Polarographic studies in aqueous hydrofluoric acid using ac and dc

May 1, 2002 - A. M. Bond , T. A. O'Donnell , and R. J. Taylor. Analytical ... Alan M. Bond , Thomas A. O'Donnell , A. B. Waugh , and R. J. W. McLaughl...
0 downloads 0 Views 779KB Size
Polarographic Studies in Aqueous Hydrofluoric Acid Using AC and DC Rapid Techniques A. M. Bond and T. A. O’Donnell Department of Inorganic Chemistry, University of Melbourne, Parkville, 3052, Victoria, Australia

Conventional polarographic procedures cannot be used in hydrofluoric acid solutions because of etching of the glass capillary of the DME. However, very rapid polarographic techniques with both ac and dc polarography have been shown to be effective because insignificant glass attack occurs in the short scan time. The necessity of constructing a DME from Teflon (Du Pont) or other hydrofluoric acid resistant material is therefore removed. The method has been applied to the systems Cd(ll), TI(I), Pb(ll), Sn(ll), Bi(lll), Mo(VI), and U(VI) in 5 to 40% hydrofluoric acid solutions. The polarographic characteristics of most electrode reactions studied in aqueous hydrofluoric acid are consistent with those expected in a solution containing a weak acid. Analytically, the most significant feature is that, for any particular concentration of metal, the value obtained in most cases is independent of wide variations in the hydrofluoric acid concentration. This simplifies analysis of mineralogical and metallurgical samples which must be dissolved in hydrofluoric acid.

To DATE polarography in hydrofluoric acid solutions has been very limited because of etching of the capillary of the conventional glass dropping mercury electrode, DME, leading to erratic drop behavior and nonreproducible results. Reported polarographic studies in concentrated hydrofluoric acid media have used DME’s constructed from Teflon (Du Pont) ( I , 2). However, these are not available commercially and are rather difficult and costly to manufacture, requiring specialized equipment and materials (3). Other workers ( 4 , 5 ) , have attempted to use alternative noncorrosive DME’s without marked success, although a glass DME coated with Tygon (U. S. Stoneware Co.) has been found to be satisfactory in acidic sodium fluoride solutions (4). Frequently in analysis, hydrofluoric acid is used to dissolve a sample because of its specific reactivity with silicate or because of the ability of fluoride to complex many cations strongly. It is commonly used for the determination of metals in glasses and in mineralogical and metallurgical samples. It is also extremely useful, and often essential, in dissolving samples such as the metals, alloys, and oxides of zirconium, hafnium, titanium, tin, vanadium, niobium, tantalum, tungsten, and aluminium. In principle, once the sample is dissolved in aqeuous H F any appropriate chemical or instrumental method of analysis can be used. However, fluoride ion interferes with many standard methods of analysis and often a second stage is necessary to re(1) H. P. Raaen, ANAL.CHEM., 34, 1714 (1962). (2) J. B. Headridge, A. G. Hamza, D. P. Hubbard, and M. S. Taylor, “Polarography 1964,” Vol. 1, Proc. 3rd International Conf., Southampton, G. J. Hills, Ed., Macrnillan, London, 1966, p 625. (3) H. P. Raaen, R. J. Fox, and V. E. Walker, AEC Report ORNL3344, Nov. 30 (1962). (4) S. S. Mesaric and D. N. Hurne, Znorg. Chem., 2,788 (1963). (5) V. S. Griffiths and W. J. Parker, Anal. Chim. Acta, 14, 194 (1956).

