Thermal conductivity measurements of Summit polycrystalline silicon.

Abstract

A capability for measuring the thermal conductivity of microelectromechanical systems (MEMS) materials using a steady state resistance technique was developed and used to measure the thermal conductivities of SUMMiT TM V layers. Thermal conductivities were measured over two temperature ranges: 100K to 350K and 293K to 575K in order to generate two data sets. The steady state resistance technique uses surface micromachined bridge structures fabricated using the standard SUMMiT fabrication process. Electrical resistance and resistivity data are reported for poly1-poly2 laminate, poly2, poly3, and poly4 polysilicon structural layers in the SUMMiT process from 83K to 575K. Thermal conductivity measurements for these polysilicon layers demonstrate for the first time that the thermal conductivity is a function of the particular SUMMiT layer. Also, the poly2 layer has a different variation in thermal conductivity as the temperature is decreased than the poly1-poly2 laminate, poly3, and poly4 layers. As the temperature increases above room temperature, the difference in thermal conductivity between the layers decreases. 4 ACKNOWLEDGMENTSThe authors thank Rosemarie Renn and Katie Francis for assistance during the design of the thermal conductivity test structures and the Microelectronics Development Lab staff for their efforts in fabricating the structures. We are also grateful for the assistance of Ted Parson in implementing the data acquisition system. We thank Blake Jakabowski and Ray Haltli for wire bonding and packaging samples for testing. We appreciate the efforts of Allen Gorby in providing general support for this project and especially for the work on the Janis cryostat, modifying it to be compatible with the samples and changing the modules between tests. We also acknowledge the contribution of Ed Piekos for his simulations of the test structures. Constructive peer reviews of this report were provided by

In-Situ Temperature Measurement on CMOS Integrated Micro-Hotplates for Gas Sensing Devices

“…The temperature rise on the surface was calculated using the steady state heat equation [27]:T(x,y)=true∬ A 8Pπd2e−(2 × 2d)2(x2+y2)12πκ(x2+y2)dx dy. where A is the absorbance of the laser, P the laser power and d the diameter of the laser spot. The material parameters of the polysilicon, like the thermal conductivity κ, but also A , are temperature-dependent [28], so the temperature increase was calculated for all the different temperatures set by the Pt-heater during calibration. In Figure 3, the Gaussian weighted average of the temperature rise can be seen as a function of each set temperature.…”