UCd$_{11}$: A strongly localized 5$f^3$ material

Abstract: UCd$<em>{11}$ is an antiferromagnetic uranium intermetallic compound ($T</em>{\rm N}$ = 5.3K) with enhanced electron mass and uranium-uranium spacings nearly twice the Hill limit, suggesting a weakly hybridized 5$f$ electronic character. Various x-ray spectroscopy techniques indicate that uranium in UCd$<em>{11}$ adopts the formal U$<sup>{3+}$ 5$f<sup>3$ configuration, while core-level photoemission spectroscopy (PES) data of UCd$</em>{11}$ reveal only a weak satellite feature, typically interpreted as a signature of itinerancy. In this work, we present density functional theory (DFT) combined with dynamical mean-field theory (DMFT) calculations of UCd$<em>{11}$, using material-specific parameters tuned to reproduce valence-band PES spectra at different photon energies, thereby exploiting the energy dependence of photoionization cross sections. Our results demonstrate that UCd$</em>{11}$ is a highly localized uranium 5$f<sup>3$ system. Furthermore, core-level spectra obtained from a DFT+DMFT Anderson impurity model reveal that, contrary to common assumptions, the presence or absence of satellite structures is not a reliable indicator of strong correlations or itinerant 5$f$ behavior.

• The paper identifies a predominant 5f3 configuration in UCd11 using multi-modal x-ray spectroscopy and theoretical modeling.
• It employs DFT+DMFT methods with fine-tuned double-counting corrections to capture localized electronic behavior.
• Results resolve discrepancies in U 4f core-level PES data, establishing new criteria for characterizing complex actinide intermetallics.

UCd$_{11}$: A Strongly Localized 5$f^3$ Uranium Intermetallic—Electronic Structure, Spectroscopic Characterization, and Theoretical Modeling

Introduction and Context

Investigations of uranium intermetallics have consistently revealed the intricate competition between localized and itinerant 5$f$ electron behavior. This duality underlies emergent phenomena such as unconventional superconductivity and magnetism. UCd$_{11}$, an antiferromagnetic (AFM) uranium compound with $T_N = 5.3$ K, exemplifies an extreme in this landscape, featuring U–U separations ($d_{UU} = 6.56$ Å) far exceeding the Hill limit and thus severely suppressed direct 5$f$ orbital overlap. The material's macroscopic properties—enhanced effective electron masses, a large specific heat coefficient, and Curie-Weiss susceptibility compatible with a Kramers doublet—implicate strong localization and a predominant U$^{3+}$ 5$f^3$ ground state. However, ambiguity persists due to conflicting signatures from core-level photoemission and x-ray spectroscopies.

Experimental Spectroscopy and Validation of a 5$f^3$ Ground State

A multi-modal spectroscopic approach demonstrates strong evidence for predominant 5$f^3$0 valence in UCd$f^3$1. RXES, PFY-XAS, HERFD, non-resonant and resonant inelastic x-ray scattering (NIXS, VB-RIXS) all display features consistent with U$f^3$2 multiplet calculations (see Fig. 1).

Figure 1: Reported x-ray scattering spectra of UCd$f^3$3, overlaid with spherical full-multiplet calculations based on a U\,5$f^3$4 configuration—demonstrating close correspondence with localized ionic multiplet theory.

Conclusive multiplet excitations support the assignment of UCd$f^3$5 as one of the rare uranium intermetallics dominated by a 5$f^3$6 manifold.

Photoemission Spectroscopy, Configuration Sensitivity, and Double-Counting Correction

Despite the consensus from x-ray scattering, core-level PES and the associated analysis of the U\,4$f^3$7 spectral line shape (Fig. 2) pose challenges for standard interpretation: strong satellites are commonly taken as hallmarks of localized, correlated 5$f^3$8 states, whereas their absence already led to misclassifications of UCd$f^3$9 as "itinerant," similar to the clear band-like UB$f$0.

