Wyckoff-Resolved Oxidation-State Atlas and Anion-Conditioned Priors for Materials Discovery

In the world of materials science, researchers are constantly trying to predict which combinations of atoms will form stable, useful crystals. These materials could be the next generation of battery components, superconductors, or solar cells. To do this, they rely on a fundamental rule of chemistry: charge neutrality. Just as a bank account must balance, the total positive and negative electrical charges within a crystal must cancel each other out to zero. While this seems simple, nature is often messy. A single element, like iron, can exist in different electrical states within the same crystal, or it might behave differently depending on exactly where it sits in the atomic structure. Traditional computer tools often struggle with this complexity, usually assigning a single, average electrical state to an element across an entire material. This works for simple cases but fails when a material requires different electrical states for the same element in different locations, leading to missed discoveries or incorrect predictions.

A researcher at New Mexico State University has developed a new, more nuanced way to handle this problem, creating a massive reference guide that respects the specific locations of atoms within a crystal. Instead of treating a material as a simple list of ingredients, this new approach looks at the crystal's internal symmetry and the specific "seats" atoms occupy, known as Wyckoff sites. By analyzing over 154,000 known crystal structures from a major scientific database, the team built a statistical atlas that maps out how often elements appear in specific electrical states at these specific sites. The result is a powerful tool that can assign electrical charges to atoms with much greater accuracy than previous methods, successfully identifying valid charge-balanced structures for over 114,000 materials, including many that were previously impossible to solve.

The core of this work is the creation of a "Wyckoff-resolved oxidation-state atlas." In chemistry, an oxidation state is a number that represents the electrical charge an atom carries in a compound. For decades, scientists have used standard lists of common charges to guess how atoms behave. However, these lists often miss the subtle variations that occur in real crystals. The new atlas goes beyond simple lists by using a two-step process to scan through known structures. First, it tries to solve the charge puzzle using only the most common electrical states for each element. If that fails, it expands the search to include less common, but still chemically known, states. This method ensures that the tool prioritizes the most likely scenarios while still having the flexibility to find solutions for more complex materials.

What makes this atlas truly unique is that it does not just look at the chemical formula of a material; it looks at the crystal structure itself. In many crystals, the same element can occupy two different types of atomic positions. For example, in magnetite, a common iron oxide, iron atoms sit in two distinct environments. A traditional tool would assign an average charge to all iron atoms, which is mathematically messy and chemically inaccurate. The new atlas, however, treats each atomic position as a separate variable. It asks, "What charge does the iron in this specific seat need to have to balance the crystal?" This allows it to assign different electrical states to the same element depending on where it is located, a capability that is crucial for understanding complex materials like battery electrodes and magnetic compounds.

The researchers tested their new system against the existing standards used by the Materials Project, a massive database of computed crystal structures. They found that their atlas could assign valid, charge-balanced electrical states to 106,053 materials when using the standard composition-based approach. This number is very close to the total count of materials the database considers possible, suggesting the new tool captures almost all the standard chemistry. However, the real breakthrough came when they used the structural, site-specific approach. By allowing atoms in different positions to have different charges, the system successfully assigned valid states to 114,403 materials. This represents a significant increase in coverage, finding valid solutions for thousands of materials that the standard methods missed.

A detailed analysis of these new discoveries revealed exactly why the structural approach was so successful. Of the 14,665 materials that were only solvable when using the site-specific method, nearly all of them—99.98%—contained at least one element that needed to have different electrical states in different parts of the crystal. This confirms that the ability to distinguish between atomic positions is not just a minor technical improvement, but a fundamental requirement for understanding a vast class of materials. The study also showed that the new tool is highly efficient, finding these solutions quickly without needing to check every single possibility, thanks to the statistical probabilities learned from the atlas.

The team also created specialized versions of this atlas for materials containing oxygen, nitrogen, or sulfur, as these elements often form complex chemical families with their own unique rules. These specialized guides showed that while oxygen almost always carries a specific negative charge, nitrogen and sulfur can vary widely depending on their partners. This level of detail allows researchers to filter out impossible chemical combinations early in the design process, saving time and computational power. The entire workflow, from the data analysis to the final tool, is open and reproducible, allowing other scientists to use the atlas to decorate new crystal structures with accurate electrical charges.

This work provides a bridge between the abstract math of crystal symmetry and the practical needs of materials discovery. By acknowledging that an atom's electrical behavior depends on its specific neighborhood within a crystal, the new atlas offers a more realistic and powerful way to model the solid world. It does not replace the need for detailed quantum mechanical calculations, but it provides a fast, reliable first step that can guide researchers toward the most promising candidates for new technologies. The ability to resolve these electrical states with such precision means that the search for better batteries, more efficient electronics, and stronger materials can proceed with a clearer map of the chemical possibilities.