Sleuthing out exotic quantum spin liquidity in the pyrochlore magnet Ce₂Zr₂O₇

Quantum spin liquids (QSLs) are exotic states of matter with topological order and fractionalized excitations, notoriously difficult to diagnose experimentally. Pyrochlore magnets are a promising platform: classical spin-ice materials exhibit a Coulomb phase, and certain realistic quantum models host U(1) QSLs. The newly discovered pyrochlore Ce₂Zr₂O₇ shows no magnetic order, a neutron-scattering continuum, and no low-temperature specific-heat anomaly, marking it as a strong QSL candidate. Unlike Yb-based systems, Ce³⁺ in Ce₂Zr₂O₇ realizes a dipolar–octupolar doublet where transverse dipolar components are suppressed and the s^z component is octupolar, rendering leading neutron coupling to octupolar fluctuations negligible. The research goal is to determine a microscopic nearest-neighbor Hamiltonian for Ce₂Zr₂O₇ by fitting thermodynamic and neutron-scattering data and to identify the resulting quantum phase. The authors fit experimental specific heat and magnetization using quantum FTLM, refine with neutron structure-factor comparisons (including weak next-nearest-neighbor couplings), and map the obtained parameters onto a known phase diagram to test for a π-flux U(1) octupolar quantum spin ice (π-OQSI).

Prior work established classical spin ice in pyrochlores and signatures of emergent electrodynamics, and proposed U(1) QSLs with gapless photons (Hermele/Fisher/Balents; Gingras/McClarty). Dipolar–octupolar doublets on pyrochlores were shown to host distinct QSLs and symmetry enrichments (Huang/Chen/Hermele; Li/Chen; Yao/Li/Chen). Experimental studies on Ce₂Zr₂O₇ observed continua without order (Gao et al.; Gaudet et al.), spurring QSL proposals. Theoretical phase diagrams for dipolar–octupolar exchange models indicate regions of π-flux quantum spin ice with characteristic gauge flux patterns (Lee/Onoda/Balents; Benton et al.; Patri/Hosoi/Kim). Prior neutron-scattering analyses identified rod-like motifs and pinch points in related systems, while emergent photon signatures were predicted to be visible in dipolar QSI but may be hidden in octupolar cases. Debates about extracting microscopic couplings from first principles persist, motivating data-driven fitting approaches.

• Model: A symmetry-allowed nearest-neighbor pseudospin-1/2 Hamiltonian on the pyrochlore lattice with local [111] axes, H_nn = Σ⟨ij⟩[J_y s_i^y s_j^y + J_x1 s_i^x s_j^x + J_x2 s_i^z s_j^z + J_xz (s_i^x s_j^z + s_i^z s_j^x)], augmented by a weak next-nearest-neighbor coupling J_nnn. Here s^z is octupolar and s^{x,y} are dipolar components in the local frame.
• Parameter extraction, stage 1 (thermodynamics): Fit experimental specific heat C_v(T) and magnetization M(H) (fields along [111], [110]) using quantum finite-temperature Lanczos method (FTLM) on a 16-site cluster (validated on 32-site). This constrains J_y, J_x (effective transverse), J_z, and finds J_xz ≈ 0. Multiple parameter sets within error bars identified.
• Parameter extraction, stage 2 (neutrons): After fixing nearest-neighbor couplings, adjust a weak J_nnn by comparing the static neutron structure factor S(q) from Self-Consistent Gaussian Approximation (SCGA, Large-N with Lagrange multiplier enforcing ⟨S_x^2+S_y^2+S_z^2⟩=1) to inelastic neutron scattering (INS) data, finding J_nnn ≈ −0.005 J_y ≈ −0.5 μeV.
• Dynamical structure factor: Use classical Monte Carlo (MC) to sample initial configurations at T = 0.06 K, then integrate Landau–Lifshitz equations of motion (molecular spin dynamics) with 4th-order Runge–Kutta to obtain S(q, E); average over N_IC = 10 configurations on lattices up to N = 8192 spins. Apply quantum–classical correspondence by rescaling intensity with βE and convolve spectra with a Lorentzian of width Γ to emulate instrumental energy resolution.
• Field coupling: Use a Zeeman term where only the dipolar components couple linearly to H; octupolar moments do not couple at leading order. g-factors fitted with g_z ≈ 2.35 and small g_x ≈ 0.2324 allowed for possible J = 7/2 admixture; g_y ≈ 0.
• Phase identification: Place fitted parameters on published phase diagrams for the dipolar–octupolar model to diagnose the ground state. Use effective U(1) gauge theory with ring-exchange J_ring ~ (J_x + J_z)/64 > 0 favoring π flux per hexagon to interpret low-energy excitations (photons, spinons, visons).
• Representative fitted set used for simulations: J_x ≈ 0.0385 meV, J_y ≈ 0.088 meV, J_z ≈ 0.020 meV, J_xz = 0, J_nnn = −0.0005 meV; g_x ≈ 0.2324, g_z ≈ 2.35.

