by the recommended procedure was investigated. The preferred order of addition of the reagents to the volumetric flasks was: Calcein Blue or zirconium, pH 5.5 solution, distilled water to volume. Under these conditions maximum fluorescence due to the zirconium complex is obtained immediately. When the Calcein Blue was diluted with pH 5.5 solution before the addition of the zirconium solution, very little fluorescence was observed initially for the complex. The fluorescence was then found to increase with time and reached a maximum more than one hour after mixing. Effect of Temperature. In all experiments, the temperature of the solutions was 21 =t3 “C. No significant variation of fluorescence intensity at 405 nm with temperature was noted, but no specific study of temperature effects was made. Precision. The recommended procedure was applied to the repetitive determination of 100 ng of zirconium over several days. The eight determinations produced a coefficient of variation of *1.8%. Effect of Foreign Ions. The effect of 30 cations and 12 anions on the determination of 50 ng of zirconium by the recommended procedure was investigated. A 500-fold molar excess of the foreign ion was employed. An ion was considered to interfere at this level when it produced an error in the fluorescence intensity compared to that of 50 ng of zirconium alone of greater than 3 times the standard deviation (5%). The effect of those ions which were found to interfere was subsequently reinvestigated at lower concentrations, uiz. 50-fold and 5-fold molar excess over 50 ng of zirconium. The results are shown in Table I. Thorium, aluminum, and fluoride cause positive interference even at the 5-fold molar excess level, while iron(II1) and tungstate completely quench the fluorescence of the zirconium complex at this level. Phosphate also causes a serious negative interference at the 5-fold molar excess level. The most unexpected interference was that of fluoride. In many other methods for the determination of zirconium, fluoride and phosphate cause serious interference by the formation of stable anionic complexes with the zirconium. In the procedure reported here, phosphate does bleach the reaction in the expected manner, but fluoride gives a positive interference.
The separation of interfering ions in the determination of trace amounts of zirconium has received considerable attention (7). A highly selective extractant for micro amounts of zirconium is thenoyltrifluoroacetone (TTA). Zirconium has been separated from aluminum, iron, the rare earths, thorium and uranium in 6M HCl by extraction with n A in xylene (8) and back-extraction of the zirconium from the organic phase with 0-1M HC1. Any small amount of residual TTA in the O.lMHC1 may then be removed by extraction with a single aliquot of xylene. Freund and Miner (9) have also separated zirconium from iron and aluminum by ion-exchange, Structure of the Complex. The nature of the complex formed between zirconium and Calcein Blue was investigated by the mole-ratio (10) method. The results, after correction for the reagent fluorescence, indicate the formation of a 1:1 complex between zirconium and Calcein Blue. Similar results were obtained by the method of continuous variations (11). CONCLUSIONS
The determination of trace amounts of zirconium by the spectrofluorimetric method with Calcein Blue reported here is very sensitive and moderate amounts of many ions may be tolerated without interference. The sensitivity compares favorably with those of other procedures reported for the spectrofluorimetric determination of zirconium with similar instrumentation. RECEIVED for review December 5, 1969. Accepted February 16, 1970. We thank the Ministry of Technology for financial support of this work, and the Science Research Council for an equipment grant for the purchase of the spectrofluorimeter used in the work. (7) R. B. Hahn, “Treatise on Analytical Chemistry,” I. M. Kolthoff and P. J. Elving, Ed., Part 11, Vol. 5, Interscience, New York, N. Y., 1961. (8) F. L. Moore, ANAL.CHEM., 28,997 (1956). (9) H. Freund and F. J. Miner, ibid., 25,564 (1953). ANAL.ED., 16, (10) J. H. Yoe, and A. L. Jones, IND.ENG.CHEM., 111 (1944). (11) P. Job, Ann. Chirn. Frame, 9, 113 (1928).
