Differential Potentiometric Determination of Tellurium in Refined Tellurium Products PETER W. BENNETT and SlLVlO BARABAS Canadian Copper Refiners, lfd., Monfreal East, Quebec, Canada
b The differential potentiometric procedure for determination of selenium in high grade selenium materials, based on controlled reduction of approximately 97% of selenium b y hydrazine sulfate and potentiometric titration of the remaining 3% b y stannous chloride, could not be used for tellurium determination in refined tellurium. Tellurium i s not reduced in sulfuric acid medium by hydrazine sulfate. The reduction in hydrochloric acid medium is neither prompt nor stoichiometric. Most satisfactory results were obtained by using strong stannous chloride solution for the controlled separation of the bulk of tellurium in hydrochloric acid and potentiometric titration o f the remaining tellurium b y a more dilute stannous chloride solution. Selenium does not interfere, as it i s volatilized with bromine-hydrobromic acid solution as SeBrr. The effect of a number of foreign ions is discussed. The standard deviation in the duplicate analysis of 12 samples on a routine basis was 0.023%.
B
in the p a d fen. years, significant progress has been made in devising numerous practical applications based on thermoelectric phenomena in d i i c h tellurium, as a seniiconductor material, plays an important role, the interest in this element has greatly increased. The extreme purity of all semiconductor materials is a prerequisite in any successful application. While the grade of high purity tellurium is necessarily established in an indirect manner from the content of total impurities, proper evaluation of refined tellurium used as a starting material must include the actual determination of tellurium. The limitations of conventional procedures for determining the main constituent in the analysis of high grade materials hare been discussed ( I ) . The successful application of the differential technique t o the analysis of refined selenium suggested a similar approach to refined tellurium. ECACSE,
ANALYTICAL BEHAVIOR
OF TELLURIUM
Tellurium does not precipitate from sulfuric acid solution by reduction with hydrazine sulfate,
I n the differential analysis of ,efined selenium, containing approximately 0.1% tellurium, selenium precipitated by a deficiency of hydrazine sulfate reagent in sulfuric acid showed no tellurium on spectrographic examination. However, if an excess of hydrazine sulfate is added, selenium will carry down some tellurium. Quadrivalent tellurium is reduced by hydrazine sulfate in hydrochloric acid medium, but the stoichiometric relationship could not be proved. Delayed separation of tellurium from filtrates was noted hours after the addition of the reagent and removal of the bulk of tellurium precipitate. Reduction with hypophosphorous acid was not satisfactory for controlled, incomplete precipitation of tellurium. Finally recourse was made t o a proved reducing agent for tellurium, stannous chloride in hydrochloric acid medium. Here the differential procedure consists in first reducing 97% of the tellurium present by reaction with a relatively strong stannous chloride solution, followed by potentiometric titration of the remaining tellurium.
