562
(Invited) Thermoelectric Properties of Semiconducting Single-Walled Carbon Nanotube Networks

Monday, 30 May 2016: 11:40
Aqua 313 (Hilton San Diego Bayfront)
A. J. Ferguson (National Renewable Energy Laboratory), A. D. Avery (Metropolitan State University of Denver), B. Norton-Baker, B. Zhou (National Renewable Energy Laboratory), J. Lee, E. S. Lee (Korea Advanced Institute of Science and Technology), E. M. Miller, R. Ihly (National Renewable Energy Laboratory), D. Wesenberg (University of Denver), K. S. Mistry (National Renewable Energy Laboratory), S. L. Guillot (University of Wisconsin-Madison), B. L. Zink (University of Denver), Y. H. Kim (Korea Advanced Institute of Science and Technology), and J. L. Blackburn (National Renewable Energy Laboratory)
Nanostructured organic semiconductors (OSCs), including single-walled carbon nanotubes (SWCNTs), offer a number of intriguing technological characteristics for thermoelectric applications, such as earth-abundant raw materials, low-cost deposition, and flexible form factors.

We will initially present a series of experiments focused on understanding the thermoelectric performance of enriched semiconducting SWCNT networks dispersed in a wide bandgap semiconducting polymer matrix, followed by more recent work aimed at understanding the role that the semiconducting polymer plays in the observed transport properties. Rational choice of the starting SWCNT material and the semiconducting polymer allows us to sensitively tune the s-SWCNT diameter and band gap distributions within the composites.

Consistent with theoretical calculations that consider the density of electronic states in individual s-SWCNTs, we observe a distinct dependence of the thermopower and thermoelectric power factor on the bandgap (or diameter) of the carbon nanotubes. We have measured large thermopower values (as high as ~2,500 µV/K for s-SWCNT networks with very low electrical conductivity) and impressive thermoelectric power factor as large as ~350 µW/m·K2.

By varying the carrier density injected into the s-SWCNT networks by a stable charge-transfer dopant, we are able to probe the relationship between the electrical conductivity and Seebeck coefficient (thermopower) in the s-SWCNT networks as a function of the carrier density and position of the Fermi energy. For carbon nanotubes prepared by high-pressure carbon monoxide (HiPCO) conversion, as we tune the carrier density, we are able to maintain a thermopower above 100 µV/K over almost the entire range of hole densities, corresponding to conductivities up to 40,000 S/m, resulting in a thermoelectric power factor of >300 µW/m·K2. These studies suggest that the low dimensionality of the SWCNTs has a stronger impact on the electrical conductivity than the thermopower, implying that they are less strongly coupled in these systems than is observed for compound inorganic semiconductors.

By modifying an approach that allows us to strip the dispersing polymer from the s-SWCNTs we were able to demonstrate that the polymer appears to play no role in modifying the barriers to electrical transport present at tube-tube junctions, and simply controls the extent of nanotube bundling and thereby the surface area available to the charge-transfer dopant. This study also indicates that densification of the s-SWCNT network results in a two-fold enhancement of the thermoelectric power factor, suggesting that careful control of the amount and nature of the matrix material is required for high-performance s-SWCNT thermoelectric materials.

Finally, we present a data from sensitive transport measurement technique, based on a microfabricated silicon nitride thermal isolation platform, to probe transport in the s-SWCNT networks, showing that the thermoelectric figure of merit (zT) is positively correlated with the measurement temperature, increasing by a factor of ~2.5 from 300 K to 350 K.

These observations demonstrate the ability to exert exquisite control of the thermoelectric performance by controlling the composition of the s-SWCNT network and tuning the carrier density (i.e., Fermi energy), and touts SWCNTs as an avenue for realizing thermally stable room temperature thermoelectric devices fashioned from inexpensive and abundant organic constituents.