V O L U M E 2 4 , NO. 6, J U N E 1 9 5 2
1025
0.9 -
10
20
20
30
30
40 5 0
MICROGRAMS Figure 4.
Quantitative Results
Log & / I is absorbancy of peaks; microgram scale is logarithmic. Adenosine run in ethyl alcohol-ammonium acetate, Readings at 260 mp. Glucose and threonine run in butanol-water, and developed as described by Partridge ( 1 0 ) and Bull e t al. (3). respectively. Readings at 400 mp (glucose) and 570 mp (threonine)
be removed with ether and the paper can be used for other chemical estimations. For instance, measurements can be made before and after treating the chromatogram with bromine vapors, thus distinguishing between uridine and adenosine derivatives. The use of solvents absorbing in the ultraviolet does not interfere with the procedure as long as they can be removed by evaporation or extraction. The authors experienced no difficulties in carrying on chromatograms with phenol, if they extracted the last traces of solvent with ether prior to examination of the paper in the spectrophotometer.
The use of the “spotometer” to obtain quantitative results is rather obvious. In Figure 4 it is shown that the absorbancy of the peaks is proportional, within certain limits, to the logarithm of the amount of substance, provided that the samples were initially deposited on the paper over equal areas. This relationship was shown by Block ( 2 ) to hold for amino acids developed with ninhydrin and by McFarren et al. (8) for sugars detected with ammoniacal silver nitrate. The applicability of the present technique to amino acids and sugars is also shown in Figure 4. One of the difficulties in obtaining quantitative results in direct photometry on filter paper chromatograms is the inhomogeneity of the paper, which introduces a certain amount of uncertainty in the measurements. Whatman No. 1 paper was used throughout. LITERATURE CITED
(1) Berosa. AI.. ANAL.CHEM..22. 1507 (1950) ~. (2j Block, R. J.; Ihid.,22, 1327(1950). 1,8) Bull, H. B., Hahn, J. W., and Baptist, V. H., J . Am. Chem. Soc., 71,550 (1939). (3) Caputto, R., J. Bid. Chem., 189, 801 (1951). (5) Carter, C. E., J . Am. Chern. Soc., 72, 1466 (1950). 16) Holiday, E. R., and Johnson, E. A., Nature, 163, 216 (1949). (7) Hotchkiss, R. D., J . Bid. Chem., 175, 315 (1948). 18) McFarren, E. F., Brand, K., and Rutkowsky, H. R., A N A L . CHEM.,23,1146 (1951). (9) Markham, R., and Smith, J. D., Nature, 163,250 (1949). (10) Partridge, S. M., Ibid., 164,443 (1949). (11) l-ischer, E., and Chargaff, E., J . Bid. Chem., 168, 781 (1947). RECEIVEDfor review September 4, 1951.
Accepted December 12, 1931.
Specific Spot Test for Antimony AND WILLIAhl C. HAMILTON Coates Chemical Laboratories, Louisiana State Uniwersity, Baton Rouge, La.
PHILIP W. WEST
HE procedures heretofore reported for the spot test detection antimony have in most c ‘dses lacked sensitivity, specificity, or both. Rhodamine B, firfit used by Eegriwe ( 1 ) for the detection of antimony, was satisfactorily sensitive, but suffered from a number of interferences. Fluorone (9-methg1-2,3,7-trihydroxy-6fluorone), originally proposed by Wenger, Duckert, and Blancpain (S),proved somewhat superior to Rhodamine B both in sensivity and selectivity, but n-as very difficult to prepare and unstable in solution. Recently JVest and Conrad ( 5 ) have utilized gossypol for the detection of antimony. Their procedure, while superior to any reported prior to it, has qeveral shortcomings: the reagent is not yet commercially available, the acidity of the test solution requires careful control, and certain interferences exist. The authors, in search of a solvent suitsble for the extraction oi the tetraiodoantimonate (111)ion, observed that the yellow color of this complex disappears when acidic solutions of it are shaken with benzene. Further investigation showed that antimony could thus be extrdctcd quantitatively and exclusivelv, and that the benzene extract would react directly with Rhodaniine B to give a sensitive and sperific teht for antimony. REAGEATS
Rhodamine B, 0.2% aqueous solution. Potassium iodide, 10% aqueous solution. Sulfuric acid, 1 to 3 aqueous solutions. EXPERIMENTAL
The general technique followed for the detection of antimony was to perform a benzene extraction of the antimony (111) iodidc, after which Rhodamine B was added to th2 benzene layer as the.