move the interfering fluoride ion. This second stage is usually tedious and introduces additional errors. There are obvious advantages in carrying out determinations directly on the hydrofluoric acid solutions. HF can be used in polarography as both solvent and supporting electrolyte. Being a weak acid, it buffers both fluoride and hydrogen ions, so that control of both of these variables in analysis need not then be critical-Le., results will not depend markedly on the amount of hydrofluoric acid used in dissolving the sample. Direct polarographic analysis in fluoride media can be performed easily by preparing standards and solutions for analysis in hydrofluoric acid, providing suitable experimental procedures are available. In the present work, the advantages of the conventional glass DME were retained by using rapid polarography. Short drop times permit fast scan rates of potential so that the glass DME has much less contact with hydrofluoric acid than in conventional polarography and etching is reduced. Cd(II), Tl(I), Pb(II), Sn(II), Bi(III), Mo(VI), and U(V1) were chosen as representative systems in H F and results are given below and can be compared with those obtained using a Teflon DME in H F ( I , 6-8) and in HF-NH4F media (2) and with those obtained in sodium fluoride solution with a glass DME (9) where no etching problems occur. Polarographic (or voltammetric) studies using HF-resistant stationary, solid electrodes would also be possible alternatives to the use of a glass DME; but the lowered reproducibility normally associated with these types of electrodes and the difficulties of correlating results with systems which have been measured with DME’s makes retention of the DME desirable in the elucidation of the polarographic characteristics of the aqueous H F system. Analytically, the present work has two separate advantages over existing approaches to the determination of samples dissolved in HF. First, a simple glass DME is quite reliable provided rapid polarographic techniques are used. Second, the analytical results obtained are independent of H F concentration over a fairly wide H F range. If this were not so, extreme care would be needed during sample dissolution and subsequent solution preparation in order that the H F concentration of the resulting solution should be known accurately. The volatility of H F itself and its ready ability to react with silicates would increase the difficulty of this task enormously. However, because the determination of most metals has been shown to be independent of H F concentration over wide ranges, it is only necessary to have sample solutions with approximately the same H F concentration as that of the standards for reliable results to be obtained. ~

~~

~

(6) H. P. Raaen, ANAL.CHEM., 36, 2420(1964). (7) Zbid.,37, 677 (1965). (8) Zbid., p 1355. (9) P. W. West, J. Dean, and E. J. Breda, Collect. Czech. Chem. Commun., 13,1 (1948).

ANALYTICAL CHEMISTRY, VOL. 41, NO. 13, NOVEMBER 1969

1801

Using DC polarography, plots of id us. concentration of metal gave satisfactory calibration curves for all systems studied. However, calibration curves based on ac techniques were acceptable only when there was no marked degree of irreversibility of the electrode process. EXPERIMENTAL Reagents. All solutions containing hydrofluoric acid were

prepared and stored in polythene containers. Standard solutions were prepared in distilled water from the fluorides of cadmium and tin, the nitrates of thallium, bismuth, uranium, and lead, and from ammonium molybdate. Sufficient perchloric acid was added to prevent hydrolysis of the bismuth solution. Stannous fluoride was dissolved in deoxygenated solution, as there was evidence of a steady decrease in diffusion current with atmospheric oxidation of any particular solution. Hydrofluoric acid solutions were prepared by dilution of 49 % Analytical Reagent (Baker Analyzed). Apparatus. All polarograms were obtained using a Metrohm Polarecord E 261 in conjunction with Metrohm Polarographie Stand E354 for rapid polarography. For ac polarography, the Metrohm AC Modulator E393, with an applied ac voltage of 10 mV at mains frequency of 50 cps was used. A controlled drop time of 0.16 sec was used with scan rates of potential of between 0.5 and 3 V per min. Scan rates of 0.5 V per min for solutions up to 20% in HF, 1 V per min for solutions 20-30% in H F and 2 V o r 3 V per min for solutions in the 30-40% H F concentration range were used unless otherwise stated. All polarograms were recorded at (25 =t1) "C. Oxygen free nitrogen was used to de-aerate the solutions. No maxima were encountered and maximum suppressors were not added. Solutions were analyzed in a polythene beaker with the glass DME dipping into the solution. A salt bridge, constructed from Kel-F, fitted with Teflon sinters, and filled with saturated potassium chloride, connected the solution for analysis with a beaker of saturated potassium chloride in which were a reference electrode (either saturated calomel or silver-silver chloride) and a tungsten auxiliary electrode to minimize cell impedance. Procedure. The junction potentials of the experimental arrangement used, which could be quite high, were unknown, and neither half wave potentials, El/*,from dc polarography nor summit potentials, E,, from ac polarography were corrected for junction potentials. Eliqvalues were usually estimated by the simple method of Frank and Hume (10) rather than by the more usual logarithmic plots of Ed e . us. log (i/id - i) because good agreement was obtained and the latter method is rather more tedious. Unless otherwise stated, the ElIz and E, values quoted are those measured against SCE. Values for diffusion currents, id, from dc polarography and wave heights, id-, from ac polarography with Ag-AgC1 reference electrodes were identical with those obtained with the SCE as expected, and either reference electrode was extremely satisfactory with the experimental arrangement used. T o obtain a polarogram for a particular solution, the solution for analysis was prepared in a polythene container, Deoxygenation and thermostating were carried out in this container before the solution was placed in contact with the glass DME. Immediately the solution was in contact with the glass DME., the polarogram was recorded. As soon as a scan was completed and while mercury was still dropping, the beaker containing the solution was replaced by one containing a saturated solution of sodium bicarbonate to neutralize any HF adhering to the DME. Before the capillary was used again with another solution, (10) R. E. Frank and D. N. Hume, J. Arner. Chem. Soc., 75, 1736 (1953).

