Data-driven discovery and rapid, direct synthesis of MXenes

Two-dimensional materials are a family of substances so thin that they are only a few atoms thick, yet they possess remarkable strength and electrical properties that make them valuable for everything from faster computers to more efficient batteries. Among these, a specific group called MXenes has captured the attention of scientists because they are made of layers of metal and carbon or nitrogen that can be peeled apart like pages in a book. For over a decade, researchers have known how to make these materials by starting with a solid block of a related compound and chemically eating away one type of atom to leave the layers behind. However, this traditional method is slow, often requires dangerous chemicals, and has only been used to make a limited number of specific types. The big question facing the field was whether the known list of these materials was truly complete, or if the vast library of scientific history held many more examples that had been made long ago but simply misidentified or forgotten.

A team of researchers has now answered that question by combining modern computer tools with old-fashioned chemistry to uncover a hidden reservoir of these materials. They began by teaching a computer program to recognize the specific atomic fingerprint of MXenes, then asked it to scan through massive databases of crystal structures that scientists have recorded over the last century. The computer found dozens of compounds that fit the description perfectly but had never been labeled as MXenes. The researchers call this collection a "Treasure Chest," consisting of thirty-eight different materials that were synthesized decades ago but overlooked because they were classified under different names. This discovery suggests that the chemical space for these materials is much larger than previously thought, containing many variations that include rare-earth metals and different types of atoms in their core, not just the standard carbon and nitrogen combinations.

To prove that these forgotten materials were real and could be made again, the team turned to a rapid and sustainable manufacturing technique called self-propagating high-temperature synthesis. Instead of using slow, energy-intensive ovens or hazardous liquids, they mixed raw metal powders with other ingredients and ignited the mixture with a brief burst of electricity. Once started, the reaction created its own intense heat and spread through the material like a wave, completing the entire process in just a few minutes without needing any external power source. Using this method, the team successfully recreated five of the materials from their "Treasure Chest" list, confirming that they could be produced quickly and cleanly. They also used the insights gained from this process to create eleven entirely new types of MXenes that had never been made before, all based on rare-earth elements like praseodymium and neodymium.

The results of these experiments revealed that these newly accessible materials have unique physical properties that differ significantly from the standard versions. The researchers found that the new rare-earth-based MXenes act as semiconductors, meaning they can control the flow of electricity in ways that are useful for electronic devices. Furthermore, because these materials contain rare-earth metals, they exhibit diverse magnetic behaviors, with some acting as magnets and others showing different types of magnetic ordering. The team verified these properties by examining the materials under powerful microscopes and measuring how they interact with light and sound. The microscopic images showed that the layers remained distinct and well-ordered, with the atoms arranged exactly as the computer models had predicted.

This work does more than just add new names to a list of materials; it demonstrates a powerful new way to accelerate scientific discovery. By using machine learning to sift through historical records, the researchers showed that the answers to modern problems might already exist in old data, waiting to be recognized with the right tools. They also proved that these materials can be made using a fast, scalable method that avoids the environmental hazards of traditional chemical processing. While the study focused on identifying and making these specific compounds, the approach itself offers a roadmap for finding other hidden materials in the vast archives of human knowledge. The findings suggest that the family of two-dimensional materials is far more extensive and versatile than previously imagined, opening the door to a wider range of applications in energy storage, electronics, and beyond.