color-developing reagent. The details of the method a w given in the procedure. The Concentrations of the aeveral reagents were varied over wide ranges without adversely affecting the test. Rhodamine B concentrations of 0.02, 0.05, 0.1, and 0.5% were tested. The potassium iodide concentration was varied between 2 and 20Yc. The acidity was varied from 4 N to 14 X without adverse effect. For acidities belop7 4 1%’ the sensitivity of the procedure was lessened, presumably because of poorer extraction, and at acidities much above 14 N the air-oxidation of potassium iodide was so much accelerated that the benzene layer became colored with free iodine, making interpretation of the test difficult. Similarly, higher concentrations of potassium iodide are more unstable and less convenient on that account. Sitrite ion and oxidizing agents int,erfered v i t h the test as clescribed, the former by- virtue of the fact that it gave a strongly fluorescent bluish color on extraction, the latter by the liberation of free iodine, which colored the benzene layer so deeply that the test color was hidden. The interference of nitrite ion was initially eliminated by evaporating the test solution to fumes of sulfur trioxide, then diluting with water to the initial volume and proceeding with the test as usual. Later it was found that this interlerence could be more conveniently circumvented by the addition of a few milligrams of solid urea t o the test solution before addition of the potassium iodide. The interference of oxidizing agents was obviated by the addition of solid sodium sulfite to the acidic solution just prior to the extraction, thus reducing any free iodine which had been liberated. The limit of identification and concentration limit of the test were measured as prescribed by Feigl ( 2 ) . The interference studies followed the procedure of West (4).The ions studied were
1026
ANALYTICAL CHEMISTRY
present in a ratio of antimony of 100 to 1, 10 micrograms of antimony being taken in each case. The scope of the studies was such that approximately one hundred of the stable inorganic ions were studied by possible interfering effects. Those ions studied in general are listed in the earlier article by West ( 4 ) . PROCEDURE
To 1drop of the solution to be tested, in a test tube, are added 5 drops of 1 to 3 sulfuric acid, followed by 1 drop of 10% potassium iodide. The solution is extracted by shaking vigorously with 1ml. of benzene. The benzene layer is removed with a pipet, and placed in the depression of a white spot plate. To the center of this depression is then added 1 drop of a 0.2% solution of Rhodamine B. The violet-colored antimony-Rhodamine B complex, diffusing into the benzene layer, indicates the presence of antimony. If nitrites are not known to be absent, a few milligrams of solid urea should be added to the test solution prior to addition of the potassium iodide. If oxidizing agents are present, solid sodium sulfite is added just prior to the extraction until the color due to free iodine is discharged. A blank of 1 to 3 sulfuric acid treated as above shows only a very faint pinkish hue in the final benzene layer. The limit of identification is 0.2 microgram of antimony. The concentration limit is essentially unrestricted, owing to the use of the extraction technique, which permits significant concentrating of the antimony into the benzene layer. REMARKS
This work illustrates the fact that the hope for specificity in spot test methods of analysis may often lie not onl? in the reagents used, but also in the procedures emplo! ed.
