Differential Determination of Vitamin B, Group SABURO FUKUI Department of Industrial Chemistry, Faculty of Engineering, Kyoto Vniversity, hryoto, Japan
A new differential assay of individual components of the vitamin Be group with Sacch. carlsbergensis 4228 has been studied. The general method consists of a combination of the separation of pyridoxaniine with synthetic cation exchanger, alkali-acetone treatment for the decomposition of pyridoxal, and microbiological assays of the fractions. The results of experiments with natural substances were similar to those of Rabinowitz and Snell. The method is useful for the differential determination of the vitamin Bs group in natural materials.
R
ABINOWITZ and Snell ( 5 ) reported that pyridoxal, pyridoxamine, and pyridoxine may be separately assayed using Saccharomyces carlsbergensis, Streptococcus faecalis, and Lactobacillus casei. The author studied a new microbiological differential determination of the vitamin B6 group, which involves the combined use of an adsorption procedure with a cation exchanger and a microbiological assay with Sacch. carlsbergensis. Sacch. carlsbergensis 4228 has been most widely used for the determination of the total vitamin Bs group. Pyridoxal, pyridoxamine, and pyridoxine are almost equally active with this organism (8). Shimizu and Shiba (6) have suggested that pyridoxamine, because it is more basic than other forms of vitamin &,, can be separated by adsorption on the R-COONa type of cation exchange resin. After the separation of pyridoxamine, pyridoxal can be inactivated by treatment with acetone and alkali ( 7 ) . The method for determining the three members of the vitamin B6 group with Sacch. carlsbergensis after fractionation by these treatments has been tested and found useful, inasmuch as it does not require so much reagent for the basal medium as the previous method.
Extraction of Vitamin Bg. For most, materials, acid hydrolysk with 0.055.1- wlfuric acid a t 20 pounds pressure for 3 hours was used. The transforniation of one form of vitamin Be to the other has heeii ,shown to be negligible under these conditions ( 3 ) . Separation of Pyridoxamine. As a n adsorbent, the following tn-o resins were found useful for the separation of pyridoxaminr from other forms of the vitamin Be group.
KH-IB (sodium form), Oda Laboratories, Department of Industrial Chemistry, Kyoto University. The rebin was synthesized by the condensation polymerization of phenoxyacetic acid and fornialdehrde ( 2 ) . Duolite C 60 (sodium form), a phosphonic cation exchanger of the Chemical Process Co., Redwood City, Calif. The exchange tube, made of brown glass 0.7 to 0.8 cm. in diameter, \vas filled with 1.5 grams of 40- to 60-mesh resin. After the extraction, the filtrate was adjusted to p H 5.0 and the volume was made to 200 ml. (Fraction -4). ;in aliquot-for example, 50 m1.-containing 0.05 to 50 of each form of vitamin Be, was allowed to pass through the resin column a t the rate of 1 to 2 ml. per minute a t pH 6 to 7. Then the column was a-ashed dova with about 50 ml. of cold water in the case of KH4B or hot water (80’ C.) in the case of Duolite C 60. A11 effluents (effluent and washing) were mixed together 9,
PROCEDURE
The general method is shown in Scheme I. Scheme I.
Table I.
General Method Fraction Standard used
1,
Value measured
(hydrolysis with 0.055 A- HzSOI a t 20 lb., for 3 hours) Extraction
(Pyridoxamine, pyridoxal, and pyridoxine)
(Pyridoxal, Pyridoxine) Acetone, alkali treatment
(Pyridoxamine)
Calculation of Values of Three Forms of Vitamin Be Total (Fraction A) Pyridoxal HCl a
Acetone-Blkali Treated Soln. (Fraction C) PyfidoxPyrime doxal HCI HC1 C b
Pyridoxine hydrochloride = b Pyridoxamine 2 hydrochloride = d Pyridoxal hydrochloride = a - (c
Eluate (Fraction D) PyridoxPyriamine doxal HCl 2HCl
e
d
+ e)
Table 11. Separation of Pyridoxamine from Pyridoxine and Pyridoxal by -4dsorption with Sodium Salt of Duolite C 60 [50 ml. of a n aqueous solution of vitamin Bb, containing 1 t o 10y oi pyridoxamine dihydrochloride, pyridoxal hydrochloride, and pyridoxine hydrochloride was passed through t h e resin column (resin 1.5 grams) washed with do ml. of hot water, a n d eluted with 50 ml. of a hot 0.2A’ NaOH-N KCl solution.] Vitamin Recovery, % Effluent plus Vitamin Ba Vitamin, y washinz Eluate Pyridoxamine 10 0 98.0 2HC1 5 0 101.0 1 0 98.0 Pyridoxal HCl 10 100.0 0
Pyridoxine HCI
5 1
101.2 101.0
0 0
10
100.0 98.5 99.0
0 0
5
1
treated solution
0
Microbiological assays were run with 5 , 10, and 20 rnillirnicrograrns of the vitamin.