+
samples containing 99 % tellurium or proportionally more for tellurium dioxide and various telluriurii salts. I n the case of refined tellurium, add 30 ml. of 1 t o 1 nitric acid, cover, and digest at approximately 80' C. to complete dissolution. Remove the cover and evaporate t o dryness. Take up the residue in a minimum amount of hydrochloric acid and evaporate to near dryness. Repeat the evaporation twice more Tvith 10-ml. portions of bromine-hydrobromic acid solution. Finally, add 150 ml. of 3-V hydrochloric acid. I n the case of tellurium dioxide and alkali tellurium salts dissolve the sample directly in 150 ml. of 3 5 hydrochloric acid. Precipitate the bulk of tellurium by adding 50 ml. of concentrated stannous chloride solution from an automatic pipet. Stir well to coagulate tellurium and proceed to potentiometric titration of the unreacted tellurium with diluted stannous chloride solution. Establish the tellurium equivalent of both diluted and concentrated stannous chloride solutions from standards of high-purity tellurium run simultaneously n ith the samples. DISCUSSION OF PROCEDURE
EXPERIMENTAL
Apparatus. T h e apparatus has been
described ( I ) . Reagents. Concentrated stannous chloride solution is freshly prepared by dissolving 35.4 grams of stannous chloride dihydrate, SnClz.2Hz0 (analytical reagent grade), in 3N hydrochloric acid and making up t o 1 liter. One milliliter of this solution reduces approximately 10 mg. of quadrivalent tellurium. Dilute stannous chloride solution is prepared by diluting 100 ml. of the concentrated solution with 3N hydrochloric acid to 1 liter, or preferably by dissolving 4.65 grams of tin metal (analytical reagent grade) in 2.5 liters of 3N hydrochloric acid against the pressure of a zinc-hydrochloric acid hydrogen generator. One milliliter of this solution reduces approximately 1 mg. of tellurium(1V). Standard tellurium solution, 0.7 mg. per ml. in 3N HCI, i s prepared from high purity tellurium. Bromine-hydrobromic acid solution contains 4 volume yo bromine in 48% hydrobromic acid. Procedure. Weigh accurately into ti 400-ml. beaker 525 to 530 mg. of
Dissolution of Sample. The sample dissolves promptly in 1 to 1 nitric acid. The subsequent removal of nitric acid, critical in the analysis of refined selenium, presents no problem here. Tellurium is considerably less volatile than selenium and even a t full hot plate heat no tellurium is lost by volatilization. I n removing nitric acid by repeated evaporations with brominehydrobromic acid solution, any selenium present is conveniently volatilized. Otherwise, if left in solution it would have interfered with the tellurium analysis. Effect of Hydrochloric Acid Concentration on Reduction Potential of Tellurium. The reduction potential of tellurium(1V) in the presence of selenium(1V) shifts toward lower values with increased acid concentration ( 1 ) . However, the total potential span a t the end point remains practically the same in the range 3 to 9N HCl. I n the present work the effect of varying hydrochloric acid concentration on potentiometric measurement VOL. 35, NO. 2, FEBRUARY 1963
139
~~
~~
Table I.
~
~
Accuracy and Precision of Differential Separation of Tellurium
(50 ml. of a 35.4 gram per liter stannous chloride solution Weight of Te H.P. Te, mg. titrated, mg. 512.85 535.7 22,85 512.70 23.10 585.8 512.87 536.8 23.93 535.4 22.70 512.70 538.2 25.31 512.89 Av. 512.81
added to each sample solution) Deviation Mg. 7% +O. 04 0,008 -0.11 0 021 +O. 06 0,011 -0.11 0.021 +o, os 0.015 0.08 0.015 Std.dev. 0.018
Table 111. Effect of Foreign Ions on Differential Potentiometric Titration on Tellurium
(All solutions contained 536.99 mg. of
Te+4
+ 10 mg. of foreign ion)
Foreign ion
Te found, mg. 537.11 536.78 537.84 537 81 539.06 537.86 549.14 547.71 538.39 538 46 538.58 542.19 538.21
Sb + 3
Bi +3 iLs + 3 rlg +
of tellurium in the absence of selenium was studied. The shape of the tellurium titration curves in 1, 2, 3, 5, 7 , and 9N acid is shotvn in Figure 1. Except for the titration in IN HCl, Fhich provided a poorly defined curve with an inverted inflection, all the other titration curves Table II. Per Cent Composition of Typical Refined Tellurium Cast
Te Se Pb Cu Ni
99.89 0.042 0.03 0.003 0.003
Fe
0.0002 A1 0.0002 Rlg 0.0002 Si 0.0001 0 0.009 Na 0.002 Total 99.98 Tellurium determined by differential potentiometric method. Selenium determined spectrophotometrically. Oxygen determined by conductometric measurement in inert atmosphere. Impurities, other than selenium and oxygen, determined spectrographically.
