Titrimetric Determination of Zirconium in Siliceous Materials SIR: A rapid method for determining zirconium in refractory and siliceous materials was needed for semiroutine work in our laboratory. Since the sample compositions were so complex and varied, a highly selective procedure was required. Milner and Phennah ( 5 ) precipitated zirconium with mandelic acid and dissolved the precipitate in boiling 4.5M sulfuric acid. After proper adjustment of p H they complexed the zirconium with excess (ethylenedinitrilo)tetraacetic acid (EDTA) and back-titrated with standard iron (111) solution. Similarly, barium sulfate can be decomposed by dissolving the precipitate directly in a n excess of ammoniacal E D T A and back-titrating with standard magnesium chloride solution ( 1 ), precluding the time needed to decompose the precipitate with a n excess of boiling strong acid and the necessity of neutralizing the large amount of acid. Therefore, i t appeared that the time required for the determination of zirconium could be appreciably shortened by dissolving the mandelate directly in a n excess of E D T A and back-titrating the free E D T A with standard zinc solution.
ferred to a 200-ml. tall beaker. Fifty milliliters of the mandelic acid solution was added, and the solution was heated to between 80" and 85' C. for 20 minutes. The resulting precipitate was filtered using a 15-ml. medium-porosity sintered-glass crucible and washed with a solution containing 2% HC1 and 5% mandelic acid. The crucible was placed in the beaker in which the precipitation was conducted. Excess standard EDTA solution, 10.00 or 15.00 ml., was added by pipet, and enough water (about 100 ml.) was added to cover the crucible. The solution was made acidic by adding 1.5 ml. of 6M hydrochloric acid. The solution was boiled for 20 to 25 minutes to form the Zr-EDTA complex. During the boiling, care was taken to wash down a n y precipitate which adhered to the wall of the beaker. The solution was cooled to room temperature, and the pH was adjusted to between 5 and 8 with 7 M ammonium hydroxide. Fifteen milliliters of the buffer solution a n d four drops of 0.1% aqueous xylenol orange indicator solution were added, and the excess EDTA was titrated with standard zinc solution until the in-
Table I. EXPERIMENTAL
Reagents. Zirconium oxide (Carborundum Co.), free of hafnium by spectrographic examination, was fused with sodium bisulfate. T h e cooled melt was dissolved in 1M sulfuric acid, and t h e solution was filtered. T h e filtrate was standardized gravimetrically b y precipitating t h e zirconium in a measured aliquot with mandelic acid and igniting to the oxide. A 0.05M zinc solution was prepared by dissolving pure zinc in hydrochloric acid a n d diluting to volume with distilled water. The disodium salt of E D T A was dissolved in distilled water. The solution (approximately 0.05M) was standardized a t p H 5 against the zinc solution using 4 drops of 0.1% aqueous solution of xylenol orange indicator. An ammonium acetate buffer solution was prepared by adjusting the p H of a 4M acetic acid solution to 5.0 with ammonium hydroxide. Fifteen grams of mandelic acid were dissolved in 100 ml. of hot distilled water a n d cooled. Procedure. Sufficient ground sample t o contain 5 to 50 mg. of ZrOs was weighed into a 150-ml. platinum dish and decomposed on a steam bath with about 3 ml. of water, 10 ml. of 9M H2SOI, a n d 10 ml. of concentrated HF. T h e solution was fumed to complete dryness on a hot plate and fused with 3 grams of fused N a H S 0 4 . The melt was dissolved with 30 to 40 ml. of 6M HCl on a hot plate and trans-
dicator changed from greenish-yellow to orange. RESULTS A N D DISCUSSION
The entire procedure was tested using aliquots of the standardized zirconium solution containing from 18.7 to 55.5 mg. of ZrOs, and in all cases quantitative recovery was obtained. If p-chloromandelic acid is substituted for mandelic acid, the corresponding zirconium precipitate is difficult to dissolve in acidic E D T A solution, but dissolution in ammoniacal E D T A occurs very rapidly. However, in this medium formation of the Zr-EDTA complex is incomplete even after 20 minutes of boiling. Siliceous materials are most easily decomposed with HF and H&h. Subsequent fusion with sodium bisulfate is necessary to remove the last traces of fluoride which will prevent complete precipitation of the zirconium. After dissolution of the melt in hydrochloric acid, insoluble sulfates are removed by filtration.
