ELECTRODILATOMETRY OF LIQUIDS, BINARY LIQUIDS, AND SURFACTANTS MANIT RAPPON*, RICHARD M.

JOHNS, and SHIH-WEI (ERWIN) LIN Department of Chemistry, Lakehead University, Thunder Bay, Ontario, P7B 5E1 Canada. *Corresponding author: manit. rappon@lakeheadu. ca

Abstract. When a liquid is subjected to high electric field, its volume change can be increased or decreased depending upon the liquid under investigation. A new technique has been developed from our laboratory to measure the relative volume change per and is known as Electrodilatometry (ED) which may be expressed as: where R is known as Electrodilatometric Effect (EDE), V and are the

volume of liquid with and without the field, respectively. ED is one of the nonlinear effects such as electro-optic Kerr effect, the electrostriction, dielectrophoresis, nonlinear dielectric effect (NDE). Ed is found to be very sensitive to hydrogen-bonded liquids. It has been applied to study pure liquids, binary mixtures, alcohols, and non-ionic surfactants such as Triton X-100. The signs of EDE (R), Kerr constant (B) and NDE are compared and contrasted. A few models have been used to calculate R with limited success. Not only can ED be used with smaller molecules but it should also be a potential tool to study polymer solutions and supramolecular assemblies.

1. Introduction The change in pressure following the impact of high electric field on a liquid is known in the literature as electrostriction and is the topic of many investigations [1,2]. This method suffers from one drawback i.e. the choice of the reference pressure varies from one laboratory to another [3]. This makes it difficult to compare experimental results. In order to avoid this, we are investigating the impact of high electric field on the change in the volume of a liquid. While there were a couple of reports in the old literature that electric fields were used to study some liquids as cited in ref.[4], these investigations did not look at the volume change as we proposed hereunder. To the best of our knowledge, the electric fieldinduced volume change is a new technique and it is proposed that this new technique, originated from our laboratory, be called Electrodilatometry (ED). The change in volume was found to vary directly as the square of the externally applied electric field strength Thus, ED is one of the nonlinear methods and is similar to the Kerr effect [5], nonlinear dielectric effect (NDE) [6,7], electrostriction, dielectrophoresis (non-uniform field) [8] and other nonlinear methods [9]. These

367
S.J. Rzoska and V.P. Zhelezny (eds.), Nonlinear Dielectric Phenomena in Complex Liquids, 367-377. 2004 Kluwer Academic Publishers. Printed in the Netherlands.

368

methods share one common feature with ED in that application of high electric fields is required. They differ only which physical parameter is used to monitor the change. For example, for ED it is the volume change; for Kerr effect, the difference between refractive indices for NDE , the change in relative permittivity etc. Thus for a given sample of liquid subjected to an applied field, ED can yield unique information on which may be integrated with information provided by Kerr effect and NDE. The presentation is organized in the following manner: the theories related to ED are presented in section 2, the experimental procedure (section 3), results (section 4), discussion (section 5), and the conclusions are drawn in section 6.

2. Theories The volume of a liquid is changed when it is subjected to high static electric field. Such a volume change has been derived. In the case of fluid flow into a region from the source of constant chemical potential, the volume change is given [10], i.e.

Where V,P,T are the volume, pressure and temperature, respectively; is the liquid volume without the field, and E is the applied electric field. Integration of equation (1) yields:

Where

is the relative permittivity of liquid. Left hand side of equation (2) where and V is the volume of liquid

may be written as

under the applied field. R represents the relative volume change per and it is henceforward called the Electrodilatometric Effect (EDE). In another derivation, attempts are made to include some geometry of a dielectric[1] as shown in equation (3):

Here, 1/K is the compressibility of the liquid, and n is the parameter depending on the geometry of the dielectric, A model which allows for change in molecular volume when a dielectric ellipsoid is placed under electric field, has been advanced [11] as shown:

369

Where is R evaluated with maximum semi-principal axis (a) parallel to the field, and is the depolarizing factor of the ellipsoid. is the internal bulk modulus of the molecule, is the internal permittivity and is the permittivity of vacuum. In fact, the observed volume change for some liquids may be due to other contributions, in addition to the classical electrostriction as in equation (5):

