Siqi Shi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Recognitions
Siqi Shi has a total of 486 co-authors. Here are the top co-authors by number of shared publications:
Maxim Avdeev
36 shared publicationsDa Wang
27 shared publicationsBing He
20 shared publicationsYue Liu
State Key Laboratory of Materials for Advanced Nuclear Energy & School of Computer Engineering and Science Shanghai University Shanghai China
20 shared publicationsZheyi Zou
School of Materials Science and Engineering Xiangtan University Xiangtan 411105 China
20 shared publicationsLiquan Chen
Beijing National Laboratory for Condensed Matter Physics Institute of Physics Chinese Academy of Sciences Beijing China
17 shared publications
Latest Publications
- Open Access
- 1 January 2027
- Published by KEAI Publishing in eScience
- Vol. 7 (1)
Enhancing high-throughput battery materials discovery and design with artificial intelligence
- 1 January 2027
- Published by KEAI Publishing in eScience
- Vol. 7 (1)
Open AccessEnhancing high-throughput battery materials discovery and design with artificial intelligence
- Open Access
- 2 November 2026
- Published by Wiley in Advanced Functional Materials
- p. e79431
A Thermodynamic‐Decoupling‐Empowered Defect Engineering Guide: Uncovering the Substantial Yet Overlooked Stress Role in Electrode Potential Regulation
Defect chemistry for improving electrochemical potential and structural stability traditionally relies on chemical bonding and charge compensation. However, it reaches a bottleneck when the beneficial defect concentration saturates under the formation‐energy constraint, because defect‐induced lattice stress generally dissipates through volume expansion, obscuring its electrochemical contribution. Herein, we establish a defect‐engineering guide that thermodynamically decouples formation energy to extract an unprecedented stress dimension alongside traditional defect concentrations. Crucially, we develop a stepwise process driven by quantitative ligand‐field and stress descriptors to simultaneously screen both dimensions to their limits. We elucidate how universally expansive electrodes like silicon can maintain structural stability while suppressing passive stress release to mitigate expansion. Furthermore, targeting the extreme archetype of non‐expansive Li 4 Ti 5 O 12 electrodes, we uncover how confined stress can be transformed into a thermodynamic gain to lower their potentials. Among the 612 dual‐doping configurations, the optimal La‐Zr doping maximizes both the defect concentration and the strain energy, reducing the lithiation potential from 1.55 to 1.00 V (vs. Li/Li + ) while preserving intrinsic the zero‐strain characteristic. This work proposes a defect‐engineering guide rooted in a four‐quadrant panorama of thermodynamically decoupled concentration and stress dimensions, breaking optimization bottlenecks by harnessing stress upon formation‐energy‐constrained defect saturation to expand the boundaries of defect chemistry.
- 2 November 2026
- Published by Wiley in Advanced Functional Materials
- p. e79431
Open AccessA Thermodynamic‐Decoupling‐Empowered Defect Engineering Guide: Uncovering the Substantial Yet Overlooked Stress Role in Electrode Potential Regulation
Defect chemistry for improving electrochemical potential and structural stability traditionally relies on chemical bonding and charge compensation. However, it reaches a bottleneck when the beneficial defect concentration saturates under the formation‐energy constraint, because defect‐induced lattice stress generally dissipates through volume expansion, obscuring its electrochemical contribution. Herein, we establish a defect‐engineering guide that thermodynamically decouples formation energy to extract an unprecedented stress dimension alongside traditional defect concentrations. Crucially, we develop a stepwise process driven by quantitative ligand‐field and stress descriptors to simultaneously screen both dimensions to their limits. We elucidate how universally expansive electrodes like silicon can maintain structural stability while suppressing passive stress release to mitigate expansion. Furthermore, targeting the extreme archetype of non‐expansive Li 4 Ti 5 O 12 electrodes, we uncover how confined stress can be transformed into a thermodynamic gain to lower their potentials. Among the 612 dual‐doping configurations, the optimal La‐Zr doping maximizes both the defect concentration and the strain energy, reducing the lithiation potential from 1.55 to 1.00 V (vs. Li/Li + ) while preserving intrinsic the zero‐strain characteristic. This work proposes a defect‐engineering guide rooted in a four‐quadrant panorama of thermodynamically decoupled concentration and stress dimensions, breaking optimization bottlenecks by harnessing stress upon formation‐energy‐constrained defect saturation to expand the boundaries of defect chemistry.
