arXiv subjects

Dibakar Datta

Publications and source records attributed to Dibakar Datta.

At least 19 recordsLinked to original sources

arXiv2026

Physics-Informed and Knowledge-Driven Generative AI for Autonomous Discovery of Porous Oxide Energy Materials: Opportunities and Challenges

The discovery of next-generation energy-storage materials is increasingly limited by the complexity of the underlying design problem rather than by computational capability alone. Porous transition-metal oxides represent a particularly challenging class of battery materials because their performance emerges from coupled interactions among crystal chemistry, pore architecture, ion transport, electrochemistry, electro-chemo-mechanics, synthesis, manufacturing, and battery-system operation. Recent advances in generative artificial intelligence (AI) have demonstrated remarkable capabilities for generating chemically plausible crystal structures. However, current approaches remain largely focused on crystallographic validity and thermodynamic stability. This perspective presents a roadmap for advancing generative AI beyond crystal generation toward physics-informed, application-aware, and synthesis-aware inverse design. Using porous oxide electrodes as a representative materials platform, we propose a seven-tier physics-informed inverse-design framework integrating chemistry, thermodynamics, transport, electrochemistry, durability, cell compatibility, and manufacturability. We further identify the "Missing Data Problem" as a fundamental bottleneck limiting application-aware AI and introduce an autonomous knowledge-generation framework supported by a Porous Oxide Energy Materials Ontology and a continuously evolving "Knowledge Base". Together, these concepts establish the foundation for Synthesis-Aware, Closed-Loop Autonomous Discovery, providing a general framework for AI-enabled autonomous materials discovery across energy-storage materials and other functional materials.

cond-mat.mtrl-sci↗

arXiv2026

An Autonomous Scientific Knowledge Generation Framework for AI-Driven Scientific Discovery

Artificial intelligence (AI) is transforming scientific discovery, but its effectiveness is fundamentally limited by the availability of structured scientific knowledge. Although existing databases have accelerated data-driven materials research, much of the knowledge needed for predictive modeling and inverse design remains embedded in unstructured scientific literature. We present an Autonomous Scientific Knowledge Generation Framework that transforms scientific publications into a Unified AI-Ready Scientific Knowledge Base. The framework integrates ontology-guided literature acquisition, hybrid scientific knowledge extraction, semantic harmonization, knowledge fusion, and validation within a unified workflow. Rather than treating literature retrieval, information extraction, and database construction as separate tasks, the framework progressively converts scientific publications into structured, semantically consistent, and provenance-preserving knowledge suitable for AI-driven reasoning. As a proof of concept, the framework was applied to electro-optic materials. Autonomous literature acquisition retrieved and validated about 1,000 publications from multiple scholarly repositories. A representative subset of eight publications was processed through the complete workflow, generating 29 structured scientific records that were harmonized into 7 canonical scientific records. The results demonstrate the complete transformation from scientific literature to an AI-ready scientific knowledge base while preserving quantitative measurements, operating conditions, provenance, and scientific context. The proposed framework provides a scalable, domain-independent foundation for predictive AI, generative AI, and closed-loop AI-driven scientific discovery.

cs.DL↗

arXiv2025

Computational Study of Li+ Solvation Structures in Fluorinated Ether, Non-Fluorinated Ether, and Organic Carbonate-Based Electrolytes at Low and High Salt Concentrations