Coulometric and Titrimetric Reduction of Iridium Fumed in Perchloric Acid Edward J. Zinser and John A. Page Departmetit of Chemistr.y, Queen’s UriiGersity, Kingston, Otitario, C‘ariadn
MOSTOF THE TITRIMETRIC methods for iridium have utilized the reduction of Ir(1V) to Ir(III), with iron(I1) the recommended titrant ( I ) . These methods have been mostly applied to the assay of pure salts, and little attention has been paid to obtaining the Ir(1V) required for the titration. Reagents that have been used to give Ir(1V) for titration include Cl., with the excess removed by boiling ( 2 ) , and aqua (1) F. E. Beamish, “The Analytical Chemistry of The Noble Metals,” Pergamon Press, Oxford. 1966. (2) D. I. Ryabchikov, Zhur. Anal. Khim., 1, 47 (1946); Chern. Abstr., 43, 2542 (1949).
regia, where the excess was removed by evaporation (3). Heating with an excess of Ce(1V) after an H2S04 fuming has also been used with conflicting claims of an Ir(1V) product ( 4 ) and a species that could be titrated through the states +4.5 to +3.5 to +3.0(5). A particularly useful oxidizing agent is fuming perchloric (3) G. Milazzo and L. Paolini, Rend. Zst. Super. Sariita (Rome), 12,693 (1949); Chem. Abstr., 44,6337 (1950). (4) N. K. Pshenitsyn, S. I. Ginzburg, and L. G. Sal’skaya, Russ. J . Inorg. Chem., English transl., 4, 130 (1959). (5) W. A. E. McBryde and M. L. Cluett, Can. J . Res., 28B, 788 (1950). ANALYTICAL CHEMISTRY, VOL. 42, NO. 7, JUNE 1970
787
1 .o
I
0.8 W
w
0.6
$
0.4
z a m
with a saturated NaCl solution. All potentials reported are referred to this reference. In general, pre-electrolysis of the electrolyte was not possible prior to coulometry; rather the fumed iridium solutions were simply diluted, deaerated, and electrolyzed, Initial currents at both Hg and Pt cathodes were of the order of 5-20 mA and these decayed to constant residual values, always less than 10 PA, in 30-45 minutes. For coulometry, a correction equivalent to this background over the entire electrolysis time was deducted from the quantity of electricity used. The correction was never greater than 0.5 The electrolyses were carried out in deaerated solutions; the Nn gas was of high purity (Matheson,
v
2
W
0.6 0.5
c
0.0 VOLUME OF Fe(ll) TITRANT (ML)
0
in the fuming system, but past this point the curve showed two inflection points. Four aliquots, each extensively fumed, gave to the first break nl = 0.26 =t 0.02 and from the first to the second nz = 1.34 f 0.03. Seven other samples were extensively fumed, cooled, diluted, and mixed with an excess of the iron(I1) solution. In these cases, back titration with K2Cr20,solution yielded an overall n = 1.67 f 0.02 for the iridium reduction. The voltammetric behavior of the fumed solutions was also investigated. A 50-ml aliquot of stock solution was extensively fumed with 25 ml of acid, then diluted to 1000 ml. Voltammetry gave a composite anodic-cathodic wave with El,* of +1.00 V, and a cathodic wave with Eliz of +0.64 V. Triplicate coulometric reductions on Pt using the fumed solutions gave nl = 0.15 =t0.01 at +0.83 V and ti2 = 1.51 i: 0.01 in a further electrolysis at 0.00 V. If the final product of the reductions is Ir(III), which is consistent with the electrochemistry (10) and the spectra, then the maximum state achieved in perchloric acid fuming of the iridium is 4.67 (3 1.67 f 0.02). This was observed only by immediate reaction with excess iron(I1). A stable oxidation 1.31 f 0.01) was also state of 4.33 (3 1.34 f 0.03 and 3 indicated by the inflection points in the iron(I1) titration and by the coulometry at 0.00 V. These values are close to fractional oxidation states of +42/3 and +4’/2 and may be rationalized in terms of iridium trimers with mixed 5,5,4 and 5,4,4 oxidation states. Analogous iridium trimers with mixed 4,4,3 and 4,3,3 oxidation states have been reported (15,16). This hypothesis is further supported by the absorption spectra (Figure 3), for that of the fumed solution (5,5,4) differs only in intensity from that of the fumed solution reduced at +0.83 V (5,4,4). This behavior is characteristic of Class 111-A mixed oxidation state compounds (17). Investigation of Bound Volatile Oxidant Postulate. The experiments of Jackson and Pantony (7) were also repeated using 25-ml aliquots of stock iridium solution. These were fumed with 20 ml of acid, cooled, and diluted to 250 ml. One 100-ml portion of this solution was immediately titrated with
+ +
+
(15) C . K. Jqrgensen, Acta Chem. Scand., 13, 196 (1959). (16) C . K. Jgrgensen and L. E. Orgel, Mol. Phys., 4, 215 (1961). (17) M. B. Robin and P. Day, “Advances in Inorganic Chemistry and Radiochemistry,” 10, 317 (1967), Academic Press, New York.
