pubs.acs.org/Langmuir © 2010 American Chemical Society
Negative Electrorheological Behavior in Suspensions of Inorganic Particles M. M. Ramos-Tejada,† F. J. Arroyo,‡ and A. V. Delgado*,§ †
Departamento de Fı´sica, Escuela Polit ecnica Superior de Linares, Universidad de Ja en, Linares 23700, Spain, Departamento de Fı´sica, Facultad de Ciencias Experimentales, Universidad de Ja en, Ja en 23071, Spain, and § Departamento de Fı´sica Aplicada, Facultad de Ciencias, Universidad de Granada, Granada 18071, Spain
‡
Received July 21, 2010. Revised Manuscript Received September 10, 2010 w This paper contains enhanced objects available on the Internet at http://pubs.acs.org/Langmuir. n
An investigation is described on the electric-field-induced structures in colloidal dispersions. Both rheological determinations and direct microscopic observations are used with that aim. The starting point of this study is the socalled electrorheological (ER) effect, consisting of the mechanical reinforcing of a fluid or suspension due to formation of chains of molecules or particles after being polarized by the action of the field. One macroscopic manifestation of this phenomenon is the transformation of the fluid from a typically Newtonian behavior to a viscoelastic material, with finite yield stress and high elastic modulus. The systems investigated were suspensions of elongated goethite (β-FeOOH) particles in silicone oils with varying amounts of silica nanoparticles. The results showed the rather unusual behavior known as “negative ER effect”, which can be best described by saying that the application of an electric field reduces the yield stress and the elastic modulus, that is, produces destruction of structures rather than their build up. The negative behavior is also found for suspensions of other inorganic powders, including hematite and quartz. On the contrary, the usual positive ER response is found for suspensions of cellulose and montmorillonite clay. The same happens if goethite suspensions are prepared in high volume fractions, high-viscosity fluids, or both. All of the results found are compatible with the so-called interfacial model of electrorheology: the reduction of the yield stress of goethite suspensions when the applied field is high enough is the consequence of particle migration toward the electrodes because of charge injection and subsequent electrophoresis. The migration leaves the gap between the electrodes devoid of particles and explains the decrease in yield stress. The addition of silica nanoparticles contributes to reduce the strength of this effect by hindering the charging and making it necessary to increase the field strength to observe the negative effect. The model appears to also be applicable to cellulose, although the positive response found for such particles is explained by their large size: larger diameters bring about larger attraction forces between particles, leading to a tendency to produce strong aggregates. This is likely to occur in suspensions of colloids which, because of their relatively high electrical conductivity, tend to acquire charge even in such nonpolar liquids as silicone oils.
1. Introduction The electrorheological (ER) effect is a phenomenon in which the structure and the rheological properties of a liquid or a dispersed system change under the application of an external electric field. This phenomenon is not fully understood yet, and several theories have been proposed to try to explain it. The most accepted approach is the so-called polarization model,1-5 which assumes that the ER effect results from the dielectric polarization of particles suspended in a nonconducting fluid, arising from the difference between the electric permittivities of the dispersed particles and the liquid medium. In the presence of an electric field, suspended particles polarize, and dipole-dipole interactions tend to align them into chains reinforcing the suspension and leading to viscoelastic behavior and field-dependent viscosities. The attractive force between the particles is proportional to the square of the product of the applied electric field Eo and the dielectric mismatch parameter βε = (εp εf)/(εp þ 2εf), where εp and εf are the real components of the electric permittivities of the particles and the host fluid, respectively. According to this model, if the electric permittivity of the dispersed material is high, then the system should show a marked ER response. However, this was found not to be fulfilled, for *To whom correspondence should be addressed. E-mail:
[email protected]. (1) (2) (3) (4) (5)
Klingenberg, D. J.; Zukoski, C. F. Langmuir 1990, 6, 15. Chen, Y.; Sprecher, A. F.; Conrad, H. J. Appl. Phys. 1991, 70, 6796. Halsey, T. C.; Toor, W. Phys. Rev. Lett. 1990, 65, 2820. Gast, A. P.; Zukoski, C. F. Adv. Colloid Interface Sci. 1989, 30, 153. Davis, L. C. Appl. Phys. Lett. 1992, 60, 319.