Understanding the solvation structure of electrolytes is crucial for optimizing the performance and stability of lithium-ion batteries. Novel electrolytes are essential for enhancing electrolyte structure and ensuring better integration with modern electrode systems. Herein, we report a novel weakly solvated ether electrolyte (WSEE) composed of a pure fluorinated ether solvent, which results in an anion-rich solvation structure even at a low salt concentration of 1 M. To explore this, we selected the advanced fluorinated solvent 2,2-difluoroethyl methyl ether (FEME) and compared it with dipropyl ether (DPE), ethylene carbonate (EC), and diethyl carbonate (DEC). The prepared electrolyte systems include DPE with 1 M, 1.8 M, and 4 M LiFSI; FEME with 1 M, 1.8 M, and 4 M LiFSI; and a 1:1 vol% EC/DEC mixture containing 1 M LiPF6. In this work, we comprehensively investigate the Li+ solvation structures using molecular dynamics (MD) simulations and density functional theory (DFT) calculations. Our computational findings indicate the presence of large ion aggregates (AGGs) in each DPE- and FEME-based electrolyte, while SSIPs (68%) are the dominant species in the mixed EC/DEC electrolyte. Notably, the formation of large ion aggregates is more pronounced in FEME-based electrolytes. We find that, similar to DPE, the FEME solvent also exhibits weak solvating power across all examined salt concentrations. More specifically, we find that FEME has weaker solvating power than DPE. This behavior is predicted by MD simulations, which indicate a strong preference for Li+ ions to coordinate with FSI- anions within the primary solvation shell. Furthermore, the quantum mechanical features of the Li+ solvation structures in these electrolytes are analyzed in detail using DFT. We anticipate that this study will provide valuable insights into the Li+ solvation structures in DPE, FEME, and EC/DEC electrolytes.

physics.chem-ph↗

arXiv2024

Generative AI for Discovering Porous Oxide Materials for Next-Generation Energy Storage

The key challenge in advancing multivalent-ion batteries lies in finding suitable intercalation hosts. Open-tunnel oxides, featuring one-dimensional channels or nanopores, show promise for enabling effective ion transport. However, the vast range of compositional possibilities renders traditional experimental and quantum-based methods impractical for large-scale studies. This work presents a generative AI framework that uses the Crystal Diffusion Variational Autoencoder (CDVAE) and a fine-tuned Large Language Model (LLM) to expedite the discovery of stable open-tunneled oxide materials for multivalent-ion batteries. By combining machine learning with data mining techniques, five promising transition metal oxide (TMO) structures are generated. These structures, known for forming open-tunnel oxide frameworks, are structurally validated through Density Functional Theory (DFT). The results show that the generated structures have lower formation energies compared to similar compositions in the Materials Project (MP) database, indicating improved thermodynamic stability. Additionally, the graph-based M3GNet model is employed to relax further generated structures, providing a more computationally efficient alternative to DFT. Machine learning-based predictions of formation energy, band gap, and energy above the hull refine the selection process, leading to the identification of materials with significant potential for real-world battery applications. This research demonstrates the power of generative AI in rapidly exploring the vast chemical space of TMOs, offering a new approach to discovering stable open-tunnel oxides for multivalent-ion batteries. The results highlight the potential of this approach to contribute to more sustainable energy storage technologies, addressing the growing concerns surrounding the scarcity of lithium.

cond-mat.mtrl-sci↗

arXiv2023

Exploring the Mechanical Behaviors of 2D Materials in Electrochemical Energy Storage Systems: Present Insights and Future Prospects

2D materials (2DM) and their heterostructures (2D + nD, n = 0,1,2,3) hold significant promise for applications in Electrochemical Energy Storage Systems (EESS), such as batteries. 2DM can serve as van der Waals (vdW) slick interface between conventional active materials (e.g., Silicon) and current collectors, modifying interfacial adhesion and preventing stress-induced fractures. Additionally, 2DM can replace traditional polymer binders (e.g., MXenes). This arrangement also underscores the critical role of interfacial mechanics between 2DM and active materials. Furthermore, 2DM can be designed to function as an electrode itself. For instance, a porous graphene network has been reported to possesses approximately five times the capacity of a traditional graphite anode. Consequently, gaining a comprehensive understanding of the mechanical properties of 2DM in EESS is paramount. However, modeling 2DM in EESS poses significant challenges due to the intricate coupling of mechanics and electrochemistry. For instance, defective graphene tends to favor adatom adsorption (e.g., Li+) during charging. In cases of strong adsorption, adatoms may not readily detach from electrodes during discharging. As a result, in such scenarios, adsorption-desorption (charge-discharge) processes govern the mechanical properties of 2DM when used as binders and current collectors. Regrettably, most existing studies on the mechanical properties of 2DM in EESS have failed to adequately address these critical issues. This perspective paper aims to provide a comprehensive overview of recent progress in the chemo-mechanics of 2DM's mechanical properties. A wide spectrum of multiscale modeling approaches, including atomistic/molecular simulations, continuum modeling, and machine learning, are discussed.