Comparison between the Titrimetric and Other Methods for Determining Zirconium in Siliceous Materials
ZrO,, % Type of sample" Titrimetric Other method Difference Si A1 Li Mg 3.40, 3.37 Zr Na K av. = 3 . 3 9 3 . 42b -0.03 Si A1 Mg Li 4.61, 4 . 6 6 av. = 4 . 6 4 4 . 71b -0.07 Zr S a Si A1 hlg Li 4.66, 4 . 6 6 av. = 4 . 6 6 4.73b -0.07 Zr Na Si A1 Mg Zr 3.70, 3.72 av. = 3 , 7 1 3.77= -. 0 . 0 6 Li Zn Na Si A1 M g Zr 3.79, 3 . 8 3 av. = 3.81 3 . 72c f0.09 Li Zn Na Si A1 B Ti 0.56, 0 . 5 6 Mg Li Zr av. = 0 . 5 6 0.55d +0.01 Si A1 Li P 2.35, 2.35, 2.34 Zr Ti K Na 0.00 Ba av. = 2 . 3 5 2 . 35e La B Ca Zr 6.12, 6 . 0 5 av. = 6 . 0 9 6.19' -0.1 Pb Zn Si Si A1 Na Li 2.06, 1 . 9 7 Ca P B Zr av. = 2 . 0 2 2.051 -0.03 Mg Si B A1 Ba 2.14 2.29 -0.06 Pb Li Zr Na Flint Clay NBS97 Si A1 Ti Fe 0.24, 0 . 2 5 0.00 Zr K Mg Ca av. = 0 . 2 5 0 .25h + P a Constituents given in decreasing order of weight percentages. * Gravimetric with cupferron ( 4 ) . c Gravimetric with cupferron, iron and zinc separated by mercury cathode electrolysis
+ + + ++ + + + + + + ++ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + ++ + + + + + + + +
+ + + + + + + + + + +
~
+ + + + + + +
(4).
Colorimetric with pararsonic acid ( 6 ) . Gravimetric with mandelic acid. f Gravimetric as phosphate (3). Standard theoretical value. NBS certificate value. Twenty milliliters of 9M HzSOl and 20 ml. of concentrated HF required to decompose this sample. Elimination of phosphorus followed by carbonate-borax fusion. d
6
VOL. 37, NO. 8, JULY 1965
1067
Materials containing phosphate must be fused with 5 grams of a 2 to 1 mixture of anhydrous sodium carbonate and sodium borate. Phosphate was removed by extraction of the cooled melt with hot water and filtration. The filter paper was burned off, the residue was fused with the carbonate-borax flux, cooled, and dissolved in hydrochloric acid. To preclude the necessity of using large amounts of flux, almost all of the silica was removed from samples containing phosphate by the HF and HSSOa treatment and evaporating to dryness before the first fusion. Minor
amounts of silica do not interfere with the filtration of zirconium mandelate or the E D T A titration. Consequently, in the presence of silica this method is superior to the gravimetric determination of zirconium by precipitation with mandelic acid and ignition t o the oxide. The method has been used to determine zirconium in a wide variety of siliceous materials. Some results are given in Table I, and they agree satisfactorily with values obtained by other accepted procedures. After dissolution of the sample, approximately 90 minutes is required For a determination.
LITERATURE CITED
(1) Belcher, R., Wilson, C. L., "New Methods of Analytical Chemistry," 2nd ed., p. 284, Reinhold, New York, 1964. (2) Eberle, A. R., Pinto, L., Lerner, M. W., ANAL.CHEM.34, 1176 (1962).
(3) Kolthoff, I. M., Elving, P. J., Sandell, E. B., "Treatise on Analytical Chemistry," part 11,Vol. 5 , p. 89, Interscience, New York. 1961. (4) Ibid., p. 95. ( 5 ) Milner, G. W. C., Phennah, P. J., A n d @ 79, 475 (1954).
