Effect of Pressure on Electrical Conduction in Ion Exchange Resins Gary Carmack and Henry Freiser" Department of Chemistry, University of Arizona, Tucson, Arizona 8572 1
The electrical conductivity of Dowex-I in the CII-, Br-, I-, and NO3- forms has been determined et pressures ranging up to 2000 atm. Activation volumes for the conduction process closely match the crystallographic volumes of the anions. The data shed light on the conductlon mechanlsm In Ion selective electrodes employing pdymeric membranes.
Table I. Activation Volume for Electrical Conduction for Various Forms of Dowex-1 Ion Exchange Resin Resin form C1-
Br-
NO;
I-
A number of important separation processes utilize an electric field as the driving force to achieve partition of solution components across a membrane. Electrodialysis is an example of such a barrier separation process in which the membrane is fabricated from an ion exchange resin. In addition to the electric current a pressure gradient can be concurrently applied, as is done in the technique of forced flow electrophoresis. Studies of the electrical conductivity of ion exchange and other electroconductive membranes which have been reported (I) generally have been restricted to verification of an Ohm's law relationship without identification of the current-carrying species. In an earlier study (2) of polymer membranes containing dissolved quaternary ammonium salts, used in coated wire and other ion selective electrodes, the determination of the pressure dependency of the conductivity of the membrane gave strong evidence of an ionic conduction mechanism. It was felt that parallel studies of ion exchange resin membranes could provide a useful, simplified model in interpreting our earlier data, inasmuch as only one ion of the pair is mobile and capable of carrying charge. EXPERIMENTAL Approximately 10 g of Dowex-1x8 (Dow Chemical) anionic exchange resin, present as the chloride form, was converted to the desired anionic form by vigorously shaking it with three successive 30-mE portions of the appropriate 1 M sodium salt. Complete exchange was judged by the absence of C1- (tested with acidic AgNOd in the aqueous phase following the final shaking. Resins were then thoroughly washed with distilled, deionized water and partially dried by lightly pressing them between sheets of filter paper. A few grams of the resin, in a KBr-type die, were pressed to a force of 8500 kg/cm2 (8260 atm), producing a "jelly-like'' product easily suspended in water. Some degradation of the resin may occur under compression in the form of disruption of crosslinkages. Films of the resins were prepared by coating glass slides with aqueous resin mixtures and allowing the water to evaporate. Circular samples (15 mm X 1 rnm) were cut from the air-dried master films then vacuum dried at room temperature for 48 h. Using this technique it has been possible to prepare the Cl-, Br-, I-, and NO; anionic forms of the resin from which uniform, pliable films could be cast. Electrical conductivity measurements were conducted using the previously described techniques and apparatus (2). RESULTS AND DISCUSSION Electrical conductivity of the films prepared from the ion exchange resins was found to obey the operational semiconductor relationship -
u
=
o0 e x p ( A G * / R T )
A
v', *
cm3/mol 19 22 23 30
I,
vln ,
a
cm3/mo1
1.81 1.95 1.93b 2.16
15.0 18.7 18.1 25.4
a AV' values have a rsd of 10% (two replicates of dupliRepresents the distance from center of cate samples). N to end of 0.
where u is the observed colume resistivity and AG' the molar free energy of activation for the conduction process, analogous to that for chemical rate processes. The preexponential factor, uo, is dependent on geometric factors and includes a term representing the initial concentration of the charge-carrying species. Similar Arrhenius-type relationships have been used to characterize the electrical conductivity in a variety of materials, including polymers (3), electrolytic glasses ( 4 ) , and many arganic semiconducting materials (5). While such expressions provide an adequate description of the observed conductivity in these materials they do not, however, distinguish between the types of charge carriers which may be electronic, ionic, or a mixture of the two. Fortunately one can determine, without prior knowledge of the conduction mechanism, the rather definitive parameter of activation volume for conduction, A V , through application of the following quasi-thermodynamic expressions
with the assumption that the sample geometry remains relatively invariant throughout the pressure range utilized in this study. Data gathered from such high pressure conductivity experiments when plotted (In u vs. P)were linear with a negative slope and were used to calculate the values of A V shown in Table I. Molar volumes, V,, for the anions presented in this table were calculated using V , = 4/&V0~r3, where No is Avogadro's number of atoms and r is the Pauling ionic radius
(6). The magnitude of the activation volume for conduction obtained from this approach has been previously demonstrated (3, 7) to be a most useful parameter in distinguishing among the several possible charge-carrying particles responsible for electrical conduction. If the conduction process was electronic, then one would expect negative AV values resulting from increased orbital overlaps between adjacent negative molecules durhg compression (7). Qn the other hand, if the movement of larger ions (or polymeric segments (8)) was the predominant mode of electrical conduction, this should be reflected through the relatively much larger activation volumes required for ion migration. ANALYTICAL CHEMISTRY, VOL. 49, NO. 6, MAY 1977
767