1802

sodium bicarbonate was removed by washing with distilled water. Using this procedure meant that for dc polarograms, the glass DME was in contact with H F for about 25-30 sec and for ac polarograms about 20-25 sec. If necessary, considerably shorter contact times could be used with ac polarograms because all the required E, and id- data can be obtained from scanning only the portion of the polarogram about the summit peak. This, however, was not done in this work and complete polarograms were always recorded When the procedures described here were adopted, there was no significant attack of the glass D M E for the concentration range 5 to 15% H F over the whole duration of the studies. After lengthy periods of use in 15 to 25% H F occasional attack of the DME was observed, indicated on polarograms by erratic drop behavior at random intervals. The DME was quickly restored to normal use simply by cutting of the tip. This indicates that H F attack occurs only at the orifice of the capillary from which the drops of mercury actually fall and that seepage up the capillary, which would cause internal damage, does not occur to any significant extent. In 25-40% HF, attack was quite rapid and frequent restoration of the glass D M E was found to be necessary with dc polarography. With ac polarography however, this was not quite so marked because of the shorter time required to record a polarogram. In view of this, except for some initial work in each of the systems, ac polarography only was used in the high concentrations of HF. Studies in 40 to 50% and greater H F concentrations were not found to be possible because attack on the glass DME was too rapid to give useful polarograms. At higher than 40% HF, drop behavior is completely erratic, and polarograms cannot be interpreted easily because of lack of reproducibility. RESULTS AND DISCUSSION

Studies of the systems Cd(II), Tl(I), Pb(II), Sn(II), Bi(III), Mo(VI), and U(V1) were carried out in the range 540% hydrofluoric acid (2.5 to 23M). Half wave potentials, Ell2, from dc polarography and summit potentials, E,, from ac polarography are reported us. SCE. These values are only reported correct to 0.01 because the conditions of operation used in this work are insufficient to allow the junction potentials present to come to equilibrium and consequently reproducibility of Eliz and Ea was not high. However, reproducibility of diffusion currents id and wave heights id- obtained respectively from dc and ac polarograms was excellent. Reproducibility of id- was found to be slightly better than for id. A major contributing factor to the greater reproducibility of idcompared with i d was the very low background current found in rapid ac polarography. Concentrations of electroactive metal ion species were M to examined polarographically over the range 5 X M and both id and id- were plotted against concentration of metal ion species in various H F concentrations. These were used as calibration graphs for analysis by comparison of unknown solutions. Aqueous H F was found to be a good conducting medium and no other supporting electrolyte was necessary. Over the concentration range studied plots of id obtained from dc polarograms and corrected for background current, us. concentration of metal were, in general, linear. This is expected for diffusion controlled reactions which follow the appropriate form of the Ilkovic Equation for conditions of rapid polarography. Plots of i d N us. metal ion, however, tended to show very slight curvature at the higher metal ion concentrations. The advantages of ac polarography over dc procedures in analytical applications reported by Breyer and

ANALYTICAL CHEMISTRY, VOL. 41, NO. 13, NOVEMBER 1969

Table I. AC and DC Data for Some Metal Systems in 5% and 25% HF Obtained by the Rapid Polarographic Method Published Eli2 values, V 0.1M HF5 Z HF 0.1M 0.5M E114 - E 3 t 4 ac half 25% HF NH4F NaF Variation of i d and System Elin V mV E. V width, mV Elil V E, V from (2) from (9) idwith concentration Cd(II)-Cd(Hg) -0.61 30 f 2 -0.61 48 f 2 -0.60 -0.60 -0.59 -0.63 idandid- constant in 5 % to 25% HF. Plots of i d andidUS. [Cd(II)] linear f

Tl(I) + Tl(Hg)

-0.47

60 f 2

-0.47

93 f 3

-0.48

-0.48

-0.46

-0.50

Pb(II)-.Pb(Hg)

-0.41

38 f 2

-0.40

56 f 2

-0.45

-0.44

-0.40

-0.41

-0.59

40 f 2

-0.59

58 rt 2

-0.66

-0.66

-0.60

-0.73

Sn(II)+Sn(IV)

-0.12

42 i 2

-0.11

60f 2

-0.06

-0.06

-0.23

-0.20

Bi(III)+Bi(Hg)

-0.41

ca.