- Open Access
- 14 October 2026
- Published by Springer Nature in npj Computational Materials
Quantifying electronic insulation of solid-electrolyte interphases through electrolyte decomposition kinetics
- 14 October 2026
- Published by Springer Nature in npj Computational Materials
Open AccessQuantifying electronic insulation of solid-electrolyte interphases through electrolyte decomposition kinetics
- 1 October 2026
- Published by Wiley in Advanced Energy Materials
A Spatially‐Resolved Approach via Quantifying the Dominant Yet Overlooked Orbital‐Weighted Energy Contribution for Breaking Solvent Oxidation Chemistry
Rational design of solvents with high oxidation chemistry stability represents an essential path toward pursuing high‐energy‐density batteries. Conventional methods assess oxidation stability using orbital energy level/ionization energetics; however, they misjudge the local nature of electron loss by ignoring the influence of collective orbitals on electron energy. Herein, we propose a spatially‐resolved energetics approach called Pseudo‐Eox, where the core average local ionization energy minimum parameter is used to measure the vulnerability of the highest‐energy electron via electron‐density‐weighted orbital energy averaging and to pinpoint its prevalent sub‐HOMO localization. This approach reproduces the oxidation stability trend of HOMO‐dominant solvents (EC/PC with quantified contributions > 60%), while mechanistically revealing why conventional models fail in most cases like DEC with HOMO contribution < 24%. Its robustness is demonstrated on 20 regular solvents with mean absolute error of only 0.38 V. Following this, taking potentially practical fluorinated solvents as a prototype, we retrieve the 2‐fluoroethyl 2,2,2‐trifluoroethyl, which will be overlooked by HOMO‐based rankings, exhibiting the highest theoretical oxidation stability (6.45 V) among candidates obtained by constructing 147 widely adopted carbonate/ether derivatives. This study overturns the single‐orbital‐only cognition of solvent oxidation, offering a spatially informed energetics perspective for investigating organic chemistry where localized electron loss governs stability and functionality.
- 1 October 2026
- Published by Wiley in Advanced Energy Materials
A Spatially‐Resolved Approach via Quantifying the Dominant Yet Overlooked Orbital‐Weighted Energy Contribution for Breaking Solvent Oxidation Chemistry
Rational design of solvents with high oxidation chemistry stability represents an essential path toward pursuing high‐energy‐density batteries. Conventional methods assess oxidation stability using orbital energy level/ionization energetics; however, they misjudge the local nature of electron loss by ignoring the influence of collective orbitals on electron energy. Herein, we propose a spatially‐resolved energetics approach called Pseudo‐Eox, where the core average local ionization energy minimum parameter is used to measure the vulnerability of the highest‐energy electron via electron‐density‐weighted orbital energy averaging and to pinpoint its prevalent sub‐HOMO localization. This approach reproduces the oxidation stability trend of HOMO‐dominant solvents (EC/PC with quantified contributions > 60%), while mechanistically revealing why conventional models fail in most cases like DEC with HOMO contribution < 24%. Its robustness is demonstrated on 20 regular solvents with mean absolute error of only 0.38 V. Following this, taking potentially practical fluorinated solvents as a prototype, we retrieve the 2‐fluoroethyl 2,2,2‐trifluoroethyl, which will be overlooked by HOMO‐based rankings, exhibiting the highest theoretical oxidation stability (6.45 V) among candidates obtained by constructing 147 widely adopted carbonate/ether derivatives. This study overturns the single‐orbital‐only cognition of solvent oxidation, offering a spatially informed energetics perspective for investigating organic chemistry where localized electron loss governs stability and functionality.
- 7 September 2026
- Published by AIP Publishing in Applied Physics Letters
- Vol. 129 (10)
Unveiling the hidden effects of local potential energy surface on ion transport in superionic conductors
The potential energy surface (PES) fundamentally governs the motion of mobile ions in superionic conductors (SICs). In particular, flat local PES around mobile ion sites, indicative of low vibration frequencies and soft lattice dynamics, are strongly associated with reduced migration barriers and fast ion transport. However, a fast and quantitative method for evaluating local PES flatness remains lacking. Here, we combine the bond valence site energy method with isosurface analysis and harmonic vibration analysis to quantify local PES flatness through two complementary descriptors: the isosurface-enclosed volume (Viso; geometric perspective) and the site-specific vibration frequency (νsite; dynamical perspective), and unveiling their hidden effects on ion transport. Analysis of representative SICs reveals that large Viso (0.019–0.084 Å3) and low νsite (2.304–5.069 THz) at specific mobile ion sites are characteristic of fast ion transport, whereas non-SICs consistently exhibit smaller Viso (0.002–0.009 Å3) and higher νsite (8.465–12.872 THz). The proposed thresholds of Viso = 0.01 Å3 and νsite = 7 THz show a clear separation between known SICs and non-SICs in the current dataset and are expected to serve as practical criteria for future high-throughput screening of new SICs.