400
700
WAVELENGTH (NM) Figure 3. Spectra of iridium solutions
Figure 2. Titration of iridium heated in perchloric acid 27 min n = 1.00; 0 34 min n = 1.11; A 75 min nl = 0.28, nz = 1.35. 5.58 mg Ir, 0.01083 M iron(I1) titrant Actual fuming began between 27 and 34 minutes heating
300
200
Solution fumed exhaustively with perchloric acid and diluted to give 0.581mM iridium A . after dilution, B. after reduction at +OS3 V, C. after reduction at 0.00 v
Table I.
Potentiometric Titration of Fumed Iridium
Immediate titration nl nz noverall 0.23 0.23 0.21
1.25 1.25 1.28
1.48 1.48 1.49
After digestion under NS nl nz {lovera11 0.07 0.04 0.06
1.35 1.33 1.43
1.42 1.37 1.49
iron(II), while a second portion was digested at 90-95 O C for 8 hours with NP gas passing through the solution at 10 liters per hour. At the end of this time, the solution was cooled and titrated. The titration curves all showed two inflection points as expected, but while digestion drastically reduced the amount of reductant necessary to reach the first equivalence point, the amount required to reach the second equivalence point remained relatively large (Table I). This is contrary to the results of Jackson and Pantony. Attempts were made to condense any volatile oxidants displaced from the solution during the digestion period using a trap at liquid Nz temperatures, but these were unsuccessful. The efficiency of the trap was confirmed in an experiment where the fumed solution transferred to the pot was made 1 M in added NaCl before digestion under NP. Digestion gave the characteristic red-brown color of IrC16*-, and titration gave one inflection point with n = 1.00, while iodometric titration of the trap contents showed oxidants equivalent to those lost from the pot. Analytical Procedure. Procedures based on the direct fuming of the iridium were found to give variable titration results, but it was found that on digestion of the fumed iridium with chloride there was a liberation of Clz and a clearly defined IrC162-species was produced for titration. Aliquots of solution containing 5 to 30 mg of iridium as Ir(1V) or Ir(II1) were transferred to a covered 150-ml tall form beaker and mixed with 12 ml of 7 2 x perchloric acid. The solution was fumed strongly for 1 to 2 hours giving about 8 ml of concentrated acid. This was cooled, transferred to a 100-ml flask and diluted almost to the mark. Transfer and dilution was effected with a 1.15M HC104-1.10M NaCl electrolyte; this gave a final 2.OM HCI04-1 .OMNaCl solution. The flasks were then heated on a steambath for 10 hours with ANALYTICAL CHEMISTRY, VOL. 42, NO. 7, J U N E 1970
789
Table II. Determination of Iridium after Fuming and Anation Sample Results Ir(1V)
11.22mg(3) 28.08 mg (2) 5.64 mg (3) 11.22 mg (6)
Ir(1V)
28.08 mg (2) 12.07 mg (4)
Ir(I1I)a
30.19 mg (2)
11.21, 11.12, 11.24mg 27.96,27.95 mg 5.64, 5.56, 5.58 mg 11.16, 11.18, 11.16, 11.10, 11.23, 11.19mg 28.01, 28.05 mg 12.01, 12.09, mg 12.03, 11.99 mg 29.88, 30.30 mg average error -0.056 mg
Method titration 0.01603M Fe(I1) coulometry
R,0.00 v
coulometry
Pt,0.00 v
