Zhe Shi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Zhe Shi has a total of 96 co-authors. Here are the top co-authors by number of shared publications:
- Ju Li
Massachusetts Institute of Technology, USA
9 shared publications Evgenii Tsymbalov
Independent Researcher, Aachen 52070, Germany
4 shared publications- Ming Dao
Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, USA
4 shared publications - Subra Suresh
Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139
4 shared publications - Chandra Veer Singh
Department of Materials Science and Engineering, University of Toronto, 184 College Street, Toronto, Ontario, M5S 3E4, Canada
3 shared publications - Alexander V. Shapeev
Skolkovo Institute of Science and Technology
2 shared publications
Latest Publications
- Open Access
- 8 October 2025
- Published by American Chemical Society (ACS) in ChemRxiv
Chemical Vapor Deposition of Uniform and Large-Scale MoS2 Using Heterogeneous Precursors
The scalable production of monolayer molybdenum disulfide (MoS2) is essential for the development of next-generation electronic devices. Current chemical vapor deposition (CVD) strategies struggle to achieve uniform and continuous monolayer MoS2 over large areas without using expensive and highly hazardous precursors. Here, we developed a CVD strategy employing heterogeneous precursors for synthesizing high-quality centimeter-scale monolayer MoS2. Different from conventional CVD strategies where growth promoters and Mo sources are mixed, our strategy employs heterogeneous precursors where the two components are introduced separately, in two layers. Using this technique, the coverage ratio of monolayer MoS2 and device yield increased significantly using a cost-effective tube furnace with powder precursors. The growth mechanism was investigated by density functional theory (DFT) calculations revealing that the pre-deposition of growth promoters significantly lowers the formation energy for in-plane MoS2 growth. As a result, after the growth of the first layer, the formation of multilayer MoS2 is suppressed causing all monolayer flakes to merge together to form a continuous single layer, as cross-verified by second harmonic generation (SHG) imaging and molecular dynamics (MD) simulations. Our method of using heterogeneous precursors enables continuous CVD growth of MoS2 through a cost-effective and operationally straightforward process, offering a promising paradigm for high-quality MoS2 fabrication towards large-scale 2D semiconductor applications.
- 8 October 2025
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessChemical Vapor Deposition of Uniform and Large-Scale MoS2 Using Heterogeneous Precursors
The scalable production of monolayer molybdenum disulfide (MoS2) is essential for the development of next-generation electronic devices. Current chemical vapor deposition (CVD) strategies struggle to achieve uniform and continuous monolayer MoS2 over large areas without using expensive and highly hazardous precursors. Here, we developed a CVD strategy employing heterogeneous precursors for synthesizing high-quality centimeter-scale monolayer MoS2. Different from conventional CVD strategies where growth promoters and Mo sources are mixed, our strategy employs heterogeneous precursors where the two components are introduced separately, in two layers. Using this technique, the coverage ratio of monolayer MoS2 and device yield increased significantly using a cost-effective tube furnace with powder precursors. The growth mechanism was investigated by density functional theory (DFT) calculations revealing that the pre-deposition of growth promoters significantly lowers the formation energy for in-plane MoS2 growth. As a result, after the growth of the first layer, the formation of multilayer MoS2 is suppressed causing all monolayer flakes to merge together to form a continuous single layer, as cross-verified by second harmonic generation (SHG) imaging and molecular dynamics (MD) simulations. Our method of using heterogeneous precursors enables continuous CVD growth of MoS2 through a cost-effective and operationally straightforward process, offering a promising paradigm for high-quality MoS2 fabrication towards large-scale 2D semiconductor applications.
