V O L U M E 2 7 , NO. 1, J A N U A R Y 1 9 5 5 and receiving the samples into a collecting rack instead of disposing of the tubes. There are many possibilities in which the sample changer and principles involved in this instrument could be incorporated to reduce the time which is used in routine analytical procedures. ACKNOWLEDGMENT
This instrument was built for use in work supported in part by a grant from the American Cancer Society as recommended by the National Research Council through the Committee on Growth. Part of the cost of the instrument was defrayed by the University of Tennessee Reserve for Research. LITERATURE CITED (1) Austin, R. R., Am. Gas Assoc. Proc., 31, 505 (1949). (2) Dunn, E. B., Melpolder, F. W., Taylor, R. C., and Young, W. S.,Proc. M i d - Y e a r Meeting Am. Petroleum Inst., JOMIII, 45 (1950).
127 Eades, C . If.,J r . , AIcKay, €3. P., and Romans, W. E., Federation Proc., 11, 205 (1952). Gapus, G . H., and Pool, M. I., Review of Scientific Instruments, 8, 197 (1937). Hawes, R . C., Strickler, A., and Petterson, T. H., Elec. Mjg., 47, 76, 212 (1951). Jacobsen, C. F., and Lkonis, JosB, Compt. rend. trav. lab. Carlsberg, SBr. chim., 27, 333 (1951). Juliard, A , , and Cakenberghe, J. van, A n a l . Chim. Acta, 2 , 542 (1948). Lingane, J. J., - ~ A L CHEM., . 20, 285 (1948). McKay, B. P., and Eades, C. H., Jr., Ibid., 27, 123 (1950). Shaffer, P. A,, Jr., Briglio, A,, Jr., and Brockman, J. -4., Jr., Ibid., 20, 1008 (1948). Wise, E. X , , Ibid., 23, 1479 (1951). Wu. C. S.,and Rainwater, James, U. S. Atomic Energy Commission, A E C Document, MDDC-1671 (1944). R E C E I V Efor D review December 31, 1952. Accepted September 29, 1954. A preliminary report ( 5 ) of a n earlier model of this apparatus mas made before t h e Federation Meetings in New York, April 14 t o 18, 1852.
Paper Chromatography of Cobalt(lll), Coppedll), and
Nickel(l1) Acetylacetonates E U G E N E W. BERG and J A C O B E. STRASSNER Coates Chemical Laboratories, Louisiana State University, Baton Rouge, f a .
Acetylacetone, one of the simplest fi-diketones, was selected for study because its availability and low cost make it a desirable reagent for chromatographic separations. Cobalt(III), copper(II), and nickel(I1) acetylacetonates were separated using mixtures of cyclohexane, dioxane, and methanol as the developing solvent. A mixture of 847” cyclohexane, 10% dioxane, and 670 methanol gave good separations. The mean R J values were reproducible to A0.02. Solubilities of the metal chelates were measured in the developing solvents. Relative adsorption affinities were obtained from dielectric constant measurements. A qualitative relation was found to exist between the relative sequence of R / values and the relative solubility and adsorption affinity (polarization) of the metal chelates.
A
LTHOUGH a large number of metal p-diketone complexes are well known (10, 16), these complexes have not been extensively used for chromatographic separations. Increasing interest in the use of chelating agents for inorganic chromatographic separations has been shown in the appearance of a number of recent articles (1-7, 9, 11-15). Acetylacetone, one of the simplest 8-diketones, was selected for this study because its availability and low cost make it a desirable reagent. Burstall et al. ( 5 ) and Pollard et al. (11) have used solvent systems containing acetylacetone in the chromatographic separation of some inorganic ions. +4number of factors may influence this type of separation-namely, the rate of chelate formation, the presence of excess chelating agent, and the possibility of chelate hydrolysis in the presence of strong acids. In order to avoid these factors the authors have preferred to spot the paper with the preformed metal acetylacetonates and to develop the chromatogram with a solvent mixture in which the chelates are stable. REAGENTS
Acetylacetone (Matheson Co.), redistilled. Methanol, C.P. Cyclohexane (practical grade), redistilled.
