IBM T.J. Watson Research Center 2D material based layer transfer Jeehwan Kim Prof. of Mechanical Engineering Prof. of Materials science and Engineering Jeehwan Kim Research Group http://jeehwanlab.mit.edu
Major bottleneck for advancing semiconductor technology Substrate: Essential building block to form Electronic/optoelectronic devices Epitaxial growth: Process for forming device film structures on the substrate FETs, LEDs, Lasers, Detectors Epitaxy of single-crystalline films is required on given available substrates Epitaxial films Substrate SiC, Sapphire Si Ge, GaAs InP Price: SiC > InP > GaAs > Ge >>> Price: Si Limited application Lattice: InP > GaAs/Ge > Si > SiC Defect generation http://www.tf.uni-kiel.de/matwis/amat/semi_en/kap_5/backbone/r5_1_4.html
2D material based layer transfer (2DLT) sp 2 -bonded graphene: No broken bonds on the surface Precise release from graphene Post-release treatment NOT required 1 sec release due to weak interaction Universal for any materials
Chemical lift-off (Epitaxial lift-off, ELO) Epitaxial Growth Lateral chemical etching of buffer Layer transfer Wafer Epilayer Buffer Wafer Buffer Wafer Epilayer Host sub Substrate reuse Pro: Control of release interface Cons: Post-treatment required Slow release Limited application mainly for GaAs & InP Current U.S. $/ W p(dc) $18.0 $16.0 $14.0 $12.0 $10.0 $8.0 $6.0 $4.0 $2.0 $0.0 For PV applications The Costs for the Epi Substrate as a Func on of Reuse Number Reference Case Repolishing Cost ($8 per repolish per 133 cm 2 wafer) and Cell Efficiency (25%) NREL Cost Analysis 9/13/2013 Reference Case Costs vs Reuses Repolishing Cost (Assuming a Repolish is Needed for Each Growth Cycle) 5 25 45 65 85 105 125 145 165 185 205 Number of Reuses http://www.nrel.gov/docs/fy14osti/60126.pdf E. Yablonovitch et al., Appl. Phys. Lett. 51, 2222 (1987). B. M. Kayes et al, IEEE J. Photovolt. 4, 729 (2014).
Graphene-based layer transfer (GBLT) sp 2 -bonded graphene: No broken bonds on the surface Precise release from graphene Post-release treatment NOT required 1 sec release due to weak interaction Universal for any materials Meet all requirements
Monolayer graphene transparency for epitaxial growth - Remote homoepitaxy Y. Kim, S. Cruz, J. Kim et al., Nature, COVER (2017)
DFT calculation HRTEM Dark field XTEM: Strain field Critical interaction gap: 1 nm In collaboration with Prof. Kolpak No sign of dislocation Remote homoepitaxy is possible through graphene Y. Kim, S. Cruz, J. Kim et al., Nature (2017)
Unversality of 2DLT (MIT) Growth of single-crystalline GaN, GaAs, InP, GaP, Ge on graphene EBSD mapping 111 001 101
Role of graphene Release layer for any semiconductors Dislocation-reducer/filter Release layer 1sec release Wafer Surface protection infinite reuse Wide application of non-si electronics/photonics Enabled heterointegration InP InGaAs GaAs GaN Si wafer
Implication for PV technology GaAs substrate
Implication for space PV technology III-V multijunction solar cells for E g oriented design Material 4 Eg~1.8eV Material 3 Eg~1.4eV Material 2 Eg~1eV Material 1 Eg~0.6eV Graphene as dislocation reducer/tunnel junction Examples GaInP(1.87eV) GaAs(1.42eV) GaInNAs(1eV) Ge(0.67eV) Si (1.1 ev) III-V substrate III-V substrate
Implication for magnetoelectric coupling Graphene transfer Epitaxy of Piezoelectric materials Exfoliation of Piezoelectric materials Piezo Bonding of freestanding Piezomaterials + magnetostrictive matreials SrTiO 3 SrTiO 3 Graphene transfer SrTiO 3 Remote Exfoliation Epitaxy of Maganetostrictive materials Magneto Piezo PDMS No substrate clamping to Maximize effect MgAl 2 O 4 MgAl 2 O 4 MgAl 2 O 4
Implication for display/lighting technology Low-cost flexible LEDs (solid state lighting/microled) Red LED (III-V) Blue LED (III-N) Dislocation-free LED on graphene Light emission from LEDs on graphene Dislocation-free GaN obtainable by GaN growth on graphene/gan High efficiency lighting High pixel density microled Y. Kim, S. Cruz, J. Kim et al., Nature cover (2017)
Implication for power electronics/heterointegration Power electronics