110

-0.41

ca.

200

-0.42

-0.42

-0.13

-0.07

Sn(I1)

-

Sn(Hg)

M O W )+ MOW)

-0.16

CU.

160

-0.16

CU.

220

-0.15

-0.15

-0.53

U(V1) + U(V)

-0.63

CU.

150

-0.69

CU.

250

-0.54

-0.58

-0.51

... .

.

I

idand id- constant in 5x to 2 5 % HF. Plots of i d and i d N us. [Tl(I)] linear dc and ac calibration curves linear for [Pb(II)J < 5 X 10-4M. Curvature observed at higher [Pb(Wl i d for anodic and cathodic wave same except for sign. dc and ac calibration curves linear. idY anodic < i d N cathodic. i d N and i d decrease slightly with increasing [HF] dccalibrationcurvelinear.

ac method not very sensitive dc calibration curve linear. ac method not very sensitive dc and ac calibration curves linear. ac method not very sensitive

Bauer (11) were found to hold in hydrofluoric acid medium. For reversible electrode reactons, therefore, ac polarography is recommended. AC polarography in high concentrations of H F was also preferred to dc for the experimental reasons given earlier. For nonreversible electrode reactions, in which the ac method is not sufficiently sensitive, dc techniques are, however, favored. One of the significant features of the work is that in general id and id- values for a given concentration of electroactive metal ion were found to be almost independent of HF concentration over quite wide ranges of change in HF concentration. Details of this constancy, and of other observations such as values of Ell2,Eli4 - E3/4,E,, and the halfband width for ac waves are given in Table I and discussed later. Criteria for Reversibility of Electrode Reactions. Electrode reactions occurring at the D M E are normally classified as reversible or nonreversible. Reversible electrode reactions are observed when Nernstian conditions prevail at the DME. Such electrode reactions produce dc polarograms which can be described by the equation:

(11) R. Breyer and H. H. Bauer, “Alternating Current Polarography and Tensammetry,” Interscience, N~~ YorklLondon,

1963.

Consequently, the dc criterion of reversibility normally used is a plot of Ed.,.us. log (ilia - i ) which should be linear with slope -2.303 RT/nF. Alternatively, and more simply, ( & 4 E3/4)values can be measured to test the degree of reversibility. For reversible electrode reactions (Eli4 - E3/4) values can be shown to be -2.303 RT/nF log 9. For reversible 1, 2, and 3 electron reduction processes at 25 ‘C,(E114 - Ella) values are thus 56, 28, and 19 mV, respectively. For nonreversible electrode processes (quasi-reversible or irreversible) (Elir E3/4) values are higher than those observed for reversible electrode reactions. Table I gives the (Ell4 - E3j4)values for the electrode reactions studied in this work. Further criteria used to examine reversibility of the electrode reactions were as follows : ( 1 ) Ell2 and E$ values should coincide and be independent of concentration of species electroactive at the DME for reversible processes. (2) AC polarograms should have a width at half the wave height of approximately 90/n mV for reversible electrode reactions. Irreversible processes have larger half widths than 90/nmV. (3) AC polarograms for reversible electrode reactions should be symmetrical and have the shape of the first derivative of the reversible dc polarogram. Nonreversible ac polarograms need not be symmetrical.