Dioxane (technical grade), redistilled. ilqueous solutions, 1%, of the metal ions prepared from: C.P. cobalt(I1) nitrate, C.P. copper(I1) nitrate, and C.P. nickel(I1) nitrate. Solution of dimethylglyoxime in ethyl alcohol, 1%. Solution of dithio-oxamide in ethyl alcohol, 0.3%. PROCEDURE
Acetylacetonates of cobalt(III), copper(II), and nickel(I1) were prepared by shaking 1% solutions of the ions, adjusted approximately to a pH of 7 with sodium acetate, with acetylacetone. The nickel acetylacetonate was extracted with n-butyl alcohol and shaken with distilled water to remove nickel ions. The copper and cobalt acetylacetonates were extracted with methyl isopropyl ketone and shaken with water to remove any ions. In the extraction of the cobalt acetylacetonate, the methyl isopropyl ketone was kept in contact with the original solution until the ketone layer developed a dark green color. This chelate corresponded to the cobalt(II1) acetylacetonate described by Gach (8). The oxidation of the cobalt(I1) to cobalt(II1) was probably due to impurities in the methyl isopropyl ketone. Final colors of the extracted and washed solutions of cobalt, copper, and nickel acetylacetonates were dark green, blue-green and yellow-green, respectively. Hydrometer cylinders, 43 cm. tall and 7 cm. in diameter, served as chromatographic chambers. A part of the cylinder was lined with filter paper soaked with the solvent in order to saturate the chamber more efficiently. Twelve hours were then alloxed for complete saturation of the chamber. Whatman No. 1 filter paper strips 2.5 inches wide were spotted with the extracted solutions of the metal acetylacetonates and dried in air for 1 hour. The strips were then placed in the chamber saturated with vapor and equilibrated for 1 hour before immersion in the solvent. The chromatograms were developed completely in 3 hours, the solvent front having ascended approximately 25 cm. Preliminary u ork indicated that methanol and cyclohexane would be desirable solvents for this study. The cobalt(III), copper(II), and nickel(I1) acetylacetonates all moved with rather large R f values in methanol, whereas only the cobalt acetylacetonate moved in cyclohexane. An appreciable amount of methanol was not soluble in cyclohexane; therefore, a third component, dioxane, was used to form a completely miscible solvent.
ANALYTICAL CHEMISTRY
128
reproducible with an average deviation of 320.02. (Cyclohexane, Dioxane, Methanol) Table I shows the average Sickel(I1) Acetylacetonate Cobalt(II1) Acetylacetonate Copper(I1) Acetylacetonate R, values of the metal acetylMethanol, % Methanol, % Methanol, 55 acetonates related to an Dio xane, Yp 3 6 9 3 6 9 3 6 9 increasing concentration of di0 . 5 8 . . . . . . 0 . 1 8 . . . . . . J . . . 0.57 . . . . . . 0.22 . . . oxane in three fixed concentra10 0.69 0.64 0.56 0.25 0.27 0.27 tions of methanol with cyclo16 0.76 0.72 0.65 0.29 0.34 0.37 20 0.79 0.77 0.73 0.34 0.39 0.44 h e x a n e complementing t h e 25 0.83 0.81 0.75 0.43 0.47 0.48 35 dioxane concentration. 0.90 0.88 0.84 0.56 0.67 O.fiO It is noted that the R, value of the copper(I1) acetylacetonate increases with increasing The position of the cobalt acetylacetonate was easily detected concentration of dioxane and/or methanol. The solubility data by its green color and by the orange coloration when sprayed for copper(I1) acetylacetonate in Table I1 show a similar relationwith dithiooxamide. The nickel acetylacetonate was detected ship between solubility and concentrations of dioxane or methby spraying the paper with an alcoholic solution of dimethylanol. Both the RI and the solubility values of the copper acetylglyoxime and then exposing it to ammonia fumes. The copper acetonate give a straight line relationship when plotted against acetylacetonate was detected last by spraying with tli thioincreasing concentration of dioxane. oxamide and exposing it to ammonia fumes. To verify that the metal acetylacetonates were being chromatographed and not the ions, a paper Table 11. Solubility of Metal Acetylacetonates in Mixed was spotted with the metal nitrates and chromatoSolvent System (Cyclohexane, Dioxane, Methanol) graphed under conditions identical to the procedure CobaltUII), Grams/Liter Copper(II), Grams/Liter Nickel(II), Grams/Litcr used with the metal acetylacetonates. When the Methanol, Methanol, % Methanol. % chromatogram was sprayed with an alcoholic solu1)ioxane. S: 3 9 3 9 3 9 tion of dithiooxamide, the metals were detected 5 5.3 ... 