In a study describing electrolytic conduction in ionic glasses, Hamman (4)demonstrated that an excellent correlation existed between the experimentally measured AV and the calculated V , of the mobile, cationic charge carriers. In a similar manner, Sasabe et al. (8) determined activation volumes for conduction which suggested that segmental motion in polymers was responsible for transport of electrical current through poly(viny1 acetate) films. As the values listed in Table I show, good correlation exists between conduction activation volumes and the molar volume of the respective anion calculated from crystallographic data. One may therefore conclude that the charge is anionically transported through the bulk resin by the anion with the experimentally measured AV corresponding to a process in which compression retards ion migration by decreasing vibrational and oscillatory motions of polymer segments resulting in an increased local viscosity. Since the cationic resin species is immobilized, it probably does not contribute to the observed conductivity. In an earlier study of poly(viny1chloride) films doped with the quaternary ammonium salt, methyltricaprylylammonium chloride, pressure-conductivity measurements yielded corrected AV values of 34 f 2 cm3/mol. In contrast to the ion exchange materials, the cation in this system is not chemically bonded to the polymer and despite its relatively low mobility should be expected to contribute to the electrical conduction process. One could very roughly estimate the contribution such as an additional charge carrier would make to the calculated activation volume by taking the summation of individual
molar volumes multiplied by their respective, weighted transport numbers (assumed to be proportional to the reciprocal of the molar volume). As a limit of relative ion sizes (cation volume >> anion volume) such a calculation would give a value of twice that of the anion volume. While this approximation is admittedly rather crude, it nonetheless suggests the earlier results are reasonable in view of this present study. The conduction process in both the poly(viny1 chloride)quaternary ammonium salt system and the ion exchange resins resembles that in liquids where activation volumes for selfdiffusion are generally one molar volume or greater (9,10). In solids and crystalline materials, it is usual to find AV < V,. A theory (11)based on defect formation in solids suggesta that AV N ‘lZV, would be reasonable.
LITERATURE CITED (1) F. Heifferich, “Ion Exchange”, McGraw-Hill, New York, N.Y., 1962. (2) G. D. Carmack and H. Freiser, Anal. Chem., 45, 2249 (1975). (3) S.Saito, H. Sasabe, T. Nakajima, and K. Yada, J. Polym. Sci., 8 , 1297 (1966). (4) S.D. Hamann, Aust. J. Chem., 18, l(1965). (5) D. D. Eley, J. Polym. Sci., Part C, 17, 73 (1967). (6) N. A. Lange, “Handbook of Chemlstry”, McGraw-Hill, New York. N.Y., 19fi7 n 129 r
(7) Prk’Datta, J. Sc/. Ind. Res., 30, 222 (1971). (8) H. Sasabe, K. Sawamura, S. Saito, arid K. Yada, Polym. J., 2, 518 (1970). (9) J. Naghizadeh and S. A. Rice, J. Chem. Phys., 38, 2710 (1962). (10) A. F. M. Barton, B. Cleaver, and G. J. Hills, Trans. Faraday Soc., 84, 208 11968). (11) R. W‘. Keyes, J. Chem. Phys., 29,467 (1958).
RECEIVED for review December 13,1976. Accepted February 4,1977. This work was supported by a grant from the Office of Naval Research.
Gas-Solid and Gas-Liquid Chromatography Using Porous Layer Open Tube Columns Made with Graphitized Thermal Carbon Black C. Vidal-Madlar, S. Bekassy,’ M. F. Gonnord, P. Arpino, and Georges Gulochon” Ecole Polytechnique, Laboratoire de Chimie Analytique Physique, 9 1128 Palaiseau Cedex, France
A slmgle method for preparatlon of porous-layer open tubular columns uslng graphltlzed thermal carbon black is described. The adsorbent is coated as a thin layer on the walls of the column. I n gas-solid chromatography the columns show high selectlvlty for geometrlcal Isomers, and allow dlfflcult separations at temperatures lower than with clasdcal columns. The columns are highly permeable and can be coated by percolation with various polar or apolar statlonary phases without dlsturblng the carbon layer. By changing the nature or the concentration of the statlonary phase, a range of selectlvltles may be obtained from that of the pure graphltlzed thermal carbon black support to that of the pure llquld statlonary phase. W4h high liquld phase loadings the columns are very selective, very stable, and accept large samples. They have also been used successfully for trace analysis In GC-MS. Present address, I n s t i t u t e f o r Organic Chemical Technology, T e c h n i c a l University, 1521-Budapest, Hungary.
768
ANALYTICAL CHEMISTRY, VOL. 49, NO. 6, MAY 1977
Porous layer open tubular columns (PLOT) with graphitized thermal carbon black (GTCB) coating present many advantages in gas chromatographic analysis over conventional columns packed with the same adsorbent (1,2):the relatively low amount of adsorbent introduced into the column reduces the analysis time. Alternatively, a given compound may be eluted in the same time at a temperature about 100 “Clower than with classical columns (I). Furthermore, high efficiencies can be achieved with long capillary columns, as the pressure drop is low because of the unrestricted flow of carrier gas. The support for the liquid stationary phase in PLOT columns is now generally introduced into the capillary column ilsing the simple methods of dynamic coating. Capillary columns with various support coatings such as silica (3), diatomaceous earths (4), inorganic salts (5), and GTCB (6) have been prepared by this method. On the other hand, PLOT columns made with GTCB and used for gas-solid chromatography are prepared by a static coating procedure similar to the one described by Halasz and
