Numerical and model predictions of the thermal conductivity of bismuth telluride nanoprism-assembled films

Yi Cheng Hou, Mei Jiau Huang*, Pi Yueh Chuang, Hsiu Cheng Chang, Chun-Hua Chen

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

2 Scopus citations


This research aims at understanding the heat transfer phenomenon in the Bi2Te3 nanoprism-assembled films. For obtaining the associated effective thermal conductivity, the effective-medium-approximation (EMA) models existing in literature were examined and properly modified for prediction and Monte-Carlo numerical experiments based on unstructured grids were performed for confirmation. Cross-plane and in-plane thermal conductivities were both explored. For model predictions, the characteristic grain size of the nanoprisms is defined as either the phonon mean free path purely due to the grain boundary scattering (i.e. excluding the intrinsic scattering) or the averaged hydraulic diameter. A combination of a non-dilute 2D EMA model with the triple bond percolation theory turns out to be the best model for predicting the in-plane thermal conductivity. On the other hand, the evaluation of the cross-plane thermal conductivity by treating the nanoprisms as thermal conductors connected in parallel is satisfactory. The investigation shows that in addition to porosity, the scattering at the grain boundaries plays a dominant role in reducing the heat transfer in the direction perpendicular to the boundaries; the in-plane thermal conductivity is therefore much smaller than the cross-plane one.

Original languageEnglish
Pages (from-to)536-543
Number of pages8
JournalInternational Journal of Heat and Mass Transfer
StatePublished - 1 Sep 2015


  • Bismuth telluride nanoprism-assembled films
  • Effective medium approximation
  • Lattice thermal conductivity
  • Monte-Carlo simulation


Dive into the research topics of 'Numerical and model predictions of the thermal conductivity of bismuth telluride nanoprism-assembled films'. Together they form a unique fingerprint.

Cite this