sciences
TABLE IV.-REACTION OF SOME PYRIDINE DERIVATIVES IN PROPOSED METHODAND IN A.O.A.C. METHOD Color Formed in Proposed Reaction a t Concentration of Derivatives
a-Substituted Picolinic acid a-Picoline 2,4-Lutidine &Substituted Kiacin Niacinamide Niacinaniide HCl Nicotinic acid ethyl ester Nicotinic acid diniethyl amitle Nicotinic acid diethyl amide Pyridine-3-sulfonic acid Nicotinic nitrile Nicotinuric acid Nicotine 7-Substituted Isonicotinic acid Isonicotinic acid amide Isonicotinic acid hydrazide "-Substituted Trigon elline "-Methyl nicotinamide Miscellaneous Pyridoxine HCl Pyridoxamine 2 HCI Pyridine a,cu'-Dipyridyl
10 mcg.
100 mcg.
nil nil nil
nil nil nil
light yellow purple purple purple purple salmon pir,k salmon pink trace pink nil yellowish
light yellow purple purple purple to red red-purple orange to red salmon pink reddish trace pink brownish
nil blue greenish
nil blue greeoish
nil nil nil nil light yellow nil
Per Cent Absorbance Comparison with Niacinam idea Proposed A.O.A.C. Method Method
0 0 0
0 3 0
0 100 100 101 69 16
99 100 97 121 102 241 36 65 6 9
7 6 2 1
0 25 2
3 0 2
nil light purple
0
0 3
nil nil light purple nil
0
2
0 1 0
0 0 40 2
a Per cent absorbance was calculated on an equimolar basis, using concentrations of 10 mcg. for the proposed method and 5 mcg. for the A.O.A.C. method.
the method, several pyridine derivatives were assayed a t 10 and 100 mcg. per test tube. Results are shown in Table IV. The a-substituted derivatives gave no reaction. The @-substituted derivatives varied widely in the color produced but only nicotinic acid ethyl ester and dimethyl amide gave the purple color characteristic of niacinamide. The ?-substituted compounds yielded blue, green, or no color. "-Methyl nicotinamide produced a purple color only a t relatively high concentrations. Pyridine produced a more stable color a t higher concentrations. No reaction was given by the two vitamin Bs compounds. Data on the relative specificity of the two reactions indicate that for the compounds listed the proposed method is a t least as specific as the A.O.A.C. method.
DISCUSSION The proposed method differs from that originally presented by Lisboa in extracting agent, p H of reaction, amounts of reagents, order of adding reagents, and absorbance maximum of reaction. Based on the absorbance per mg. of niacinamide, the proposed was about 10 times as sensitive as the former procedure. Interference of niacin in the recovery of niacinamide was evident only when niacin was present in three times the amount of niacinamide, but even a t that level the effect was not serious. Examination of results published by Lisboa suggested greater interference of niacin a t about the same proportions. Commercial preparations containing the amide ordinarily have little, if any, niacin. Assays and recovery tests with representative preparations indicated that the proposed method compared favor-
ably with the A.O.A.C. method. The proposed method was highly reproducible and was applicable to all preparations while the A.O.A.C. method gave unsatisfactory results with a syrup and a liver elixir preparation because of interfering colors produced by hydrolysis. I n the proposed method sample blanks were often not necessary because dilute assay solutions were clear and contained no interfering color. The reaction with barbituric acid is much more specific for the determination of the amide than the previous methods using hydrolysis. Of the several compounds tested, only nicotinic acid ethyl ester and nicotinic acid dimethyl amide reacted similarly t o niacinamide, but their presence would not be expected in multivitamin preparations. The time required for assay by the proposed method was shorter than by the A.O.A.C. method since hydrolysis and p H adjustments were not required. The main advantages of the proposed method are that it deterniines niacinamide directly and is simpler than available methods.
REFERENCES (1) "Official Methods of Analysis," 8th ed., Association of Official Agricultural Chemists, Washington, D. C., 1955, p. 3845. (2) Mueller, A., and Fox,S . H . , J . B i d . Chem., 167, 291 (1947). ' (3) Wollish, E. G.,Schmall, M , and Shafer, E. G. E., Anal. Chem., 23, 768(1951). (4) Sweeney, J. P.,and Hall, PI. I-., J. Assoc. Oiftc. Agr. Chemists, 36, 1918(1953). (5) Sweenev. _ .-T. P.., and Hall. W. L.. Anal. Chem.. 23. 983 (1951). (6) Lamb, F. W., Ind. Eng. Chem. Anal. E d . , 15, 352 (1943). (7) Melnick D. and Oser B. L. ibid.. 15 355(1943). (8) Ciusa W.Ann. chim. bpgl. j9 93(1969). (9) Chaukhu;i, D.K., and Kodicek, E., Biochem. J . , 44, 343(1949). (10) Lisboa, B. P . , Natzrrwissenschaffen, 44, 617(1957).