(Pyridoxine)
1884
1885
V O L U M E 2 5 , NO. 12, D E C E M B E R 1 9 5 3 and made up to a definite volume (Fraction B). The solution was considered to contain pyridoxal and pyridoxine. Pyridoxamine adsorbed on the resin was eluted with 50 or 100 ml. of a hot 0.2N sodium hydroxide-11%- potassium chloride aolution a t 80' C. a t the rate of 1 ml. per minute. The eluate was neutralized with 1N hydrochloric acid and made up to a definite volume (Fraction D). Destruction of Pyridoxal. According to Snell ( 7 ) , pyridoxal is inactivated in solution by treatment with equal volumes of acetone and 5 s sodium hydroxide. In this experiment, an aliquot' of the effluent (Fraction B), containing more than 20 millimicrograms of pyridoxine and appropriate amounts of pyridoxal, was mixed with 0.5 volume of acetone and 5N sodium hydroxide, respectively, alloffed to stand for 4 hours a t room temperature, neutralized with hydrochloric acid, and made up to a definite volume, after removal of acetone under vacuum (Fraction C). Microbiological Assay of Vitamin Bs. The vitamin contents of the fractions were determined with Sncch. carlsbergensis according to the nwthod of Atkin et al. ( f ) , modified by Rabinowitz and Snell ($1. The stunple solution for microbiological away was prepared from rat-h fraction to contain 1 to 20 niilliniicrograms. The averages of the values measured with the sample solutions diluted iu three wags from earh fraction were indicated as the amounts of the vitamin in the fraction. F3ec.su.e the responses of Sacch. ccidsberyensis to pyridoxal
Table 111.
RESULTS AND DISCUSSION
The results of the tests for the separation of pyridoxamine from pyridoxal and pyridoxine by the adsorption are shown in Table 11. Pyridoxamine is adsorbed on the resin, while neither pyridoxine nor pyridoxal is adsorbed. It would be expected that pyridoxamine could be separated from the others by this procedure. When applied to natural products, however, some loss of pyridoxamine may occasionally be observed. This might be accounted for by interferences of adsorbed contaminating substances. In such cases, better results should be obtained by passing the influent through the two resin column^ successively.
Table VI.
Separation of Pyridoxal from Pyridoxine by Acetone-Alkali Treatment Ainounts of Vitamin Treated,
Vitamin E6 Pyridoxal HC1
Destruction of Pyridoxal, a.
Y
/a
5 1 0.1
Pyridoxine, HCI
Pyridoxine Recovery,
100 0 100.2
... ...
5 1
0.1
...
0.05
95.0 100.0 105.1 102.0 97.5 97.0 102.5 99.0
5+5 102.5 Pyridoxal HCl plus 1 + 1 103 0 Pyridoxine 0.1 + O . l 97.5 0.05 0.05 101.0 HC1 Microbiological assays were carried out with 2 and 4 millimicrograms of vitamins.
+
Recoveries of Pyridoxal, Pyridoxamine, and Pyridoxine
Fraction Effluent (Fraction B) Acetone-alkalitreated solution (Fraction C) Eluate (Fraction
Recoveries of Vitamins iidded, % StandardMean Deviation
Vitamins . ._ _.......
Expected Pyridoxal, pyridoxine
104,3
5
5.3
Pyridoxine
99.8
7.5
Pyridoxamine
98.2
3.0
Pyridoxal 9 68
Vitamin Bsa pyridoxamine
Pyridoxine
12.9 0.20
0.40 2 28
4.59 9.33 5.98 0.62 7.38 0 17
4 . .56 2 05
0 92 3 97 0 13
26.2 0 42 0.45 2.49 1.04 0.13 0 57 0.57
For extraction, 1 gram of rice bran and 20 grams of fresh spinach were autoclaved with 0.44N sulfuric acid a t 20-pound pressure for 3 hours; 2 grams of pressed yeast, beef liver, beef heart, and hog liver was hydrolyzed nith 0.055N sulfuric acid a t 20-pound pressure for 3 hours; 100 mi. of sake was k e d with concentrated sulfuric acid and water, made up to 180 ml., and after adjustment of the acid concentration to 0.055N, was autoclaved as usual. For the recovery tests, 5r of pyridoxal hydrochloride, 57 of pyridoxamine hydrochloride, and 10-/ of pyridoxine hydrochloride were added to 50 ml. of the extract.
