Historical perspectiveEvaluation of particle charging in non-aqueous suspensions
Graphical abstract
The stability of non-aqueous suspensions (sols) is dependent on sufficient charging of dispersed particles. Charges are created by electron, anion and proton adsorption. Sources for charged species are suspension liquids as well as dissolved ions and impurities. Zeta potential is one of the most sensitive experimentally available parameters for charging in non-aqueous suspensions. Changes in zeta potential of particles can be related to proton dissociation, acid-base adduct formation and hydrogen bonding energies (enthalpies) of suspension liquids. Moreover, it may equally well be related to proton reduction and water electrolytic dissociation potentials, as well as to Fermi energies for electron exchange. The repulsive energy model developed for aqueous electrolyte suspensions can be substantially simplified when dealing with non-aqueous suspensions. The exponential ratio between distance to repulsive maximum ensuring suspension stability and diffuse charge layer thickness can be used to evaluate the range of applicability of each model. The attractive energy may be neglected, but it is in most cases smaller but significant in comparison to repulsive energy.
Introduction
Particle charging in non-aqueous liquid suspensions (sols) has received increasing attention due to enhanced technological interest. Investigations on dispersant aided stabilization of carbon particles in benzene [1, 2] showed that the basic requirement was a sufficient surface charge. In fully or almost non-aqueous suspensions zeta potential is practically the only parameter available for characterization of particle charging. Electric charges and surface potentials depend not only on the nature of solid particles but also on the nature of suspension liquid and on soluble residues (potential determining ions, impurities, etc.). Verwey [3] found for some polar liquids that SiO2 was most acidic (strongest proton donor, most negative ζ-potential) followed by TiO2 and ZrO2. For any of these solids water was the strongest proton acceptor (Brϕnsted acid, most negative ζ-potential) followed by acetone, ethanol and methanol. Rutile became positive in xylene, butanol and heptanol and negative in butylamine and nitrobenzene (stronger bases). A distinction is therefore made between acidic (SiO2), neutral (TiO2, ZrO2) and basic (Al2O3, CaCO3, MgO) particles in very polar (εr > 40), polar (20 < εr < 40), low-polar (20 < εr < 10), non-polar (5 < εr < 10) and apolar (εr < 5) liquids.
Due to lack of charges it is very difficult to stabilize particles in apolar liquids. In fact, it is difficult to disperse ions and particles at all. Measurements of the effective surface (zeta) potential proves, however that some charges develop despite these expectations. Large ions are dissolved and large particles are suspended more readily than small ones. Therefore they tend to appear as characteristic clusters. Due to low dielectric constants the surface potential decays very slowly and can safely be replaced by zeta potential. The generally accepted Deryagin-Landau and Verwey-Overbeek (DLVO) model for suspension stability in aqueous systems [[4], [5], [6]], is replaced by Coulomb type models [6, 7]. As a general trend both repulsive and in particular attractive interaction energies are expected to be weakened.
Surface charging in non-aqueous suspensions is considered to occur through:
- 1.
Electron transfer due to extreme Lewis type of electron acceptor (acid) – electron donor (base) interactions.
- 2.
Proton transfer due to extreme Brϕnsted type of hydrogen bond (acid–base) interactions. This is typical in the presence of moist which results in hydrolysis to surface hydroxyl groups and in dissolution of ions as solvated complexes.
- 3.
Adsorption of surface active solutes (surfactants), ion transfer of dissolved ions or presence of liquid or solid impurities. This enlarges the investigation to three-component systems.
In order to match the classification of surface charging dispersion liquids could be classified as:
- 1.
Aprotic liquids, which cannot exchange protons, but can exchange electrons with the solid or form Lewis type adducts with surface sites.
- 2.
Protic liquids, which can react with surface sites followed by Bϕnsted type proton uptake or release. Alternatively they may solvate released protons or dissolved ions.
- 3.
Amphoteric liquids such as water, which can donate and accept protons at surface sites and in suspension. Weak acids may be activated by surface induced electrolytic dissociation (SIED) [8].
