In the early 2000’s, nanocomposites were the big field of research in materials science (not to say that it has died down, although it can be reasonably claimed that the hype phase has passed). There were some remarkable findings in the previous decade that motivated anyone loosely associated with the field to seek to capitalize.
One of the big motivators was a material called Ube nylon. Around 1996, Japanese researchers at Ube Industries working with Toyota created a wonder material. They took nylon-6 and mixed in a clay called montmorillonite. This type of clay, an aluminosilicate, consists of irregular sheet-like particles loosely held together by surface forces. When the clay is mixed in water, the layers swell and you get mud. It turns out, when you mix it with nylon, the layers separate and become individual sheets that distribute throughout the plastic matrix (or the pre-cursor to nylon, caprolactam, which is reacted with the clay to form the nylon nanocomposite). The individual platelets interact strongly with the nylon molecules, providing significant mechanical reinforcement and improved thermal stability. The big news was that this improvement allowed for nylon to be used as an engine for the first time, which gives big weight and cost savings. This had never been done before.
According to Ube Industries, Ube nylon-6 has the following claims:
Toughest among all engineering plastics
Excellent resistance to heat, oils, solvents, and chemicals
Low friction coefficient and excellent abrasion resistance with self-lubricating properties
Excellent processability — suitable for molding parts designed wtih complex shapes
Good compatibility with other fillers and stiffeners, including glass fiber
Superior oxygen gas barrier properties; suitable for food packaging.
Challenges of Clay Composites
Suddenly every research group was mixing clay in different plastics to repeat the success. But, despite the hype, the performance improvements were modest at best and in many cases, performance decreased with the filler. The work highlighted the need for fundamental knowledge regarding how the structure of the system (clay type, surface chemistry, and size, along with how it was distributed in the polymer) translated to end user performance. As a result, the field was effectively reduce to trial-and-error attempting to replicate an isolated event.
The major challenge was that the system was complicated, in both microstructure and in testing methods. Nanocomposites research requires an entirely new suite of tools that bridge across length scales to provide meaningful results. Predicting performance based on molecular structure isn’t new — there has been detailed knowledge of how carbon atoms “fit” within iron crystals to create steel for over a century. The challenge is finding the right properties to describe performance, and to see how these properties are connected to structure at different levels.
One big problem is that “clay” is a natural material, and montmorillonite particles are irregular in size, with variable spacing and complicated chemistry. When mixed in a polymer, the “nanoparticles” are not so nano, tending instead to distribute as clumps and agglomerates. In other words, the reinforcing particles are more like flaws that cause failure at lower stresses than the polymer alone (albeit at a slightly higher stiffness, which was generally regarded as reason enough to publish a paper). Alternatively, the clay may be intercalated, meaning that the layers are swelled by the polymer, but still not uniformly distributed. In rare cases, the swelling would be enough to overcome the interaction between layers and enable a true nanocomposite structure.
To understand these systems, extensive imaging of the individual particles at different positions was needed (such as the embedded video below, which is made from a collection of hundreds of individual microscope images!), along with mechanical, fracture, thermal, and other methods. Toughness was a particular problem because of the complex nature of failure in polymers. Most methods to evaluate fracture behavior provide little insight into how a filler particle actually contributes to toughness.
Synthetic Clays for Model Studies
One of the major contributions to the field came from a random meeting in an elevator between two professors at Texas A&M University, Prof. Abraham Clearfield and Prof. Hung-Jue Sue. In the late 1960’s, Prof. Clearfield had discovered a way to synthesize a material called alpha-zirconium phosphate (ZrP), which had a strikingly similar structure to clays like MMT, with the advantage that it was grown under controlled conditions and had much more uniform chemical and physical properties. Prof. Sue was investigating clay-filled polymers and quickly realized the value of having a “model clay” to help clear the ambiguity in the field.
Long story short, it was demonstrated that the unique properties of Ube nylon come from specific chemical interactions between nylon and the surface of the clay. This provided a strong interface and enabled the clay to be uniformly distributed in the polymer, both of which are necessary for effective reinforcement.
Some major findings from this research:
To get the biggest gains in performance at low concentration, uniform dispersion is needed
Aggregation is the main reason for poor performance of most nanocomposites
Individual clay particles are not effective at improving the toughness of most polymers — changes in toughness are actually due to aggregations
Larger platelets can be more effective at reinforcement at lower concentration
Extreme size platelets can lead to processing challenges due to increase in viscosity, and show higher flexibility that limit reinforcement effect
Dispersion state can be tuned by grafting different molecules on surface of nanoplatelets
In some cases, maximum concentration of nanoplatelets can be increased due to self-assembly of large particles, providing substantial improvements in mechanical properties and gas barrier characteristics
The moral of the story is that experiment and theory need to go hand-in-hand. Nanoparticles alone are not enough — aggregated particles will hurt performance. Effective testing methods are needed to understand the structure across multiple length scales, which requires collaboration between multiple disciplines and clear communication of results.
Beyond Uniform Dispersion
In addition to improved fundamental understanding, the use of model ZrP nanoplatelets has lead to a number of discoveries due to the unique ability to control the particle structure. One interesting phenomena observed with ZrP nanoplatelets is self-assembly. If a small concentration of plate-like nanoparticles (nanoplatelets) is well dispersed in an amorphous (non-crystallizing) polymer such as epoxy, they will be randomly positioned without long-range order. As concentration increases, there is a "choice" faced by the particles - they can either jam together into a gel, resulting in a semi-frozen glass-like material, or they can start to orient in a similar direction, increasing their packing efficiency. If the size of the particles is well controlled, the latter route can be promoted, resulting in systems with long-range order similar to nacre. In this case, the organization occurs due to the properties of the particle and is not due to any fancy deposition technique, which is a more scalable strategy. An additional advantage is processability - typically, the viscosity of this type of system is thousands of times higher than the epoxy at relatively low contents of nanoparticles, but if long-range order is achieved, the viscosity may be similar to the unfilled material.
Minhao Wong and co-workers reported the formation of a self-assembled mesophase of nanoplatelets in an epoxy. They were able to spray-coat the material to form a mechanically robust coating with excellent gas barrier properties due to the highly aligned and overlapping structure of the nanoplatelets. The aligned nanoplatelets also provided excellent anti-corrosion characteristics when applied to an aluminum substrate. The work was later extended to graphene to prepare high strength nanocomposites with excellent barrier properties at high humidity.
A second case was the use of nanoplatelets as solid dispersants for other nanoparticles, including zinc oxide quantum dots and carbon nanotubes. Interactions between two types of nanoparticles resulted in improvements in dispersion, and provided a route to tailor nanocomposite behavior [Sun et al., Small 2009, vol. 5, p 2692-2697]. The result was a remarkable 40% increase in stiffness and 50% in strength compared to an unfilled epoxy with only 1% addition of filler. In similar materials, a 10% improvement is considered significant, and generally only at much higher concentrations. The remarkable improvements were connected to a synergistic interaction between the nanoparticles, with the nanoplatelets serving as “electrostatic bridges” and generating an overall stress shielding effect in the nanocomposite [White et al., Polymer 2016, vol. 84 p 223-233].