ANALYTICAL CHEMISTRY, VOL. 41, NO. 13, NOVEMBER 1969

1803

Discussion of Individual Systems Studied. CADMIUM(II) THALLIUM(~). The Cd(I1) and Tl(1) electrode reactions were observed to be polarographically reversible by the agreement of El 2 and E,, the values of (Eli4 E3l4)and the ac half-width (See Table I). Furthermore, Ell2 and E, were found t o be independent of metal ion concentration and ac polarograms were symmetrical. The virtual independence of i d and id- upon HF concentration, over the range 5 t o 40% is consistent with the formation of only very weak fluoro-complexes of Cd(I1) (12-14), and of Tl(1) (15,16) and the buffering action provided by HF as a weak acid. E112 and E, values were observed to vary only slightly with change in H F concentration, which is again consistent with only very weak fluoride complexing. Linear ac and dc calibration curves make these both ideal systems for polarographic analysis. LEAD(U). The Pb(I1) electrode reactions showed a fair degree of reversibility. E1/2 and E, values were observed t o coincide and symmetrical ac waves were obtained. (Eli4 E314)and ac half widths were, however, slightly higher than theoretically expected for a reversible electrode reaction. The very short drop times used in rapid polarography (0.16 sec) may not be sufficient for complete attainment of equilibrium of the electrode reaction. This could cause the slight departure observed from polarographic reversibility. Even at longer drpp times, however, equilibrium conditions may still not be attained as Raaen (8), using a Teflon D M E with drop times of 3.5 to 4.0 sec observed that slopes of Ed.e.us. h g ( i / i d i ) were 36 mV in 1M HF and 33 mV in 12M HF, compared with the theoretical value of 30 mV at 25 "C. Both dc and ac calibration curves were found to be linear up to about 5 x 10-4M. Above this concentration pronounced curvature occurs. This may well be due t o the insplubility of lead fluoride at the higher lead concentrations. Mesaric and Hume (4) have evaluated polarographically the fluoride complexes of the lead(I1)-fluoride system. Their work shows that lead(I1) fluoride complexes are considerably more stable than either cadmium(I1) or thallium(1) fluorides and the observed negative shifts of Ellz and E$ and the very slight lowering of id and idCV with increased HF concentration may be explained in terms of complex formation. This interpretation, however, is not unambiguous because, as has been mentioned previously, there is some doubt that equilibrium conditions are obtained for lead(1I) under conditions of rapid polarography, and changes in kinetic parameters at the electrode with changes in H F concentration could also account for the variation in these values. TIN(II). For the tin(I1) system, twQ very well defined waves were observed in both ac and dc polarograms. The i d values of both the cathodic and anodic dc waves were found t o differ only in sign, which provides evidence for both electrode reactions involving the same number of electrons. With ac polarography, no such distinction between anodic and cathodic waves is possible as id- has the same sign in both cases. The ac anodic wave was, however, observed t o have a slightly lower i d N value than the ac cathodic wave. Analytically, the occurrence of two tin(I1) waves in HF solution is quite useful. Either wave can be used t o give two independent analyses of the ane sample. Alternatively, if a n AND

-

(12) D. R. Crow, J . Elecrroanal. Chem., 16, 137 (1968). (13) S.S.Mesaric and D. N. Nume, Inorg. Chem., 2,1063 (1963). (14) A. M. Bond, J . Efectroaiiaf. Chem., 20, 223 (1969). (15) R. P. Bell and J. H. B. George, Trans. Farpday SOC.,49, 619 (1953). (16) A. M. Bond, unpublished work, University of Melbourne, Australia, 1968.

1804

interfering species is present with a wave overlapping one of the tin(I1) waves, the other can be used because of the large separation of anodic and cathodic waves. Schaap, Davis, and Nebergall (17) have shown polarographically that both Sn(I1) and Sn(1V) form very stable fluoride complexes. These very high stabilities may well explain the observed shifts of Eliz and E, and the slight lowering of i d and i d N with increasing HF concentration. Schaap et a/. also reported that the anodic wave was not as reversible as the cathodic wave and approached reversibility only in acidic fluoride solutions. This may well explain why in ac polarography i d - for the anodic wave is slightly less than for the cathodic wave. The height of the ac wave is particularly sensitive toward the reversibility of the electrode reaction and probably in aqueous HF, the anodic wave is less reversible than the cathodic wave. Both electrode reactions were observed t o show a high (but not complete) degree of polarographic reversibility as was previously described for the lead(I1) electrode reaction. (Ell4and ac half widths are slightly higher than theoretically predicted for reversibility (see Table I), but other criteria such as symmetrical shape of ac wave, independence of Ellzand Es of tin(I1) concentration and coincidence of El/*and E8are satisfied. BISMUTH(III).The bismuth(II1) standard solutions used in this work were prepared by dissolving bismuth(II1) nitrate in perchloric acid, as described earlier, thus the medium for this system is, strictly speaking, HC104/HFrather than pure aqueous HF, as used with all other systems. The wave observed was found t o be irreversible. The assignment of the electrode reaction as a three electron reduction Bi(II1) 3e e Bi(ama1gam) was made by comparison with results from other electrolytes where almost invariably the bismuth(II1) reduction wave has been shown t o be a three electron process. The decrease in (Ell4 - E3i4)and ac half widths with increase in HF concentration suggests that the rate of the electrode reaction or reversibility increases with increasing H F concentration. This idea is also supported by the increased symmetry of the ac wave. In 5 HF the ac wave is nonsymmetrical, but at 2 5 x HF, complete symmetry is observed. The possible explanation of this may lie in the increased acidity of the higher HF concentrations suppressing hydrolysis reactions of bismuth(II1) or of forming monomeric fluorocompIexes of Bi(II1). In weakly acidic solutions, bismuth (111) exists as BiO+ but more particularly as polymeric oxoand hydroxy-species. Only in very highly acidic solutions is there any evidence for a n aquo-bismuth(II1) cation. Polarographic reduction of BiO+ and of polymeric ions in particular is likely to be a slower electrode process than reduction of Bi3+ or of simple bismuth-fluoride complexes. Hence any conditions, such as a n increase in concentration of HF, which increases the concentration of acidic species and favors formation of Bi3+,or of simple fluoro complexes, may lead t o a n observed increase in apparent polarographic reversibility. Insufficient data is available in the literature on the fluoro complexes of bismuth(II1) to quantitatively assess the likely effect of change in fluoride concentration with increasing H F concentration. However, polarographic studies by Bond (18) have indicated that the stability is not sufficiently great t o or id with change in HF concause significant changes in centration and this was found experimentally.

