Polarographic Behavior of Organic Compounds Efect of Ionic Strength, B u f e r Nature and Concentration, and pH PHILIP J. ELVING, JOSEPH C. KOMYATHY, ROBERT E. VAN ATTA, CHING-SUNG TANG, AND ISADORE ROSENTHAL The Pennsylcunia State College, State College, P a . In the development of analytical procedures and in general polarographic investigation involving organic compounds, the importance of calibration and study at the same level of ionic strength and buffer characteristics is indicated as a result of a survey of the factors affecting polarographic behavior. In order to evahtate the effects of the medium on polarographic behavior, a study has been made of a-bromo-n-butyric acid over the pI1 range of 1 to 12 and the normal working range (0.1 to 3 M) of ionic strength, using most of the commonly encountered buffer systems. The change in E0.6 was found to be dependent upon the nature of the buffer components, the concentration of the buffer components, the pH of the solution, and the ionic strength of the solution. Regions of ionic strength exist in which the measurement of Ea.&will yield values which are more valid when used to compare the results of various investigators than Eo.5 values nieasured in other regions. In order to obtain consistent results which will lend themselves to duplication and to comparison upon theoretical bases, the test solutions should be made up of definite ionic strength, the value of which should he stated in the same manner as the routine statement of capillary constants. The exact natiire and concentration of buffer coniponents sh011lrf be specified.
LTHOUGH the importance of the effwt of pH on the polarographic reducibility of organic compounds has been demonstrated in recent years, there has been no systematic study of the effect of ionic strength and its variat,ion due to the use of variws buffers in the polarographic invest,igation of organic compounds. The u d practice in buffering solut,ionr for organic polarographic study is to u,w a standard buffer. or set of buffers at the concentrations listed in t'he literature--e.g., the handbook. However, any one buffer, even if used over a pH range of only 2 units, c m vary tremendously in iouic strength. For. example, in the cme of a phosphate or citrate buffer uscd in the region of pH equal to the pK, of the third ionization, t,he ionic strength CFQ change by a factor of 9 / 4 over the recommended pH interval of 2 ( p K f 1). Such changes in ionic strength with one buffer over a pH range or between two buffers at any one pH are probably responsible for much of t,he confusion arising when attempts u e made to duplicate reported value. In addition, even if two different buffers at the same ionic strength are used, polarographic results may vary considerably, depending on the nature of the buffer components. These variations can lead to distortion that will completely mask the pH-dependence of the process. This type of confusion has led to the attempt made in this paper a t evaluating the phenomena of variation of polarographic behavior with ionic strengt,h and buffer composition. One of the earliest observations of the effect of specific buffer components on the half-wave potential, E0.6, was made by Furman and Stone ( 5 )in the course of the investigation of the polarographic behavior of several anthraquinones in various commonly used buffering media; shifting of the Eo.5 to more negative values was taken as evidence of complex formation with borate and phosphate buffers. On further investigation, Stone (9) found that buffer constituents may play a vital role in the polarographic reduction, if a relatively stable species is formed between the buffer anion and the reducible material or one of its reduction products. DeFord and Andersen ( 2 ) have studied the variation of the Eo.sof cadmium as a function of ionic strength in various supporting electrolytes of ionic strengths extending up to 12 M ; flo.' a t first became more negative as t'he ionic strength increased