D)
Table V.
Sample Rice bran (air-dried) Spinach (fresh) Baker's yeast (fresh) Fodder yeast (fresh) Beef liver (fresh) Beef heart (fresh) Hog h e r (fresh) Sake Free vitamin B6 base.
All three compounds were adPorbed rimilarly, by the COOH type of KH4B and Duolite C 60 resins. More leakage of pyridoxamine was observed with amberlite IRC 50. The results of experiments on the decomposition of pyridoxal are given in Table 111. By acetone-alkali treatment, pyridoxal was completely decomposed; pyridoxine was not affected. The recoveries of the three members of vitamin Be during the total process are shown in Table IV. The distributions of pyridoxal, pyridoxamine, and pyridoxine in some natural materials are given in Tables V and VI.
Tests were run a t concentrations of 1, 0.2, 0.1,0.02,a n d 0.01~per ml. of ,sample solution.
Fraction A, Pyridoxal HC1 Assay Recoverya, 7% value ((I)
Distribution of Pyridoxal, Pyridoxamine, and Pyridoxine in Natural Materials
%
97.5 99.3
0.05
Table IV.
pyridoxamine, and pyridoxine are not exactly identical, standards for the three forms must be run. The values of the three forms of vitamin Bs were calculated as shown in Table I. Recovery Test. To an aliquot of the extract (Fraction A), definite amounts of pyridoxal, pyridoxamine, and pyridoxine were added. Assays were run just as for the sample solution.
The contents of the three forms of the vitamin were indicated
Differential Assay of Vitamin BF,in Natural Materials
Fraction C Pyridoxine HC1 Assay Recovery, value ( b ) %
Pyridoxal HCl, assay value (e)
Fraction D Pyridoxamine Pyridoxal 2HC1 HCI, Assay Recovery, assay value ($1 7% value (e)
Sample Rice bran 54.3 96 31.5 98 28.3 19.1 Fresh baker's yeast 8.07 108 0.50 103 0.54 6.42 Fresh beef liver 9.80 101 1.06 101 1.25 8.37 Fresh beef heart 1.79 100 0.16 107 0.14 0.87 Vitamin content, micrograms per gram of sample. a Amounts of pyridoxine HC1 and pyridoxamine 2HC1 added were converted to pyridoxal HCI.
97
14.3
Pyridoxal HCl ( a )
Pyridoxine HC1 (b)
Pyridoxamine 2HC1 ( d )
11.7
31.5
19.1
(c
+ e) -
95
4.83
2.74
0.51
6.42
92
6.28
2.46
1,25
8.37
98
0.65
1.00
0.16
0.87
ANALYTICAL CHEMISTRY
1886 as the amounts of free bases (micrograms per gram of material, for sake, milliliters) (Table VI). The distributions of pyridoxal, pyridoxamine, and pyridoxine in the natural materials so far tested are similar to the results of Ribinowitz and Snell. Consequently, the method has been considered practicable, inasmuch as less reagents are required for basal medium than in the previous method. ACKNOWLEDGMENT
The author wishes to express his appreciation to Tadanobu Kishibe and Hiromichi Saito for the technical assistances, to Hiroshi Shimizu for helpful suggestions, and to Ryohei Takata for his encouragement.
LITERATURE CITED
(1) Atkin, L., Schults, A.
S.,Williams, W. L., and Frey, C. N., 1x11. ENG.CHEM.,ANAL.ED., 15,141 (1943). (2) Oda, R., Shimisu, H., and Nakayama, Y., Chem. High Polymers ( J a p a n ) ,5,21 (1948). (3) Rabinowitz, J. C., Nondy, K. I., and Snell, E. E., J . Bid. Chem., 175,147 (1948). (4) Rabinowitz, J. C., and Snell, E. E., ANAL.CHEM.,19,277 (1947). (5) Rabinowitz, J. C., and Snell, E. E., J . Bid. Chem., 176, 1167 (1948). Shimisu, H., and Shiba, H., J . Chem. SOC.( J a p a n ) ,72,442(1951). (7) Snell, E. E., J . Bid. Chem., 157,491 (1945). (8) Snell, E. E., and Rannefeld, 4.N., Ibid., 157, 475 (1945). f6)
RECEIVED for reriew February 9, 1953.