- Open Access
- 14 February 2024
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 121 (8)
Phonon stability boundary and deep elastic strain engineering of lattice thermal conductivity
Recent studies have reported the experimental discovery that nanoscale specimens of even a natural material, such as diamond, can be deformed elastically to as much as 10% tensile elastic strain at room temperature without the onset of permanent damage or fracture. Computational work combining ab initio calculations and machine learning (ML) algorithms has further demonstrated that the bandgap of diamond can be altered significantly purely by reversible elastic straining. These findings open up unprecedented possibilities for designing materials and devices with extreme physical properties and performance characteristics for a variety of technological applications. However, a general scientific framework to guide the design of engineering materials through such elastic strain engineering (ESE) has not yet been developed. By combining first-principles calculations with ML, we present here a general approach to map out the entire phonon stability boundary in six-dimensional strain space, which can guide the ESE of a material without phase transitions. We focus on ESE of vibrational properties, including harmonic phonon dispersions, nonlinear phonon scattering, and thermal conductivity. While the framework presented here can be applied to any material, we show as an example demonstration that the room-temperature lattice thermal conductivity of diamond can be increased by more than 100% or reduced by more than 95% purely by ESE, without triggering phonon instabilities. Such a framework opens the door for tailoring of thermal-barrier, thermoelectric, and electro-optical properties of materials and devices through the purposeful design of homogeneous or inhomogeneous strains.
- 14 February 2024
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 121 (8)
Open AccessPhonon stability boundary and deep elastic strain engineering of lattice thermal conductivity
Recent studies have reported the experimental discovery that nanoscale specimens of even a natural material, such as diamond, can be deformed elastically to as much as 10% tensile elastic strain at room temperature without the onset of permanent damage or fracture. Computational work combining ab initio calculations and machine learning (ML) algorithms has further demonstrated that the bandgap of diamond can be altered significantly purely by reversible elastic straining. These findings open up unprecedented possibilities for designing materials and devices with extreme physical properties and performance characteristics for a variety of technological applications. However, a general scientific framework to guide the design of engineering materials through such elastic strain engineering (ESE) has not yet been developed. By combining first-principles calculations with ML, we present here a general approach to map out the entire phonon stability boundary in six-dimensional strain space, which can guide the ESE of a material without phase transitions. We focus on ESE of vibrational properties, including harmonic phonon dispersions, nonlinear phonon scattering, and thermal conductivity. While the framework presented here can be applied to any material, we show as an example demonstration that the room-temperature lattice thermal conductivity of diamond can be increased by more than 100% or reduced by more than 95% purely by ESE, without triggering phonon instabilities. Such a framework opens the door for tailoring of thermal-barrier, thermoelectric, and electro-optical properties of materials and devices through the purposeful design of homogeneous or inhomogeneous strains.
- 5 February 2024
- Published by American Physical Society (APS) in Physical Review Materials
- Vol. 8 (2) , 025201
Thermalization of electron-hole pairs in and CLLB: Monte Carlo simulationA Monte Carlo model is used to investigate electron-hole (e-h) generation created by incident gamma-ray radiation in , and (CLLB) scintillators. Our approach follows the detailed energy loss mechanisms and describes the microscopic structure of ionization tracks in order to address differences in the scintillation properties of these three materials. The mean energy required to create an e-h pair, , theoretical light yield, and the spatial distribution of e-h pairs are determined. We found that approaches constant and similar values at high incident energies for and CLLB, suggesting that these materials should have similar light yields. However, the experimental light yield of CLLB is almost half that of . Unlike for of increases with increasing energy excitation and shows a nonlinear behavior in e-h creation, potentially explaining slightly lower light yield and worse energy resolution of . Furthermore, we observed that the spatial distributions of electron-hole pairs in and are very similar, while the number of high-density e-h domains in CLLB is greater in comparison. This discrepancy could explain the lower light yield of CLLB. The thermalization model of e-h pairs showed that the longitudinal optical phonon energy has a profound effect on the thermalization time and distance of e-h pairs, leading to a much higher density of excitation in but a more diffuse one in CLLB. This effect leads to steeper gradients in , resulting in varying density effects and worse proportionality, while CLLB suffers from more uniform but more pronounced quenching. The fraction of nonradiatively recombined electrons in and CLLB was estimated to be 30% and 45%, respectively. These results correlate well with experimental observations of the scintillation properties of these materials. The approach can be used to predict the expected properties of new materials and support further development of existing materials.- 5 February 2024
