Scott T. Huxtable

Thermal conductivity of Si/SiGe and SiGe/SiGe superlattices

The cross-plane thermal conductivity of four Si/Si0.7Ge0.3 superlattices and three Si0.84Ge0.16/Si0.76Ge0.24 superlattices, with periods ranging from 45 to 300 and from 100 to 200 A, respectively, were measured over a temperature range of 50 to 320 K. For the Si/Si0.7Ge0.3 superlattices, the thermal conductivity was found to decrease with a decrease in period thickness and, at a period thickness of 45 A, it approached the alloy limit. For the Si0.84Ge0.16/Si0.76Ge0.24 samples, no dependence on period thickness was found and all the data collapsed to the alloy value, indicating the dominance of alloy scattering. This difference in thermal conductivity behavior between the two superlattices was attributed to interfacial acoustic impedance mismatch, which is much larger for Si/Si0.7Ge0.3 than for Si0.84Ge0.16/Si0.76Ge0.24. The thermal conductivity increased slightly up to about 200 K, but was relatively independent of temperature from 200 to 320 K.

SiGeC/Si superlattice microcoolers

Monolithically integrated active cooling is an attractive way for thermal management and temperature stabilization of microelectronic and optoelectronic devices. SiGeC can be lattice matched to Si and is a promising material for integrated coolers. SiGeC/Si superlattice structures were grown on Si substrates by molecular beam epitaxy. Thermal conductivity was measured by the 3omega method. SiGeC/Si superlattice microcoolers with dimensions as small as 40×40 µm^2 were fabricated and characterized. Cooling by as much as 2.8 and 6.9 K was measured at 25 °C and 100 °C, respectively, corresponding to maximum spot cooling power densities on the order of 1000 W/cm^2.

The influence of interface bonding on thermal transport through solid–liquid interfaces

We use time-domain thermoreflectance to show that interface thermal conductance, G, is proportional to the thermodynamic work of adhesion between gold and water, WSL, for a series of five alkane-thiol monolayers at the gold-water interface. WSL is a measure of the bond strength across the solid-liquid interface. Differences in bond strength, and thus differences in WSL, are achieved by varying the terminal group (ω-group) of the alkane-thiol monolayers on the gold. The interface thermal conductance values were in the range 60–190 MW m−2 K−1, and the solid-liquid contact angles span from 25° to 118°.

Thermal contact conductance of adhered microcantilevers

The thermal contact conductance G for polycrystalline silicon cantilever beams that are adhered to an underlying substrate is examined using two different optical techniques. Using time-domain thermoreflectance, we measure G=9±2 MW m−2 K−1 at 25 °C and G=4±1 MW m−2 K−1 at 150 °C. The room temperature value is confirmed using a modified Angstrom method, which establishes a lower limit of G>5 MW m−2 K−1. This contact conductance is a factor of 10–105 greater than values reported for metal–metal and ceramic–ceramic interfaces. The large interfacial conductance is consistent with the presence of a thin layer of water trapped between the cantilever and the substrate. The thermal conductivity Λ of the phosphorus doped polysilicon cantilever is nearly isotropic with Λcross plane=65 W m−1 K−1, and Λin plane=70 W m−1 K−1 at room temperature.The thermal contact conductance G for polycrystalline silicon cantilever beams that are adhered to an underlying substrate is examined using two different optical techniques. Using time-domain thermoreflectance, we measure G=9±2 MW m−2 K−1 at 25 °C and G=4±1 MW m−2 K−1 at 150 °C. The room temperature value is confirmed using a modified Angstrom method, which establishes a lower limit of G>5 MW m−2 K−1. This contact conductance is a factor of 10–105 greater than values reported for metal–metal and ceramic–ceramic interfaces. The large interfacial conductance is consistent with the presence of a thin layer of water trapped between the cantilever and the substrate. The thermal conductivity Λ of the phosphorus doped polysilicon cantilever is nearly isotropic with Λcross plane=65 W m−1 K−1, and Λin plane=70 W m−1 K−1 at room temperature.