physics.app-ph↗

arXiv2023

Electro-Chemo-Mechanical Modeling of Multiscale Active Materials for Next-Generation Energy Storage: Opportunities and Challenges

The recent geopolitical crisis resulted in a gas price surge. Although lithium-ion batteries represent the best available rechargeable battery technology, a significant energy and power density gap exists between LIBs and petrol/gasoline. The battery electrodes comprise a mixture of active materials particles, conductive carbon, and binder additives deposited onto a current collector. Although this basic design has persisted for decades, the active material particle's desired size scale is debated. Traditionally, microparticles have been used in batteries. Advances in nanotechnology have spurred interest in deploying nanoparticles as active materials. However, despite many efforts in nano, industries still primarily use 'old' microparticles. Most importantly, the battery industry is unlikely to replace microstructures with nanometer-sized analogs. This poses an important question: Is there a place for nanostructure in battery design due to irreplaceable microstructure? The way forward lies in multiscale active materials, microscale structures with built-in nanoscale features, such as microparticles assembled from nanoscale building blocks or patterned with engineered or natural nanopores. Although experimental strides have been made in developing such materials, computational progress in this domain remains limited and, in some cases, negligible. However, the fields hold immense computational potential, presenting a multitude of opportunities. This perspective highlights the existing gaps in modeling multiscale active materials and delineates various open challenges in the realm of electro-chemo-mechanical modeling. By doing so, it aims to inspire computational research within this field and promote synergistic collaborative efforts between computational and experimental researchers.

cs.CE↗

arXiv2023

Exploring Thermal Transport in Electrochemical Energy Storage Systems Utilizing Two-Dimensional Materials: Prospects and Hurdles

Two-dimensional materials and their heterostructures have enormous applications in Electrochemical Energy Storage Systems (EESS) such as batteries. A comprehensive and solid understanding of these materials' thermal transport and mechanism is essential for the practical design of EESS. Experiments have challenges in providing improved control and characterization of complex structures, especially for low dimensional materials. Theoretical and simulation tools such as first-principles calculations, boltzmann transport equations, molecular dynamics simulations, lattice dynamics simulation, and non-equilibrium Green's function provide reliable predictions of thermal conductivity and physical insights to understand the underlying thermal transport mechanism in materials. However, doing these calculations require high computational resources. The development of new materials synthesis technology and fast-growing demand for rapid and accurate prediction of physical properties require novel computational approaches. The machine learning (ML) method provides a promising solution to address such needs. This review details the recent development in atomistic/molecular studies and ML of thermal transport in EESS. The paper also addresses the latest significant experimental advances. However, designing the best low-dimensional materials-based heterostructures is like a multivariate optimization problem. For example, a particular heterostructure may be suitable for thermal transport but can have lower mechanical strength/stability. For bi/multilayer structures, the interlayer distance may influence the thermal transport properties and interlayer strength. Therefore, the last part addresses the future research direction in low-dimensional materials-based heterostructure design for thermal transport in EESS.

cond-mat.mtrl-sci↗

arXiv2023

Effect of Graphene Interface on Potassiation in a Graphene- Selenium Heterostructure Cathode for Potassium-ion Batteries