- Published by American Physical Society (APS) in Physical Review Materials
- Vol. 8 (2) , 025201
Thermalization of electron-hole pairs in and CLLB: Monte Carlo simulationA Monte Carlo model is used to investigate electron-hole (e-h) generation created by incident gamma-ray radiation in , and (CLLB) scintillators. Our approach follows the detailed energy loss mechanisms and describes the microscopic structure of ionization tracks in order to address differences in the scintillation properties of these three materials. The mean energy required to create an e-h pair, , theoretical light yield, and the spatial distribution of e-h pairs are determined. We found that approaches constant and similar values at high incident energies for and CLLB, suggesting that these materials should have similar light yields. However, the experimental light yield of CLLB is almost half that of . Unlike for of increases with increasing energy excitation and shows a nonlinear behavior in e-h creation, potentially explaining slightly lower light yield and worse energy resolution of . Furthermore, we observed that the spatial distributions of electron-hole pairs in and are very similar, while the number of high-density e-h domains in CLLB is greater in comparison. This discrepancy could explain the lower light yield of CLLB. The thermalization model of e-h pairs showed that the longitudinal optical phonon energy has a profound effect on the thermalization time and distance of e-h pairs, leading to a much higher density of excitation in but a more diffuse one in CLLB. This effect leads to steeper gradients in , resulting in varying density effects and worse proportionality, while CLLB suffers from more uniform but more pronounced quenching. The fraction of nonradiatively recombined electrons in and CLLB was estimated to be 30% and 45%, respectively. These results correlate well with experimental observations of the scintillation properties of these materials. The approach can be used to predict the expected properties of new materials and support further development of existing materials.- 2 August 2021
- Published by Wiley in Advanced Materials
- Vol. 33 (38) , 2102034
Low‐Density Fluorinated Silane Solvent Enhancing Deep Cycle Lithium–Sulfur Batteries’ Lifetime
The lithium metal anode (LMA) instability at deep cycle with high utilization is a crucial barrier for developing lithium (Li) metal batteries, resulting in excessive Li inventory and electrolyte demand. This issue becomes more severe in capacity‐type lithium–sulfur (Li–S) batteries. High‐concentration or localized high‐concentration electrolytes are noted as effective strategies to stabilize Li metal but usually lead to a high electrolyte density (>1.4 g mL−1). Here we propose a bifunctional fluorinated silane‐based electrolyte with a low density of 1.0 g mL−1 that not only is much lighter than conventional electrolytes (≈1.2 g mL−1) but also form a robust solid electrolyte interface to minimize Li depletion. Therefore, the Li loss rate is reduced over 4.5‐fold with the proposed electrolyte relative to its conventional counterpart. When paired with onefold excess LMA at the electrolyte weight/cell capacity (E/C) ratio of 4.5 g Ah−1, the Li–S pouch cell using our electrolyte can survive for 103 cycles, much longer than with the conventional electrolyte (38 cycles). This demonstrates that our electrolyte not only reduces the E/C ratio but also enhances the cyclic stability of Li–S batteries under limited Li amounts.
- 2 August 2021
- Published by Wiley in Advanced Materials
- Vol. 33 (38) , 2102034
Low‐Density Fluorinated Silane Solvent Enhancing Deep Cycle Lithium–Sulfur Batteries’ Lifetime
The lithium metal anode (LMA) instability at deep cycle with high utilization is a crucial barrier for developing lithium (Li) metal batteries, resulting in excessive Li inventory and electrolyte demand. This issue becomes more severe in capacity‐type lithium–sulfur (Li–S) batteries. High‐concentration or localized high‐concentration electrolytes are noted as effective strategies to stabilize Li metal but usually lead to a high electrolyte density (>1.4 g mL−1). Here we propose a bifunctional fluorinated silane‐based electrolyte with a low density of 1.0 g mL−1 that not only is much lighter than conventional electrolytes (≈1.2 g mL−1) but also form a robust solid electrolyte interface to minimize Li depletion. Therefore, the Li loss rate is reduced over 4.5‐fold with the proposed electrolyte relative to its conventional counterpart. When paired with onefold excess LMA at the electrolyte weight/cell capacity (E/C) ratio of 4.5 g Ah−1, the Li–S pouch cell using our electrolyte can survive for 103 cycles, much longer than with the conventional electrolyte (38 cycles). This demonstrates that our electrolyte not only reduces the E/C ratio but also enhances the cyclic stability of Li–S batteries under limited Li amounts.
