MODELING NON-LINEAR OSCILLATORY AND COLLECTIVE PHENOMENA IN ACTIVE COLLOIDAL ASSEMBLIES
Open Access
- Author:
- Sanchez Farran, Maria Antonieta
- Graduate Program:
- Chemical Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- May 29, 2019
- Committee Members:
- Ali Borhan, Dissertation Advisor/Co-Advisor
Ali Borhan, Committee Chair/Co-Chair
Antonios Armaou, Committee Member
Themis Matsoukas, Committee Member
Ayusman Sen, Outside Member - Keywords:
- autocatalysis
colloids
diffusiophoresis
nanomotors
oscillations - Abstract:
- Active colloids exhibit self-powered motion and collective phenomena that resem- ble biological systems such as bacterial colonies. This “intelligent” behavior makes colloid functionalization an attractive alternative for the design of smart micro- machines. In this thesis, we developed an active colloidal system that displays oscillations and local synchronization. Here, colloidal assemblies (clusters) of sil- ver phosphate microparticles undergo surface reactions under the presence of UV illumination and hydrogen peroxide. These reactions involve the production and consumption of ions with different diffusivities, which leads to local chemical gra- dients and self-generated electric fields. A non-uniform particle distribution leads to a net electric field, to which the particles respond. Self-motility arises from self-diffusiophoresis while overlapping electric fields lead to diffusiophoretic inter- actions between neighboring particles. The particles oscillate between two behav- iors, schooling and dispersion; we provide evidence for incipient synchronization in the transition from schooling to dispersion. Such collective behavior is observed within H 2 O 2 concentrations of 0.1%-0.25% w/w and the oscillation frequency in- creases with hydrogen peroxide concentration. Starting from our experimental findings, our objective was to develop non-linear models that would qualitatively capture the oscillations and local synchronization in experiments. To model the dynamics of interacting clusters, we developed two approaches that successfully capture the qualitative features of the experimental system. The first approach adopts the Boissonade and De Kepper model, a generic non-linear model used for oscillatory reactions in solution, to include inter-cluster interactions in our heterogeneous system. As mentioned above, the diffusiophoretic interactions between clusters are driven by a reacting cluster surface. Therefore, the coupling term involves the rate at which neighboring clusters change their surface composition, while the strength of interaction is a function of inter-cluster distance. When the interactions are sufficiently strong, the clusters oscillate and synchronize. We also explore the transition from a non-oscillatory to an oscillatory regime and learn that fluctuations in hydrogen peroxide may drive a quiescent system into oscillations and local synchronization. In addition to a generic model, we developed a non-linear system that is based on a simple kinetic model in which the relevant surface species are ∗ OOH – and ∗ O. Multi-cluster simulations based on this model predict different dynamical regimes: synchronization with zero phase offset, synchronization with a stable phase offset, and oscillations without a stable phase relationship. The richer and more exotic behavior comes from frustrated interactions within the system. An interesting finding is that, at moderate interactions, the dynamics of a cluster triplet with equal interactions satisfies a rule in which two clusters oscillate with large ampli- tude and a π relative phase shift, while the third cluster has a small oscillation amplitude and phase locks to the large amplitude clusters. This rule is similar in spirit to how magnetic spins order in anti-ferromagnetic systems. Such systems exhibiting interesting properties such as ground state degeneracy: when the spins interact within a lattice, there are several spin conformations that minimize un- favorable interactions. Studies have shown that ground state degeneracy may be exploited towards encoding information. As opposed to frustrated magnets, our dynamical model has no associated energy. However, we can relate the concept of ground state degeneracy to a number of persistent dynamical states that sat- isfy the triplet rule in a network of interacting triplets. If this is true, a triplet network could also be used to encode information. We performed a simulation- based investigation and identified several persistent dynamical states that satisfy the triplet rule at moderate interactions. These dynamical states share bound- aries, a required property for assembling elementary logic circuits. We illustrate the design of a two-input/one-output AND logic gate, and we initiate a discussion on the practical considerations for their implementation.