Selenium (Se) cathodes are an exciting emerging high energy density storage system for Potassium ion batteries(KIB), where potassiation reactions are less understood. Here, we present an atomic-level investigation of KxSe cathode enclosed in hexagonal lattices of carbon(C) characteristic of multilayered graphene matrix and multiwalled carbon nanotubes (MW-CNTs). Microstructural changes directed by graphene substrate in KxSe cathode are contrasted with graphene-free cathode. Graphene's binding affinity for long-chain polyselenides (Se-Se-Se = -2.82 eV and Se-Se = -2.646 eV) and ability to induce reactivity between Se and K are investigated. Furthermore, intercalation voltage for graphene enclosed KxSe cathode reaction intermediates are calculated with K2Se as the final discharged product. Our results indicate a single-step reaction near a voltage of 1.55 V between K and Se cathode. Our findings suggest that operating at higher voltages (~2V) could result in the formation of reaction intermediates where intercalation/deintercalation of K could be a challenge, and therefore cause irreversible capacity losses in the battery. Primary issues are the high binding energy of long-chain polyselenides with graphene that discourage K storage and Se-Se bond dissociation at low K concentrations. A comparison with graphene-free cathode highlights the substantial changes a van der Waals (vdW) graphene interface can bring in atomic-structure and electrochemistry of the KxSe cathode.

physics.app-ph↗

arXiv2023

Transferable and Robust Machine Learning Model for Predicting Stability of Si Anodes for Multivalent Cation Batteries

Data-driven methodology has become a key tool in computationally predicting material properties. Currently, these techniques are priced high due to computational requirements for generating sufficient training data for high-precision machine learning models. In this study, we present a Support Vector Regression (SVR)-based machine learning model to predict the stability of silicon (Si) - alkaline metal alloys, with a strong emphasis on the transferability of the model to new silicon alloys with different electronic configurations and structures. We elaborate on the role of the structural descriptor in imparting transferability to the model that is trained on limited data (~750 Si alloys) derived from the Material Project database. Three popular descriptors, namely X-Ray Diffraction (XRD), Sine Coulomb Matrix (SCM), and Orbital Field Matrix (OFM), are evaluated for representing Si alloys. The material structures are represented by descriptors in the SVR model, coupled with hyperparameter tuning techniques like Grid Search CV and Bayesian Optimization (BO), to find the best performing model for predicting total energy, formation energy and packing fraction of the Si alloy systems. The models are trained on Si alloys with lithium (Li), sodium (Na), potassium (K), magnesium (Mg), calcium (Ca), and aluminum (Al) metals, where Si-Na and Si-Al systems are used as test structures. Our results show that XRD, an experimentally derived characterization of structures, performs most reliably as a descriptor for total energy prediction of new Si alloys. The study demonstrates that by qualitatively selection of training data, using hyperparameter tuning methods, and employing appropriate structural descriptors, the data requirements for robust and accurate ML models can be reduced.

cond-mat.mtrl-sci↗

arXiv2023

Open-tunneled oxides as intercalation host for multivalent ion (Ca and Al) batteries: A DFT study

Lithium-ion batteries (LIBs) are ubiquitous in everyday applications. However, Lithium (Li) is a limited resource on the planet and is therefore not sustainable. As an alternative to lithium, earth-abundant and cheaper multivalent metals such as aluminum (Al) and calcium (Ca) have been actively researched in battery systems. However, finding suitable intercalation hosts for multivalent-ion batteries is urgently needed. Open-tunneled oxides are a particular category of microparticles distinguished by the presence of integrated one-dimensional channels or nanopores. This work focuses on two promising open-tunnel oxides, viz: Niobium Tungsten Oxide (NTO) and Molybdenum Vanadium Oxide (MoVO). We find that the MoVO structure can adsorb greater numbers of multivalent ions than NTO due to its larger surface area and different shapes. The MoVO structure can adsorb Ca, Li, and Al ions with adsorption potential at around 4 to 5 eV. However, the adsorption potential for hexagonal channels of Al ion drops to 1.73 eV because of less channel area. NTO structure has an insertion/adsorption potential of 4.4 eV, 3.4 eV, and 0.9 eV for one Li, Ca, and Al, respectively. In general, Ca ion is more adsorbable than Al ion in both MoVO and NTO structures. Bader charge analysis and charge density plot reveals the role of charge transfer and ion size on the insertion of multivalent ions such as Ca and Al into MoVO and NTO systems. Our results provide general guidelines to explore other multivalent ions for battery applications.

