Platy clay crystals, for example of bentonite, are common functional fillers in filled polymer systems, among other things because they decrease permeability (if well bonded to the polymer matrix) they provide enormous strength. Each clay grain has quartz-like mechanical properties in its two long dimensions, due to its phyllosilicate structure. The individual clay grains have very large aspect ratios, which allows them to transfer the large loads they can withstand to the polymer matrix over a large contact area. However, they are normally limited to fairly low volume fractions in the filled system because their orientation is random, which limits the strength that can be achieved by the resulting composite material.
One researcher whose name I can’t recall at the moment (though he was polite enough to answer my email, years ago) was able to deposit grains of bentonite all in parallel planes, bound together by something like limpet glue, with the result that the resulting composite material had a very high volume fraction of clay particles, and thus a very high strength and stiffness; he had to devise a new means for testing the stiffness by measuring small displacements under a microscope, using a bunch of glass spheres stuck to the film of material as the load was applied, then writing software to analyze the images to measure the displacement. But his procedure was extremely slow, depositing one layer of clay on the surface at a time.
I just thought of a process which might be faster, although it still deposits layer by layer.
The clay particles with appropriate surface treatment are suspended in a resin dissolved in a solvent. This is spin-coated onto a substrate, which orients the clay particles parallel to the surface of the thin layer, then heated to drive off the solvent, leaving a layer of solid resin and clay particles. The heating time needs to be long enough for the solvent to diffuse out from underneath clay particles. Another layer is applied on top of the first, and similarly dried, and the process is repeated many times. If the clay particles are 10% of the original dispersion, but the solvent is 85% of it, then after the solvent is removed, the clay particles will have a volume fraction of some 67%, which should be enough to have clay particles on each layer overlapping particles on the previous and next layers enough to form a continuous filler network.
Two possible improvements are relevant.
First, after evaporating each solvent layer, the new surface is washed before the next spin-coating step long enough to remove some of the resin just deposited. This will preferentially remove resin that is not underneath a clay particle, although if continued long enough it will remove clay particles too. So there’s a certain range of washing intensity within which this change will give a higher volume fraction of filler in the final product.
Second, although the resin used is still solid when the solvent evaporates, it is a photopolymerizable precursor to another resin. This permits a higher-strength final product by photopolymerizing the whole thing at the end, as well as 3-D printing by selective photopolymerization of each layer, which in that case would also need to include UV absorbers to keep the photopolymerization at the surface. See Powder-bed 3-D printing with a sacrificial binder for related 3-D printing processes.