A thermodynamic reference for field-induced phenomena in alpha-RuCl 3: two coexisting polymorphs account for multiple magnetic phase transitions

Magnetic fields in monoclinic α-RuCl3 reveal rhombohedral inclusions underlying apparent oscillations. This study establishes the intrinsic magnetic phase diagram of monoclinic (C2/m) RuCl3 by resolving multiple magnetic transitions arising from the coexistence of two structural polymorphs, providing a framework to interpret previously reported oscillatory features in thermal transport.

Structural Characterization and Phase Diagram Comparison

The research focuses on clarifying the crystal structure of α-RuCl3 as a function of temperature and sample size, which is crucial for interpreting magnetic properties. The authors demonstrate that while the monoclinic structure can be isolated in nanogram-scale crystals, mapping the AFM phase boundary reveals that it closely resembles that of rhombohedral crystals, but is systematically shifted to higher transition temperatures and critical fields. This comparison highlights a key finding: features once attributed to a QSL are instead symptomatic of an incomplete transition from the high-temperature monoclinic structure to the low-temperature rhombohedral structure.

Resolution of Multiple Magnetic Transitions

The core contribution lies in using high-resolution magnetotropic susceptibility measurements to map the AFM phase boundary across 12 crystallographic planes. For specific field orientations, such as Ba, the study observes a two-step suppression of AFM order, indicating an intermediate ordered phase analogous to the ZZ2 phase reported in rhombohedral samples. Furthermore, when rotating magnetic fields within the ac-plane at high fields (14 T), four transitions are observed as sharp minima in k, where two correspond to the monoclinic (C2/m) phase and two are attributed to a minor rhombohedral (R¯3) contribution. These additional features, labeled B˜c1 and B˜c2, are explicitly linked to the ZZ1 and ZZ2 transitions of the coexistence of rhombohedral (R¯3) domains.

Origin of Oscillatory Signatures

The paper provides a mechanism for understanding oscillatory features previously reported in thermal transport experiments. The authors propose that these oscillations arise not from a distinct new phase, but rather from the interplay of monoclinic and rhombohedral stacking environments, where the coexistence of these closely-spaced instabilities produces a cascade of thermodynamic anomalies that manifests as oscillatory behavior in other angle-dependent measurements. This suggests that the observed features are linked directly to magnetic phase boundaries rather than emergent spinon Fermi surfaces.

Implications for Quantum Magnetism

The findings suggest that structural inhomogeneity, specifically the coexistence of monoclinic and rhombohedral domains, is a dominant factor in interpreting field-induced phenomena in α-RuCl3 and related two-dimensional quantum magnets. The work clarifies why experiments performed on crystals containing mixtures of C2/m and R¯3 stacking often report oscillations over a broader range of field directions; these measurements effectively average over domains with different stacking orientations, bringing multiple phase boundaries into the same field window. This provides a natural explanation for oscillatory features by assigning each anomaly to a well-defined thermodynamic phase transition associated with either the C2/m structure or the residual R¯3 fraction.

Conclusion and Future Directions

The study establishes that monoclinic crystals host essentially the same microscopic exchange interactions as the rhombohedral phase, with only a uniform enhancement of the magnetic energy scales. By establishing this intrinsic phase diagram, the work provides a clear thermodynamic reference against which previously-reported anomalies can be understood, suggesting that future experiments controlling distortions may offer a route to identify conditions under which a Kitaev spin liquid may emerge. The results challenge the perception that monoclinic crystals represent lower-quality realizations of RuCl3, instead showing they possess the same underlying magnetic physics with systematically enhanced energy scales.

The gist: The intrinsic magnetic phase diagram of monoclinic (C2/m) α-RuCl3 is determined by resolving multiple magnetic transitions arising from the coexistence of two structural polymorphs, which provides a framework to interpret previously reported oscillatory features in thermal transport.

How it works

1. The authors map the AFM phase boundary across 12 crystallographic planes at low temperature (T = 1.5 K) and high magnetic fields (up to 14 T) using resonant torsion magnetometry (RTM).

2. They observe that the AFM transition in the monoclinic structure is systematically shifted compared to rhombohedral crystals, suggesting unequal Ru–Ru bond lengths in the C2/m structure partially relieve exchange frustration.

3. For specific field orientations, such as Ba, a two-step suppression of AFM order is observed, indicating an intermediate ordered phase analogous to the ZZ2 phase found in rhombohedral samples.