cond-mat.mtrl-sci↗

arXiv2022

Turbostratic Orientations, Water Confinement and Ductile-Brittle Fracture in Bi-layer Graphene

Bi-layer graphene (BLG) can be a cheaper and more stable alternative to graphene in several applications. With its mechanical strength being almost equivalent to graphene, BLG also brings advanced electronic and optical properties to the table. Furthermore, entrapment of water in graphene-based nano-channels and devices has been a recent point of interest for several applications ranging from energy to bio-physics. Therefore, it is crucial to study the over-all mechanical strength of such structures in order to prevent system failures in future applications. In the present work, Molecular Dynamics simulations have been used to study crack propagation in BLG with different orientations between the layers. There is a major thrust in analyzing how the angular orientation between the layers affect the horizontal and vertical crack propagation in individual layers of graphene. The study has been extended to BLG with confined water in interfaces. Over-all strength of graphene sheets when in contact with water content has been determined, and prominent regional conditions for crack initiation are pointed out. It was seen that in the presence of water content, graphene deviated from its characteristic brittle failure and exhibited the ductile fracture mechanism. Origin of cracks in graphenes was located at the region where the density of water dropped near the graphene surface, suggesting that the presence of hydroxyl groups decelerate the crack formation and propagation in straining graphenes.

physics.app-ph↗

arXiv2022

Engineering frictional characteristics of MoS2 structure by tuning thickness and morphology- An atomic, electronic structure, and exciton analysis

We performed atomic and electron dynamics analysis to study the impact of morphological and thickness changes of a MoS2 system on its tribological properties through a diamond tip. We had considered 4 cases: variable layers (1-4 layers) and number (2-8 indents), radius (12Å, 16Å, 20Å, 24Å), and pattern of indents (0°, 25°, 30°, 35°, 45°, 60°) resulting into 18 subcases. MD results showed changing the radius and number of indents were the most, and number of layers and indents' pattern were the least effective way to tune the frictional characteristics. Ground state ab-initio study demonstrated an increase in the number and radius of indents, raising the number of stretched bonds. Consequently, the volume covered by the HOMO iso-surface increases, and that of LUMO decreases. That makes higher area/volume available to lose/share electrons, resulting in stronger interlocking between layers and tip. And TD-DFT calculation proves the existence of interfacial excitons, resulting in stronger interlocking between the layer's surface and tip despite a contraction in the LUMO iso-surfaces' area/volume. We believe these interlayer excitons result in higher average Z-axis(hence frictional force) reaction forces for the indents number subcases and lower for indents radius subcases as the number and radius of indents increase.

cond-mat.mtrl-sci↗

arXiv2022

Developing Potential Energy Surfaces for Graphene-based 2D-3D Interfaces from Modified High Dimensional Neural Networks for Applications in Energy Storage