- 7 September 2026
- Published by AIP Publishing in Applied Physics Letters
- Vol. 129 (10)
Unveiling the hidden effects of local potential energy surface on ion transport in superionic conductors
The potential energy surface (PES) fundamentally governs the motion of mobile ions in superionic conductors (SICs). In particular, flat local PES around mobile ion sites, indicative of low vibration frequencies and soft lattice dynamics, are strongly associated with reduced migration barriers and fast ion transport. However, a fast and quantitative method for evaluating local PES flatness remains lacking. Here, we combine the bond valence site energy method with isosurface analysis and harmonic vibration analysis to quantify local PES flatness through two complementary descriptors: the isosurface-enclosed volume (Viso; geometric perspective) and the site-specific vibration frequency (νsite; dynamical perspective), and unveiling their hidden effects on ion transport. Analysis of representative SICs reveals that large Viso (0.019–0.084 Å3) and low νsite (2.304–5.069 THz) at specific mobile ion sites are characteristic of fast ion transport, whereas non-SICs consistently exhibit smaller Viso (0.002–0.009 Å3) and higher νsite (8.465–12.872 THz). The proposed thresholds of Viso = 0.01 Å3 and νsite = 7 THz show a clear separation between known SICs and non-SICs in the current dataset and are expected to serve as practical criteria for future high-throughput screening of new SICs.
- 17 August 2026
- Published by American Chemical Society (ACS) in Journal of the American Chemical Society
- Vol. 148 (33) , 36175-36183
Revealing Excited-State Electrochemical Reaction Pathways at Solid Electrolyte Interphases by Ab Initio Nonadiabatic Molecular Dynamics
Electrode potential controls electrochemical reactivity by governing interfacial electron-transfer processes at electrified interfaces. Under varying electrochemical potentials, injected electrons may access different electronic states of interfacial electrolyte complexes, potentially driving distinct nonequilibrium excited-state reaction pathways. Conventional electrochemical frameworks primarily focus on equilibrium reaction energetics, while the nonequilibrium dynamics following interfacial electron transfer remain largely unexplored. Here, we investigate such ultrafast interfacial reaction dynamics using ab initio nonadiabatic molecular dynamics (NAMD) simulations within the Ehrenfest dynamics framework. Using ethylene carbonate (EC) as a prototypical electrolyte solvent in Li-ion batteries, we show that electron injection into different electronic states of solvated Li+ complexes can drive qualitatively distinct nonadiabatic reaction pathways on femtosecond-to-picosecond time scales. Furthermore, local solvation environments and interactions between neighboring reaction units substantially modify the reaction landscape and the resulting decomposition channels relevant to solid electrolyte interphase formation. This work establishes a microscopic nonequilibrium perspective on electrochemical reactivity and suggests that excited-state dynamics may play an important role in interfacial electrochemical reactions beyond conventional ground-state descriptions.
- 17 August 2026
- Published by American Chemical Society (ACS) in Journal of the American Chemical Society
- Vol. 148 (33) , 36175-36183
Revealing Excited-State Electrochemical Reaction Pathways at Solid Electrolyte Interphases by Ab Initio Nonadiabatic Molecular Dynamics
Electrode potential controls electrochemical reactivity by governing interfacial electron-transfer processes at electrified interfaces. Under varying electrochemical potentials, injected electrons may access different electronic states of interfacial electrolyte complexes, potentially driving distinct nonequilibrium excited-state reaction pathways. Conventional electrochemical frameworks primarily focus on equilibrium reaction energetics, while the nonequilibrium dynamics following interfacial electron transfer remain largely unexplored. Here, we investigate such ultrafast interfacial reaction dynamics using ab initio nonadiabatic molecular dynamics (NAMD) simulations within the Ehrenfest dynamics framework. Using ethylene carbonate (EC) as a prototypical electrolyte solvent in Li-ion batteries, we show that electron injection into different electronic states of solvated Li+ complexes can drive qualitatively distinct nonadiabatic reaction pathways on femtosecond-to-picosecond time scales. Furthermore, local solvation environments and interactions between neighboring reaction units substantially modify the reaction landscape and the resulting decomposition channels relevant to solid electrolyte interphase formation. This work establishes a microscopic nonequilibrium perspective on electrochemical reactivity and suggests that excited-state dynamics may play an important role in interfacial electrochemical reactions beyond conventional ground-state descriptions.
