Thermodynamics of the adsorption of organic cations on kaolinite : temperature dependence and calorimetry

The present work is aimed at understanding the interactions involved in the adsorption of cationic surfactants on heterogeneous surfaces. The relevance of the study derives from the environmental aspects of the adsorption of small organic molecules onto soil constituents. This thesis emphasizes the experimental aspects.

In order to achieve a better understanding of the driving forces involved in the adsorption process, classical equilibrium thermodynamics is used to estimate the energetic and entropic parameters of the system.

The main experimental systems were a homo-ionic kaolinite in an aqueous electrolyte solution which contained a cationic surfactant with a dodecyl tail and either a pyridinium chloride (DPC), or a trimethylammonium bromide head group (DTAB).

In Chapter 2 a comprehensive study of the physico-chemical properties of the adsorbent has been carried out. Several techniques, such as X-ray diffractometry, electron microscopy, Ar/N2 adsorption calorimetry, and BET gas adsorption have been used to characterize the kaolinite surface. The collected evidence shows that our kaolinite is free of 2:1 clay contaminants. Combination of electrophoresis measurements and potentiometric titration results gives information on the properties of our representative of the group of 1:1 clay minerals. Contrary to the idea of some soil scientists, about 50% of our kaolinite surface charge stems from isomorphic substitution which is mainly exposed to the basal surfaces. These surfaces are homogeneous with respect to Ar and surfactant adsorption. The particles posses a variable charge (pH-dependent) on the edges. The edges have a more heterogeneous character. Constant hysteresis observed between back- and forward potentiometric titrations confirm the binary nature of the surface charge and the possibility of edge-plate interactions resulting in a so-called card-house packing. The ratio found for the edge/plate surface area depends on the method used, viz. 0.35 from DPC adsorption, 0.55 from potentiometric titration, 0.20 from Ar adsorption calorimetry and 0.41 from electron microscopy. The CEC measured by the silver-thiourea method amounts to 57 μmole/g whereas that determined by using the ammonium acetate method is 30 μmole/g. Potentiometric titrations show that the clay surface is not covered by any oxide coating. Over the experimental pH range of 4 to 10, neither an isoelectric point (iep), nor a zero point of charge (pzc) has been found for the entire particles. Extrapolation of the electrophoretic data suggests a pzc of about 2. The zero point of charge of the edges (epzc) is estimated from the inflection point of charge-potential curves which is located in the region of minimum electrolyte effect. Of the three cations Li ⁺, Na ⁺and Cs ⁺, Cs ⁺adsorbs most strongly, resulting in a lower electrophoretic mobility and a slightly lower epzc for Cs-kaolinite. The epzc is 6.7 for Li- and Na- kaolinite and 6.0 for Cs-kaolinite.

Protons are specifically adsorbed, not only on the functional edge groups but also on the plates. Adsorption of protons on Na-kaolinite is exothermic, with the enthalpies increasing when the surface charge becomes more negative. The electrolyte effect on the proton adsorption enthalpy is very small. This suggests that the pH has a more pronounced effect on the surface potential than the indifferent electrolyte. From the fact that electrophoretic mobilities change more strongly with pH than with the electrolyte concentration we come to the same conclusion. The proton adsorption enthalpies at each pH show a qualitative similarity to that of oxides with similar pzc's.

Chapter 3 deals with the properties of the adsorbate. Here the thermodynamic properties of a homologous series of surfactants are studied. The micellization enthalpies of three surfactants with C10, C12 and C14 tails and pyridinium head groups are directly measured as a function of electrolyte concentration and temperature. At a certain temperature and electrolyte concentration each surfactant has a characteristic critical micelle concentration (cmc) value, which is some 60-80, 10-20 and 5-10 mmoles per litre for C ₁₀ - C ₁₂ - and C ₁₄ -pyridinium chloride, respectively. Enthalpies of micellization of the surfactants (Δ _mic H _m ) are temperature dependent. They change sign at a certain temperature, T _trans . T _trans is dependent on the chain length and to a lesser extent on the electrolyte concentration. The (extrapolated) Δ _mic H _m (T) curves of the three surfactants cross each other at - -12°C. The enthalpy at this temperature is attributed to the head group contribution; here the enthalpic contribution of the tail is zero.

Increasing the temperature decreases the structuring of water around the tails and, consequently, the entropy rises upon association of tails, but this effect is more than compensated by the decrease in enthalpy.

An attempt has been made to break down the Gibbs energy, enthalpy and entropy of micellization into their electrical, hydrophobic and chemical constituents. Using the mass action model, Gibbs energies of micellization are estimated. In the presence of 0.1 M NaCl, the thermodynamic parameters of micellization of the surfactants expressed per CH ₂ group are very close to the corresponding parameters of transporting a methyl group from water to a hydrophobic phase. According to our calculations, hydrophobic bonding contributes most to the micellization Gibbs energy. The electrical contribution is rather small and unfavourable. The constant chemical contribution is also unfavourable and can perhaps be attributed to a decrease in hydration forces of the solvent around the head group.

The enthalpy of adsorption may be derived from adsorption isotherms determined at different temperatures or may be directly measured by using a microcalorimetric technique. The former method is discussed in Chapters 4 and 5, the latter in Chapter 6.