Mixed-dimensional heterostructures composed of two-dimensional (2D) and three-dimensional (3D) materials are undisputed next-generation materials for engineered devices due to their changeable properties. The present work computationally investigates the interface between 2D graphene and 3D tin (Sn) systems with density functional theory (DFT) method. It uses computationally demanding simulation data to develop machine learning (ML) based potential energy surfaces (PES). The approach to developing PES for complex interface systems in the light of limited data and transferability of such models has been discussed. To develop PES for graphene-tin interface systems, high dimensional neural networks (HDNN) are used that rely on atom-centered symmetry function to represent structural information. HDNN are modified to train on the total energies of the interface system rather than atomic energies. The performance of modified HDNN trained on 5789 interface structures of graphene|Sn is tested on new interfaces of the same material pair with varying levels of structural deviations from the training dataset. Root mean squared error (RMSE) for test interfaces fall in the range of 0.01-0.45 eV/atom, depending on the structural deviations from the reference training dataset. By avoiding incorrect decomposition of total energy into atomic energies, modified HDNN model is shown to obtain higher accuracy and transferability despite limited dataset. Improved accuracy in ML-based modeling approach promises cost-effective means of designing interfaces in heterostructure energy storage systems with higher cycle life and stability.

cond-mat.mtrl-sci↗

arXiv2022

Drug repurposing for SARS-COV-2: A high-throughput molecular docking, molecular dynamics, machine learning, & ab-initio study

A molecule of dimension 125nm has caused around 479 Million human infections (80M for the USA) & 6.1 Million human deaths (977,000 for the USA) worldwide and slashed the global economy by US$ 8.5 Trillion over two years. The only other events in recent history that caused comparative human life loss through direct usage (either by (wo)man or nature, respectively) of structure-property relations of 'nano-structures' (either (wo)man-made or nature, respectively) were nuclear bomb attacks of Japanese cities by the USA during World War II and 1918 Flu Pandemic. This molecule is SARS-CoV-2, which causes a disease known as COVID-19. The high liability cost of the pandemic had incentivized various private, government, and academic entities to work towards finding a cure for these & emerging diseases. As result, multiple vaccine candidates are discovered to avoid the infection in first place. But so far, there has been no success in finding fully effective therapeutics candidates. In this paper, we attempted to provide multiple therapy candidates based upon a sophisticated multi-scale in-silico framework. We have used the following robust framework to screen the ligands; Step-I: high throughput docking, Step-II: molecular dynamics, Step-III: density functional theory analysis. In total, we have analyzed 2.2 Million unique protein binding site/ligand combinations. The proteins were selected based on recent experimental studies. Step-I had filtered that number down to 10 ligands/protein based on molecular docking binding energy, further screening down to 2 ligands/protein based on drug-likeness analysis. Additionally, these two ligands/proteins were investigated in Step-II with a molecular dynamic based RMSD analysis. It finally suggested three ligands (ZINC1176619532, ZINC517580540, ZINC952855827) attacking different binding sites of the protein(7BV2), which were further analyzed in Step III.

q-bio.BM↗

arXiv2020

Variation in interface strength of Silicon with surface engineered Ti3C2 MXenes

Current advancements in battery technologies require electrodes to combine high-performance active material such as Silicon (Si) with two-dimensional materials such as transition metal carbides (MXenes) for prolonged cycle stability and enhanced electrochemical performance. More so, it is the interface between these materials, which is the nexus for their applicatory success. Herein, the interface strength variations between amorphous Si and Ti3C2Tx MXene are determined as the MXene surface functional groups (Tx) are changed using first principle calculations. Si is interfaced with three Ti3C2 MXene substrates having surface -OH, -OH and -O mixed, and -F functional groups. Density functional theory (DFT) results reveal that completely hydroxylated Ti3C2 has the highest interface strength of 0.6 J/m2 with amorphous Si. This interface strength value drops as the proportion of surface -O and -F groups increases. Additional analysis of electron redistribution and charge separation across the interface is provided for a complete understanding of underlying physico-chemical factors affecting the surface chemistry and resultant interface strength values. The presented comprehensive analysis of the interface aims to develop sophisticated MXene based electrodes by their targeted surface engineering.