Where is the volume change due to interactions; and is the volume change due to structural interactions. is the volume change due to the deformation of the dielectric without changing its shape [1], or electricallyinduced distortional strain [8]. is calculated from (2) and the net contribution of this term will be small and negative [4]. is the volume change due to, e.g. the induced dipole-induced dipole, dipole-induced dipole interactions. It has been shown that for the case of ionic interactions involving central metal ion and the surrounding solvent molecules, this term is large and negative [12]. is the volume change due to structural interactions. It is caused by molecules with permanent dipole moments, which respond to an external electric field by realignment so as to reduce their potential energy, and sometimes known as orientational polarization [8]. This process may lead to several volume changes as a consequence, e.g. inter- and intra- steric interactions, forming and breaking of H-bonds, changes in molecular conformation, etc. For a hydrogenbonded system, is likely to be large and positive due to the net change in the number of hydrogen bonds as the molecules are forced to realign with the field.

3. Experiment The detailed construction of ED cell and its assembly was provided earlier [13] and is briefly described Two stainless steel cylindrical electrodes, inner one was solid and the outer one hollow. They were assembled concentrically leaving an electrode gap of 0.105 cm. They were sufficiently large in order to sink any Joule heat generated during a measurement. Two electrically insulating plates were used to cap securely both ends of the electrodes. Liquid sample was introduced through an inlet port located at the bottom of the cell and it moved up the electrode gap and out through the outlet port at the top of the cell into a connecting line. The latter was connected to a glass capillary. The liquid sample partially filled the capillary. Water jackets were used for the cell and the capillary so that the temperature was kept constant and controlled to 0.02K by a circulating bath. The apparatus was assembled as shown schematically in Figure 1. Laser light (Spectra Physics) at 632.8 nm, passed through a cylindrical lens and was converted

370

to a vertical line and focused on to the sample liquid level in the capillary. The image of the level was captured by a linear charge-coupled device, CCD (Texas Instruments, TC-104). The output from the board was fed into an A/D board of a PC. Regulated high DC voltages were generated from a power supply (Bertan, 205A-20R). Typical field strength used was in the range depending upon the sample used. The

Figure 1. A diagram of ED apparatus (See text for details).

change in pixel number was converted to physical height by calibration. The net change in height was used to calculate the The system is sensitive to change in Pure liquids used were of spectroscopic grade and were dried by molecular sieves (heptane and cyclohexane by sodium wire) and filtered. Triton X-100 [TX100] (t-octylphenoxypolyethoxyethanol)(Aldrich) with and the water content was 0.3% by weight. The reduced form of TX-100 was obtained from the same source whose structure is similar to the regular TX-100, except the phenyl ring was hydrogenated to cyclohexyl.

Figure 2. Sample output of cyclohexanol, field off (lower) and on (upper).

371

4. Results Figure 2. shows a typical output of a signal in a measurement in which the pixel number is plotted against the time (arbitrary unit). Figure 3. shows a sample plot of the difference in pixel number against the square of the applied voltage, Table 1. is a collection of R values for various liquids at 298 K, unless stated otherwise. For some hydrogen-bonded liquids, e.g. some pentanols, the binary mixtures with n-heptane are also shown because of the availability of the Kerr constants. In addition, the values of calculated R from equation (2) are shown for some liquids where the data for are available in

the literature. Some data for the Kerr constants (B) and the nonlinear dielectric effect

Figure 3. Plot of difference in pixel number (arbit.) vs.

are also collected in the same Table for comparison. Table 2. shows the various R values at 293 K for each isomer of pentanols mixed with n-heptane at various mole fraction (F). Table 3. is a collection of R at 298 K values for a series of cyclohexanol mixed with n-heptane. The values of R for TX-100 in cyclohexane at various concentrations (mole/kg) and at 298 K are collected in Table 4.

5. Discussion The discussion is divided into 5 sections: 5.1 General aspects, 5.2 Pure and hydrogen-bonded liquids, 5.3 Pentanols and cyclohexanol, and 5.4 Triton-X 100 solutions.

