Single-atom catalysts (SACs) offer distinct advantages for the direct dehydrogenation of ethane (EDH), primarily attributed to their atomically dispersed active sites, near-perfect atomic utilization, uniform active centers and superior catalytic performance^[1]. While Pt-based catalysts are still predominant in this area^[2], recent research has revealed promising light-alkane dehydrogenation activity for non-noble metal systems such as V^[3], Cr^[4] and Ni^[5], expanding the pool of potential catalytic materials and enabling greater flexibility in the design of active metal centers. Notably, a key challenge for EDH, which typically operates under harsh high-temperature and strongly reducing conditions, is preserving the structural stability of isolated metal atoms.

Studies have shown that both optimizing support properties^[6−8] and fine-tuning the local coordination environment of the active sites^[9] can substantially improve the stability and service life of SACs. Graphene, as a typical two-dimensional support, presents a challenge for anchoring metal atoms due to its intrinsic chemical inertness. However, its tunable electronic structure and well-defined geometry make it an excellent model system for investigating structure-performance relationships, offering unique value for theoretical studies and rational catalyst design.

Current research on graphene-supported SACs has largely focused on symmetric M-N₄ coordination structures. Yet, their uniform electron distribution often limits charge-transfer efficiency and the tunable adsorption of reaction intermediates, leading to relatively high energy barriers and compromised catalytic performance^[10−11]. To address this, researchers have introduced heteroatoms like B, O, S and P to construct asymmetric M-N_xY coordination environments. This strategy effectively modulates the electronic properties of the active center, optimizes intermediate adsorption, and thereby significantly enhances both catalytic activity and stability^[12−17]. Nevertheless, systematic studies on how different coordination structures and active metal centers synergistically promote the EDH reaction remain scarce. This gap hinders a deeper understanding of the structure-performance relationship for graphene-supported SACs in EDH. Therefore, conducting systematic theoretical investigations covering a wide range of coordination environments paired with various active metals is of great importance for the rational design of high-performance EDH catalysts.

The ongoing advancement of computer hardware capabilities and the widespread adoption of high-throughput computational techniques have collectively driven the rapid development of computational materials science. With the sustained promotion of the “Materials Genome Initiative”, a series of efficient high-throughput computing frameworks and material screening workflows have been successfully established^[18−20], leading to the emergence of numerous material databases. However, there is no systematic database specifically dedicated to heteroatom-doped graphene-supported SACs for the EDH reaction.

Constructing a SACs database that covers diverse heteroatom-doped graphene supports and different metal single atoms holds dual significance. On the one hand, this database will provide critical data to uncover novel material properties and to rationalize the design of new catalysts. On the other hand, databases integrating comprehensive information, including atomic structures, electronic properties and computational parameters, will greatly facilitate data sharing and reuse within the scientific community. This would effectively minimize redundant computational efforts and significantly conserve valuable computational resources.

Therefore, establishing a specialized database for heteroatom-doped graphene-supported SACs will provide a solid theoretical foundation for the screening and preparation of high-efficiency EDH catalysts. Furthermore, this database can also serve as a representative model of doped material databases. The open sharing of its data is expected to advance research in EDH catalysis and related scientific fields significantly.

This database is centered on the study of heteroatom-doped graphene-supported SACs for the EDH reaction. It systematically integrates parameter data, structural information, software environment details and physical property data. All calculation files are organized by computation type and structured within a unified, standardized directory hierarchy. Using Pt SACs supported on graphene with the -N₃B coordination structure (Pt-N₃B SAC) as an example, the file storage structure is illustrated (see Figures 2−4 for details). All data files follow a consistent two-level naming convention: Level-1 folders are named by metal type (Cr SACs, Ni SACs, Pd SACs, Pt SACs and V SACs), and Level-2 folders are named by coordination structure type (A₃B, A₂B₂c-a, A₂B₂c-b, A₂B₂t, AB₃ and A₄). The specific data contents are as follows:

(1) MS Calculation Data (Figure 2)

The file “DFT calculation data of SACs by Materials Studio” contains computational files for 162 stable SACs (across different combinations of coordination structures and metal single-atom), covering: Geometry optimization (slab), Electronic energy (DMol³ Energy) and Density of States (DOS), Structures of adsorbed reactants, intermediates, and products (adsorption and co-adsorption), Transition-state search (TS) results. Additionally, it includes files for 93 unstable SACs (across different combinations of coordination structures and metal single-atom), covering: Geometry optimization (slab), Electronic energy (DMol³ Energy) results. Here, unstable SACs refer to systems where the binding energy of the metal single-atom (E_b=E_M/sub−(E_M+E_sub)) is lower (more negative) than the corresponding metal’s cohesive energy (E_coh-M). E_M/sub is the total energy of the metal-single-atom-on-graphene system; E_M is the energy of an isolated metal single-atom, and E_sub is the energy of the graphene support.