Accepted July 30, 1953.
Quercetin as Colorimetric Reagent for Determination ot iirconium FRANK S. GRIMALDI, United States Geological Survey, Wushington, D . C., AND CHARLES E. WHITE, University of Murylund, College Park, Md.
Methods described in the literature for the determination of zirconium are generally designed for relatively large amounts of this element. A good procedure using colorimetric reagent for the determination of trace amounts is desirable. Quercetin has been found to yield a sensitive color reaction with zirconium suitable for the determination of from 0.1 to 507 of zirconium dioxide. The procedure developed involves the separation of zirconium from interfering elements by precipitation with p-dimethylaminoazophenylarsonic acid prior to its estimation with quercetin. The quercetin reaction is
A
LTHOCGH numerous excellent reagents and procedures are available for the determination of macro amounts of zirconium, the situation is far from satisfactory with respect to the determination of microgram amounts. Only a few reagents have been developed for trace analysis and the reactions are not altogether ideal from the standpoints of sensitivity and selectivity. Alizarin (or Alizarin Red S) (4-7, 9, I O , I S , 1 4 ) is probably the most important reagent for the colorimetric determination of zirconium, but the color reaction is not too sensitive. p-Dimethylaminoazophenylarsonic acid (8, 11, 12) is second in importance to dlizarin Red S. The procedures are indirect, light absorption measurements being made on the dye solution obtained by decomposing the zirconium azo-arsonate with ammonia. Stehney and Safranski ( 1 2 ) determined microgram amounts of zirconium in this manner. Flavonol, introduced by .%lford and coworkers ( I ) , is important in zirconium analysis by fluorescence. This study, made in part on behalf of the Atomic Energy Commission, was undertaken with two objectives in mind: First to find a colorimetric reagent sensitive to small concentrations of zirconium and second, to apply it to the determination of microgram amounts of zirconium in siliceous materials. Quercetin was selected from more than 100 compounds tested, because of its high sensitivity, a nearly colorless blank, stable color over a wide acidity range, and availability in a pure state. EXPERIMENTAL
Factors Affecting the Zirconium-Quercetin Color System. When quercetin is added to an acid solution of zirconium an intense yellow color is obtained. Various factors affecting the
carried out in 0.5N hydrochloric acid solution. Under the operating conditions it is indicated that quercetin forms a 2 to 1 complex with zirconium; however, a 2 to 1 and a 1 to 1 complex can coexist under special conditions. Approximatevalues for the equilibrium constants of the complexes are KI = 0.33 X 10-6 and Kz = 1.3 X lovg. Seven Bureau of Standards samples of glass sands and refractories were analyzed with excellent results. The method described should find considerable application in the analysis of minerals and other materials for macro as well as micro amounts of zirconium.
zirconium-quercetin color system were studied to establish optimum working conditions. Preliminary experiments indicated that a certain amount of alcohol was necessary to prevent the precipitation of quercetin and this variable was included in the study. I n the experiments, all the solutions were made to a total volume of 25 ml. The order of addition of the reagents was always the same. The zirconium solution was added first, acid second, alcohol third, and an alcoholic solution of quercetin last. Absorbancies were determined with a Beckman spectfophotometer, Model DE, uEing 1-cm. cells and distilled water-as reference solution. The slit width was 0.05 mm. except for the spectral transmittancy data below 420 mp where 0.1 mm. was used. Unless otherwise indicated all solutions were 0.5B in hydrochloric acid and contained 3 mg. of quercetin and 8 ml. of alcohol. Spectral transmittancy data for the reagent blank and 307 of zirconium dioxide are given in Figure 1 . The optimum wave length was taken as 440 mp because at this wave length the absorption given by the blank is small and that by zirconium still large. In the work that follows, all absorbancies were measured at 440 mp. Figure 2 illustrates the effect of alcohol concentration on the absorbancy of 54.47 of zirconium dioxide. A precipitate of quercetin was obtained almost immediately from solutions containing less than 6 ml. of alcohol; with 6 ml. of alcohol some quercetin precipitated after 10 hours. It is desirable to keep the alcohol concentration to a minimum because of the smaller solubility of salts in alcoholic media. For this reason 8 ml. of alcohol in 25 ml. of solution was taken as optimum.