Langmuir 2010, 26(22), 16833–16840
instance, for suspensions of barium titanate (a ferroelectric material, whose relative permittivity can be as high as 2000).6 In fact, experiments have demonstrated that barium titanate only displays ER response in the presence of ac fields or after water adsorption.6-9 This is not the only reason why the polarization model is considered to be incomplete because some facts such as the frequency (ω) dependence of the ER behavior or the role of particle and fluid electric conductivities cannot be explained exclusively on its basis. Modifications to the original polarization model appeared to solve these limitations by using the complex, frequency-dependent permittivities (εp*,εf*) instead of its low-frequency, real values.10-14 If, in the range of frequencies of interest, neither the fluid nor the particles experience any dielectric relaxation, then the MaxwellWagner interfacial polarization can describe the process by which particles get polarized.15 Under such conditions, the permittivity (6) Hao, T. Adv. Colloid Interface Sci. 2002, 97, 1. (7) Otsubo, Y.; Watanabe, K. J. Soc. Rheol., Jpn. 1990, 18, 111. (8) Rankin, P. J.; Klingenberg, D. J. J. Rheol. 1998, 42, 639. (9) Zukoski, C. F. Annu. Rev. Mater. Sci. 1993, 23, 45. (10) Anderson, R. A. Proceedings of the 3rd International Conference on Electrorheological Fluids; World Scientific: Singapore, 1992; p 81. (11) Davis, L. C. J. Appl. Phys. 1992, 72, 1334. (12) Chen, Y.; Conrad, H. In Developments in Non-Newtonian Flows; Siginer, D., Van Arsdale, W., Altan, M., Alexandrou, A., Eds.; ASME: New York, 1993;Vol. 175, p 199. (13) Parthasarathy, M.; Klingenberg, D. J. Mater. Sci. Eng. 1996, R17, 57. (14) Wu, C. W.; Conrad, H. J. Phys. D: Appl. Phys. 1996, 29, 3147. (15) Maxwell, J. C. A Treatise on Electricity and Magnetism; Dover Publications: New York, 1954.
Published on Web 10/12/2010
DOI: 10.1021/la1029036
16833
Article
Ramos-Tejada et al.
can be written
εp, f ¼ εp, f - i
σ p, f ω
ð1Þ
√ where i = -1 and σp(σf) is the particle (fluid) conductivity. From eq 1, it is found that at high frequency it is the permittivity mismatch that controls the polarization, whereas at low frequencies the conductivity mismatch βσ = (σp - σf)/(σp þ 2σf) is the determinant quantity. The characteristic frequency separating both domains is the Maxwell-Wagner relaxation frequency ωMW ¼
2πðσp þ 2σf Þ εp þ 2εf
ð2Þ
In fact, whereas the dielectric polarization model seems to explain well the ER behavior at high-frequency fields, it has been recognized that in the case of dc or low frequency ac applied field the conducting properties of the solid and suspending liquid play an important role in the ER effect.16,17 The conduction model18-20 considers that in dc fields the effect is induced by the mismatch of the conductivity of the particles and host fluid, Γ = σp/σf. Qualitative analyses20,21 suggest that an aggregation of the particles to form chains between the two electrodes is possible only if Γ > 1; in this case, the dipole moment is parallel to the field, and two particles attract each other and tend to align with the field. In this way, stable chains of particles can easily form. The attractive force between the particles is then proportional En, where n f 2 at low fields and n f 1 at high fields, the latter being a consequence of the field dependence of the liquid conductivity.14,21 However, when the conductivity of the particles is too high, complete chains do not form when an electric field is applied so that the ER effect may disappear.22,23 When σp < σf, the apparent viscosity of the whole suspension decreases as the external electric field increases. This phenomenon is totally opposite to the ER effect, and it has been termed the negative ER effect.6,21 Boissy et al.21 suggested that the mechanism responsible for this response is an electrophoretic effect: this induces a segregation of the fluid into two adjacent phases of high and low volume fractions, respectively, under the action of the field on the charged particles that migrate toward an electrode. Wu and Conrad16 provide additional information