Abstract: Ideal symmetry is known to break down under almost any noise. One measure of asymmetry in a periodic crystal is the relative multiplicity $Z'$ of geometrically non-equivalent units. However, $Z'$ discontinuously changes under almost any displacement of atoms, which can arbitrarily scale up a primitive cell. This discontinuity was recently resolved by a hierarchy of invariant descriptors that continuously change under all small perturbations. We introduce a Continuous Invariant-based Asymmetry (CIA) to quantify (in physically meaningful Angstroms) the deviation of a periodic crystal from a higher symmetry form. Our experiments on several Crystal Structure Prediction datasets show that about a half of simulated crystals have high values of CIA, while all experimental structures in these datasets have CIA=0. On another hand, many crystals with high values $Z'$ in the Cambridge Structural Database (CSD) turned out to be close to more symmetric forms with $Z' \le 1$ due to low values of CIAs.
Introduction: The Fragility of Ideal Symmetry
In crystallography, symmetry is a fundamental property used to classify and understand materials. Traditionally, the relative multiplicity ($Z'$)—the number of geometrically non-equivalent units in a crystal's asymmetric unit—serves as a primary indicator of complexity. Ideally, most crystals prefer a simple $Z'=1$ arrangement.
However, ideal symmetry is mathematically fragile. A single atom nudged by a fraction of an Angstrom can technically strip a crystal of its high-symmetry classification. This causes traditional discrete metrics like $Z'$ to jump discontinuously, even if the physical structure remains nearly identical to the eye. This sensitivity creates significant challenges in computational materials science, where simulations often produce "noisy" structures that are nearly symmetric but technically classified as highly complex.
The Problem of Discontinuous Symmetry
The core issue is that any small displacement of atoms can cause the primitive cell of a crystal to explode in size. As shown in the figure above, a perfectly symmetric grid (center) can be transformed by "tiny noise" into a structure that requires a much larger unit cell (left and right) to describe its periodicity. In these cases, $Z'$ increases from 1 to 2 or 3, despite the underlying geometry being almost unchanged. This is a version of the Sorites paradox: at what point does a slightly distorted "symmetric" crystal become truly asymmetric? Traditional tools cannot distinguish between a crystal that is fundamentally complex and one that is merely a noisy version of a simple form.
Methodology: Continuous Invariant-based Asymmetry (CIA)
To resolve this discontinuity, Majumder et al. introduced the Continuous Invariant-based Asymmetry (CIA). This metric provides a physically meaningful value, measured in Angstroms, that quantifies exactly how far a "noisy" crystal is from a perfectly symmetric state. The methodology relies on three key components:
- Pointwise Distance Distributions (PDD): The authors utilize PDD, an isometry invariant that records distances to the $k$-nearest neighbors for every point in a crystal. To neutralize the effect of crystal density, they subtract an asymptotic curve to create the Pointwise Deviation from Asymptotic (PDA).
- Earth Mover’s Distance (EMD): To compare two geometric blocks (such as two molecules in a unit cell), the algorithm uses EMD. This calculates the minimum "cost" to transform the distance profile of one molecule into another.
- The CIA Metric: CIA is defined as the minimum-maximum distance between these blocks. It identifies the "most central" molecule and measures the geometric deviation of the most distant "equivalent" molecule from it.
Because this foundation is based on continuous distance distributions, a small change in atomic position results in only a small, proportional change in the CIA value, ensuring stability under noise.
Experimental Validation and Insights
The researchers evaluated CIA across two major domains: Crystal Structure Prediction (CSP) datasets and the Cambridge Structural Database (CSD).
CSP Simulations vs. Nature
In CSP datasets containing over 50,000 simulated crystals, the authors found that approximately 50% of simulated structures exhibited high CIA values. In stark contrast, all experimental structures within these same datasets had a CIA of exactly 0. This validates the "Symmetry Principle" in nature: while simulations may explore vast asymmetric landscapes, physical reality strongly favors perfectly symmetric arrangements.
Discoveries in the CSD
Analysis of 69,196 periodic crystals in the CSD revealed that many structures labeled with high $Z'$ values (indicating high complexity) actually possessed very low CIA values. For example, the crystal IYIWIY was recorded with high $Z'$ but had a CIA of 0.000 Å. This suggests such crystals are geometrically identical to $Z' \le 1$ structures and may have been assigned the wrong space group or represent minor experimental distortions rather than true asymmetry.
Conclusion
CIA treats symmetry as a spectrum rather than a binary state. By providing a fast, continuous metric (calculable in seconds even for complex cells), it allows researchers to screen simulated materials more effectively and identify potential errors in experimental databases. While currently optimized for chemically identical blocks, future work aims to extend this metric to co-crystals, further refining our ability to quantify the "almost symmetric" nature of the crystalline world.
Summary updated