Back to Table of Contents
43672-AC10
Thermal Transport across Organic/Inorganic Interfaces
Simon R. Phillpot, University of Florida
While most
nanostructures investigated to-date have been fabricated from purely inorganic
materials, the study of organic/inorganic and bio/inorganic hybrid
nanostructures is an important research frontier with enormous potential
applications in areas such as bioMEMs and microelectronics. A key maintenance
task in all nanostructures is to efficiently move this heat away from the
active components into heat-management structures such that the functionality
and structural integrity of the nanostructure are not compromised Indeed, there
is a growing recognition that in many cases heat-management structures have to
be designed into devices with the same care that is given to the design of the
primary functionality. The issue of thermal management in organic/inorganic
systems is potentially even more critical since the very structure of many
organic materials can be irreversibly compromised by even rather modest
temperatures increases.
We have addressed the question as to how high the interfacial conductance
of an organic/inorganic interface could be under optimum conditions. To address
this question, we used atomic-level simulation methods to explore the thermal
transport properties across highly idealized polyethylene/diamond interfaces,
which we take as a prototypical organic/inorganic system. We have found that,
in principle organic/inorganic interfaces should be able to conduct heat as
well as inorganic/inorganic interfaces if there is strong bonding at the
interface itself.
Before
addressing the interface itself, we determined the thermal conductivity of the
polymer alone. We determined the axial thermal conductivity for a polyethylene
crystal of infinite molecular weight to be 310 ± 190 W/m•K. This is
considerably larger than either the extrapolated values of 40 and 70 W/m•K
determined experimentally. One of the reasons for the discrepancy between our
calculated values and the experimental values is probably the absence of
imperfections in the simulated PE structure. We thus characterized the effects
of adding defects in the form of unsaturated carbons. Because of the shorter,
and thus stronger C-C bond lengths, these segments have a higher intrinsic thermal
conductivity than the polyethylene itself. It turns out that the detrimental
effect of the defects is almost exactly balanced by the increase in the thermal
conductivity due to the superior intrinsic thermal transport properties of the
segments added. By contrast, when polyethylene defects are added to
polyacetylene, both the defects themselves and their lower intrinsic thermal
conductivity lead to a strong decrease in the overall thermal conductivity.
The calculated values for the polyethylene/diamond interfacial
conductance range from 690 to 930 MW/m2K, depending on the surface
orientation of the diamond crystal.
Although these values are extremely large, they are not without
precedent. In particular, strongly bonded inorganic interfaces give interfacial
conductances that reach values of hundreds of MW/ m2•K. Also, from the
earlier molecular-dynamics simulations (partly funded by this project), the
interfacial conductance in diamond was predicted to be 9-17 GW/m2•K depending on
the type of grain boundary. Moreover, a simulation of a model diamond
nanocrystal containing a number of high-angle (001) grain boundaries yielded an
average interfacial conductance of ~4.5 GW/m2•K, a result that
was reasonably consistent with the experimentally derived value ~3 GW/m2•K for
ultrananocrystalline diamond thin films.
Our
simulations demonstrate the possibility of extremely high thermal transport in
polyethylene and across polyethylene/diamond interfaces. Adhesion of adsorbates
to the diamond surface can be dramatically increased by surface dehydrogenation
and creation of reactive sites such as C· radicals, polar C=O and π-
bonded C=C groups. Thus polyethylene bonding to the diamond may be viable but
will require additional chemical treatment. It may thus be anticipated that the
increasing ability to functionalize surfaces may lead to methods for the
systematic development of inorganic/organic interfaces with conductances
similar to those of interfaces between inorganic materials.
Back to top