- Open Access
- 28 May 2021
- Published by Springer Nature in npj Computational Materials
- Vol. 7 (1) , 1-10
Machine learning for deep elastic strain engineering of semiconductor electronic band structure and effective mass
The controlled introduction of elastic strains is an appealing strategy for modulating the physical properties of semiconductor materials. With the recent discovery of large elastic deformation in nanoscale specimens as diverse as silicon and diamond, employing this strategy to improve device performance necessitates first-principles computations of the fundamental electronic band structure and target figures-of-merit, through the design of an optimal straining pathway. Such simulations, however, call for approaches that combine deep learning algorithms and physics of deformation with band structure calculations to custom-design electronic and optical properties. Motivated by this challenge, we present here details of a machine learning framework involving convolutional neural networks to represent the topology and curvature of band structures in k-space. These calculations enable us to identify ways in which the physical properties can be altered through “deep” elastic strain engineering up to a large fraction of the ideal strain. Algorithms capable of active learning and informed by the underlying physics were presented here for predicting the bandgap and the band structure. By training a surrogate model with ab initio computational data, our method can identify the most efficient strain energy pathway to realize physical property changes. The power of this method is further demonstrated with results from the prediction of strain states that influence the effective electron mass. We illustrate the applications of the method with specific results for diamonds, although the general deep learning technique presented here is potentially useful for optimizing the physical properties of a wide variety of semiconductor materials.
- 28 May 2021
- Published by Springer Nature in npj Computational Materials
- Vol. 7 (1) , 1-10
Open AccessMachine learning for deep elastic strain engineering of semiconductor electronic band structure and effective mass
The controlled introduction of elastic strains is an appealing strategy for modulating the physical properties of semiconductor materials. With the recent discovery of large elastic deformation in nanoscale specimens as diverse as silicon and diamond, employing this strategy to improve device performance necessitates first-principles computations of the fundamental electronic band structure and target figures-of-merit, through the design of an optimal straining pathway. Such simulations, however, call for approaches that combine deep learning algorithms and physics of deformation with band structure calculations to custom-design electronic and optical properties. Motivated by this challenge, we present here details of a machine learning framework involving convolutional neural networks to represent the topology and curvature of band structures in k-space. These calculations enable us to identify ways in which the physical properties can be altered through “deep” elastic strain engineering up to a large fraction of the ideal strain. Algorithms capable of active learning and informed by the underlying physics were presented here for predicting the bandgap and the band structure. By training a surrogate model with ab initio computational data, our method can identify the most efficient strain energy pathway to realize physical property changes. The power of this method is further demonstrated with results from the prediction of strain states that influence the effective electron mass. We illustrate the applications of the method with specific results for diamonds, although the general deep learning technique presented here is potentially useful for optimizing the physical properties of a wide variety of semiconductor materials.
- 17 May 2021
- Published by Wiley in Advanced Materials
- Vol. 33 (26) , 2008723
Dense All‐Electrochem‐Active Electrodes for All‐Solid‐State Lithium Batteries
The energy density presents the core competitiveness of lithium (Li)‐ion batteries. In conventional Li‐ion batteries, the utilization of the gravimetric/volumetric energy density at the electrode level is unsatisfactory (−1 and 1945 W h L−1 at the electrode level based on the total cathodes and anodes when at 70 °C.
- 17 May 2021
- Published by Wiley in Advanced Materials
- Vol. 33 (26) , 2008723
Dense All‐Electrochem‐Active Electrodes for All‐Solid‐State Lithium Batteries
The energy density presents the core competitiveness of lithium (Li)‐ion batteries. In conventional Li‐ion batteries, the utilization of the gravimetric/volumetric energy density at the electrode level is unsatisfactory (−1 and 1945 W h L−1 at the electrode level based on the total cathodes and anodes when at 70 °C.