on the negative ER effect and suggest that the electrophoretic or dielectrophoretic-like effect may not exist or that it is very small. They studied Teflon/silicone oil systems and found that the particles acquired a negative charge probably because of some interaction between the oil and the Teflon particles. Their theoretical analysis shows that if both σp > σf and εp > εf, then the ER response is always positive. In the opposite case (σp < σf, εp < εf), a negative ER response can be expected. For other combinations, the frequency of the electric field must be considered: if σp , σf and εp > εf, then the ER effect changes from positive to negative upon decreasing the frequency; if σp > σf and εp < εf, then the same kind of change will be observed if the frequency is increased. In summary, in dc fields as those considered (16) Wu, C. W.; Conrad, H. J. Rheol. 1997, 41, 267. (17) Atten, P.; Foulc, J. N.; Felici, N. Int. J. Mod. Phys. B 1994, 8, 2731. (18) Tang, X.; Wu, C.; Conrad, H. J. Rheol. 1995, 39, 1059. (19) Conrad, H.; Li, Y.; Chen, Y. J. Rheol. 1995, 39, 1041. (20) Felici, N.; Foulc, J. N.; Atten, P. In Electrorheological Fluids; Tao, R. Roy, G. D., Eds.; World Scientific: Singapore, 1994; p 139. (21) Boissy, C.; Atten, P.; Foulc, J. N. J. Electrost. 1995, 35, 13. (22) Boissy, C.; Atten, P.; Foulc, J. N. In Electro-Rheological Fluids, MagnetoRheological Suspensions and Associated Technology; Bullough, W. A., Ed.; World Scientific: Singapore, 1995; p 753. (23) Block, H.; Rattray, P.; Watson, T. Proceedings of the 3rd International Conference on Electrorheological Fluids; World Scientific: Singapore, 1992; p 93.
16834 DOI: 10.1021/la1029036
in the present investigation, the ER response will be negative if σp < σf and εp < εf or if σp , σf and εp > εf. However, this model still fails to explain why some systems such as magnesium hydroxide/silicone oil suspensions24 and ZnO nanowires/silicone oil suspensions25 display a clear ER negative effect while not fulfilling the above-mentioned criteria. Trlica et al.24 suggested that the shape of the suspended particles may play a significant role in this effect, and Feng et al.25 found that ZnO nanowires migrated toward both electrodes, which was different from the usual case (migration toward one electrode, i.e., the particles acquire a single charge sign). They speculate that the aggregation of ZnO nanowires on both electrodes originates from an electron transfer among ZnO nanowires under the dc electric field. Other models have been proposed that try to overcome the limitations of the conduction and polarization approaches just described. Hao et al.6,26-28 introduced the so-called dielectric loss model, according to which the formation of particle chains is possible only if the polarized particles can reorientate in the field direction. This requires a high dielectric loss tangent (>0.1) and hence a comparatively high density of bound surface charge. These authors quantitatively found that a negative ER effect would require εp > 4εf dεp dεf =dT 27φ2 ε2f ðεp - 4εf Þ 2
εf ðε2p
- 1:5εp εf - ε2f Þ
although these conditions are somewhat difficult to test experimentally because we seldom have so much information about the effect of temperature on the permittivities of both phases. If εp < 4εf, then a weakly positive or clearly negative ER effect can be predicted. The effect of mixing particles of different sizes on the ER response has been the object of only a few investigations.29-34 Ota and Miyamoto29 carried out a series of computer simulations using chains of cubic-shaped particles, and they disclosed that a homogeneous ER fluid should give the largest yield stress. Wu and Conrad30 supported these conclusions by experiments with mixed glass spheres of two sizes (6 and 100 μm) in silicone oil and found that the yield stress decreased to a minimum value when the volume fraction of the small particles equaled that of the large ones. Tam et al.31 explained the reduction of yield stress of a bidisperse fluid as follows: the heterogeneous chains are weak because the smaller particles embedded in the chains have weaker dipole interactions, producing an easier breaking or deformation. They found a drastic improvement in yield stress of bare-glass ER systems upon addition of increasing concentrations of nanoparticles of lead zirconate titanate or lead titanate, with a fixed total solid/liquid volume fraction. Their experimental results suggested (24) Trlica, J.; Quadrat, O.; Bradna, P.; Pavlinek, V.; Saha, P. J. Rheol. 