pooled standard deviation 0.116 mg 14.6 mg Ir coulometry 14.8 mg Ir FY, 0.00 v 15.2 mg Ir 15.3 mg Ir 15.8 mg Ir calculated 24.6 Ir found 24.0 i O . O % lr Iridium(II1) prepared from iridium(1V) by reduction with Hz02. Excess HZOZ was destroyed by acidifying and boiling the iridium containing solution. b Vaska’s Compound (18)-Supplied C. V. Senoff. 61.1 mg 61.1 mg 63.3 mg 64.2 mg 65.8 mg
0
Ir(1V) 5.64 mg 11.22 11.22 11.22 11.22 11.22 11.22 28.08 28.08 11.53 11.53 11.53 28.84 28.84
Table 111. Determination of Iridium in Presence of Other Platinum Metals Sample Found Rh(I1I) Ru(II1) Os(1V) Ir 3.21 mg ... ... 5.47 mg 1.61 3.21 3.21 3.21 3.21 8.02 8.02 8.02
... ... ... ... ... ... ... ...
... ... ... ... ...
11.12 11.29 11.01 11.42 11.20 11.24 ... 28.05 ... 28.03 11.46 ... 5 mg 10 mg ... 10 5 11.52 11.48 ... 10 10 5 5 ... 28.61 ... 10 5 28.82 Ir determined by coulometry at 0.00 V on Pt. Rh subsequently determined by coulometry at -0.30 V on Hp. Average error in Ir found -0.048 mg. Pooled standard deviation 0.146 mg.
the stopper in place. The solution was then transferred to a covered 150-ml tall form beaker and gently boiled for just one hour to displace any dissolved Clz before titration. The results, shown in Table 11, exhibit a small negative bias for the amount of iridium found relative to the coulometric standardization. This procedure was applied to the analysis of an iridium(1) organic complex. Here the samples were first treated with 12 ml of 72 perchloric acid together with 5 ml of concentrated nitric acid. The mixture was gently heated for 30 to 40 minutes to effect decomposition; then the excess HNO, was boiled off before the strong HC104 fuming. The results are shown in Table 11. The effect of added ruthenium, osmium, and rhodium was also investigated and the results are shown in Table 111. The first two metals, would be expected to be volatilized as tetroxides from the fuming HClOa solution, and no interferences were detected. The rhodium, added as Rh(III), would not (18) L. Vaska and J. W. Diluzio, J. Amer. Chem. SOC.,83, 2784 (1961). 790
ANALYTICAL CHEMISTRY, VOL. 42, NO. 7, JUNE 1970
...
Rh 3.23 mg 1.58 3.15 3.17 . . *
3.14 8.04 7.96 7.95
... *.. ...
...
...
be expected to be oxidized by the fuming and did not interfere in the coulometric determination. The rhodium was subsequently determined by coulometry at -0.30 V on a Hg cathode. The procedure as outlined was found to be critical. If the flask was not kept closed in the anation step, significantly low results were obtained. Blank titrations indicated that this was due to reducing impurities in the reagent NaC1, and sealing of the flask evidently results in their oxidation without reduction of Ir(1V). The digestion step was slow and a period of approximately 8 hours was necessary for the full development of the characteristic IrClsz- spectra. The Clz produced in this step would interfere in the subsequent titration, but it was removed by gentle boiling. RECEIVED for review February 3, 1970. Accepted April 15, 1970. This study was supported by operating grants from the National Research Council of Canada and The Province of Ontario. The award of a National Research Council Fellowship to E. J. Zinser is gratefully acknowledged.