and then shifted to lays nrgativc values with continued increase of ionic strength, the magnitude of the shift depending on the nature and total ionic strength of the supporting electrolyte. Elving, Rosenthal, and Kramer (3),in a polarographic investigation of iodoacetic acid and the bromoacetic acids, obtained significant variations in the Eo1 in rome cases where two or more buffers at the same pH and ionic strength were used. The variation of the buffer component concentration has been found to affevt the Eo of reducible organic compounds (I,$, 4 , 8 ) . -4s such phenomena arc of great importance to the interpretation of the data obtained in the polarography of organic compounds, tJie polarographic behavior of a-bromo-n-butyric acid has been studied over the pII range of 1.0 to 12.4and the normal norking range (0.1 t o 3 M ) of ionic strength, using m w t of the commonly encountered buffer systems. Ionic strength as varied in two ways: by altering the buffer component concentration, and by adding potassium chloride. The concentration of the buffer components was varied in order to determine the magnitude of this effect on the Eo.&;it was hoped to obtain some indication as to whether interaction between the electroreducible substance and the buffer components occurred. Any variation of Eo6 under identical conditions including similar ionic strength, except that of buffer component concentration, could he interpreted as being due to the formation of complex species. It was felt that from the data obtained in this study, the optimum operating conditions for each buffer could be determined and Borne insight be gained into the nature of the reduction pr0c-s and the factors influencing it. The selection of a-bromo-n-butyric acid as the substmce to be used in this investigation was based on several factors. The Eo is within the readily measurable range of -0.3 to -1.4 volts for the pII range of 1 to 13; this potential range permit8 the use of all the common buffers, some of which give decomposition waves beginning at relatively low potentials. The compound itself is stable over the time required. In addition, the compound, containing a four-carbon chain, is not the introductory member of the series. It was felt best to avoid the introductory members of series, as they usually behave in an anomalous manner. The reduction of the a-bromo-n-butyric acid involves the fission
V O L U M E 23, NO. 9, S E P T E M B E R 1 9 5 1 of tlw carl)on-bromine bond and the form:ttion of ri-but \ ri( acid. The clwtrodc reaction is irreversible, as indicated by the slope of tlic w a w , the IlkoviE equation indicates thc tiio-electron reducbtion procws which results in current flow. The latter is a diffusion-controlled process, as indicated by the effects of variation of concentration of reducible \pecies, of head of mercury, and of temperature variation upon the current flow. While not of absolutely symmetrical charact cr, as iz the wave obtained in the case of a reversible speries, the wave is suficiently well defined to leave little question as to the determination of the polarographic values from it. R liile this particular compound yields the behavior reported, all other reducible organic species do not necessarily show the same type of behavior. However, the results reported indicate a behavior which may be found with other substances. I n particular, similar effects have been observed in the polarography of other halogenated compounds where the electrode reaction involves the fission of carbon-halogen bonds. Studies similar to the one described should be made of other irreversible organic systems as ne11 BS of reversible organic Bystems. It would be of considerable interest to investigate irreversible and reversible redox systems in which ionization does not occur, in order to determine whether the sigmoid curve and iso-Ep point phenomena subsequently discussed are associated with pHdependence. Such studies are nom in progress in the authors' laboraton
1219 I esistance
of the solutions was measured by means of a General Radio Co. Type 650-A impedance bridge. Preparation of Buffer Solutions. The buffers used in this investigation are described in Table I. Buffer A was prepared by f m t diluting the calculated amounts of hydrochloric acid and of potassium chloride to approximately 250 to 400 ml. The p H of the solution was then measured continuously during dilution t o approximately 475 ml. ; standard solution of hydrochloric acid or of sodium hydroxide was added to adjust any variation in p€I due to dilution. The solution was then transferred to a t500-ml. volumetric flask and diluted to mark. Buffers B, F, €I, and .J were similarly pi epared. Buffer C was prepared as needed from stock solutions of the tuc) components (0.22 and 0.67 .If citric acid and 0.11 and 0.33 'Id disodium hydrogen phosphate). The test solutions were prcpared by mixing stork solutions in the required proportions, diluting with an equal volume of a solution of a-bromo-n-butyric acid, adding the calculated amount of potassium chloride to adjust the ionic strength, adding small amounts of the required
~ _ _ _ _ _ _ _ _ _ ~ _ _ Table I.