0.64 ... 0.63 ... only in the position of the original spot. X o migra10 ... 17.4 0.82 2.02 ... 1.35 I5 10.1 ... 1.24 2.60 1.07 ... tion of ions was observed. 20 ... 27.3 1.64 3.05 ... ... 2d 17.5 ... 2.13 3.76 3.13 8.6 The applicability of this technique to the separa35 ... ... 3.08 5.08 . . . ... tion of a mixture of cobalt(II), copper(II), and nickel(I1) ions was determined in the following manner. Table I.
Rt Values of Metal Acetylacetonates in Mixed Solvent System
An equimolar mixture of the three ions was converted to their metal chelates using the following procedure, Sixty milliliters of a 2% solution of cobalt, copper, and nickel nitrates &-eretreated with 20 ml. of 3% hydrogen peroxide, and heated on :t s t e m bath for 15 minutes. The solution was then adjusted to R pH oi 7 Lyith sodium acetate, and 20 ml. of a 1 to 1 mixture of acetylacetone-ethyl alcohol was added to form the metal che1atc.s. The solution was heated for a few minutes and then treated with 50 ml. of a I to 1mixture of methyl isopropyl ketone and n-butyl alcohol, heated on a steam bath for 30 minutes, and allowed to stand for several hours or until the organic layer was dark green. The organic layer was then removed and used to spot the chromatographic paper. The chromatographs of the mixture w r c identical with those of the individual chelates spotted together. Sufficient amounts of solid metal chelates for solubility deterininations were obtained by the following procedures. Copper(I1) acetylacetonate was precipitated by shaking acetylacetone with a 1% ' solution of copper(I1) nitrate, adjusted approximately to a pH of 7 with sodium acetate. The pale blue precipitate was washed with hot water, dissolved in hot methanol and recrystallized by addition of water. The cobalt(II1) and nickel( 11) acetylacetonates were obtained by evaporating the solutions of extracted chelates to dryness in a vacuum desiccator. The solid cobalt( 111) and nickel(I1) acetylacetonates were, respectively, dark green and pale green. The solubility of each metal acetylacetonate in various compositions of the developing solvent was determined by saturating the solution with the solid chelate, evaporating an aliquot of the saturated solution to dryness in a vacuum desiccator, and weighing. A Sargent Model V chemical oscillometer with cell holder-Type .4 and a fixed frequency of 5 X 1 0 6 cycles per second was used for the relative dielectric constant measurements of 0.01M solutions of the metal chelates in methanol.
The R, values of the cobalt acetylacetonate increased with increasing concentration of dioxane but decreased slightly with increasing concentration of methanol over the concentration range studied. The R, values of the nickel acetylacetonate are very small, but seem to increase slightly with increasing concentration of dioxane and/or methanol. If solubility of the metal acetylacetonates in the developing solvent were the governing factor in this system, then the solubility of the metal acetylacetonates would be expected to be cobalt > copper > nickel. Table I11 shows that cobalt acetylacetonate is the most soluble chelate as one might anticipate, but also shows that the solubilities of the copper and nickel acetylacetonates are too similar for these compounds to be separated on the basis of solubility alone.
Table 111. Solubility of Cobalt(III), Copper(II), and Niclcel(I1) Acetylacetonates in Pure Solvents Chelate co cu Ni
Table IV.
