trace components. If the concentration of a component for which spectra are desired is low, it is our practice to apply several parallel spots of the sample to the chromatograni to increase the amount available for recovery. R e have not found it necessary to use a "vacuum cleaner" device for collecting the adsorbent from the chroniatogram as described by Rlillett, Moore, and Saeman (I) because it is not difficult to scrape the adsorbent onto a piece of glazed paper, transfer it to the elution column, and pack it. However, if desired, a fitting similar to that described could be used on the upper end of the capillary pipet. Choice between the two methods seems to be a matter of personal perference.
An example of a spectrum obtained by the technique is shown in Figure 3. Approximately 100 pg. of the antioxidant 1,3,5-trimethyl-2,4,6-tris(3,5ditert - butyl - 4 - hydroxybenzyl) benzene was migrated on a 2- X %inch silica gel G chromatogram with cyclohexane: benzene (4: 1 v./v.) solvent and the section containing the compound was removed and eluted as described. The spectrum was obtained in carbon tetrachloride solvent using a PerkinElmer Model 21 infrared spectrophotometer equipped with a beam condenser and using a n approximately 0.05-mm. path cell. A spectrum of the corresponding blank is also shown and is identical for that obtained with the cell filled with carbon tetrachloride.
An alternative, but even simpler, version of the technique is used for obtaining ultraviolet spectra. I n this case, evaporation of solvent is unnecessary and greater dilution is acceptable. A straight elution tube is inserted directly into a 1-ml. volumetric flask and the compound is eluted into the flask for subsequent measurement. LITERATURE CITED
( 1 ) Millett, A4. A,, Moore, U'. E., Saeman, J. F., ANAL.CHEM.36, 491 (1964).
R. N . McCoy E. C. FIEBIC
Shell Development Co. Emeryville, Calif.
Chemical Analysis of Thorium Phosphides Sir: I n the study of the synthesis and phase relations of the thorium phosphides, one of the points of interest was the possible existence ranges of both Th3P4and ThP1- (I). The latter compound, ThP1-., is a rock salt structure phase of extended existence range, where z variei from 0.04 to 0.45 a t 1000' C. The phaze boundary compositions were determined by the change in lattice parameter with composition, where the lattice parameters were determined from x-raj powder diagrams of quenched samples and the compositions were determined by the analytical method presented here. The detection of the changes in stoichiometry by the weight change resulting from the reactions was not desirable because of the unfavorable ratio of atomic weight between thorium and phosphorus. Although 27" T h o 2 inipurity by weight could be detected by x-ray powder diffraction, this would not include any oxygen possibly dissolved in the phosphide lattice. A chemical method was found that n a s used to determine the P / T h ratio of the phosphides. The samples were dissolved in concentrated nitric acid with the addition of potassium permanganate to ensure complete oxidation of phosphorus to phosphate. Thorium was separated as the oxalate (2) and determined by E D T A titration ( 3 ) . Phosphate was precipitated with zirconium (IT') and determined gravimetrically as ZrP207,a modification of the well known phosphate method for determining zirconium(1V) (4). EXPERIMENTAL
Materials. All chemicals were reagent or analytical grade. For the dissolution, 69Yc HSOs, 1 S KMnOa, and 30Tc HzOz are needed. Reagent was used for the grade H2C2O4,2Hz0
separation of thorium(1V). Determination of thorium(1V) requires 0.1M E D T A standardized against high purity bismuth and pyrocatechol violet indicator. Analytical grade Zr(SO&. 