Coniposition of Buffers
(Potassium chloride added as ionic strength component DesigBuffer nation 1" Ionization Conatants (8) HC1-KCl 1.0 1.0, 2.0. 3.0, 4.5 Ha citrate: 8 . 4 X IO-' Nar citrate-€IC1 1.8 x 10-6 4 0 x lo-; 2.2, 3.6. 5.7, 7.9 C HaPOd: 1.1 X 10HI citrste-XazliP0, 7 . 5 x 10-6 4 . 8 x 10-1: I) 3.5, 3.6, 3.7, 4.3, HOAc: 1.86 X lo-* HOAc-NaOAc 5.0, 5.2, .5.3 E 6.0, 6.1 F HaBOa: 6 . 4 X 10-10 7.8 ?;HIOH: 1.8 X 10-8 8.2, 8.5, 8.8 G H 9.9 10.4. 11 5 , 12.4 J W 3.5
Materials. All buffer components were analytical reagent grade chemicals. The a-bromo+-butyric acid (Eastman Iiodak white label) was recrystallized several times from pentane. The values of the melting point and the density agreed with those reported in the literature. The compound was polarographically pure. h solutions of a-bromo-n-butyric acid stood, slow decomposition of the ac!d Table 11. Polarographic 1alues and Conbtants of a-Hromo-n-butyric Acid occurred-e.g., a decrease of 25% in in Yarious Buffer Media the diffusion current of 0.5 millimolar a-bromo-n-butyric acid in buffer C a t (Potassium chloride added a$ ionic strength component) pH 5.7 nnd 7.9 was observed over :i Concentration Averaged Values of period of one month. On titration of Buffer the solution, the total acidity was found First Second a-Bromo-n!2 m component component butyric actd ~ 1 1 td tlt.viwd ~ O (X2 ) employed a dual titmtion where one titration na? kept .;lightly in advance of the other. -i more practic:tl method dewloped by MacInnes and Jonea (10) utilized the concentration vo1t:igt~obtained by removal of a mll portion of thv titrated solution 1x.forrS w c h addition of reagent. Miiller ( 1 1 ) employcd a simihr cchtmie. These last two whernes require a change ot rclfewnce solution before each addition of reagrnt. Some dihsinlilar bimetallic electrode eyPtems give results t h a t simuhte these differential potentiomctric titr:ttions ( I , $ , 17). Delahay ( - 5 ) has n w n t I \ discushetl the Foulk and Bawden 3'dc~;d-stop''inc.chanisni (6) by coneideration of tht. Phaptv O f various polarization curves. ti polarized bimetallic system using two similar electrodes n as suggested by Willard and Fenwick (18). Van Same and Fenwick (16) discussed VOLTbDL the behavior of two platinum electrodes polarized by R 0.5-volt sourcc. The method de-rrihctl in this paper is a polarized system similar t o thnt (Jf Fenwick e l al., but cam is taken to polarize the electrodes with constant current. Tlie electrochemical phenomenon giving rise to the curves i c esplained in tern- of polarographic behavfor. I t is hoped that the esplanation will rliminatr the, hcretoforc empirical nature of the method and assist the analyst in Rpplying this technique to new situations
points w i l l be r r a l i m c i in t h e titration of a succession of substance&. If one of the systems is irreversible, the electrodes w i l l s h o w a sharp increase or decrease in e.1n.f. at the end point. The conclusions we= verified experinlentally. This method provides simplicitj of apparatus, con tinirous indication, sharper breakfi mer the conventional potentiometric metho d s in mine cahes, rapid attainment of stable readings, eliniinatioo of a reference half-cell, and ability to detect a siicce4on of end points simply and witho u t plotting.
Figure 1. Typical Polarograms of a Solution of Ferrous Ion Being Titrated by Ceric Ion A. E.
Initial solution of ferrous ion Solution half titrated
Solution at end point Solution after addition of excess d
Typical polarograms taken a t a platinum electrode in a stirred solution of a polarographically reversible system (ferric-ferrous) titrated by a reversible system (ceric-cerous) are shown in Fig-