physics.app-ph↗

arXiv2020

Understanding the Strength of the Selenium -- Graphene Interfaces

We present a comprehensive first-principles Density Functional Theory (DFT) analyses of the interfacial strength and bonding mechanisms between crystalline and amorphous selenium(Se) with graphene(Gr), a promising duo for energy storage applications. Comparative interface analyses are presented on amorphous silicon(Si) with graphene and crystalline Se with aluminum(Al) substrate. The interface strength of monoclinic Se (0.43 J/m2) and amorphous Si with graphene (0.41 J/m2) is similar in magnitude. While both materials (c-Se, a-Si) are bonded loosely by van der Waals (vdW) forces over graphene, interfacial electron exchange is higher for a-Si/Gr. This is further elaborated by comparing potential energy step and charge transfer (delta q) across the graphene interfaces. The delta q for c-Se/Gr and a-Si/Gr are 0.3119 e-1 and 0.4266 e-1, respectively. However, the interface strength of c-Se on the 3D Al substrate is higher (0.99 J/m2), suggesting stronger adhesion. The amorphous Se with graphene has comparable interface strength (0.34 J/m2), but electron exchange in this system is slightly distinct from monoclinic Se. The electronic characteristics (density of states analysis) and bonding mechanisms are different for monoclinic and amorphous Se with graphene and they activate graphene via surface charge doping divergently. Our findings highlight the complex electrochemical phenomena in Se interfaced with graphene, which may profoundly differ from their 'free' counterparts.

physics.app-ph↗

arXiv2020

Machine Learning in Materials Modeling -- Fundamentals and the Opportunities in 2D Materials

The application of machine learning in materials presents a unique challenge of dealing with scarce and varied materials data - both experimental and theoretical. Nevertheless, several state-of-the-art machine learning models for materials have been successfully developed to predict material properties for various applications such as materials for photovoltaic cells, thermoelectric materials, dielectrics, materials for batteries, fuel cells, etc. The setup of comprehensive materials databases, and openly accessible algorithm frameworks have also spurred the usage of machine learning for solving some of the most pressing problems in materials science. Some such recent implementations are discussed in this book chapter. A multitude of two-dimensional (2D) materials exist with the potential to replace the conventional materials for energy storage and nanodevices. The challenges faced in designing batteries and how machine learning tools can help in screening and narrowing down on the best composition, as well as the synthesis of air-stable 2D materials, are also discussed.

cond-mat.mtrl-sci↗

arXiv2020

Comprehensive understanding of water-driven graphene wrinkle life-cycle towards applications in flexible electronics: A computational study

The presence of wrinkles in Graphene Nanoribbons (GNR) and other two-dimensional (2D) materials significantly alter their mechanical, electronic, optical properties, which can be either beneficial or detrimental. Experimentally, it has been observed that during the commonly used growth process of GNR, water molecules, sourced from ambient humidity, can be diffused in between GNR and the substrate. The water diffusion causes wrinkle formation in GNR, which influences its properties. Furthermore, the diffused water eventually dries, creating the alteration not only in the geometry of Wrinkled Graphene Nanoribbons (WGNR) but also its features. Computational analysis of these phenomena can provide an atomistic-level understanding of the phenomena. Therefore, in this work, Molecular Dynamics (MD) simulations are performed to model the water diffusion and evaporation in between GNR and its substrate, and their effect on wrinkle formation and dynamics. Additionally, Density Functional Theory (DFT)-based analysis is used to characterize the difference in the electronic structure of WGNR caused by the change in wrinkle geometry. Our study reveals that the initially distributed wrinkles tend to coalesce to form a localized wrinkle whose configuration depends on the initial wrinkle geometry and the amount of diffused water. The wrinkle configuration changes upon drying, while it remains static until the complete drying. The movement of the localized wrinkle is the combination of three fundamental modes - bending, buckling, and sliding. The stress analysis reveals that the maximum stress is at the base of the wrinkle, and its magnitude is always below the plasticity limit. The DFT results provide insight into the potential of using the wrinkles to control the direction of electron flow for the applications in flexible electronics.

cond-mat.mtrl-sci↗