MATERIALS SCIENCE AND CHEMISTRY

The polarization of nano-objects in salt solution (electrolytes) in alternating electric fields (AC fields) is the result of a complex interplay of competing processes, involving, among other, the motion of ions in free solution, the motion of ions along the surface of the nano-objects, the motion of nano-objects, the deformation of nano-objects. All these processes have different speed, i.e., different characteristic time scales. At very high AC frequencies, they are all suppressed and the polarizability drops to zero. However, at intermediate frequencies, they contribute with different weights. From an application point of view, this opens the interesting perspective that the AC frequency can be used to manipulate and tune the local charge structure in the vicinity of the nano-object and the resulting polarization.

The goal of the project was to disentangle and analyze the key processes contributing to the polarizability of nano-objects, in order to obtain basic insights into the dynamics of complex charged fluids and to provide guidance in the design of future manipulation strategies. Two types of nano-objects were studied: Nanocolloids and charged polymers (polyelectrolytes). Even though coarse-grained models were used, the study required extensive simulations and would not have been possible without the use of supercomputer resources.

The simulations were based on ‘’Dissipative particle dynamics’’ (DPD), and the fluid particles are represented by simple ‘’DPD’’-beads which interact with each other only by friction. Such models have proven to be efficient for simulating fluid flow phenomena on nanometer scales or higher. Additionally, a method was developed which allows avoiding the use of overly bulky explicit salt ions [10]. The use of this method turned out to be crucial for the success of the polyelectrolyte studies [1].

Put simply, the processes determining the polarizability of nano-objects – as observed in the simulations - can be summarized as follows: First, if electric fields are applied, both charged nano-objects and the ions in solution are set into motion, in opposite directions depending on their charge. Nanocolloids are then obstacles for the salt ions. As a result they are polarized even if they are not charged! This is because the ions accumulate at the front and back side (Fig. 2, top).

Second, if the colloids are charged, another mechanism comes into play and dominates: Charged particles are surrounded by a layer of oppositely charged ions (the electric double layer, EDL). This layer gets distorted in the presence of electric fields. The resulting polarization is opposite to that found for uncharged colloids (Fig. 2, bottom).

Third, ions are also driven from regions with high salt concentrations (thick EDL) to regions with low salt concentrations (thin EDL), which counteracts the distortion of the EDL. However, this can only happen if the ions have time to diffuse along the side of the colloid, i.e., at low frequencies. Therefore, the polarizability initially increases with the frequency until, at very high frequencies, it vanishes again because the ions don’t have time to follow the electric field at all (Fig. 2, bottom).

In the case of colloids, analytical theories can be found in the literature which can be tested against the simulations. In particular, the electrokinetic theory accounts for all mechanisms of polarization described above and shows very good agreement with the simulation results (grey lines in Figs. 2 and 3). For flexible polyelectrolytes, the situation is much more complicated and developing analytical theories is more difficult. Nevertheless, the simulation results show that the behavior of polarizability as a function of the frequency is very similar to that observed for colloids, suggesting that the main mechanisms of polarization are the same (Fig.3). Moreover, the simulations show that the polarizability scales in a slightly superlinear manner with the size of the molecule in agreement with theoretical expectations for short chains. Theoretically, one expects linear scaling in the limit of very long chains.