- 11 August 2026
- Published by American Chemical Society (ACS) in The Journal of Physical Chemistry B
- Vol. 130 (33) , 8503-8513
Specific Cation Effects on Structure and Dynamics of Carbonyl with Water in N -Methylacetamide Aqueous Salt Solutions
Specific ion effects are typically manifested as differences in the behavior of salts in solution; these effects have largely been focused on preferential binding interactions at specific sites. However, the direct interactions between metal cations and functional groups were confirmed to be far weaker, most cations were not observed to exhibit any binding. It has therefore been proposed that the indirect role of water is equally important in understanding the specific cation effects, yet this significant contribution of the solvent has received less attention. Herein, we performed molecular dynamics simulations to investigate how cations affect the structural and dynamical properties of the solvation shell at the carbonyl–water interface in N-methylacetamide aqueous chloride salt solutions. Our simulations reveal that the distinct solvation behaviors of the carbonyl group exhibit a pronounced dependence on the specific cation present. Moreover, different cations produce specific effects on the structure and dynamics of the solvation of the carbonyl group, which well correlate with both the size and valence of the cation. It has been also observed that salt ions present the considerable effects on the structure and dynamics of the solvation of the carbonyl group via the salt concentration. These findings combined with established theories of the rigid solvation shell induced by cation could provide a molecular basis for understanding the specific cation effects.
- 11 August 2026
- Published by American Chemical Society (ACS) in The Journal of Physical Chemistry B
- Vol. 130 (33) , 8503-8513
Specific Cation Effects on Structure and Dynamics of Carbonyl with Water in N -Methylacetamide Aqueous Salt Solutions
Specific ion effects are typically manifested as differences in the behavior of salts in solution; these effects have largely been focused on preferential binding interactions at specific sites. However, the direct interactions between metal cations and functional groups were confirmed to be far weaker, most cations were not observed to exhibit any binding. It has therefore been proposed that the indirect role of water is equally important in understanding the specific cation effects, yet this significant contribution of the solvent has received less attention. Herein, we performed molecular dynamics simulations to investigate how cations affect the structural and dynamical properties of the solvation shell at the carbonyl–water interface in N-methylacetamide aqueous chloride salt solutions. Our simulations reveal that the distinct solvation behaviors of the carbonyl group exhibit a pronounced dependence on the specific cation present. Moreover, different cations produce specific effects on the structure and dynamics of the solvation of the carbonyl group, which well correlate with both the size and valence of the cation. It has been also observed that salt ions present the considerable effects on the structure and dynamics of the solvation of the carbonyl group via the salt concentration. These findings combined with established theories of the rigid solvation shell induced by cation could provide a molecular basis for understanding the specific cation effects.
- Open Access
- 10 August 2026
- Published by Elsevier in Journal of Materiomics
- Vol. 12 (6)
Dual-descriptor-guided electrolyte additive screening strategy for achieving highly reversible aqueous zinc-ion batteries
- 10 August 2026
- Published by Elsevier in Journal of Materiomics
- Vol. 12 (6)
Open AccessDual-descriptor-guided electrolyte additive screening strategy for achieving highly reversible aqueous zinc-ion batteries
- Open Access
- 23 July 2026
- Published by Wiley in Advanced Materials
- Vol. 38 (49) , e74207
Direct Modeling of the Interfacial Resistance in All‐Solid‐State Battery
Interfacial reconstruction and its associated high resistance govern the performance of all‐solid‐state batteries (ASSBs). However, indirectly inferring interfacial potentials from bulk band alignments masks the true solid–solid electrochemistry, causing orders‐of‐magnitude discrepancies in predicting space‐charge layer (SCL) resistances and impeding interface screening. Herein, by traversing 310 distinct interfaces from ∼29,000 literatures, we develop a non‐empirical numerical procedure that directly maps lithium‑ion redistribution to interfacial resistance by integrating ligand‑field theory with the SCL model. Considering electric potential differences and intrinsic carrier properties during interfacial reconstruction via a modified ligand‐field splitting strength (MLFSS) descriptor yields unprecedented bridging between modeling and measurement, reducing predicted resistance discrepancies from over ten orders of magnitude to within two. On this basis, we resolve the highly system‐dependent controversy over oxide interfacial resistances by identifying extreme MLFSS disparities (>3.5 eV) as the decisive factor, while emphasizing ion‑intercalation sulfides (0.5 Cr 1.5 S 4 /Li 10 GeP 2 S 12 /75% Li 2 S‐24% P 2 S 5 –1% P 2 O 5 /Li prototype. The resulting ultralow interfacial resistance of 8.8 Ω cm 2 ensures superior cycling stability at an active‐material energy density of 562 Wh kg −1 , establishing a practical paradigm for breaking the energy and kinetics trade‐off in ASSBs.