THERMAL CONDUCTIVITY OF INDIUM PHOSPHIDE-BASED SUPERLATTICES

Semiconductor superlattice structures have shown promise as thermoelectric materials for their high power factor and low thermal conductivity. While the power factor of a superlattice can be controlled through band gap engineering and doping, prediction and control of thermal conductivity has remained a challenge. The thermal conductivity of three different InP/InGaAs superlattices was measured to be between 4 and 9 W/m-K from 77-320 K using the 3! method. Although the thermal conductivity of InP is an order of magnitude higher than that of InGaAs, we report the intriguing observation that as the fraction of InP is increased in InP/InGaAs superlattices, the thermal conductivity decreases. For one superlattice, the thermal conductivity was even below that of InGaAs. These observations are contrary to predictions of effective thermal conductivity by the Fourier law.Semiconductor superlattice structures have shown promise as thermoelectric materials for their high power factor and low thermal conductivity. While the power factor of a superlattice can be controlled through band gap engineering and doping, prediction and control of thermal conductivity has remained a challenge. The thermal conductivity of three different InP/InGaAs superlattices was measured to be between 4 and 9 W/m-K from 77-320 K using the 3! method. Although the thermal conductivity of InP is an order of magnitude higher than that of InGaAs, we report the intriguing observation that as the fraction of InP is increased in InP/InGaAs superlattices, the thermal conductivity decreases. For one superlattice, the thermal conductivity was even below that of InGaAs. These observations are contrary to predictions of effective thermal conductivity by the Fourier law.

Nanowire arrays for thermoelectric devices

This study reports on the fabrication and characterization of two prototype thermoelectric devices constructed of either silicon (Si) or bismuth telluride (Bi2 Te3 ) nanowire arrays. The growth mechanisms and fabrication procedures of the Si and Bi2 Te3 devices are different as described in this paper. To characterize the thermoelectric device components, current-voltage (I-V) characteristics were first used to estimate their performance. For the Si device, the I-V characteristics suggest ohmic contacts at the metal-semiconductor junction. For the Bi2 Te3 device, the I-V characteristics curve showed a rectifying contact. Either low doping of the Bi2Te3 or surface contamination, i.e. native oxide, may cause the rectifying contact. The reversible Peltier effects occurring within the Si device were analyzed using a micro-thermocouple. Results indicated possible limitations of using Si nanowire arrays for the thermoelectric device.Copyright

Nanowire composite thermoelectric devices

This paper discusses the design, fabrication and testing of a novel thermoelectric device comprised of arrays of silicon nanowires embedded in a polymer matrix. By exploiting the low thermal conductivity of the composite and presumably high power factor of the nanowires, a high figure of merit should result. Arrays were first synthesized using a vapor-liquid-solid (VLS) process leading to one-dimensional growth of single-crystalline nanowires. To provide both structural support and thermal isolation between nanowires, parylene, a low thermal conductivity and extremely conformal polymer, was embedded within the arrays. Mechanical polishing and oxygen plasma etching techniques were used to expose the nanowire tips and a metal contact was deposited on the top surface. Scanning electron microscopy pictures illustrate the results of the fabrication processes. Using the 3ω technique, the effective thermal conductivity of the nanowire matrix was measured.Copyright

THE EFFECT OF DEFECTS AND ACOUSTIC IMPEDANCE MISMATCH ON HEAT CONDUCTION SIGE BASED SUPERLATTICES

The cross-plane thermal conductivity of four Si/Ge , Si/Si0.4Ge 0.6, and Si0.9Ge 0.1/Si0.1Ge 0.9 superlattices was measured using the 3ω technique. All four superlattices were found to have thermal conductivity values between 1.8 and 3.5 W/m-K, which are below the values of typical SixGe 1-x alloys. The growth quality of these superlattices was evaluated qualitatively through the use of x-ray diffraction and transmission electron microscopy. These studies indicated that the superlattices contained a relatively high density of defects. The low thermal conductivity values are presumed to be due in large part to these defects.