Introduction to the Thesis 2009

When I finished this thesis in early 2004 I regarded it as the best scientific introduction to EPD at the time, and to my knowledge it still is.

However, even though I thought it would be very helpful to people first exploring EPD, I have been slow in posting it to the web.  This is due in part to having little time to revisit the thesis, but in larger part due to my dissatisfaction with the completeness of explanation it offered.  I continued to see effects in my own work and in the work of others which were not covered by the by the theories presented in the thesis.  It gradually became clear to me that to understand the results of the wide variety of experiments that have been reported under the heading of EPD, it is vital to understand the behavior of colloids which are neither fully dispersed nor rigidly deposited.

In my thesis I adopted a fairly strict definition of EPD.  This definition requires that the system begin as a stable dispersion of particles which can move independently, and that the end result be a rigid deposit of particles contacting each other at their primary minimum.  It was not a conscious decision at the time, but by adopting this definition I was able to avoid making explicit two very significant assumptions.

The first of these assumptions regards dispersion.  In the thesis section on dispersion I listed the standard methods for keeping particles far enough apart that they are not pulled together irreversibly by the London-van der Waals force.  The implicit assumption was that if the attractive energy between two particles was less than a small multiple of the thermal energy of the solvent, Brownian motion would be sufficient to assure that the particles do not become irreversibly attached to each other, and the particles could therefore be considered to move independently.  However, as the volume fraction of particles in suspension rises, the interaction between particles changes from simple two particle interactions to multiparticle interactions where crowding and momentum transfers become significant.   At the extreme, particles can pack together to the point that the suspension becomes a pseudoplastic solid.  These particles are not deposited, the energy barrier between them is not overcome, and with an adequate volume of solvent they will spontaneously diffuse apart.  These particles technically remain dispersed - that is all of the particles are separated by a layer of solvent - even while the dispersion behaves as a solid.

The second implicit assumption is that if the barrier between the particles does not keep them far enough apart, the L-vdW force will inexorably pull the particles into contact at their primary minimum, a contact that is rigid and brittle.  However in some cases the interaction between the solvent and the surface is stronger than the attraction between the particles.  A layer of solvent then remains between the particles.  In this case, even though the particles are bound together by the L-vdW force, the contact between the particles is lubricated.  The particles are able to rotate and slide past each other.  At high volume fractions they will form a solid capable of large plastic deformation without losing cohesion (i.e. without cracking).  At low volume fractions they can form a gel which does not sediment or phase separate but which flows at such a low applied shear force that it can be mistaken for a fully dispersed colloid.

So what is needed to complete the scientific background necessary for understanding EPD?  It is the science necessary to describe the behavior of these “Condensed Colloids”, colloids with a particulate volume fraction generally greater than 30% and with non-Newtonian shear behavior.

The first step in explaining these colloids is to extend the discussion of the van der Waals forces to include quantitative approximations.  A quantitative understanding is vital because the behavior of these colloids depends not just on the nature of these forces but on the ratios of these forces on the angstrom to nanometer scale. 

This then leads into a discussion of the solvent/particle interface.  The Keesom and Debye vdW forces at a solid interface can alter the structure of a solvent.  In polar solvents, this alteration can be so significant that the nanometer of solvent at the interface will have significantly different properties from the bulk solvent.  These solvent/solid interactions have been most extensively considered in the literature of wetting, but the focus needed here is primarily on wetting considered on the nanoscopic scale.

The next step is to consider how this altered solvent layer reacts as two particles come together.  This should be divided into three cases: when the solvent is completely squeezed out of the contact point, when a monolayer of solvent remains and when the system is stable with a layer several molecules thick.  If the solvent is completely removed, the dry contact points between the particles will have mechanical properties approaching those of the material making up the particles.  If a monolayer of solvent remains, the contact point will be much weaker and particles will be better able to roll and possibly slide past each other.  If several molecular layers remain between the particles at the contact point, the contact will be lubricated allowing the particles to easily roll and slide past each other while also providing an elastic connection between the particles.

The micromechanics of the contact point between particles then needs to be translated into the mechanical behavior of the colloid.  Since the usual objective of EPD is to maximize the consolidation of the particles, we are most interested in the compressive strength of the particulate component of the colloid.  However, this property is rarely measured.  Shear strength is more easily and far more often measured for colloids, therefore it would be more useful if a connection between shear and compressive strengths could be drawn.

The compliment to the compressive strength of the particulate component of a colloid is the compressive force acting on those particles.  In EPD the particles move independently.  Their migration in an electric field toward the deposition electrode is well described by straightforward and well-tested electrokinetic theories.  In a condensed colloid, the particles by definition do not move independently.  Electrokinetic effects chain up to act on multiple particles which also act mechanically on each other.  The high viscosity of the condensed colloid will stabilize the solvent against large scale convection.  This allows ionic concentration gradients to develop, modifying electric fields and osmotic pressures.

It is because the consolidation of condensed colloids is so fundamentally different from EPD that it should be separated from EPD and referred to as either electrophoretic consolidation (EPC) or simply electrocasting.  However, because many processes are a combination of EPD and EPC and because even processes which are purely electrocasting have been identified in the literature as EPD, it is not possible to fully understand EPD without understanding EPC as well.

Finally, there is the subject of drying.  It is rare for EPD or EPC to produce a fully dense deposit prior to drying.  If a low density, rigid deposition is formed, ≈30vol%, it will almost inevitably crack during drying.  If a high density, rigid deposit is formed, 40-50vol%, cracking can frequently be avoided.  However, the highest density coatings are formed when the particles are not rigidly deposited.  It occurs when the particles are consolidated on the deposition electrode in a solvated pseudoplastic slurry.  This condensed colloidal slurry has enough shear strength to allow it to be removed from the deposition apparatus, but the particles are still able to re-arrange during drying, achieving near theoretical maximum random packing density.  However, this is not a panacea.  If the shear yield strength of the drying deposit is too high in relation to the crack initiation and propagation energy, the coating will achieve high density as a collection of cracked flakes.

EPD and EPC are surprisingly complex phenomena.  Understanding all of the phenomena exhibited by experiments in this field requires an understanding of phenomena in a wide variety of scientific subject areas.  As time allows, I will elaborate on the topics above and more, both on this site and through additional journal publications.  The long term goal is a complete introduction to the relevant existing science.  This will firstly promote novel, clearly designed experiments focused on gaps in understanding, and secondly, provide those tasked with developing new EPD/EPC systems a logical basis for designing and improving processes.