• Effective exchange parameters from thermodynamics: dominant antiferromagnetic J_y (octupolar–octupolar) with values clustered around J_y = 0.08 ± 0.01 meV (up to ~0.1 meV), subleading J_x ≈ 0.05 ± 0.02 meV and J_z ≈ 0.02 ± 0.01 meV, and vanishing J_xz.
• Next-nearest-neighbor coupling: small ferromagnetic J_nnn ≈ −0.005 J_y ≈ −0.5 μeV is required to reproduce the ring-like S(q) patterns in neutron data.
• Neutron structure factor: The model reproduces energy-integrated ring-like intensity around Γ and enhanced intensity near X points, consistent with INS. Powder-averaged spectra and momentum–energy cuts show broadly matching energy scales across the Brillouin zone.
• Phase diagnosis: All fitted parameter sets lie deep in the π-flux octupolar quantum spin-ice (π-OQSI) regime of the known phase diagram, implying an emergent U(1) gauge field with π flux per hexagonal plaquette.
• Excitations and energy scales: Gapless emergent photons (octupolar, neutron-invisible), gapped spinons (magnetic monopoles) with scale ~J_y ~ 1 K producing a Schottky-like C_v feature, and low-energy visons with scale J_ring ~ 0.01 K likely thermally populated at base temperatures and mixing with photons.
• Octupolar invisibility: Pinch-point features reside in S_zz (octupolar) correlations and drop out of the neutron cross-section due to g_z-only coupling, explaining the absence of conventional photon pinch points in INS despite QSI behavior.
• Field response prediction: For H ∥ [001], the ring-like S(q) weakens at small fields (~0.01 T) and evolves into sharp Bragg peaks at high fields (~4 T), testable by future INS.

The central question—what microscopic Hamiltonian and quantum phase describe Ce₂Zr₂O₇—is addressed by a multi-pronged fit to thermodynamics and neutron scattering. The dominance of antiferromagnetic octupolar coupling J_y drives a classical 2-in/2-out manifold, while finite J_x and J_z generate quantum dynamics favoring a π-flux U(1) QSL. The fitted parameters consistently locate Ce₂Zr₂O₇ in the π-OQSI phase, whose neutron signatures differ from dipolar QSI because octupolar correlations do not couple at leading order to neutrons, accounting for missing pinch points and the observed ring-like intensity (modulated by small J_nnn). The identification of low-energy visons and neutron-invisible photons rationalizes the lack of sharp dispersive modes in INS at current resolutions and suggests very low temperature/large-Q probes or thermodynamic measurements for photon detection. The octupolar character also implies reduced sensitivity to magnetic disorder and fields, potentially enhancing the robustness of the QSL compared to dipolar counterparts.

By quantitatively fitting specific heat, magnetization, and neutron-scattering data with a symmetry-allowed dipolar–octupolar Hamiltonian, the study identifies Ce₂Zr₂O₇ as a strong candidate for a π-flux octupolar U(1) quantum spin ice. The extracted couplings (dominant J_y, subleading J_x and J_z, negligible J_xz, weak J_nnn) reproduce key experimental features and place the system deep in the π-OQSI regime with emergent photons, spinons, and visons. The octupolar nature naturally explains the muted neutron visibility of photon pinch points and suggests enhanced stability against disorder. Future directions include field-dependent INS to test the predicted suppression of ring-like features and emergence of Bragg peaks, high-momentum or higher-order neutron probes to access octupolar photons, ultra-low-temperature thermodynamics to detect photon and vison contributions, and additional measurements to further constrain exchange parameters and g-factors.

• FTLM calculations were performed on finite clusters (16-site, checked against 32-site), leading to finite-size effects, especially in low-temperature specific heat where fits are less accurate.
• The g-factors were taken as effective parameters; high-field magnetization discrepancies may arise from field-dependent g-factors not included.
• The determination of J_nnn is phenomenological via SCGA fits to INS and may be influenced by instrumental resolution and background subtraction uncertainties.
• Classical MD with quantum-classical rescaling approximates quantum dynamics; octupolar photon visibility in neutrons requires higher-order couplings not fully treated.
• Limited experimental data (temperature range, resolution, and field-dependent INS) constrain the precision of parameter extraction.