1996, 40, 943. (25) Feng, P.; Wan, Q.; Fu, X. Q.; Wang, T. H.; Tian, Y. Appl. Phys. Lett. 2005, 87, 033114. (26) Hao, T.; Kawai, A.; Ikazaki, F. Langmuir 1998, 14, 1256. (27) Hao, T.; Kawai, A.; Ikazaki, F. Langmuir 1999, 15, 918. (28) Hao, T.; Kawai, A.; Ikazaki, F. Langmuir 2000, 16, 3058. (29) Ota, M.; Miyamoto, T. J. Appl. Phys. 1994, 76, 5528. (30) Wu, C.; Conrad, H. J. Phys. D: Appl. Phys. 1998, 31, 3403. (31) Tam, W. Y.; Wen, W. J.; Sheng, P. Physica B 2000, 279, 171. (32) See, H.; Kawai, A.; Ikazaki, F. Rheol. Acta 2002, 41, 55. (33) Jun, J. B.; Uhm, S. Y.; Cho, S. H.; Suh, K. D. Langmuir 2004, 20, 2429. (34) Dassanayake, U. M.; Offner, S. S. R.; Hu, Y. Phys. Rev. E 2004, 69, 021507.
Langmuir 2010, 26(22), 16833–16840
Ramos-Tejada et al.
that nanoparticles not only modify the electric permittivity of the liquid but also coat the micrometer glass spheres to form, effectively, particles with higher permittivity. Other papers show trends opposite to the latter observations: See et al.32 found that the field-induced yield stress of the mixed size system was slightly higher than that of the two original monodisperse suspensions. They argued that a possible reason for the different behaviors observed could be that the size heterogeneity influences the packing of the particles and hence the microstructure of the columns or particle aggregates responsible for the flow resistance. Jun et al.33 found that in most cases bidisperse ER fluids exhibited shear yield stresses lower than those corresponding to the homogeneous ER fluid composed of just the larger particles. However, a great enhancement in the yield stress was peculiarly obtained when a low fraction of small hydrolyzed styrene-acrylonitrile copolymer particles were mixed with large particles of the same copolymer.34 The reason suggested for this synergistic ER effect was that the presence of two different particle sizes strongly affects the ordering and packing in the particle microstructure. This brings about as a consequence an increased stiffness of the chains or columns responsible for the stress buildup. In the present Article, we describe experimental results on the electrorheology of suspensions of different solids in silicone oils of several viscosities. It will be shown that the negative ER effect is in fact more frequent than expected and appears for low volume fractions and high field strengths and is also more likely when the viscosity of the medium is low. The response is negative for goethite, goethite/silica, hematite, and quartz (all inorganic with relatively high conductivity) but positive for cellulose and montmorillonite particles. Rheological data will be completed by microscopic observations of the structures presented by the suspensions in the presence of the (dc) electric field, and the whole set of results will be discussed in the light of existing theories.
2. Experimental Section Materials. The goethite (β-FeOOH) particles used in this investigation were purchased from Bayer (Barcelona, Spain) under the trademark BayFerrox 920. According to the manufacturer, the mass density of the particles is 4.1 g/cm3, and the average semiaxes of these rodlike particles are 50 and 400 nm, respectively (Figure S1, Supporting Information). In some cases, goethite was mixed with silica nanoparticles to analyze the possible role of the nanoparticle inclusions on the rod-rod interactions. The silica nanoparticles employed (Aerosil 300) were 7 nm in diameter and manufactured by Evonik (Hanau-Wolfgang, Germany). Two hematite samples were also tested: one was from Bayer (BayFerrox 130) and consisted of approximately spherical particles 170 nm in diameter. The other was from Sigma-Aldrich (St. Louis, MO) (size fraction