+

(17) W. B. Schaap, J. A. Davis, and W. H. Nebergall, J . Amer. Chem. SOC.,76, 5226 (1954). (18) A. M. Bond, J . Electroanal. Chem., in press.

ANALYTICAL CHEMISTRY, VOL. 41, NO. 13, NOVEMBER 1969

Linear dc calibration curves were found, which makes the dc polarographic method of analysis very suitable for determination of bismuth(II1). Id- values from ac polarography were rather low, and the ac method was not very sensitive for analysis of Bi(II1). Until recently it was believed that ac waves were not observed for nonreversible electrode processes. However, Smith and McCord as well as others (19,20) have examined theoretical aspects of ac polarography and have shown that ac waves can be expected for quasi-reversible and irreversible electrode reactions. The sensitivity of the ac method of analysis, however, is not very great for irreversible electrode reactions and for irreversible processes the dc method is to be preferred. Molybdenum(V1). One reasonably well defined wave was observed in dc and ac polarograms. (Eli4 - E3/4) and ac half widths observed (Table I) show that the electrode reaction is nonreversible. The ac wave showed slight departure from symmetry, but was very broad and only extremely small idvalues were observed as expected from an irreversible ac wave. E, and Ellzvalues in the various concentrations of H F were found to lie in the range -0.14 to -0.18 V, the absolute value depending upon concentration of both Mo(V1) and HF. In most other media, Mo(V1) can produce two waves attributable to consecutive reduction of Mo(V1) to Mo(V) and then of Mo(V) to Mo(II1) (21). In acidic supporting electrolytes, the Mo(V1) to Mo(V) wave shows E112 values in the range -0.1 to -0.3 V (21) and as the wave observed in H F was also found in this range, the electrode reaction has been designated as a one electron reduction step Mo(V1) e MOW). This assignment is consistent with the work of Headridge et al., who used a Teflon DME to study the polarographic behavior of Mo(V1) in a number of mixed H F electrolytes ( 2 ) and polarographically determined molybdenum in niobium-base alloys using HF/HzSO4 as the supporting medium (22). In 0.5M HF-O.5M NHaF, they found Ell2 for the Mo(V1) to Mo(V) wave to be -0.53 V, and for the Mo(V) to Mo(II1) wave, to be -1.22 V. The second wave coalesced with the wave for reduction of hydrogen ion. In the considerably more acidic medium of 0.1M H2S040.5M H F , Headridge and coworkers observed only the one wave for reduction of Mo(V1) to Mo(V) with an Ellz of -0.13 V. Their results also show that the more acidic the media, the less negative the El;zvalues for this wave. In the pure aqueous H F media used in this present work, the wave due to hydrogen ion was observed at about - 1.O V and presumably the wave obscures the wave for reduction of Mo(V) to Mo(II1) which is observed at -1.22 V in the less acidic HF/NH4Fmedia ( 2 ) . D C analytical calibration curves were found to be linear. AC calibration curves, however, showed curvature, and because of this nonlinearity and the low sensitivity of the ac method toward irreversible electrode reactions, dc polarography is more suitable for analysis of molybdenum. Uranium(V1). One fairly well defined wave was observed in dc polarograms of uranium(V1). In ac polarography, the wave observed was very broad and nonsymmetrical with very These observations show that the elecsmall values of id-. and ac trode reaction is nonreversible as do the (E1 half-widths given in Table I.

+

Table 11. Dependence of E, for Uraniurn(V1) on Fluoride ConcentrationQ Concentration of sodium fluoride, M 0 2 4

6 8 1.4 2.0

US.

