0
Research Papers: Conduction

Temperature-Dependent Thermal Boundary Conductance at Metal/Indium-Based III–V Semiconductor Interfaces

[+] Author and Article Information
LeighAnn S. Larkin

Department of Mechanical and
Aerospace Engineering,
University of Virginia,
Charlottesville, VA 22904-4746
e-mail: LSL9HD@virginia.edu

MacKenzie R. Redding, Nam Q. Le

Department of Mechanical and
Aerospace Engineering,
University of Virginia,
Charlottesville, VA 22904-4746

Pamela M. Norris

Fellow ASME
Department of Mechanical and
Aerospace Engineering,
University of Virginia,
Charlottesville, VA 22904-4746

Contributed by the Heat Transfer Division of ASME for publication in the JOURNAL OF HEAT TRANSFER. Manuscript received June 9, 2016; final manuscript received October 5, 2016; published online November 22, 2016. Assoc. Editor: Alan McGaughey.

J. Heat Transfer 139(3), 031301 (Nov 22, 2016) (5 pages) Paper No: HT-16-1369; doi: 10.1115/1.4034938 History: Received June 09, 2016; Revised October 05, 2016

At the microscale length and smaller, solid–solid interfaces pose a significant contribution to resistance, resulting in a build-up of energy carriers, in turn leading to extreme temperature gradients within a single electronic component. These localized temperature gradients, or “hot spots,” are known to promote degradation, thus reducing device longevity and performance. To mitigate thermal management issues, it is crucial to both measure and understand conductance at interfaces in technologically relevant thin film systems. Recent trends in photonic devices have been pushing the consumption of indium in the U.S. to grow exponentially each year. Thus, we report on the temperature-dependent thermal boundary conductances at a series of metal/In-based III–V semiconductor interfaces. These measurements were made using time-domain thermoreflectance (TDTR) from 80 to 350 K. The high-temperature thermal boundary conductance results indicate, for these interfaces, that interfacial transport is dominated by elastic transmission, despite varying levels of acoustic mismatch. There is a strong direct correlation between the interfacial bond strength, approximated by the picosecond acoustics, and the thermal boundary conductance values. Both the interfacial bond strength and the overlap in the phonon density of states (PDOS) play significant roles in the magnitude of the thermal boundary conductance values. Measurements are compared against two separate predictive models, one for a perfect interface and one which accounts for disorder, such as interfacial mixing and finite grain sizes.

FIGURES IN THIS ARTICLE
<>
Copyright © 2017 by ASME
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Fig. 1

TDTR experimental results of thermal boundary conductance as a function of temperature for the various interfaces listed in Table 1. Data points are the average fit of ten separate scans collected at random locations across the sample surface, and error bars are the normal standard deviation. Solid lines represent diffuse mismatch model (DMM) predictions using isotropic (111) phonon dispersion for the metal films and isotropic (100) dispersion for semiconductor substrates [3437]. Dashed lines represent hBD(T) predictions using the δ-DMM, a modified DMM that explicitly accounts for disorder at an interface [38,39].

Grahic Jump Location
Fig. 2

A plot of the measured thermal boundary conductance at 300 K (open circles) and 80 K (filled diamonds) as a function of α, given by Eq. (1). When considering all the interfaces measured, the trend is unclear and inconclusive. The Al-containing interfaces are highlighted with a dotted line.

Grahic Jump Location
Fig. 3

Thermal boundary conductance, hBD at 300 K, as a function of damping coefficient, Γ, of the picosecond ultrasonics, where Pd-interfaces are shown with a pentagon and Al-interfaces with a diamond. The value of Γ was determined by fitting the first 50 ps of each of the ten TDTR scans to a damped simple harmonic oscillator function. Error bars in Γ therefore represent the normal standard deviation between the ten individual fits. Error bars for hBD values include both the normal standard deviation among ten individual scans, as well as the effects of a 2 nm uncertainty in the thickness of the metal film. The inset graph shows the picosecond ultrasonics with the general TDTR temperature decay subtracted out from the signal. Symbols are reused from Fig. 1.

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In