372

5.1 General aspects It may be seen from Figure 3. that the plot of the change in height (hence, the volume) is directly proportional to This shows that Electrodilatometry (ED) is one of the nonlinear techniques similar to Kerr effect and nonlinear dielectric effect. 5.2 Pure and hydrogen bonded liquids From Table 1., if the signs are neglected and only the magnitude of R is inspected by normalizing against the carbon disulfide, the normalized magnitude is comparable to when the Kerr constant (B) is normalized against its carbon disulfide, for most liquids. However, there are two liquids on the list which do not follow such a trend, i.e. nitrobenzene and 2-pentanol in n-heptane at F=0.601. While nitrobenzene has very large B value, the 2-pentanol mixture has very large R value. It shows that for some hydrogen-bonded systems, ED is a very sensitive technique. If the attention is now paid to the signs of R (excluding the t-pentanol), it may be seen that the sign of experimental R value is opposite to the sign of B and the sign of for those liquids (and mixtures) where these values are available. While the exact reasons are not known, only plausible suggestions to this are hereby given. At this time, only the signs of R and B are discussed first, and the signs of are to be discussed later. Take the case of a non-hydrogen-bonded liquid, the negative value of R suggests that the volume decreases in the presence of the applied field. The existence of the high electric field gives rise to high pressure. The latter forces the molecules to be closer, hence the observed volume decrease. This is accompanied by the B value being positive, which suggests that the refractive indices On the other hand, for a hydrogen-bonded liquid, the volume is increased under the field (R is positive). To account for this, it is plausible that the electric field forces the molecules to reorientate and in so doing, the system experiences the net loss of the number of hydrogen bonds. This lessened intermolecular forces, could translate into the observed expansion of the volume under the field. This is similar to the term structure breaking effect which was used in the magnetooptical investigations of associated liquids [19]. This positive R is accompanied by (B being negative). The sign of NDE is the same as the sign of B, for all the liquids where these values are available. While some forms of explanations may be offered to account for the sign of R and B, it is, however, more subtle to pin down the origin of the sign of NDE. This is because the sign of NDE depends upon an intricate balance of several

373

competing factors [18] which are briefly summarized. (a) At large field the contribution from the higher terms of the Langevin function must be taken into account. This leads to the normal saturation and its contribution to is negative. (b) In an equilibrium, the field shifts the equilibrium in favour of the more polar species. This is called the anomalous saturation and makes a positive contribution to the (c) For species with anisotropic polarizability, the axis of highest polarizability is forced to align in the direction of the field (which is related to the Kerr effect). Its contribution to the is positive. (d) Contributions from the first and second hyperpolarizabilities should also be considered and their contributions to can be either negative or positive depending on the parameters and Some of these factors were used to explain the anomalous behaviour of NDE for solutions of 1-penatnol [20]. For the t-pentanol, R is positive (like other hydrogen bonded liquids) and is the same sign as B (contrary to other hydrogen-bonded liquids on the list.). This unique situation may be due to the unique ability of t-pentanol to form a cyclic trimers (not excluding higher cyclic multimers). These trimers can also form stacking structure in which one ring stacks on top of another and many rings may be involved. Each ring is staggered from the lower one by about 60. This facilitates two things; firstly, the formation of inter-ring hydrogen bonds by the bifurcated hydrogen bonds of O and H atoms; secondly, the bulky t-pentyl groups are displaced further away from each other. This structure is consistent with the results of our studies from other techniques: i.e. Kerr effect [21], viscometry [22], NMR [23], and photochromic reaction probe [24]. The calculated R [from equation (2)] for some liquids and their values are approximately 2 orders of magnitude lower than the experimental ones. For all the liquids calculated, all of them gives negative sign. This is the same as the experimental ones, except for chloroform for which the sign is positive. The latter is

374

explained in the next section. While the signs of experimental and calculated R [based on equation (4)] are in agreement, the discrepancy in their magnitude has yet to be resolved. The discrepancy in magnitude may be partly due to the contribution to volume change from other factors as shown in equation (5), i.e. and in addition to

