Computational method predicts thousands of new exfoliable 2D materials from ordinary crystals

Two-dimensional materials have dazzled researchers ever since graphene was first isolated in 2004, offering electronic, magnetic and chemical behaviors that simply do not exist in their bulky, three-dimensional parent crystals. Yet the hunt for these ultrathin substances has long been constrained to a narrow corner of the materials world: crystals that were already built from loosely bonded layers, which can be peeled apart much like pages of a book. Now a team at the Helmholtz-Zentrum Dresden-Rossendorf (HZDR) and TU Dresden has broken through that constraint, presenting a computational method that can predict separable atomic layers in any type of crystal — and applying it to more than 24,000 known crystal structures to identify 2,531 particularly promising candidates for new two-dimensional materials.

The findings, published in Nature Communications, mark the first time an efficient computational approach has been able to search systematically for exfoliable layers in crystals whose atoms form three-dimensional networks of strong bonds. In such non-layered solids, unlike the well-known van der Waals crystals, it has until now been extremely difficult to predict which materials could be split into stable layers. Dr. Rico Friedrich of HZDR’s Institute of Ion Beam Physics and Materials Research and TU Dresden explains the problem that motivated the work: given how young this class of materials still is, selecting suitable candidates for experiments has often relied on experience and intuition rather than a systematic, physics-based procedure.

At the heart of the new approach lies what the researchers call the eXfoliation and Cleavage Potential, or XCP. This computational model takes the chemical elements involved in a crystal and the distances between them, and from that information calculates very precisely the strength of the bonds between individual atoms. Two additional algorithms then go to work on the resulting picture: one searches for easily separable crystal planes, while the other gradually removes the weakest bonds from the model until two-dimensional substructures remain. Crucially, the method also detects atomically uneven separation surfaces.

Efficiency is the decisive advantage. As lead author Tom Barnowsky points out, calculating the electronic structure for every conceivable surface of a crystal would be incredibly time-consuming. The XCP model, by contrast, can sweep through large materials databases with minimal computational effort, reserving detailed calculations for the particularly promising hits that emerge from the initial screen. This tiered strategy is what made it feasible to analyze tens of thousands of crystal structures in a single study, rather than examining a handful of materials case by case.

The method has not merely been tested on hypothetical examples. Wide-ranging checks against two-dimensional materials that have already been produced experimentally showed that XCP can reliably determine the preferred exfoliable planes. To make the tool useful to the wider community, the team has released it as a freely available web tool at xcp.hzdr.de, allowing interested researchers anywhere to probe crystals of their own choosing for separable layers.

When the team applied the method to more than 24,000 known, non-layered crystal structures drawn from one of the largest experimental materials databases, 2,531 candidates emerged as particularly promising and were selected for more precise calculations. In these materials, the chemical composition remains unchanged upon detachment, while the calculated bonding ratios strongly favor the formation of a two-dimensional structure. Further investigations reinforced the selection’s potential: in most of the candidates the team calculated comparatively low exfoliation energies, meaning relatively little energy would be needed to separate the layers, and many of the structures proved both dynamically and thermodynamically stable — a basic precondition for any technological application.

The researchers are careful to stress what these predictions do and do not establish. The stability calculations identify which crystals are worth pursuing, but the experiments still have to demonstrate whether and under what conditions the predicted materials can actually be produced. As Friedrich summarizes, the results vastly expand the horizon for two-dimensional materials, but the method does not guarantee that every candidate can be made in the lab. What it does show is which crystals are actually worth the experimental effort — a significant improvement over intuition-based selection in a field where the candidate space is enormous.

Perhaps the most striking aspect of the predicted materials is how markedly they differ from the entries already present in existing databases. According to the calculations, some of the candidates might even undergo a fundamental change in their electronic properties during the transition from bulk to thin layer. Among the predicted cases are compounds in which the three-dimensional crystals act as semiconductors but become metallic in two dimensions — a dramatic transformation of electronic character driven purely by dimensional reduction. In other candidates, a direct band gap is created in the two-dimensional form, a property that may be of interest for optoelectronic devices, where direct band gaps are highly valued.

Magnetic candidates also appear among the predictions, and the researchers suggest they could open up new possibilities for spintronics in data processing and storage, a field that seeks to exploit electron spin alongside electronic charge.

For a field that has grown explosively since graphene’s debut, the study offers a practical roadmap for the next phase of discovery. By replacing intuition with a universal, bond-based criterion, the XCP method turns the entire body of known crystal structures into a searchable space of potential two-dimensional materials, complete with promising candidates, calculated exfoliation energies and stability assessments. The next step belongs to experimentalists: synthesizing or exfoliating the most promising of the 2,531 predicted structures and confirming whether the calculated transformations from semiconductor to metal, the newly created direct band gaps and the magnetic candidates behave in the laboratory as the computations predict. If they do, the family of two-dimensional materials could soon expand far beyond the van der Waals crystals that have dominated the field for two decades.

Subject of Research: Computational prediction of exfoliable two-dimensional materials from non-layered crystals

Article Title: From crystal to new 2D materials space

Article References: From crystal to new 2D materials space. Original publication (Bibliographic details unavailable.)

Image Credits: AI Generated

DOI: Not provided

Keywords: two-dimensional materials, graphene, exfoliation, crystal structures, computational materials science, van der Waals crystals, semiconductors, band gap, spintronics, materials databases, Nature Communications, HZDR