Spot Test Detection of Antimony by Means of Gossypol PHILIP W. W E S T AND LOUIS J. CONRAD, Louisiana State University, Baton Rouge, La. PROCEDURE
NUMBER of spot tests for antimony have been reported in
A the literature (4, 6, 7-10, I I ) , but they often leave much to
be desired, because of lack of specificity, poor sensitivity, instability or unavailability of the reagent, or a combination of these. The test which showed the most promise was based on the red color formed between trivalent antimony and 9-methyl-2,3,7trihydroxy-6fluorone. This test, as originally proposed by Wenger and Blancpain ( l a ) ,lacked specificity, but the changes in procedure recently reported by Gillis, Hoste, and Claeys ( I f ) have overcome this objection to some extent. Despite the improvements in the procedure several disadvantages remain: A number of ions still interfere, the reagent is not commercially available, the synthesis of the reagent is not satisfactory, and the solution of the reagent is not stable. Boatner et al. (a) recently reported the use of antimony trichloride for the spectrophotometric determination of gossypol in cottonseed extract, based on the red color of chloroform solutions of the reaction product. The structure of the gossypol as reported by Adams and co-workers ( 1 ) contains two aromatic o-dihydroxy groupings. Proceeding from the gossypol-antimony trichloride spectrophotometric test and the aromatic o-dihydroxy group action displayed toward antimony by such compounds as pyrocatechol, pyrogallol (7, 9)' and 9-methyl-2,3,7-trihydroxy-6 fluorone ( l a ) ,an investigation was undertaken to study the use of gossypol as a spot test reagent for antimony. REAGENTS
Gossypol Solution. Purified goss pol obtained from the Southern Regional Laboratory, New &leans, La., was made up to a strength of 0.1% in reagent grade acetone. Phosphoric Acid Solution. One volume of 85% reagent grade phosphoric acid was diluted with 4 volumes of distilled water. EXPERIMENTAL
The spot plate as a medium for carryin out the test was found t o be far superior to paper. The procefure followed in eneral was to adjust the antimony solution to between 0.4 N an30.7 N with respect to hydrochloric acid and then add this solution to the spot late, followed by 2 drops of gossypol solution per aqueous test &op. The evaluation of the test was made after 30 seconds had elapsed. Antimony in the range of 0.5 to 10 micrograms per drop gave an orange to red preci itate. Greater than 10 micrograms per drop gave a di'stinct r e f precipitate.
The solution to be tested must be acidic (0.4 N to 0.7 N with respect to hydrochloric acid) and should be gently warmed prior to making the spot test. On a spot plate, 1 drop of the test solution is placed and to it is added 1 drop of the phosphoric acid solution, followed by 4 drops of the gossypol solution. An orange or red precipitate indicates the presence of antimony. If the antimony in solution is in the quinquevalent state, it must be reduced to the trivalent form by means of sodium sulfite prior to carrying out the test. REMARKS
The gossypol reaction, when used as a spot test, has a limit of identification of 0.5 microgram of antimony per drop of solution a t a limiting concentration of 1 part in 100,000. KOpositive interferences were found when the interference tests were compared with controls containing 1 microgram of antimony per drop. Vanadate, dichromate, iodate, bromate, perosmic, and molybdate ions give negative or masking interferences. The interfering action of the oxidizing agents is due to the formation of quinquevalent antimony and/or the reaction of gossypol with the excess oxidizing agent. Molybdate, dichromate, vanadate, and perosmic ions react with gosaypol alone to form colored precipitates: molybdate yellow, dichromate brown, and vanadate and perosmic green. Iodate and bromate ions do not form colors with the reagent alone, but in the presence of gossypol and antimony they cause the test color to be a faint green. Attempts to reduce these oxidizing agents in the presence of antimony in the spot plate were not satisfactory. The colors produced in these interfering reactions as well as the inherent colors of such ions as permanganate, chromous, and various platinum metal complexes tend to complicate the interpretations of test results. The phosphoric acid used in the test prevents the interfering effects of ferric, quadrivalent titanium, stannous, stannic, and tungstate ions which would normally give the following colors with the reagent: ferric green, quadrivalent titanium and stannous orange, stannic red, and tungstate reddish brown. Zirconium, fluorides, thiosulfates, oxalates, and tartrates are masking interferences, but the interference due to fluoride can be removed successfully by the addition of boric acid which sequesters the fluoride as the tetrafluoroborate complex. Sulfide is not compatible with antimony. Nitrite ion oxidizes the trivalent antimony to the quinquevalent state which does not form a colored product with gossypol. The excess nitrite ion can be removed by warming the acidic solution and then the quinquevalent antimony can be reduced by the addition of sodium sulfite. The acidity of the test should be controlled carefully. With hydrochloric acid concentrations below 0.4 N the antimony oxy-