-411prior investigators of the antimony-Rhodamine B reaction had observed that a high concentration of chloride ion was requisite to the success of the test. This was usually obtained by making the test solution strongly acidic with hydrochloric acid. The present authors have not used chloride ion in their procedure. Apparently iodide ion form8 an adequate substitute. The antimony as involved in the authors’ procedure appears to react with Rhodamine B while in the trivalent condition. All investigators in the past have maintained that it was necessary to have antimony in the pentavalent state for the reaction with Rhodamine B to occur. The authors believe that this is true when the test is carried out in an aqueous system, but not when their extraction procedure is employed. They have obtained conclusive evidence that in their procedure antimony is extracted into the bnrzenr in the trivalent state and presumably reacts as such. ACKNOWLEDGMENT
The authors acknowledge the aid of the Office of Naval Research, under Those program this research was conducted. LITERATURE CITED
(1) Eegrine, E., 2. m a l . Chem., 70, 400 (1927). ( 2 ) Feigl, F , “Qualitative Analysis by Spot Tests,” 3rd ed., p. 4, S e w York, Elsevier Publishing Co., 1946. (3) Wenger, P., Duckert, R., and Blancpain, C. P., HcEu. Chim. Acta, 20, 1427-45 (1937). (4) West, P. W., J Chenz. Education, 18, 528 (1941). (5) West, P. W., and Conrad, L. J., A N A I .CHEM.,22, 1336 (1950). HEci:I!-t.u
for ieview .June 8 , 1951. Accepted August 1, 1951
Dimethylglyoxime for Determination of Nickel in large Amounts E. L. BICKERDIKE, Santa Barbara College, University of California, Santa Barbara, Calif., A N D H. H. WILLARD, C’niwersity of Michigan, Ann Arbor, Mich. HE gravimetric determination of nickel as nickel dimethylTglyoxime is generally limited to samples containing not more than 30 mg. of nickel because of the great bulkiness of the precipitate generally formed ( I , 3). I t is possible, however, by changing the conditions of precipitation, to handle nickel dimethylglyoxime precipitates containing up to 100 mg. of nickel easily and with satisfactory accuracy, This method uses the principle of precipitation from a homogeneous solution developed by Willard (4, 5 ) . The precipitate of nickel dimethylglyoxime obtained from a homogeneous solution is coarsely crystalline, compact, and darker in color, and does not adhere to glass, in contrast to the precipitate of nickel dimethylglyoxime as it is usually prepared. The precipitate may be completely and easily washed from glass surfaces without “cropping.” It is easily and rapidly filtered and washed on a coarse filtering crucible. I t has a tendency toward creeping while being washed. Sterling ( 2 ) also gives a method for determining large amounts of nickel by precipitation w the oxime. Nickel dimethylglyoxime is precipitated by heating an acidified solution containing nickel, urea, and a sufficient amount of a 1% solution of dimethylglyoxime in 1-propanol to precipitate the nickel. Heating for an hour is required to hydrolyze the urea and to raise the p H to the point of complete precipitation of the nickel dimethylglyoxime. The use of 1-propanol as the solvent for the dimethylglyoxime is necessary in order to prevent evaporation of the solvent during heating, with subsequent loss of precipitating reagent due to its coming out of solution. This would result in the precipitates containing dimethylglyoxime as an impurity. The precipitate is read) foi filtering as soon as it has cooled to room temperature.
The following pioceduie \+ill work on 3nmples containing approvimately 100 mg. of nickel. Dilute the solution to 200 nil. and adjust the pH to 2 to 3 with hydrochloric acid. Add 20 grams of urea. Add 50 ml. of a hot (50’ C.) solution of dimethylglyoxime in 1-propanol (1 gram of reagent per 100 ml. of solvent), and heat on a water bath for an hour. Cover the beaker or Erlenmeyer flask with a watch glass to prevent excessive loss of solvent. At the end of this time the solution should be alkaline to litmus. Add a few drops of the dimethylglyoxime solution to test for completeneas of precipitation. Cool to room temperature and filter into a coarse filtering crucible (Selas No. 2010). %-ash with cold watef and dry to constant weight a t 110” to 120” C. The nickel dimethylglyoxime contains 20.32%nickel. This procedure was tested on samples containing cobalt-free nickel sulfate. Five samples containing 0.0996 gram of nickel, determined electrolytically, yielded the following results: 0.0996, 0.0997, 0.0995, 0.0996, and 0.0996 gram; average, 0.0996 gram. LITERATURE CITED
( 1 ) Rieman, W., Neuss, J. D., and Naiman, B., “Quantitative Analysis,:’ 3rd ed., p. 367, New York, MoGraw-Hill Book Co., 1981. (2)
Scott-Furman, “Standard Methods of Chemioal Analysis,” Vol. 1, 5th ed., footnote p. 619, New York, D. Van Nostrand Co., 1939.
Willard, H. H., and Diehl, H., “Advanced Quantitative Analysis,” p. 384,New York, D. Van Nostrand Co., 1950. (4) Willard, H. H., and Fogg, H. C., J. Am. C h a . SOC.,59, 1197
(3)
(1937).
(5)
Willard, H. H., and Tang, S . K., IND.ENB.CREM.,ANAL.ED.,9, 357 (1937).
RECEIVED for review October 10, 1951. Accepted December 17, 1951.