Here are the specific improvements that can be made to AI systems by leveraging the insights from this scientific paper, along with what those improved AI systems could achieve:

) Specific Improvements for AI Systems:

1. Predictive Phase Diagram Generation for Complex Materials: Integrate the derived phase boundary relationships (e.g., how monoclinic vs. rhombohedral domains shift transition temperatures and critical fields) into machine learning models trained on material structure/composition data.

2. Structural Inhomogeneity Signature Detection: Develop deep learning classifiers capable of analyzing experimental spectroscopic or scattering data (like X-ray diffraction RSMs or diffuse scattering patterns) to automatically identify the presence, size, and orientation of minor structural inclusions (e.g., rhombohedral domains within a monoclinic matrix).

3. Anisotropy-Aware Field Response Modeling: Build advanced simulation models that incorporate the highly anisotropic magnetic Hamiltonian derived from structural distortions (bond length variations) to accurately predict field-induced transitions in materials like RuCl3, moving beyond isotropic approximations.

4. Feature Attribution for Quantum Signatures: Create an AI framework that can map specific experimental anomalies (e.g., oscillatory features in thermal transport) directly back to the underlying physical cause—distinguishing between intrinsic quantum spin liquid signatures and artifacts arising from structural phase coexistence or domain averaging.

5. Cross-Modal Data Fusion for Spin-Liquid Hunting: Implement multimodal AI systems that fuse magnetic susceptibility data, thermal transport measurements, and structural characterization (XRD/VSM) to provide a holistic assessment of candidate materials, allowing the system to diagnose whether observed phenomena are due to intrinsic quantum phases or material complexities.

) Capabilities of the Improved AI System:

1. Automated Material Screening for QSL Candidates: The improved system can rapidly screen vast libraries of candidate 2D quantum magnets by predicting their most likely structural symmetry (C2/m vs. R3) and quantifying the expected shifts in magnetic transition temperatures based on predicted bond-length variations, drastically reducing experimental screening time.

2. High-Fidelity Simulation of Field Effects: The system can simulate complex field-orientation dependence for anisotropic materials, allowing researchers to predict whether a specific field configuration will induce a single transition or multiple transitions (like the observed Bc1/Bc2 and B˜c1/B˜c2 features), guiding experimental design toward optimal measurement angles.

3. Automated Interpretation of Ambiguous Experimental Data: When presented with complex, noisy data sets exhibiting multiple competing signatures (e.g., oscillatory thermal transport near an AFM boundary), the AI can autonomously attribute these features to their most probable physical origin—either a true spin-liquid signature or a manifestation of structural domain averaging/coexistence.

4. Diagnosis of Structural Disorder from Scattering Data: The system can analyze reciprocal space maps (RSMs) and definitively distinguish between ideal lattice structures and those containing small twin domains or stacking faults by analyzing the projection-dependent behavior of diffraction peaks, providing quantitative measures of material quality relevant to magnetic ordering.

5. Discovery of Hidden Phase Boundaries: By learning the systematic relationship between structural polymorphs (monoclinic/rhombohedral) and magnetic phase boundaries, the AI can predict where hidden transition lines exist in a material that may be obscured by sample inhomogeneity, potentially leading to the discovery of previously unobserved intermediate ordered phases (like the X phase).

The majority of research on α-RuCl 3 has focused on applying in-plane magnetic fields to suppress the antiferromagnetic order and induce a quantum spin liquid (QSL). However, interpreting field-induced phenomena has been complicated by the material's temperature- and sample-size-dependent crystal structure and its sensitivity to strain-induced stacking disorder. Here, we isolate the monoclinic (C2/m) phase in nanogram-scale single crystals, providing the opportunity to study Kitaev physics in a new setting. We perform high-resolution magnetotropic susceptibility measurements on a structurally well-defined crystal within several crystal planes. Mapping the antiferromagnetic (AFM) phase boundary as a function of temperature, magnetic field, and field orientation, we find that the monoclinic phase diagram closely resembles that of rhombohedral crystals, but is systematically shifted to higher transition temperatures and critical fields. We find that the critical fields coincide quantitatively with anomalies observed across the about 7-11 T field range, where Kitaev spin liquid physics has been widely reported across numerous independent studies and experimental techniques. Our study shows that these anomalies arise from the magnetic phase boundaries of coexisting monoclinic and rhombohedral structural domains, rather than an exotic phase. These results provide a thermodynamic reference against which past studies can determine their structural composition, and toward which future studies can look for genuine signatures of spin-liquid phenomena beyond the close vicinity of the AFM boundary under in-