The determination of the limiting concentration and limit of identification was performed in accordance with the procedures described by Feigl (6). Interference studies followed in general the procedure discussed by West ( I S ) , except that the concentration of the antimony ion in solution was 0.01% and that of the ions used in the interference studies was 1%. A d d i t i o n a l i n t e r f e r e n c e studies were made against a control Table I. Ions I n v e s t i g a t e d in Interference Studies having a concentration of antimony (Periodio table grouping) in solution of 0.002%. The interference I I1 I11 IV V VI VI1 VI11 studies were carried out on solutions FFe++ Li Be BOz cot - acidified with hydrochloric acid in order Fe+++ c1N8++ + B407-to duplicate test conditions. Clot - co;: K A1 zr+t++ Si cu++ Zn++ clodThe ions investigated in the interfer++ M n + + RuCls-Sr++ Rb IMnOlRhCls--Cd+' La+++ ence studies (Table I ) are given below Br PdClr-Bat+ TI in their more common forms. It is BrOtOSOS - AuClrHg + IrCls--I H g + + realized that in many instances the ions rot - PtCls-ReOC concerned are present as complexes, but where structures of such complexes may be in doubt, only the valence of the central atom is indicated. +
6;:
+
+E:
++++
+p;
++++++
+
1336
e?+-+-+
i$++ ;bh:+ ce+++
Miscellaneous
I337
V O L U M E 22, NO. 10, O C T O B E R 1 9 5 0 chloride begins to precipitate, and above 0.7 N the color formation between the antimony and gossypol is suppressed. Gossypol is reported to be unstable when stored a t room temperature (3): I n order to determine the effect of the decomposition of the gossypol on its use as a spot test reagent, the dry powder and an acetone solution of the material were stored in the’ dark a t an average temperature of 80 F. At the end of 4 months the powder and the solution had darkened somewhat but both were satisfactory for use in the spot test. O
CONCLUSION
Considering all the requirements for a good spot test, gossypol appears to be superior to any other reagent for antimony which has been previously reported in the literature. It is sensitive, highly selective, stable, and readily available. The spot test procedure is very simple, requiring no elaborate conditioning treatments or specialized techniques. ACKNOWLEDGMENT
The authora wish to express their appreciation for financial
assistance given them under a contract with the Office of Naval Research. LITERATURE CITED
(1) Adams, R.,Geissmann, T. A., Dial, W. R., and Fitzpatrick, J. T.,J . Am. Chem. Soc., 63,2439 (1941). (2) Boatner, C. H.,Caravella, M., and Kyame, L., INO. ENO. CHEM.,ANAL.ED.,16,566 (1944). (3) Castillion, L. E., Hall, C. M.. and Boatner, C. H., J . A m . Oil Chemists’ Soc., 25,233 (1948). (4) Denigbs, G.,Chem. Zenlr., 1901, 11, 1214. (5) Eegriwe, E., 2. anal. Chem., 70,400 (1927). (6) Feigl, F.,“Chemistry of Specific, Selective, and Sensitive Reactions,” pp. 8-10, New York, Academic Press, 1949. (7) Feigl, F., Mikrochemie, 1, 74 (1923). (8) Feigl, F., and Neuber, F., Z . anal. Chem.. 62,382 (1923). (9) Feigl, F., and Ordelt, H., Ibid., 64,41 (1924). (10) Fresenius, C.R., Ibid., 1, 444 (1862). (11) Gillis, J., Hoste, J., and Claeys, A., Anal. Chim. Acta, 1, 291 (1947). (12) Wenger, P.,and Blancpain, C. R., Helv. Chim. Acta, 20, 1427 (1937). (13) West, P.W., J. Chem. Education, 18,528 (1941). RECEIVED January 16, 19.50.
Introduction of liquid Samples into the Mass Spectrometer K. M. PURDY AND R. J. HARRIS, Esso Laboratories, Esso Standard Oil Company, Baton Rouge, La.
,/I
announced by the Atlantic Refining Company ( 2 ) . A sinteredglass disk mercury valve is employed in a majority of the techniques now being used for the
/ / Figure 2.
TO SAMPLE N TRODU CT IO N MAN IFOLD
BASKET HEATER WITH N IC H ROME W I RE HEAT1NG ELEMENT ON INNER WALL OF HEATER
Modified Apparatus for Viscous Samples
pool of mercury used in the sintered-glass valve, thus eliminating the introduction of air with every sample. A short length of capillary glass tubing is used in making this sampling device. It must, necessarily, be small because of pressure limitations in the mass spectrometer inlet system. The pipet shown in Figure 1 will deliver a volume of about 0.001 ml., if made according to the dimensions shown. A constant-volume capillary pipet was developed in the authors’ laboratories to eliminate disadvantages of other methods of liquid sample introduction investigated. As shown in Figure 1, the pipet is so constructed as to be completely immersed in the
Figure 1. ConstantVolume Capillary Pipet
The bottom of the capillary tube is carefully ground to a point, so that ositive contact with a single point of the fritted plate may be ma&. The top of the capillary is ground to a conical or rounded shape to prevent the collection of a small pool of liquid on top when the capillary is filled with sample. A glass rod of small diameter is secured to the middle of the pipet as a handle to facilitate handling the small pipet. Emery paper has been found to be the most satisfactory medium for grinding these pipets. Emery paper No. 1 is used for
