The properties of an edge of a transition metal dichalcogenide (TMDC) sensitively depend on its atomic structure; therefore, knowing the exact structure is a precondition for predicting the applications of TMDC, such as in catalysis. Although some edge reconstructions of TMDCs have been reported in previous studies, a global search of TMDC edge structures and calculations of their stabilities are still absent. Here, we propose an approach to explore all possible edge reconstructions of a monolayer TMDC by employing the particle swarm optimization algorithm and first-principles calculations. Taking the most studied TMDC material, MoS₂, as a representative, we have built a database of the edge structures of TMDCs. Including the five experimentally observed edges, the thirty-four most stable edges with various sulfur concentrations are predicted for the first time. The most stable edge structures of 1H–MoS₂ in different environments along both the armchair and zigzag directions have been predicted and agree well with experimental observations. A specific edge of 1H–MoS₂ in the database can be stabilized by modulating the chemical composition of the atoms, which offers an efficient way to tune the properties of TMDC nanostructures, such as TMDC quantum dots and/or nanoribbons.

过渡金属二硫化碳（TMDC）边缘的性质敏感地取决于其原子结构；因此，了解确切的结构是预测TMDC在催化中的应用的先决条件。尽管在以前的研究中已经报道了TMDC的一些边缘重构，但是仍然缺乏对TMDC边缘结构的整体搜索及其稳定性的计算。在这里，我们提出一种方法，通过使用粒子群优化算法和第一性原理计算来探索单层TMDC的所有可能的边缘重构。采用研究最多的TMDC材料MoS ₂作为代表，我们建立了TMDC边缘结构的数据库。包括五个实验观察到的边缘在内，首次预测了具有各种硫浓度的34个最稳定的边缘。已经预测了在不同环境中沿着扶手椅和锯齿形方向的1H–MoS ₂的最稳定边缘结构，并与实验观察结果非常吻合。可以通过调节原子的化学组成来稳定数据库中1H–MoS _2的特定边缘，这为调节TMDC纳米结构（如TMDC量子点和/或纳米带）的性质提供了一种有效的方法。