Individual carbon nanotubes (CNTs) can conduct heat effectively, but in large collections known as “forests,” their thermal conductivity can be drastically reduced. This phenomenon has been a matter of curiosity for researchers for some time; now, new research from researchers at the Stewart Blusson Quantum Matter Institute (SBQMI) and collaborators proposes a simple microscopic picture of this reduced thermal conductivity. The work, published this week in ACS Nano, can be applied to a number of materials and complex microstructures and could be used to explain experiments and devise design principles for advanced functional materials with tunable thermal properties.
“It has been commonly observed that while an isolated CNT is an extremely good heat conductor, the same CNT in bundles, sheets, or ‘forests’ shows a large drop in heat conductivity,” said Debashish Mukherji, SBQMI Research Associate and last author.
CNTs are cylinder-shaped molecules comprised of carbon atoms and, due to their promising physical properties, are excellent candidates for electronics, biosensors, and energy devices such as lithium-ion batteries. In a CNT “forest,” CNTs self-assemble in vertically oriented structures entwined with other CNTs.
“While the reduced heat conductivity was initially reported in CNT forests, a number of recent works have shown that a rather broad range of quasi-one-dimensional materials (Q1DMs) show this behavior as well,” said Mukherji. Q1DMs are considered promising potential building blocks for nanoscale electronic devices, and are thought to use less power than more complex materials.
“Here, the complexity of molecular assemblies with large entropic disorders has hindered the emergence of a clear microscopic picture of this intriguing effect,” said Mukherji.
Traditionally, all-atom simulations have been used to investigate the properties of Q1DMs, a daunting task even with modern advanced computational facilities. Additional challenges arise when dealing with complex molecular assemblies.
“To circumvent the difficulties associated with all-atom simulations, we take a completely different path and use a multiscale (coarse-grained) model that considers a Q1DM as a linear polymer chain and thus a Q1DM forest as a polymer brush, where certain properties of the native Q1DMs, such as flexibility, are inherently incorporated in our model via the system parameters. The use of such multiscale simulation methods is quite common in the soft condensed matter community,” said Mukherji. Soft condensed matter includes polymers, proteins and even products such as edible gels. “However, our work is the first in which a multiscale model in combination with the conventional polymer physics concepts is used to explain a counterintuitive hard condensed matter problem.”
“The advantage of such a minimalistic model is that it can capture a broad range of systems within one unified physical framework and provides a simple generic picture of an otherwise complex problem,” said Mukherji.
Citation: Aashish Bhardwaj, A. Srikantha Phani, Alireza Nojeh, and Debashish Mukherji. Thermal Transport in Molecular Forests. ACS Nano 2021 15 (1), 1826-1832. DOI: 10.1021/acsnano.0c09741