Seminar & Colloquium
[세미나: 11월 17일(금), 오전 10시 30분] Prof. QIYE ZHENG, Hong Kong University of Science and Technology
Title
Advancing Thermal Metrologies and Materials for Thermal Management in Microelectronics and Energy Technologies
Speaker
Prof. QIYE ZHENG, Hong Kong University of Science and Technology
Education
- 2012.08 - 2017.12 PhD, University of Illinois at Urbana-Champaign in Materials Science and Engineering
- 2008.09 - 2012.07 BS, Peking University in Physics
Professional Experience
- 2022.08 - present Assistant Professor at Mechanical and Aerospace Engineering, Hong Kong University of Science and Technology
- 2019.01 - 2022.07 Postdoc in Mechanical Engineering Lawrence, Berkeley National Lab & University of California
- 2018.01 - 2018.12 Postdoc in Materials Science and Engineering, University of Illinois at Urbana-Champaign
- 2015.01 - 2015.05 Co-op in Materials Research Division, Toyota Technical Center, Michigan
| Date | Friday, November 17th, 2023
| Time | 10:30 ~
| Venue | https://snu-ac-kr.zoom.us/j/91867259933?pwd=czgzdU1ZQkdyOWFQZU9mRERhQmp3UT09
ID: 918 6725 9933 PW: 1010
[Abstract]
Thermal management challenges in microelectronics and energy technologies necessitate advanced thermal metrologies and materials. On the one hand, as feature sizes shrink toward the nanoscale in current and emerging semiconductor devices, thermal bottlenecks have emerged as performance limitations due to the high, non-uniform power densities over the past decades. Nanoscale thermometry for temperature mapping is crucial to improve chip-level thermal design and study non-diffusive transport phenomena and device compatible heat spreading materials with ultra-high thermal conductivity (k) are desired for efficient chip-level heat dissipation. For diverse energy technologies, on the other hand, rapid screening of material candidates demands high-throughput characterization.
Firstly, I will present our development of a novel electron-beam (e-beam) Reflection Electron Elastic Energy Thermometry (REEET) to address the need for nanoscale thermometry. REEET leverages the Doppler broadening of quasi-elastically scattered electrons in reflection electron energy loss spectroscopy to quantify the mean atomic kinetic energy and thus the local phonon temperature within the 5 nm e-beam probe area. Our calibration experiments compared with first-principles modeling confirmed the physics of REEET, showing a relative temperature sensitivity of ~0.09%/K for light elements like carbon and silicon between 300-600 K. We used REEET for far-field temperature mapping of 300-400 nm nanostructured graphitic thin films and a 250-nm diameter multiwalled carbon nanotube under Joule heating. Local hotspots were visualized in-situ with 30-50 nm resolution using a machine learning denoiser. Suppressed in-plane/axial k’s were observed via numerical modeling, ascribed to size effects and interface scattering. REEET overcomes the optical diffraction-limits and scanning probe complexity to enable flexible characterization of nanoscale temperature fields without the sample thinning requirement in TEM.
Secondly, if time allows, I will briefly share our work on experimental discovering ultrahigh thermal conductivity in cubic boron pnictides with isotope engineering towards unconventional ultrahigh k using time domain thermoreflectance (TDTR). We measured an k≈1600, 540, and 1200 W/m-K in isotopically enriched BN, BP and BAs near room temperature which rivals the performance of diamond heat spreaders. The anomalous trend agrees with first-principles predictions and elucidates the roles of high-order phonon interactions.
Finally, I will introduce our development of high-throughput photothermal metrology for materials screening leveraging structured illumination with thermal imaging (SI-TI). Traditional contact thermal measurements have challenges with throughput and the lack of spatially resolvable property mapping, while laser methods require serial raster scanning to achieve property mapping and frequently need mirror smooth sample surfaces. In comparison, SI-TI method is demonstrated to (1) enable paralleled measurement of multiple regions and samples without raster scanning; (2) allow dynamic adjustment of the heating pattern in software to optimize the measurement sensitivity for anisotropic materials; and (3) tolerate rough (~3 μm) and scratched sample surfaces, via measurements of dense and porous materials such as mica and 3D printed thermoelectric thick films. This highlights SI-TI as a new avenue in adaptivity and throughput for thermal characterization of diverse materials.
| Host | 장혜진 교수(02-880-7096)