5.3 Pentanol and cyclohexanol For pentanols in n-heptane as shown in Table 2., all R values are positive, i.e. the volume is increased under the applied fields. Some of the concentrations could not be measured due to difficulties in obtaining concordance readings, and for tpentanol, the changes at low concentrations are too small for the measurements. At low concentrations the changes in R values are small and tend to increase with increasing concentration. This suggests that for 1-, 2- and 3- pentanol, the linear multimers tend to dominate. However, for 2-pentanol at F=0.800, R= 5305 which is the greatest among the measured values. To account for this, it is necessary to use the linear model of multimers where O of one molecule is connected to H of another by H-bond. This situation is repeated to form a chain whose pendant groups are the various alkyl groups of the alcohols, i.e. n-pentyl, and for 1-,2-and 3-pentanol, respectively. It is likely that the unsymmetrical alkyl groups attached to the C with OH group, interact with each other by way of the intra- and inter- chain interactions as the molecules are forced to realign by the field. Such interactions make positive contribution to the terms A similar contribution for 1- and 2- pentanol is expected to be smaller. For the case of t-pentanol, only measurements at high concentrations are obtained and their values are small. This information together with results from other techniques as mentioned earlier, indicates that the cyclic multimers are dominant. One of the possibilities is the cyclic trimers . However, it does not exclude other higher cyclic multimers. For the cyclic trimers, a model may be visualized by taking the chair form of a cyclohexane and replace all the six C atoms with alternating O and H atoms. Each of the three t-pentyl groups is attached to each O atom. These rings can form stacking structure as mentioned in section 5.2. To account for such a low magnitude of R, it is plausible that each stacking structure responds to the applied field as a unit with minimum disruption of the bifurcated H-

375

bonds. Consequently, the is small and the volume increase is found to be the smallest among the pentanols at the same concentrations.

For cyclohexanol, R is positive and small and remains unchanged until F=0.60. This suggests that probably a similar stacking associates may be formed similar to the case of t-pentanol, except that the t-pentyl group is replaced by a bulky cyclohexyl one. At higher concentrations, R increases with increasing concentration until F=1.0. Such a rapid increase in R implies that a different type of associates may be formed. 5.4 Triton X-100 ED has been extended to hydrogen-bonded surfactants: Triton X-100 (regular) and Triton X-100 (reduced) [25]. Due to limited space, only the regular form of Triton X-100 (TX-100) is hereby given. From Table 4, R values may be divided into 3 groups.

376

(a) From m=0.2-0.42, here R increases with increasing concentration. It is plausible that reverse micelles are formed in this region. When more surfactant is added, more micelles formed and led to the increase in R. This view is in agreement with QELS [26] and dye probe [27]. (b) In the region m=0.69-1.6, R values are greater than region (a) but their values are hardly changed. If the associates were to be of the reverse micellar type as in (a), we should have observed the increase in R with increasing concentration. It is likely that the associates change to a new type with a rod-like structure. This may be formed from a cyclic multimers, which may be stacked together similar to the t-pentanol case. As more surfactant is added, the rods are extended in a one-dimensional growth. In the presence of a field, only the and are the major contributions to R. So only small change in R with increasing concentration is observed, (c) From m=3.3-12, here R values increase from region (b) but stay more or less the same with increasing concentration. It is likely that the associates change to an inverted bilayers shape. One possibility would be the association of the OH groups in one plane with their tails perpendicular to it. The OH groups of such a plane can be realigned with OH groups of another plane. The growth can be extended in two directions (twodimensional growth). When the field is applied the dipolar groups such as OH and are forced to realign and this should make a positive contribution to the in addition to and

6. Conclusions

It has been shown that the new technique of ED can provide unique and valuable information on liquids. The information can be used to integrate with other nonlinear techniques such as Kerr effect and NDE in order to enhance our understanding of liquids at the molecular level. ED is particularly sensitive to Hbonded systems. It has been used to investigate simple liquids, binary mixtures and surfactants. It should also be very useful to study supramolecular assemblies [28] and polymer solutions.