were most satisfactory. I n view of this, all subsequent titrations were carried out in 3 N HC1. Accuracy and Precision. The accuracy and precision of the differential potentiometric procedure for tellurium were established as for selenium. Varying amounts of accurately weighed high purity tellurium were dissolved and approximately 97% of the total tellurium was precipitated by adding from an automatic buret 50 ml. of concentrated stannous chloride solution to each sample solution. The remaining tellurium was titrated potentiometrically (Table I). The reproducibility of tellurium precipitation, as indicated by the standard deviation of O.OlS%, is superior to any other procedure for tellurium reported in the literature. This standard deviation is one half of that reported for selenium by the same differential method. The apparent superiority of the tellurium procedure over that for
4 u +3 r\;i+2 c u +z Fe+3 Cd .L2 Pb +2 Zn -2 Pt + 4
Pd +4
Te recovered,
% 100.02 99.96 100.16 100.15 100.39 100.16 102.26 102.00 100.26 100.27 100.30 100.97 100.23
Table IV. Effect of Foreign Ions on Direct Potentiometric Titration of Tellurium
(All solutions contained 21.48 mg., of Te+4 10 mg. of foreign ion)
+
Foyeign ion None Sb +3
Bi + 3
k q +3 ._
Ag
Au+3 Ni +2 Cut2 Fe +a Cdf2 Ph +2 Zn +2
SnCl?,
ml . 14.328 14.35 14.26 14 23 14.22 17.44 14.48 20.42 20.45 14.31 14.26 14.30 ~~
~
Deviationw MI. /O 0.02 0.14 0.03 0.21 0.42 0.06 0.09 0.63 0.70 0.10 21.8 3.12 1.1 0.16 42.6 6.10 6.13 42.8 0.01 0 06 0.02
0.07
0.42 0.14
Average of 3 determinations.
rnv
Figure 1. curves
Effect of hydrochloric acid concentration on shape of tellurium titration 21.3 mg. of Te present in each case
140
ANALYTICAL CHEMISTRY
selenium might be due to more efficient removal of residual nitric acid. I n fact, the tellurium-nitric acid solution can be taken down t o dryness and the residue treated repeatedly with hydrochloric acid without fear of volatilization. The same treatment of selenium solution would be unthinkable. Interferences. For better evaluation of the possible interferences from foreign ions commonly present in tellurium, complete analysis of a typical refined tellurium cast is reported in Table 11. The main impurity in refined tellurium is selenium, followed by lead and oxygen. Selenium, being volatilized in the course of the dissolution of the sample, can cause no error in tellurium determination. Oxygen, in any concentration, has no effect on tellurium determination. The average of all known impurities combined, excluding selenium and
oxygen, is 0.06%, half of which is lead. Even if all these impurities had a biased, cumulative effect proportionate to their individual concentrationh, which is not the case, the differential method should still be considered most satisfactory. Because of the possibility of applying the differential method to other than refined materials, such as crude tellurium and tellurium dioxide muds, the effect of significant amounts of foreign ions 011 tellurium analysis was considered (Table 111). Each impurity added represented 1.86% of the tellurium present. Antiniony and bismuth do not interfere, while the average interference for arsenic, silver, nickel, cadmium, lead, zinc, and palladium is less than 0.25aj, of the total tellurium present. In the case of lead, the main interfering impurity in rcfmed tellurium, the per cent error due to its presence is only one sixth its concentration. For an actual concentration of o.O3y0 lead in refined tellurium, the error in tellurium analysis n-ould be about 0.005%, Fhich is insignificant. For estimating tellurium in a number of tellurium refinery intermediate products containing relatively moderate amounts of tellurium and significant amounts of foreign ions, direct potentiometric titration of tellurium was considered (Table IV) . Each impurity added represented 4G.570 of the tellurium present; the
Table
V.