In apolar solvents the dispersive interaction between solids and dissolved Lewis or Brϕnsted active solutes is neutralized. This gives an opportunity to extract the specific interaction contribution from the overall interaction. There is, however no self-consistent parameter sensitive to all these properties. Since the relative permittivity (dielectric constant) has proven to be insufficient for liquid classification the evaluation has been extended to Hamaker constants and solubility parameters [9]. Thermodynamic characterization based on Gutmann's donor and acceptor numbers (acid–base enthalpies) has been extended to correlations with hydrogen ionization (protolysis) enthalpy as well as with hydrogen bond enthalpy and equilibrium constants [10].
In this evaluation published zeta potentials of SiO2, glass, TiO2, Al2O3, CaCO3, MgO and metallic gold (Ag) have been related to the introduced thermodynamic parameters. Finally, electronic scales for electron (Lewis activity), proton (Brϕnsted activity) and ion exchange are considered. Publications reveal a surprisingly large scatter in measured zeta potentials. Particles suspended in equal non-aqueous liquids are reported to possess both positive and negative surface potentials of considerable magnitude. The key aim of this review is therefore to evaluate the source of these differences. Finally the expected stability of suspensions are calculated from a Coulomb version of DLVO model [6, 7]. This provides an opportunity to estimate the distance of the repulsive maximum resulting in stability. For the sake of consistency only binary solid–liquid interactions are evaluated, including to some extent the influence of moist and impurities.
Section snippets
Experimental
Models used to calculate liquid and solid properties are presented in the next section. Moreover the relevant properties of liquids and solids are collected for internal comparison purposes. Finally, the conditions of experiments producing original data are presented.
Molecular interaction
The interaction energy of liquids can be expressed in terms of molar refraction and molar polarization. One may define the electronic molar refraction Rm (Table 2) [26] as:
This is Lorenz-Lorentz equation. Alternatively we may define the total molar polarization Pm (Table 2) [5, 26] as:
This is Clausius-Mossotti equation. The molar refraction and molar polarization are related through Debye-Langevin expression:
Zeta-potentials
As stated earlier this evaluation is restricted to binary (particle–solvent) systems. However, considering DLVOE model the influence of dry content (particle fraction) and thermal treatment of particles (number of surface hydroxyls, adsorbed water (moisture) and impurities are evaluated to some extent. The limiting properties of liquids adsorbed from fully dispersive apolar solvents on particles may be used as useful references for the extent and strength of specific solid – probe liquid
Interaction energy
Since concentration of dissolved charged species is unknown the repulsive internal energy is calculated from Morrison's expression for overall (repulsive) energy (Eq. (50)). All interaction energies are from now on expressed as kJ/mol. Although internal energy of attraction is assumed to be negligible as compared to the repulsive energy (ΔAUm < < ΔRUm), it is calculated for comparison using Eq. (36) for uniform particles (Table 5). Eq. (48) is used to calculate the distance to repulsive maximum
Conclusions
Charging in non-aqueous suspensions remains quite unclear and controversial. As stated in introduction the following three topics have been focused on in this review:
The charging seems to be initiated by both electron and proton exchange as well as by adsorption of liquid molecules and charged impurities. It was found that Gutmann's donor number (DN = negative Lewis type molar acid-base adduct formation enthalpy) was successfully used to exhibit the charging in non-aqueous dispersions, likewise
Acknowledgment
This investigation was supported by the Academy of Finland to Center of Excellence of Functional Materials, Laboratory of Physical Chemistry at Åbo Akademi University.
References (57)
- et al.
J Colloid Interface Sci
(1952) - et al.
J Colloid Interface Sci
(1953) - et al.
Adv Colloid Interface Sci
(2015) Adv Colloid Interface Sci
(2010)Adv Colloid Interface Sci
(2017)- et al.
Adv Colloid Interface Sci
(2008) - et al.
J Colloid Interface Sci
(1996) - et al.
J Colloid Interface Sci
(1984) - et al.
J Colloid Interface Sci
(1987) - et al.
J Electroanal Chem
(1976)