y J. ^T /.• Y/^/ //^ ■'$:/ rr' MDDC - 1015 UNITED STATES ATOMIC ENERGY COMMISSION POLAROGRAPHY WITH STATIONARY ELECTRODES by L. B. Rogers Oak Ridge National Laboratory Date Declassified: May 29, 1947 Issuance of this document does not constitute authority for declassification of classified copies ofthesameor similar content and title and by the same author. ^ UjlS'lV. OF FL Lis — -DOkl.! Mfc.N.TS DEPT U.S. DEPOSITORY Styled, retyped and reproduced from copy as submitted to this office Technical Information Division, ORE, Oak Ridge, Tennessee AEC, Oak Ridge, Tenn., 5-5-50— 900-A18980 Printed in U.S.A. PRICE 10 CENTS Digitized by tine Internet Arclnive in 2011 witli funding from University of Florida, George A. Smathers Libraries with support from LYRASIS and the Sloan Foundation http://www.archive.org/details/polarographywithOOoakr POLAROGRAPHY WITH STATIONARY ELECTRODES By L. B. Rogers ABSTRACT Mixtures of ions which react rapidly with mercury usually cannot be analyzed polarographically with the dropping mercury electrode. However, these polarographic analyses appear to be feasible using solid electrodes made from noble metals. The general features of their behavior are covered in this report. My collaborators in carrying out this work were A. F. Stehney, R. B. Goodrich, and H. H. Miller. INTRODUCTION Polarographic reactions are fundamentally electrolytic processes although polarography as a name attached to the use of micro electrodes to obtain current -voltage curves for analytical purposes was first proposed in 1922. Until just before the last war, polarographic studies were confined almost entirely to the use of the dropping mercury electrode. This was natural because many elements are readily reduced from one oxidation state to another and because mercury with its high hydrogen over- voltage enables one to examine many of these cathode reaction potentials to -1.5 v or more (vs a saturated calomel electrode). However, the reactions which can be studied are usually limited to those which occur at a potential more negative than the one at which mercury itself begins to dissolve. Therefore, we have examined the possibility of making polarographic analyses in the anodic region by substituting a platinum electrode for the dropping mercury electrode. Platinum and other noble metals having high oxygen overvoltages are, for anodic studies, in a position comparable to that of mercury for cathodic studies (see Figure 1). The range of potential covered by each type of electrode is altered by changes in the acidity of the solution. It might be advisable at this time to point out the incompleteness of the common conception of the terms anode and cathode. In an electrolytic cell containing two similar platinum electrodes at different potentials, one of them is of necessity more negative than the other. If a reaction such as Ag+ + e ->-AgO is possible, one would expect it to take place faster though not necessarily exclusively at the electrode having the higher concentration of electrons, i.e., the more negative potential. One may draw an analagous picture for oxidation in such a cell. In contrast to this set-up, the polaro- graphic cell usually consists of one electrode, dipping into the cell, whose potential is measured through a salt bridge against a reference electrode. The potential on the polarizable electrode within the polarographic cell can be varied continuously from strongly positive to strongly negative. This electrode is not in competition with another during a reaction. Thus in a polarographic cell, an elec- trode might be sufficiently negative for reduction to take place while its potential would make it an anode in the usual electrolytic cell. Therefore, "anodic" reductions and "cathodic" oxidations are quite possible in polarography. Thus, if one wishes to study the reduction of some of the stronger oxidizing agents such as permaganate, dichromate, eerie, argentous, cupric, and ferric ions, one is forced into studies of the anodic region. All of these ions are usually reduced at zero potential by MDDC - 1015 1 2 MDDC-1015 the dropping mercury electrode unless they are in the form of stable complexes which, being more difficult to reduce, require more negative potentials. To date, several stages of development have been reached in studies of anodic reactions using stationary electrodes of noble metals. As early as 1900 electrolytic studies of the reduction of per- manganate showed that as the potential became less positive a reduction current began to flow, increased with further lowering of potential, and finally flattened off in a manner exactly like the polarographic diffusion current. Other investigators showed later that many redox pairs behaved similarly, their efforts being aimed primarily at the study of reversibility of reactions or of over- voltage of different metals. The third stage was marked by the use of the polarograph by Walen and Haissinski to record the following electrolytic data automatically rather than by tediously recording each point. 4 , ' 1 ' ■ 1 1 1 3 2 — / ""/-^"z / — 1 in UJ £ a. S < o ^J__,^ / — -2 -3 0H° ► Og / Hg°-^Hg,^^ -4 1 1 1 1 \ 1 — iO 0.5 -0.5 -\.Q H.5 -2.0 POTENTIAL VS SATURATED CALOMEL ELECTRODE Figure 1. Ranges of potential covered by platinum and dropping mercury electrodes in neutral solutions. In 1940, Laitinen and Kolthoff pointed out the usefulness of stationary platinum electrodes for quantitative polarography, but they used a tedious method which enabled them to proceed slowly under conditions approaching equilibrium. Esxh point on the polarographic curve was obtained by approach- ing a given potential slowly and then waiting 2 to 3 minutes for the deposition current to reach a con- stant value. To obtain ten points for a curve required a minimum of thirty minutes. Laitinen and Kolthoff also reported experiments with a rotating platinum electrode which had two advantages over stationary electrodes. A constant diffusion current was reached at once, and, for a given concentration of ion, the current was much larger. Thus, the rotating electrode shortened the time required to make a curve and it reached a lower concentration of reducible