Typical Analyses of Refined Tellurium Products
%
Material Lot Tellurium 218
yo Te
99.93 99.91 219 99.84 99.75 220 99.93 99.85 221 99.95 100.00 222 99.61 99.64 223 99.61 99.64 224 99.59 99.29 225 99.r9 99.78 Tellurium R550-25 80.03 dioxide 79.99 R550-27 79.95 79.98 R550-29 80.11 80.11 Telluric R501-1654 68.24 acid 68.27 cast
deviation 0.01
0.045
tellurium present. Gold, copper, and iron interfere and some kind of complexing or masking agents mould have t o be introduced in their presence. However, for a higher telluriumforeign ion ratio than that considered, the effect of foreign ions would be less pronounced. RESULTS
0.04 0.025 0.015 0.015 0.00
0.005 0.02 0.015 0.00
0.015
>4v. 0.017 Std. dev. 0.023
tellurium-foreign ion ratio was approximately 2 to 1. Antimony, cadmium, and zinc had practically no effect on tellurium analysis. Bismuth, arsenic, lead, silver, and nickel had only a limited effect, causing an average error of 0.65% of the total
The differential potentiometric procedure for tellurium has been in daily use in our laboratory for almost a year and its reliability proved to our best satisfaction. The results from duplicate runs agree very closely, generally closer than in selenium analysis by the same technique. Some typical duplicate results for refined tellurium, tellurium dioxide, and telluric acid are shown in Table V. Eight of the 12 duplicate results reported agree within 0.015%, the mean deviation for all duplicate runs being 0.017% and the standard deviation 0.023%. The results shown were obtained by skilled technicians in regular routine analysis of current production lots. LITERATURE CITED
(1) Barabas, Silvio, Bennett, P. If7.,. I s a ~ . CHEM.35, 135 (1963). RECEIVEDfor review July 23, 1962. Accepted November 20, 1962. Thirteenth Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Pittsburgh, Pa., March 1962.
Spectrophotometric Determination of Iron with 2,6-Pyrid inedicarboxylic Acid and 2,4,6Pyrid inetrica r boxy1ic Acid ICHIRO MORIMOTO and SUSUMU TANAKA 1aboratory o f Chemistry, Gifu University, Gifu, Japan
b Color reactions between ferrous iron and 2,6-pyridinedicarboxylic acid, and between ferrous iron and 2,4,6pyridinetricarboxylic acid have been studied to determine the optimum conditions for the analytical use of these reagents. The iron complexes themselves are unstable in aqueous solutions, but could be stabilized for several days in the presence of hydroxylamine hydrochloride. Strongest color developed within the pH range from 5 to 6. Ferrous iron complexes have an absorption maximum at 4 8 5 mp in the case of 2,6-pyridinedicarboxylic acid, and at 520 mp in the case of 2,4,6-pyridine-tri-
carboxylic acid. The mole ratio of iron to the reagent in these complexes was 1 :2. The absorbance of these solutions at 5 17 mp obeyed Beer’s law in the ferrous iron concentration range of 1 to 2 0 p.p.m.
T
REDDISH-YELLOW COLOR produced by the interaction of a solution of ferrous salt with various pyridine-2-carboxylic acids was first observed by Skraup (8), and the color reaction was used for the determination of iron with picolinic acid (7) and quinolinic acid (4). We found that the HE
color reactions with 2,B-pyridinedicarboxylic acid and 2,4,6-pyridinetricarboxylic acid were more sensitive than those with picolinic acid and quinolinic acid in the paper chromatography of pyridinecarboxylic acids (6). 2’6-Pyridinedicarboxylic acid and 2,4,6pyridinetricarboxylic acid form soluble ferrous iron complexes of reddish-yellow or reddish-violet color. These complexes exhibit maximum absorption a t 485 and 520 mp between pH 5 and 6 . The reagents are specific for ferrous iron and hence can be used for the spectrophotometric determination of iron. The complexes obey Beer’s law a t 517 mp for ferrous iron concentraVOL. 35, NO. 2, FEBRUARY 1963
141