Ag/AgCI(V) 0.14 0.16 0.20 0.24 0.26 0.27 0.28 0.35 0.59 0.70

x 10-4 x 10-4 X

x 10-4

x x 4.6 x 1.0 x 2.0 x

a

-E,

10-3 10-3

10-3 10-2

10-2

Uranyl nitrate = 10- ‘M. Sodium perchlorate = 1.00M.

Using a Teflon DME, Raaen ( I ) has similarly found that in aqueous H F one well-defined dc wave is obtained for U(V1). Similarly in 0.1M HF-O.1M NH4F (2) only one wave was found. However, in neither of the above studies, was an assignment of the electrode reaction made. Kolthoff and Lingane (23) have summarized the results of polarographic reduction of U(V1) in various media. In weakly acidic to highly acidic solutions, U(V1) exists as the uranyl cation, UOz2+and in perchloric or hydrochloric acid one of the waves observed is the reversible one electron reduction step UOz2+ e e UOz+. Bond and O’Donnell (24) have observed that Eliz values, and the nature of the electrode reactions of uranium are particularly sensitive toward small concentrations of fluoride and it was thought that the wave abserved could also quite likely involve the electrode reaction U(V1) e eU(V). Table I1 shows that in 1M NaC104, Egvalues of the urae U(V) change markedly with nium(V1) wave, U(V1) added fluoride ion. In the absence of fluoride at pH 5 (Eli4E S l 4was ) found to be 60 mV for the reduction wave and the ac half width 90 mV, confirming the one electron reversible reduction step mentioned above. However, when 0.02M NaF was added, E, shifted considerably and had approximately the same values as observed in HF. Furthermore, with this concentration of fluoride the ac id- value decreased drastically, compared with the reversible electrode reaction observed in the absence of fluoride, and a broad, nonsymmetrical wave was observed with similar characteristics to the wave found in HF. Thus the assignment of the electrode reaction observed for e S U(V) wave seems satisU(V1) in H F as the U(V1) factory. Linear dc calibration curves were obtained and were very suitable for analysis of uranium(V1). AC calibration curves obtained in 5 and 25 H F were also linear, but because of the low currents produced, little sensitivity or accuracy was possible by the ac method and the ac method is not recommended. Analytical Implications. Polarographic analysis in aqueous H F provides a convenient and simple method for direct analysis of metal ion species soluble in the solvent. Because

+

+

+

+

~

(19) D. E. Smith and T.G. McCord, ANAL.CHEM., 40, 475 (1968). (20) T. G. McCord and D. E. Smith, ibid., p 289. (21) I. M. Kolthoff and J. L. Lingane, “Polarography,” Vol. 11, 2nd ed., Interscience, New YorklLondon, 1952, pp 457-60. (22) J. B. Headridge and D. P. Hubbard, Atzalyst, 90, 173 (1965).

(23) I. M. Kolthoff and J. L. Lingane, “Polarography,” Vol. 11, 2nd ed. Interscience, New York/London. 1952, pp 462-7. (24) A . M. Bond and T. A. O’Donnell, ANAL.CHEM.,40, 1405 (1968).

ANALYTICAL CHEMISTRY, VOL. 41, NO. 13, NOVEMBER 1969

1805

HF is a weak acid and effectively buffers the solutions being studied toward both hydrogen and fluoride ion, strict control of the concentration of HF, H+, and F- used in both calibration and unknown solutions is not necessary. This factor, added to the fact that HF is a particularly good solvent for many materials, and can be used as the supporting electrolyte, makes aqueous H F a particularly useful medium in the field of

polarographic analysis, particularly where mineralogical or metallurgical samples are involved. RECEIVED for review March 3,1969. Accepted September 10, 1969. The authors thank the Australian Research Grants Commission for funds made available to purchase the Polarecord and auxiliary apparatus used in this work.

Chronopotentiometric Determination of Halides and Analysis of Halide Mixtures with Silver Anodes Dennis G. Peters and Akio Kinjo Department of Chemistry, Indiana University, Bloomington, Ind.

Chronopotentiometry with silver anodes is suitable for the determination of halides both individually and in mixtures. Iodide concentrations ranging from 4.8 x 10-4 to 2.5 X 10-2M in 1F sodium nitrate medium have been determined, the analytical error being about &0.5%at the higher concentration and +4% at the lower limit. If a correction for charging of the double-layer is applied, 1.0 x 10-4Miodide is determinable with an error of less than 5%. Solutions containing either bromide or chloride in the concentration range from 5.0 x to 2.5 X 10-2M can be analyzed with uncertainties between 1 and 5%. Analysis of mixtures of two or three halides may be accomplished with uncertainties for the individual species ranging from 3 to 25%, the magnitudes of the errors depending on the relative concentrations.