YAO-SINSu Glass Research and DeveloDment Corning Glass Works Corning, N. Y. 14832
Gas Chromatographic Analysis of Pyrethrin-Type Esters and Dimethrin Isomers SIR: For the analysis of pyrethrintype esters of chrysanthemumic acid and related species, colorimetric procedures are normally employed ( 1 , 5-7) ; in all instances, these being either qualitative, concentration-limited, or time consuming. I n addition to these disadvantages or limitations, all are incapable of identifying (a) each species in multicomponent mixtures without prior separation and (6) the predominant isomeric forms of some of these materials. I n toxicological studies, a method for isomer separation and analysis would be desirable truly to evaluate a compound's activity. I n some cases, its activity is attributed to a specific isomeric form. I n the colorimetric procedure of Schreiber and McClellan ( 7 ) suitable for pyrethroid compounds such as allethrin, dimethrin, and pyrethrins, the Denigh reagent is reduced by chrysanthemumic acid, yielding various shades of color such as the salmon red, turning to a raspberry red, in contrast to the pink to stable purple obtained for allethrin and chrysanthemumic acid (racemic mixture of the dl-cis and dltrans acids), respectively. To circumvent problems associated with Denighs reagent instability, dependency of color intensity with p H , and time-consuming steps related to sample dilution, saponification by refluxing with ethanolic NaOH, alcohol evaporation, sulfate anion removal by barium cation precipitation, and filtration, extraction, reagent addition, and color formation reactions, etc., the present gas chromatographic investigations were conducted primarily for the analysis of dimethrin [2,4-dimethylbenzyl - 2,2 - dimethyl - 3 - (2 - methylpropenyl)cyclopropanecarboxylate] in addition to feasibility studies for allethrin and the four active ingredients in pyrethrum extracts -pyrethrin I and 1068
0
ANALYTICAL CHEMISTRY
I1 and cinerin I and 11. Whereas the four latter compounds have been chromatographed using either isothermal (4) or programmed-temperature (3) column conditions coupled with colorimetry, infrared spectroscopy, and liquid partition column chromatography for identification purposes (S), this report shows that the cis- and transisomers of dimethrin can be easily determined a t concentration levels, for example, ranging from 0.01 to 1.80 pg. for the trans-isomer per pl. injection of solution. Beckman et al. (3) also pointed out that there were several additional unidentified gas chromatographic responses observed other than the four principal esters. On the other hand, Donegan et al. (4) made assignments to observed gas chromatographic peaks (chromatogram not shown) without any proof to identity, although it has been determined that the trans-isomer is the naturally occurring compound in the pyrethrum esters. EXPERIMENTAL
The Jarrell - Ash Model 28-710 equipped with a flame ionization detector and the Bristol Dynamaster Model 1P12H560, 11-inch strip-chart recorder, 0 to 10 mv., was used for this investigation. The chromatographic column was made of borosilicate glass (6 ft. by a/,,-inch 0.d.) packed with either (a) General Electric's SE-30 methyl silicone gum rubber or (6) Dow Corning's QF-1 trifluoropropylmethyl silicone polymer; each prepared by standard coating techniques to yield 5% by weight concentrations on 80/90 mesh Anakrom AS (Analabs, Inc., Hamden, Conn.). Prior to use, the packed glass columns were conditioned at 300" C. for 72 hours. For the dimethrin analysis, the gas
chromatographic operating conditions were maintained as follows : Column temperature, "C.
128 and 168 Injector temperature, "C. 180 Detector temperature, "C. 220 1.25 cu. Air flow rate ft./hr. Hydrogen gas pressure 6.0 p.s.i. Sensitivity setting 1 x 10-9 and 1 x 1O-IO A. Nitrogen carrier gas flow- 6 8 . 3 rate, cc./min. Recorder chart speed, 2.0 min./inch I n subsequent chromatographic separations of other pyrethrin-type esters, all the above conditions remained constant with the following exceptions : column temperature, 206" C.; injector temperature, 290" C.; detector temperature, 260' C. For calibration purposes, 1 pl. Hamilton microsyringe injections of varying dimethrin concentrations dissolved in benzene as solvent were made directly onto the coated, silanized column solid support through a rubber septum. Benzene was also used as solvent for the other pyrethrin-type esters investigated. DISCUSSION OF RESULTS
Initially, the purpose of this study was (a) to examine the chromatographic behavior of dimethrin and (6) to establish the operating parameters required for its analysis. Using either the SE-30 or QF-1 column, the dimethrin compound [purity reported as 98Oj, by colorimetry (2) ] in all chromatograms yielded two major components as shown in Figure 1; these chromatograms obtained with the QF-1 column operated a t 168" and 128" C., respectively. T o establish that peak A was not a thermal degradation product of the ester, the injector (180-250" C.) and column