(2) VASP frequency analysis data (Figure 3)

The file “Frequency analysis data of SACs by VASP” contains frequency analysis results for six elementary reactions on the 162 stable SAC structures: C₂H₆→C₂H₅+H, C₂H₅→CHCH₃+H, C₂H₅→CH₃+CH₂, C₂H₅→C₂H₄+H, C₂H₄→CHCH₂+H and C₂H₄→CH₂+CH₂. The data for each reaction step includes the structures for the initial state (IS), transition state (TS) and final state (FS).

(3) VASP Bader charge data (Figure 4)

The file “Bader charge data of SACs by VASP” contains the Bader charge calculation results, along with the corresponding processed data, for all 162 stable SAC structures.

Throughout the development of this database, rigorous convergence tests were conducted on the key parameters that influence computational outcomes—total energy, maximum atomic displacement and maximum force—ensuring that all computational parameters meet strict convergence criteria. These tests ensured that all computational settings met stringent convergence criteria, thereby guaranteeing the stability and reliability of the results, with a typical energy error of ±0.15 eV.

All calculations were performed using first-principles methods within the DFT framework. This theoretical approach offers significant advantages for analyzing material structural stability, evaluating catalytic performance, and interpreting electronic structure. It can accurately describe experimental phenomena and provide profound theoretical insights into reaction mechanisms.

The complete computational source files provided in this database hold significant value for in-depth structural analysis and subsequent research. Researchers can directly access the original files for data interpretation or secondary processing to validate existing conclusions or extract specific physicochemical parameters. Moreover, the standardized and comprehensive calculation files can serve as a basis for further studies, facilitating tasks such as transition-state searches and electronic structure analysis.

Data users can also directly use the pre-optimized structure files in the database to quickly initiate simulations of related properties, substantially improving research efficiency. This database systematically integrates 255 heteroatom-doped graphene-supported SACs, comprising 51 distinct coordination environments with five different metal single atoms. This well-structured dataset can provide reliable structural references and a solid theoretical foundation for future theoretical calculations and experimental explorations.

The files in this database can be directly downloaded and are formatted for compatibility with both MS and VASP software. Due to copyright restrictions on pseudopotential files, they are not included in the distributed package.

All first-principles calculation data presented in this paper—including optimized structures, electronic energies, DOS, Bader charges and frequency analysis results—are stored in the Science Data Bank (SDB) hosted by the Computer Network Information Center, Chinese Academy of Sciences. The dataset is titled “A dataset of heteroatom-doped graphene-supported SACs for EDH” and is assigned the following persistent identifiers: Database DOI:https://doi.org/10.57760/sciencedb.32498; Database CSTR:https://cstr.cn/31253.11.sciencedb.32498. The dataset is available for access and download upon request via the SDB platform, and may be used for subsequent non-commercial scientific research.

Example 1 (Exploration of Structure-Performance Relationships): Researchers can use this database to systematically extract Bader charges or d-band centers (which can be computed from the provided DOS files) for the active metal centers across all catalyst systems. These electronic structure descriptors can then be correlated with the overall energy barriers for EDH reaction, helping to reveal universal structure-performance relationships within heteroatom-doped catalytic systems.

Example 2 (Screening of High-Performance Catalysts): By screening the database for systems that exhibit both low overall energy barriers for the EDH reaction and a thermodynamic preference for ethylene formation, researchers can rapidly identify several potential high catalytic activity, ethylene selectivity candidate catalysts. For instance, the Pt-N₃B SAC emerges from such an analysis, providing a clear and promising target for subsequent experimental synthesis and validation.

Through systematic first-principles calculations, this study constructed a comprehensive dataset of heteroatom-doped graphene-supported SACs for the EDH reaction. It systematically reveals variations in catalytic performance among different heteroatom-doped structures across individual elementary reaction steps. Based on electronic structure analyses involving DOS and Bader charge calculations, the physical mechanisms underlying these performance differences were elucidated at the electronic level, providing a theoretical foundation and screening criteria for the rational design of efficient EDH catalysts. This database not only presents trends in reaction energies but also establishes a structure-performance relationship characterized by “dopant type → electronic state of active metal center → catalytic activity” through its electronic structure data. This deepens the atomic-scale understanding of the EDH reaction mechanism on SACs. The research demonstrates how high-throughput calculations can generate well-structured, information-rich, and mechanistically explicit domain-specific data. To advance data-driven research paradigms, we plan to compile lightweight feature extraction code in future work, enabling the direct export of feature-value tables in .csv format to support the training of various machine-learning models. Furthermore, we will carry out experimental validation to continuously enrich and enhance the content and reliability of this scientific database. We firmly believe that the value of data lies in its circulation and sharing. Therefore, we fully embrace data openness principles to promote the development of a new data-driven paradigm in the EDH research field.