- Top Cited Papers
- 25 March 2021
- Published by Springer Nature in Nature Energy
- Vol. 6 (5) , 495-505
Ultra-high-voltage Ni-rich layered cathodes in practical Li metal batteries enabled by a sulfonamide-based electrolyte
- 25 March 2021
- Published by Springer Nature in Nature Energy
- Vol. 6 (5) , 495-505
Top Cited PapersUltra-high-voltage Ni-rich layered cathodes in practical Li metal batteries enabled by a sulfonamide-based electrolyte
- Open Access
- 5 October 2020
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 117 (40) , 24634-24639
Metallization of diamond
Significance: Identifying the conditions for complete metallization of diamond solely through mechanical strain is an important scientific objective and technological demonstration. Through quantum mechanical calculations, continuum mechanics simulations validated by experiments, and machine learning, we show here that reversible metallization can be achieved in diamond deformed below threshold elastic strain levels for failure or phase transformation. The general method outlined here for deep elastic strain engineering is also applicable to map the strain conditions for indirect-to-direct bandgap transitions. Our method and findings enable extreme alterations of semiconductor properties via strain engineering for possible applications in power electronics, optoelectronics, and quantum sensing.
- 5 October 2020
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 117 (40) , 24634-24639
Open AccessMetallization of diamond
Significance: Identifying the conditions for complete metallization of diamond solely through mechanical strain is an important scientific objective and technological demonstration. Through quantum mechanical calculations, continuum mechanics simulations validated by experiments, and machine learning, we show here that reversible metallization can be achieved in diamond deformed below threshold elastic strain levels for failure or phase transformation. The general method outlined here for deep elastic strain engineering is also applicable to map the strain conditions for indirect-to-direct bandgap transitions. Our method and findings enable extreme alterations of semiconductor properties via strain engineering for possible applications in power electronics, optoelectronics, and quantum sensing.
- 8 October 2019
- Published by American Chemical Society (ACS) in Inorganic Chemistry
- Vol. 58 (20) , 13782-13787
Inserting Amide into NOTT-101 to Sharply Enhance Volumetric and Gravimetric Methane Storage Working Capacity
Amide-functionalized NJU-Bai45 exhibits high volumetric and gravimetric CH4 working capacities of 190 cm3 cm–3 and 298 cm3 g–1 (from 5.8 to 65 bar), higher than that of the similar NJU-Bai43 (187 cm3 cm–3 and 292 cm3 g–1). This finding indicates that nitrogen-functionalized aromatic ring may not always play a positive role in amide-MOF for CH4 storage.
- 8 October 2019
- Published by American Chemical Society (ACS) in Inorganic Chemistry
- Vol. 58 (20) , 13782-13787
Inserting Amide into NOTT-101 to Sharply Enhance Volumetric and Gravimetric Methane Storage Working Capacity
Amide-functionalized NJU-Bai45 exhibits high volumetric and gravimetric CH4 working capacities of 190 cm3 cm–3 and 298 cm3 g–1 (from 5.8 to 65 bar), higher than that of the similar NJU-Bai43 (187 cm3 cm–3 and 292 cm3 g–1). This finding indicates that nitrogen-functionalized aromatic ring may not always play a positive role in amide-MOF for CH4 storage.
- Open Access
- 15 February 2019
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 116 (10) , 4117-4122
Deep elastic strain engineering of bandgap through machine learning
Significance: Deforming a material to a large extent without inelastic relaxation can result in unprecedented properties. However, the optimal deformation state is buried within the vast continua of choices available in the strain space. Here we advance a unique and powerful strategy to circumvent conventional trial-and-error methods, and adopt artificial intelligence techniques for rationally designing the most energy-efficient pathway to achieve a desirable material property such as the electronic bandgap. The broad framework for tailoring any target figure of merit, for any material using machine learning, opens up opportunities to adapt elastic strain engineering of properties and performance in devices and systems in a controllable and efficient manner, for potential applications in microelectronics, optoelectronics, photonics, and energy technologies.
- 15 February 2019
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 116 (10) , 4117-4122
Open AccessDeep elastic strain engineering of bandgap through machine learning
Significance: Deforming a material to a large extent without inelastic relaxation can result in unprecedented properties. However, the optimal deformation state is buried within the vast continua of choices available in the strain space. Here we advance a unique and powerful strategy to circumvent conventional trial-and-error methods, and adopt artificial intelligence techniques for rationally designing the most energy-efficient pathway to achieve a desirable material property such as the electronic bandgap. The broad framework for tailoring any target figure of merit, for any material using machine learning, opens up opportunities to adapt elastic strain engineering of properties and performance in devices and systems in a controllable and efficient manner, for potential applications in microelectronics, optoelectronics, photonics, and energy technologies.