ion than the station- MDDC-1015 3 ary electrode. The sole disadvantage of the rotating electrode appeared to be the greater difficulty in adapting it to the analysis of small volumes of solution. Our interest in the anodic region of potential and the desirability of using small volumes of solu- tion prompted us to combine and to extend the work of Laitinen and Kolthoff and that of Walen and Haissinski. To date we have examined the behavior only of stationary electrodes but half -wave po- tentials obtained for these electrodes should be essentially the same as those for the more versatile rotating electrodes. After our work had been in progress for several months, other investigations with stationary electrodes came to our attention. Last fall Dr. O. H. Mueller reported that he had used automatic recording with stationary electrodes for his studies on mixtures of quinone and hydro- quinone. By flowing the solution past the electrode he approached conditions similar to those obtained with a rotating electrode in which equilibrium is reached instantaneously. One would expect automatic recording to be feasible under these conditions. A short time ago, Dr. L. A. Matheson mentioned that he had carried out some unpublished experiments several years ago using stationary electrodes but that his studies had not proceeded beyond the preliminary stage. We have compared the limits of reliability of data obtained with stationary electrodes using the usual automatic recording apparatus with data obtained manually according to the method of Laitinen and Kolthoff. The reliability was judged by: (1) the constancy of the halt -wave potential, (2) the re- producibility of the diffusion current for a given concentration, and (3) the linearity of the relation between diffusion current and concentration. We have studied the variations introduced by using differ- ent rates of change of potential, by increasing the area of the electrode, and by stirring the solution. We selected silver ion because of its known electrochemical simplicity and reversibility, and because we expected that a reduction involving precipitation might introduce complications resulting from changes in the surface of the electrode which would be quite small if a soluble ion were produced (i.e., Ag+->.Ag° vs Fe3+.>Fe^"^). APPARATUS In our studies, we have used three different polarographs, Sargent Models Xn and XX and a manual set-up consisting of a potentiometer, slide -wire and Rubicon galvonometer. Our polarographic cell was conventional in that the potential of the polarized electrode was measured through an agar bridge saturated with potassium nitrate against a large saturated calomel electrode. Our polarographic solution was O.IM potassium nitrate at pH 4 containing 5 x 10-4m silver nitrate. RESULTS In general, the curves had rounded maxima whose height appeared to increase with current (i.e., from higher concentrations of silver ions, or from larger electrodes). (See Figure 2.) These maxima did not appear to be of the same nature as those encountered with the dropping mercury electrode and this impression was strengthened by finding that the presence of gelatin appeared to have no ef- fect on them. They probably result from the time-lag in reaching diffusion equilibrium. Attempts to study these maxima by varying the rate of polarization were unsuccessful because duplicate runs sometimes showed differences in behavior ranging from the usual maximum to the extreme of no maximum at all. Although we were unable to reach a definite conclusion concerning the effect of the rate of polar- ization on the height of the maximum, we did find that the faster rates produced markedly larger diffusion currents (see Table 1). However, the half-wave potential appeared to be essentially un- changed regardless of the rate of polarization or the direction of polarization. MDDC-1015 +0.7 16 12 ^^_- in 8 UJ q: UJ 0. s < o q: o S 4 : 1 ' 1 1 1 1 1 1 + 0.6 +0.5 +0.4 +0.3 POTENTIAL +0.2 + 0.i Figure 2. Reaction of silver ion at a stationary platinum electrode; recorded automatically. Table 1. Effect of rate of polarization on diffusion current. Diffusion current Rate of polarization (Ma) (mv/sec) Quiet solutions Stirred solutions Manual 3.2 usually no i^; 7.8 1.46 3.0 7.8 - 8.4 2.92 3.9 4.38 4.1 16.8 - 17.6 MDDC-1015 5 Our studies which began with the micro platinum electrode of Laitinen and Kolthoff extended to include those up to two sq cm in area. As expected, a roughly linear relation was found between elec- trode area and diffusion current for a given solution, but our data indicated that the behavior of elec- trodes became increasingly erratic with increased size. If actual unknowns were to be analyzed, it appeared to be more reliable to calibrate each electrode with a known solution rather than to depend upon calculations based on the measurement of area and current from another electrode. Despite the MOLAR CONCENTRATION Figure 3. Relation between concentration of silver ion and diffusion current for 1 mm electrode (0.02 cm -2). somewhat lower accuracy found for larger electrodes, the larger diffusion current which results from the increased area enables one to examine a lower range of concentration. This is closely analogous to the findings of Laitinen and Kolthoff concerning the better applicability to more dilute solutions of a rotating micro electrode as compared to a stationary one. Therefore, in order to cover a wider range of concentration (10~2 to lO'^M) one should have at least two sizes of electrodes (see Figures 3, 4, and 5). Stirring the solution completely eliminated the maximum and, as one would expect, increased the diffusion current many fold. However, while it was possible to attain 5 to 10% accuracy and precision in quiet solutions, motor stirred solutions gave results in which 20 to 30% variations were common. Thus, the method fell from semiquantitative to crudely approximate when the solution was stirred MDDC-1015 ■a c a o •i-t u > z o o ^- H < ^H (M 1- z UJ o o o c ii (C tU 0) < -5 o o 0) =M 5 ^ t, c S 2 ■" S c rt 0) '-^ t^ "U u ffj 3 u lo c o fc. w 3 3 bX) ^ •- a &< T3 saaadwvoaoiw T3 C ni a o ti 01 > ^ S eo W 1 ^ s o o 2 c o O O 'I *J o