More specifically, in Chapter 4 the adsorption of DPC and DTAB on Nakaolinite is studied as a function of electrolyte concentration and pH at two temperatures. Adsorption isotherms have steep initial slopes reflecting the high affinity of the adsorptives for the surface. For both surfactants and at both temperatures, all isotherms, if measured at different salt concentrations, exhibit a common intersection point (cip) roughly around the iep. Below the cip adsorption is reduced by electrolyte addition; beyond it electrolyte promotes adsorption. Below the cip adsorption is to a large extent electrostatically driven; addition of salt reduces the attraction. However, beyond the cip association of adsorbed surfactant molecules takes place, which occurs despite the repulsion between the head groups; now electrolyte reduces this repulsion and hence promotes adsorption. The pyridinium head group shows a slightly higher affinity for the kaolinite surface than the trimethylammonium head group.

On kaolinite, adsorption of both surfactants takes place mainly on the plates. With increasing pH the (total) adsorption increases, but not so much that all adsorption sites on the edges become covered. On the plates adsorbed protons are, at least partly, exchanged against the surfactant molecules, as the pH is increased. A bilayer adsorption model has been developed on the basis of the Frumkin-Fowler-Guggenheim isotherm equation fits the adsorption data well.

Chapter 5 compares the temperature dependence of the adsorption of organic cations on Na-kaolinite with that on AgI. The former adsorbent is hydrophilic, the latter hydrophobic. On AgI the adsorption of tetrabutylammonium nitrate (TBAN) proceeds up to a monolayer, whereas on the hydrophilic kaolinite the adsorption of DPC and DTAB continues up to a bilayer. On AgI the adsorption of TBAN exhibits a maximum as a function of temperature. For amphiphilic molecules on kaolinite, the enthalpies of the formation of the first layer show hardly any temperature dependence, whereas those of the formation of the second layer again pass through a maximum as a function of temperature. The adsorption of surfactant molecules increases the hydrophobicity of kaolinite, hence, around the region of completion of the first layer, the surface becomes hydrophobic. Adsorption enthalpies on AgI and on the hydrophobic kaolinite show the same trend as those for the micellization of the surfactants used. This indicates that all these processes are determined by the same mechanism, viz. hydrophobic bonding.

Chapter 6 discusses the micro calorimetric measurements of the enthalpies of adsorption of DPC on Na-kaolinite at different salt levels and temperatures. Adsorption enthalpies are temperature dependent; they are positive at low temperatures, reduce to zero at about 24°C and turn to negative at T>24°C. A break in the plots of the cumulative adsorption enthalpy as a function of adsorbed amount is detected under most experimental conditions. This break, the iep and the cip of the adsorption isotherms at different electrolyte concentrations are all identical within experimental error. Below and above the break, the adsorption enthalpies are fairly constant, suggesting that the surface is homogeneous. This is in accordance with the earlier conclusion that
adsorption takes place mainly on the basal surfaces. The change of the slope of the ΔH(Γ) curves upon completion of the first layer shows again the different adsorption mechanism for the first and second layer. This part of the study confirms that among the different contributions to the enthalpy the hydrophobic part is the most important for the second layer formation. However, for the first layer the head group affinity for the surface dominates.

Comparison of the isosteric adsorption enthalpy and the directly measured heat of adsorption leads to the next two conclusions: (a) For the formation of the first layer the isosteric heat of adsorption deviates somewhat from that measured by calorimetry. (b) For the formation of the second layer, good agreement has been found between the two. Both values are rather close to the enthalpy of micelle formation. Probably an important reason for the discrepancy found for the first layer is the variation of the concentrations of the other adsorbed cations at the interface as a consequence of a temperature change. Since the formation of the second layer starts when the charge of the kaolinite is compensated, the concentration of these ions at the interface is of minor importance. Therefore, beyond the iep a complete agreement between the two techniques is found.

With regard to the adsorption mechanism our results can be summarized as follows:
- Adsorption of cationic surfactants on charged clay surfaces takes place with a Gibbs energy of few tens of kJ/mole, implying a physical adsorption.
- Surfactants adsorb primarily on the basal surfaces of kaolinite which possesses a negative charge. Whether the adsorption also takes place on the edges is uncertain. On the plates first a monolayer is formed, then a bilayer. For the former, electrolytes act as competitors, for the latter, where hydrophobic bonding is the driving force, as promoters.
- Formation of the second layer is similar to micellization. The enthalpies of both processes have the same trend as a function of temperature. They are close to each other and their transition temperature agrees within a few degrees.
- The nature of the head group does not play a dominating role in micelle formation or in adsorption.

The adsorption of amphiphiles is sensitive to parameters such as the nature of the surface charge, electrolyte concentration, pH, and temperature. Therefore, in different climates one must expect different adsorption capacities of soils and hence, different rates of water pollution. For example, in natural systems consisting mainly of particles with a permanent charge, the adsorption capacity is probably much higher than in those systems which contain relatively large amounts of particles with variable charge. Hence, from this point of view, tropical soils may be more prone to pollution caused by the transport of organic compounds to ground water.
The adsorption of amphiphiles is not limited to charge compensation. Therefore the value of the CEC may underestimate the adsorption capacity of soils.
As the hydrophobic character of the charged particles changes with the amount adsorbed, there may be a large effect of soil pollution on the permeability of soils.
The overall conclusion is that the present study not only contributes to our knowledge about adsorption forces but may also be helpful in understanding and predicting processes which occur in natural systems.