(Grama per liter) Cyclohexane Dioxane 1.0 89.8 0.2 16.3 0.2 26.0
Methanol 58.0 3.5 78.2
Capacitance Measurements of 0.01M Solutions of Metal Acetylacetonates in Methanol Chelate cu co Ni
Capacitance 29,800 30,140 >32,000
RESULTS
The most effective separations occurred with the lower concentrations of methanol and dioxane. Mixtures of cyclohexane and dioxane did not give chromatograms suitable for measurement. Solvent mixtures near 6% methanol, 10% dioxane, and 84y0 cyclohexane gave good separations. The R, values were
Capacitance measurements for the oscillometer cell filled with 0.01M solutions of the various metal acetylacetonates in methanol are given in Table IV. The relative dielectric constants of the solutions are directly proportional to the measured capacitance. I t is also recognized that the relative dielectric constants of the
V O L U M E 2 7 , NO. 1, J A N U A R Y 1 9 5 5 equal molar solutions are a measure of the relative total polarization of the solute molecules. Thus the relative total polarization of the various metal chelates should be in the order, nickel>> cobalt> copper. The greater the polarization the greater the strength of adsorption. Comparison of the solubility and polarization data indicate. that the relative R/ values fall in a logical sequence: Solubility Polarization R/ values
Co >> Cu, Ni Ni > > Co > Cu Co > Cu > h-i
The small difference in solubility of the copper and niclwl chelates coupled with the large polarization of the nickel chelatv gives a Rj sequence copper > nickel, whereas the much large] solubility of the cobalt chelate coupled with only a slightly largrr polarization than the copper chelate gives a R/ sequence cobalt > copper. Thus the relative Rj values can be explained qualitatively by these two factors. Another p-diketone chelate system has been studied by the authors (not yet published) in which five chelates have been measured with a similar correlation hetwren solubility, polarization, and R, values. LITERATURE CITED (1) -41-3lahdi. A. K.. and Wilson. C. L.. M i k r o c h m i t cer. M i k r o c h i t ~ r . Acta, 36/37, 218 (1951): 40, 138 (1952).
129 (2) .Inder8oti. J. 11. .I..and Lederer, M., Anal. Chim. Scta, 5, 396
(1951). (3) -\shiaawa, T., Repts. Balneol. Lab. Olcayama Univ., KO.5, 1-12 (1951). (4) Berg. E. W., and McIntyre, It. T., ANAL.CHEX, 26, 813(1954). ( 5 ) Burstall, F. IT., Davies, G. R.. Linstead, R. P., and Wells, R . d., J. Chem. Soc., 1950,51G. ( 6 ) Erlenmeyer, €1.. and Dahn, IT., Helv. Chim. iicta. 22, 1369 (1939). (7) Fernando, Q., and Phillips. J. P., .\NAL. CHEM..25, 819 (1953). (8) Gach, F., Monatsh., 21,98 (1900). (9) Laskowki, D. E.. and hlcCrone, W . ’ C . ,ANAL.C m x . , 23, 1579 (1951). (10) llartell. d.E., and Calvin, 31.. “Chemistry of the Metal Chelate Compounds.” Kew York, Prentice-Hall, Ino., 1952. (11) Pollard. F. H., AIcOmie, J. F. W., and Elbeih, I. I. l I . , , J , L / I P M . SOC..1951.460.470. (12) Pollard, F. H., lIcOmie, J. F. W..and Elbeih, I. I. M., Dtscicssions Faraday Soc., No. 7, 183 (1949). (13) Pollaid, F. H., llcOmie, J. F. W., and Stevens, H. XI., J . C h ~ m . Soc., 1951,771; 1952,4730. (14) Reeves, W. -1..and Crumuler. T. n.. ANAL. CHEM.. 23. 1576 (1951). (15) Venturello, G., and Ghe. A . 11.. B n n . chim. (Rome), 43, 267 (1963). (16) Welcher, F. J., “Organic .\nalytical Reagents,” New T o i k . n. Van Nostrand Co., 1947 I