- 23 July 2026
- Published by Wiley in Advanced Materials
- Vol. 38 (49) , e74207
Open AccessDirect Modeling of the Interfacial Resistance in All‐Solid‐State Battery
Interfacial reconstruction and its associated high resistance govern the performance of all‐solid‐state batteries (ASSBs). However, indirectly inferring interfacial potentials from bulk band alignments masks the true solid–solid electrochemistry, causing orders‐of‐magnitude discrepancies in predicting space‐charge layer (SCL) resistances and impeding interface screening. Herein, by traversing 310 distinct interfaces from ∼29,000 literatures, we develop a non‐empirical numerical procedure that directly maps lithium‑ion redistribution to interfacial resistance by integrating ligand‑field theory with the SCL model. Considering electric potential differences and intrinsic carrier properties during interfacial reconstruction via a modified ligand‐field splitting strength (MLFSS) descriptor yields unprecedented bridging between modeling and measurement, reducing predicted resistance discrepancies from over ten orders of magnitude to within two. On this basis, we resolve the highly system‐dependent controversy over oxide interfacial resistances by identifying extreme MLFSS disparities (>3.5 eV) as the decisive factor, while emphasizing ion‑intercalation sulfides (0.5 Cr 1.5 S 4 /Li 10 GeP 2 S 12 /75% Li 2 S‐24% P 2 S 5 –1% P 2 O 5 /Li prototype. The resulting ultralow interfacial resistance of 8.8 Ω cm 2 ensures superior cycling stability at an active‐material energy density of 562 Wh kg −1 , establishing a practical paradigm for breaking the energy and kinetics trade‐off in ASSBs.
- 18 June 2026
- Published by American Chemical Society (ACS) in The Journal of Physical Chemistry Letters
- Vol. 17 (26) , 7474-7484
Predicting Li-Ion Migration Energy Barriers in Battery Cathode Materials via Convolutional Neural Network Model Based on Descriptors Divide-and-Conquer Strategy
Predicting the Li-ion migration energy barrier in battery cathode materials via machine learning has attracted increasing attention. However, the capture of complex material features via the accurate definition of structural descriptors and cross-scale information integration for neural network modeling remains challenging. Benefiting from the reported divide-and-conquer strategy, we propose a convolutional neural network (CNN) model employing hybrid geometric-topological descriptors for the efficient prediction of Li-ion migration energy barriers, in which the geometric descriptors capture local Li-O polyhedra of initial and transition states, as well as the topological descriptors derived from persistent homology characterize structural connectivity and ring channels along the migration pathway. Compared with the recurrent neural network (RNN) and Fourier-feature network (FFN) models, the CNN model optimized via residual block structures and L2 regularization achieves a mean absolute error (MAE) of 0.0589 eV. The minimum Li-O distance during the Li-ion migration is identified as the most critical factor affecting the migration barrier, and the comparable importance scores of dmin, dstd, and H0 bars suggest a synergistic effect of multiple structural descriptors, highlighting the necessity of adopting hybrid geometric-topological descriptors. The present work provides an efficient and accurate approach for high-throughput screening of materials with rapid Li-ion diffusion.
- 18 June 2026
- Published by American Chemical Society (ACS) in The Journal of Physical Chemistry Letters
- Vol. 17 (26) , 7474-7484
Predicting Li-Ion Migration Energy Barriers in Battery Cathode Materials via Convolutional Neural Network Model Based on Descriptors Divide-and-Conquer Strategy
Predicting the Li-ion migration energy barrier in battery cathode materials via machine learning has attracted increasing attention. However, the capture of complex material features via the accurate definition of structural descriptors and cross-scale information integration for neural network modeling remains challenging. Benefiting from the reported divide-and-conquer strategy, we propose a convolutional neural network (CNN) model employing hybrid geometric-topological descriptors for the efficient prediction of Li-ion migration energy barriers, in which the geometric descriptors capture local Li-O polyhedra of initial and transition states, as well as the topological descriptors derived from persistent homology characterize structural connectivity and ring channels along the migration pathway. Compared with the recurrent neural network (RNN) and Fourier-feature network (FFN) models, the CNN model optimized via residual block structures and L2 regularization achieves a mean absolute error (MAE) of 0.0589 eV. The minimum Li-O distance during the Li-ion migration is identified as the most critical factor affecting the migration barrier, and the comparable importance scores of dmin, dstd, and H0 bars suggest a synergistic effect of multiple structural descriptors, highlighting the necessity of adopting hybrid geometric-topological descriptors. The present work provides an efficient and accurate approach for high-throughput screening of materials with rapid Li-ion diffusion.