7. References
1. Landau, L.D., Lifshitz, E.M. and Pitaevskii, L.P. (1984) Electrodynamics of Continuous Media Ed., Butterworth-Heinemann, Oxford, pp. 51 -54. 2. Corson, D.R. and Lorrain, P. (1962) Introduction to Electromagnetic fields and Waves, Freeman San Francisco, p.117. 3. Guggenheim, E.A.(1967) Thermodynamics, Ed., North Holland, Amsterdam, p.335. 4. Rappon, M. (1985) Electrodilatometry of simple liquids, Chem. Phys. Lett. 118, 340-344. 5. Buckingham, A.D. (1976) Electric birefringence in gases and liquids, in OKonski, C.T. (ed.), Pt I. Molecular Electro-Optics, Dekker, New York, pp. 27-62. 6. Kielich, S. (1972) in Dielectric and related molecular processes, Davies M.(ed.), Vol.1,The Chemical Society, London, p. 192. 7. Hellemans, L. and DeMaeyer, M.(1982) High electric field effects and permittivity changes in nondipolar liquids, J. Chem. Soc. Faraday Trans. II 78, 401-416. 8. Pohl, H.A. (1978) Dielectrophoresis, Cambridge University Press, Cambridge. 9. Bloembergen, N. (1982) Nonlinear optics and spectroscopy, Rev. Mod. Phys. 54, 685-695. 10. Buckingham, A.D. (1964) The Laws and Applications of Thermodynamics, Pergamon Press, London, p.168. 11. Brevik, I. and Hye, J.S. (1988) Note on the electro-dilatometric effect, Physica 149A, 206-214.

377
12. Borsarelli, C.D., Corti,H., Goldfarb, D. and Braslavsky, S.E. (1997) Structural volume change in photoinduced electron transfer reactions, J. Phys. Chem. A. 101,7718-7724. 13. Rappon, M. and Johns, R.M. (1999) Molecular association of pentanols in n-heptane V: electrodilatometric effect, J. Mol. Liquids 80, 65-76. 14. Crossley, J., Morgan, B.K. and M. Rujimethabhas (Rappon) (1979) New Kerr cell for lowtemperature measurements, Rev. Sci. Instrum. 50, 1400-1402. 15. Krupkowski, T., Jones, G.P. and Davies, M. (1974) Permittivity increments in non-dipolar solvents due to high electric fields, J. Chem. Soc. Faraday Trans. II 70, 1348-1355. 16. Rujimethabhas (Rappon), M. and Crossley, J. (1980) Temperature dependence of electro-optic Kerr effect for liquids at 632.8 nm, Can. J. Phys. 58, 1319-1325. 17. Piekara, A. (1962) Dielectric saturation and hydrogen bonding, J. Chem. Phys. 36,2145-2150. 18. Bttcher, C.J.F.(1973) Theory of Electric Polarization, Vol.1, Elsevia, Amsterdam, pp. 289- 326. 19. Dawber, J.G., (1984) Magneto-optical rotation studies of liquid mixtures, J. Chem. Soc. Faraday Trans. I 80, 2133-2144. 20. Malecki, J. (1962) Dielectric saturation in aliphatic alcohols, J. Chem. Phys. 36, 2144-2145. 21. Rappon, M. and Greer, J. M. (1987) Molecular association of pentanols in n-heptane I: Temperature dependence of Kerr effect, J. Mol. Liquids 33, 227-244. 22. Rappon, M. and Kaukinen, J.A. (1988) Molecular association of pentanols in n-heptane II: Viscosities as a function of temperature covering the low temperature range, J. Mol. Liquids 38, 107-133. 23. Rappon, M. and Johns, R.M. (1989) Molecular association of pentanols in n-heptane III: Temperature and concentration dependence of proton NMR chemical shift of hydroxyl group, J. Mol. Liquids 40, 155179. 24. Rappon, M., Syvitski, R.T. and Ghazalli, K.M. (1994) Molecular association of pentanols in nheptane IV: A photochromic reaction probe, J. Mol. Liquids 67, 159-179. 25. Rappon, M. Electrodilatometry of Triton X-100 (reduced form), to be submitted for publication. 26. Zhu, D.-M. Feng, K.-I. and Schelly, Z.A. (1992) Reverse micelles of Triton X-100 in cyclohexane, J. Phys. Chem. 96, 2382-2385. 27. Zhu, D.-M. and Schelly, Z.A. (1992) Investigation of the microenvironment in Triton X-100 reverse micelles in cyclohexane, using methyl orange as a probe, Langmuir 8, 48-50. 28. Lehn, J.M. (1995) Supramolecular Chemistry, VCH, Weinhein

This page intentionally left blank