4H20 was used to prepare the zirconium (IV) solution for phosphate separation. Procedure. A 300-mg. thorium phosphide sample was covered with 50 ml. of 69Yc nitric acid a n d 5 nil. of 1 N potassium permanganate in a 250-ml. beaker and brought to boil. As the phosphide dissolved, more permanganate was added and the mixture was stirred. I t was possible to keep the solution clear enough to see the remaining phosphide, yet maintain the color of permanganate in the solution. Precipitated manganese dioxide and excess permanganate were destroyed by the addition of hydrogen peroxide, and then the excess hydrogen peroxide was removed by boiling. The remaining solution was colorless but often contained residual T h o s in amounts of 1 to 5 mg., assumed to be the T h o 2 impurity content of the sample. This was removed by filtration. '4 quantitative separation of thorium from phosphorus was achieved by precipitation of thorium oxalate. The solution of thorium(1V) and phosphate was transferred to a 500-ml. beaker, evaporated to 10 ml., and cooled. A drop of 30Oj, hydrogen peroxide was added and then boiled away. When the solution was a t room temperature, 300 ml. of 1N sulfuric acid were added and the mixture was heated l o boiling. Five grams of oxalic acid dehydrate and some filter paper pulp were stirred into the solution to effect the precipitation. This mixture was digested for 15 minutes a t 95" C., cooled on an ice bath, and set aside overnight. The precipitated thorium oxalate was filtered and washed with a dilute oxalic acid solution. The filtrate and washings were retained for phosphorus analysis. The thorium oxalate precipitate was dissolved in 100 nil. of 10% boiling nitric acid. The filter paper pulp was
removed by filtration, the solution was heated, and enough permanganate solution was added to destroy any oxalate. The volume of the solution was made u p to 500 ml. in a volumetric flask and the thorium content was determined by titration (3) against standard 0.01.11 EDTA solution using pyrocatechol violet indicator a t a pH of 2. Phosphorus analysis was made by precipitating zirconiuin phosphate from the filtrate solution retained a t the point of thorium-phosphorus separation, igniting the precipitate to zirconium pyrophosphate, and weighing the white residue ZrPz07. The oxalate in the filtrate solution was destroyed by the addition of permanganate, and precipitation was carried out by the addition of a slight excess of zirconium sulfate solution to the filtrate heated to 80' C. After maintaining a temperature of 50' C. for 2 hours, the precipitate was set aside to digest overnight and then gathered by filtration using ashless analytical filter paper. The precipitate and filter paper were ignited in a platinum crucible over a Meker burner flame and finally fired for 12 hours a t 1000' C. in an alumina boat. The boat and contents were cooled and weighed to determine the amount of ZrPzO,. DISCUSSION
A series of control experiments to test the method was carried out by mixing known amounts of phosphate and thorium(1V) in the dissolution medium and then carrying out the analytical procedure. The results are shown in Table I. Samples 1 and 2 were thorium(1V) determinations in the absence of phosphate, and 3 and 4 were gravimetric determinations of phosphorus from thorium-free phosphate solutions. Samples 5, 6 , and 7 were mixtures simVOL. 37, NO. 4, APRIL 1965
595
Table I.
Taken, mg. Th P
No. 1 2 3
251.7 185 5
4 5 6 7
318.1 206.0 217.8
38.6 39.9 30.5 21.2 27.2
Control Experiments on Analytical Method
Recovered, mg. Th P250.1 185.0 316.0 205.1 216.2
38.2 39.9 30.2 21.0 26.9
ulating dissolved phosphides. The precision of the thorium determinations ’ but was better than + 0.5 relative % the determinations were also approximately 0.5 relative yo too low. The phosphorus values were between 0 and 1 relative yo by weight too low. Repeated analyses on several phosphide samples demonstrated that the results were reproducible for actual samples. For instance, a sample of ThPl-. was analyzed three times to yield ThPo 749*0 013 where the uncertainty represents the standard deviation. Three different samples of the Th3P4 phase were ThPl 35+0 03, ThPl 32 and ThP133t001. The mean value of ThPl 33*0 o2 was taken as an indication that this phase was stoichiometric Th3P4. The difference between analyzing actual phosphides and mixtures of thorium(1V) and phosphate is mainly in the step where the phosphide is dissolved, where complete oxidation of phosphorus
Formula -Calculated Analyzed
ThPo.718 ThPo.771 ThPo. 086
ThPo.7ia ThPo.767 ThPo 912
Weight lost, yo 0.64 0.27 1,03 0 0.69 0.48 0.73
