Direct thermal to electrical energy conversion

Direct Thermal to Electric Energy Conversion

DARPA Project:
Metal-Semiconductor Nanocomposites for High Efficiency Thermoelectric Power Generation

UC Santa Cruz, UC Santa Barbara, UC Berkeley, MIT, Purdue University, University of Delaware, BSST.

Program Manager
Vye Greanya, DARPA

For a full list of the team members and recent publications related to the project, please see below.


Direct thermal to electrical energy conversion systems that could operate in the 100-700°C temperature range with high efficiencies (>30%) provide an attractive compact alternative to internal combustion engines for many military applications. They will also expand the possibilities for waste heat recovery applications. The core of the solution we are proposing is a metal/semiconductor nanocomposite that will allow us to modify four intrinsic material properties in order to fabricate more efficient thermoelectric systems. The concept of a metal/semiconductor nanocomposite as a solid-state thermionic material represents a radical alternative to conventional homogeneous thermoelectric materials. Instead of focusing on semiconducting materials with a highly asymmetric density of states about the Fermi level yielding an optimal doping level of about 10^19/cm^3 at room temperature, the metal/semiconductor nanocomposite concept utilizes the Schottky barrier to filter the electron energy distribution, creating a large difference between the average energy of the conduction electron and the Fermi energy. The high electron density in the metal (>10^22/cm^3) compensates for the negligible contribution to the conductivity from the majority of the electrons in the metal that have energies below the top of the barrier. The high interface density and/or nanoscale embedded nanoparticles in a metal/semiconductor nanocomposite are expected to suppress the transport of mid-long wavelength phonons. This is very important to reduce the lattice contribution to thermal conductivity below the alloy limit. The ability to tune the properties by controlling layer thicknesses and nanoparticles sizes as well as manipulating lattice mismatch and barrier height by alloying offers additional degrees of freedom for materials design. Modeling of the transport properties of metal/semiconductor superlattices suggested that ZT values in excess of 4-5 should be possible if the barrier height is adjusted to be in the range of 4-5 kT.

A unique team of researchers experienced in nano-engineered semiconductor materials, physics, electrical and mechanical engineering has been assembled to address fundamental limits to existing materials. This work is closely coordinated with our industrial partner, BSST, who will investigate reliability, manufacturing and scale up production for military systems. BSST is the world’s leading consumer of TE material and is already working with the Department of Defense on the development of several state-of-the-art solid-state thermoelectric systems.

Team Members

  • Ali Shakouri, Nobby Kobayashi and Zhixi Bian (EE, UC Santa Cruz)
  • John Bowers and Art Gossard (ECE & Materials, UC Santa Barbara)
  • Arun Majumdar (ME, UC Berkeley)
  • Rajeev Ram (EE, MIT)
  • Tim Sands (Materials & ECE, Purdue University)
  • Joshua Zide (ECE, University of Delaware)
  • Lon Bel (BSST)

Recent Papers

Subscribe to Quantum Electronics Group News