SEVERAL INVESTIGATIONS of the chronopotentiometric behavior of halides have been repprted, including work in which a halide is oxidized to the halogen and in which a mercury or silver electrode is polarized anodically in a halide solution to produce an adherent film of the corresponding mercury(1) or silver salt. Well defined chronopotentiograms can be obtained for the oxidation of either bromide or iodide to the respective halogen at platinum, gold, and carbon paste anodes in acidic solutions (1-5). Oglesby, Anderson, McDuffie, and Reilley (6) examined the use of mercury electrodes in conjunction with thin-layer chronopotentiometry for the determination of bromide and chloride in potassium nitrate media, and Delahay, Mattax, and Berzins (7) proposed that halides might be determined chronopotentiometrically with silver anodes. Recently, chronopotentiometry was employed by Meyer, Posey, and Lantz (8) to study the kinetics of the anodic formation of silver chloride films and to determine chloride ion in various aqueous media. Because extensive data pertaining to the chronopotentio(1) F. C. Anson and J. J. Lingane, J . Amer. Chem. SOC.. 79. 1015 (1957). (2) R. N. Adams, J. H. McClure, and J. B. Morris, ANAL.CHEM., 30, 471 (1958). (3) R. T. Iwamoto, R. N. Adams, and H. Lott, Anal. Chim. Acta, 20, 84 (1959). (4) A. J. Bard, ANAL.CHEM., 35, 340 (1963). ( 5 ) D. G . Davis and M. E. Everhart, ibid., 36, 38 (1964). (6) D. M. Oglesby, L. B. Anderson, B. McDuffie, and C. N. Reilley, ibid., 37, 1317 (1965). (7) P. Delahay, C. C. Mattax, and T. Berzins, J. Amer. Chem. SOC., 76, 5319 (1954). (8) R. E. Meyer, F. A. Posey, and P. M. Lantz, J . Electroanal. Chem., 19, 99 (1968).

1806

metric determination of halides are not available, the present study was undertaken to assess the usefulness of chronopotentiometry with silver electrodes for this purpose. Silver wire anodes were employed exclusively in this work because of their relative ease of preparation and manipulation. Analytical results are tabulated for the determination of halides, both individually and in mixtures, over a wide range of concentrations, and some comparisons between chronopotentiometry and other electroanalytical procedures are presented. EXPERIMENTAL

Reagents. All chemicals were of reagent-grade quality and were used without additional purification. Solutions were prepared with deionized distilled water. Stock solutions of sodium chloride, sodium bromide, and sodium iodide, with 1 F sodium nitrate as supporting electrolyte, were prepared by weight. Aliquots of these solutions were diluted with 1F sodium nitrate medium for each series of experiments. Apparatus and Procedure. A conventional chronopotentiometric cell, which was thermostated at 25 f 0.1 “C, and an electrolysis circuit similar to those described by Lingane (9) were employed. Pure nitrogen was used to stir the sample solution and to remove dissolved oxygen, and the solution was kept under a nitrogen gas atmosphere during the trials. All chronopotentiograms were recorded with a Varian F-80 X-Y recorder in the time-base mode, and transition times were determined manually by means of a method outlined in earlier publications (10, 11). Working electrodes were fabricated from five-inch lengths of pure silver wire, coated with Tygon paint (U. S. Stoneware Co., Akron, Ohio) in such a manner that an approximately 1.5-cm portion of wire close to one end remained exposed (10). The area of each electrode was computed from the diameter of the wire (measured with micrometer calipers) and the length of the uncoated part (determined with the aid of a binocular microscope). After every fourth or fifth transition-time measurement, the working electrode was removed from the cell, dipped into 1Fsodium cyanide medium to dissolve the silver halide film, washed thoroughly with distilled water, and returned to the cell. This procedure was followed to avoid the possibility that the silver halide film might become sufficiently thick to affect the transition times. (9) J. J. Lingane, J . Electroanal. Chem., 1, 379 (1960). (10) D. G. Peters and W. D. Shults, ibid., 8, 200 (1964). (11) D. G. Peters in “Standard Methods of Chemical Analysis,” 6th ed., Vol. 111, Part A, F. J. Welcher, Ed., Van Nostrand, Princeton, N. J., 1966, pp 404-427.

ANALYTICAL CHEMISTRY, VOL. 41, NO. 13, NOVEMBER 1969