to is required. Losses of phosphorus would be variable from sample to sample, and, because the precision of determinations was relatively high, it can be concluded that phosphorus loss was negligible. Some Th3P4 samples were dissolved by adding standardized K M n 0 4 from a buret to the phosphide in the hot acid medium. The equivalence point was determined by back-titration of excess permanganate with sodium oxalate after all solid phosphide had dissolved. If all phosphide phosphorus was assumed to be P-3, the analyses by titration along with the usual thorium analysis gave compositions close to Th3P4. The equation used to calculate the phosphorus was
24H+
+ 8Mn04- + 5P-3 = 5P04-3 + 8iUn+2 + 12H20
This titration method was attempted for several compositions of the ThPl-,
phase, but the results were not consistent with any one hypothetical oxidation state for phosphorus, and, therefore, a large excess of 1InO4- was used in the oxidation of the thorium phosphide samples. If the systematic errors in the thorium and phosphorus determination are taken into account, the mean error of the determination of the atomic P / T h ratio of thorium phosphide samples in the composition range ThPo.5 to ThP1.33is estimated to be with in ~ t 0 . 0 2 . LITERATURE CITED
( 1 ) Gingerich, K. A , , Wilson, D. W., un ublished data, 1964. ( 2 ) olthoff, I. ?*I.,Elving, P. J., Sandell,
IF
E. B.. “Treatise on Analvtical Chemistry,” Part 11, Yol. 5 , p. 159, Interscience, New York, 1961. (3) Schxartzenbach, G., “Complexometric Titrations,” p. 75, Interscience, New York, 1957. ( 4 ) Wilson, C. L., Wilson, D. W., “Comprehensive Analytical Chemistry,” Yol. IC, p. 310, Elsevier, New York, 1962. DANIELW. WILSON~ KARLA . GINGERICH~
Department of Chemistry The Pennsylvania State Cniversity University Park, Pa. WORKsupported by the U. S. Atomic Energy Commission under Contract Xo. AT(30-1)-2541. 1 Present address, Department of Animal Science, Colorado State I‘niversitg, Fort Collins, Colo. * Present address, Battelle Memorial Institute, 505 King Ave., Columbus, Ohio
Determination of Zinc Additives in Lubricating Oil by a Dithizone Spectrophotometric Method SIR: The use of zinc additives in lubricating oils has necessitated the development of a rapid and precise method for controlling the additive concentration. Zinc may be determined directly in lubricating oils, without prior treatment, by chelating the metal with (ethylened i n i t r i l o ) t e t r a a c e t i c a c i d (EDTA) in an oil-acetone medium and titrating the excess EDTA with standard magnesium chloride solution (1). The EDTA titration, however, lacks the sensitivity necessary for the determination of very low additive concentrations. T h e use of dithizone as an analytical reagent for zinc is well documented (2) and need not be described here. The method presented is an adaptation of the dithizone method wherein the zincdithizone complex is formed directly in a n oil-carbon tetrachloride solution; t h e excess dithizone is destroyed with 596
e
ANALYTICAL CHEMISTRY
dilute ammonium hydroxide and the absorbance of the zinc dithizonate is measured at 530 mp. Because all prior treatment of the sample is eliminated, the time of analysis is suitably short for routine determinations. EXPERIMENTAL
Apparatus and Reagents. All absorbance measurements were made on a Cary recording spectrophotometer using fused silica cells of 20-mm. light path. Standard Zinc Solution. Zinc cyclohexane butyrate (National Bureau of Standards Sample 1073A) in a mineral oil of the lubricating type (acid-treated napthenic base from Gulf Coast Crude) was used. T h e standard zinc-oil solution was prepared according to the instructions accompanying the IGBS sample. All other reagents were analytical reagent grade.
Procedure. A standard curve was prepared by transferring weighed quantities of the standard zinc-oil solution (148 fig. of zinc per gram of oil) to 60-ml. separatory funnels containing 15 to 20 ml. of carbon tetrachloride. The contents were thoroughly mixed and a solution of dilhizone in carbon tetrachloride (0.01 gram per 100 ml.) was added dropwise (with intermediate shaking) until an excess of dithizone was present, as shown by the appearance of a green tint in the solutions. The excess dithizone was destroyed by adding 10 ml. of ammonium hydroxide solution (1 : 200) and shaking vigorously for 1 minute. The two phases were allowed to separate and the oil-carbon tetrachloride phases were transferred to 100ml. volumetric flasks. The separatory funnels were washed with two 10-ml. portions of carbon tetrachloride and the washings were added to the volumetric flasks.