Figure 2: Integral-type ("Shirley") background-corrected U\,4f core-level PES spectra—UCd$f$1 exhibits a broad main line and only a weak satellite, counter to typical expectations for strong localization.

To unravel this, valence band (VB) PES at both soft and hard x-ray energies were employed, exploiting cross-section variations to disentangle U\,5$f$2 and ligand contributions. Rigorous background subtraction and photon energy dependence (see Fig. 3) reveal sharp U\,5$f$3-derived spectral structures near $f$4.

Figure 3: VB-PES spectra of UCd$f$5 measured at 600 eV and 6000 eV incident energies, providing direct sensitivity to U\,5$f$6 and ligand-derived electronic states.

Consistent with strongly localized uranium, these features are well-captured by DFT+DMFT calculations when the double-counting correction $f$7 is fine-tuned via direct spectral comparison. The theoretical spectra reproduce the experimental line shapes only for $f$8 placing the system deep in the 5$f$9 regime.

Figure 4: Orbital-resolved DFT+DMFT spectral densities for various $_{11}$0 indicating spectral weight transfer driven by 5$_{11}$1 filling and the emergence of the upper Hubbard band upon localization.

DFT+DMFT: Ground-State Configuration, Localization, and Correlation

Applying DFT+DMFT with $_{11}$2 fixed by experiment, the 5$_{11}$3 valence histogram (see Fig. 5) demonstrates a highly monodisperse distribution for UCd$_{11}$4, sharply peaked at 5$_{11}$5 (occupancy $_{11}$6), in contrast to UGa$_{11}$7 (correlated 5$_{11}$8) and UB$_{11}$9 (itinerant 5$T_N = 5.3$0).

Figure 5: (a)-(c) Weights of 5$T_N = 5.3$1 configurations in the ground state, highlighting the almost pure 5$T_N = 5.3$2 character of UCd$T_N = 5.3$3; (d)-(f) show sensitivity of these weights to $T_N = 5.3$4.

Correlation strength and localization are further quantified by the imaginary-time charge correlation function, with UCd$T_N = 5.3$5 displaying the lowest instantaneous charge fluctuation amplitude and slowest decay (Fig. 6).

Figure 6: Logarithm of the charge correlation function $T_N = 5.3$6, indicating suppressed charge fluctuation and strong localization for UCd$T_N = 5.3$7 relative to UGa$T_N = 5.3$8 and UB$T_N = 5.3$9.

Detailed Modeling and Interpretation of Core-Level Spectral Features

High-fidelity DFT+DMFT/AIM calculations show that the broad main emission and weak satellite in the UCd$d_{UU} = 6.56$0 U\,4$d_{UU} = 6.56$1 spectrum are intrinsic and consistent with a strongly localized system (Fig. 7), not a consequence of itinerancy.

Figure 7: Experimental U\,4f core-level spectra of UCd$d_{UU} = 6.56$2 matched with DFT+DMFT/AIM calculations, confirming the origin of line broadening and satellite suppression without invoking itinerancy.

A simplified two-level Anderson impurity model clarifies this "inverted" satellite phenomenology: for a formal $d_{UU} = 6.56$3 ground state, final-state configuration mixing does not generate strong satellites—the intensity and energy window for the shake-up peaks are fundamentally distinct from $d_{UU} = 6.56$4 cases (see Fig. 8).

Figure 8: Two-level model simulation of core-level PES spectra for UCd$d_{UU} = 6.56$5 ($d_{UU} = 6.56$6 regime), UB$d_{UU} = 6.56$7, and UGa$d_{UU} = 6.56$8 ($d_{UU} = 6.56$9 regime), elucidating how satellite position and intensity depend on the occupancy and hybridization.

Thus, the absence of satellites in UCd$f$0 is directly traced to its $f$1 character—a robust theoretical result that resolves the empirical contradiction between PES and other spectroscopies.

Broader Implications and Outlook

This work establishes that UCd$f$2 is an archetypal strongly correlated, strongly localized uranium compound with a dominant $f$3 configuration. It exposes the inadequacy of satellite intensity as a universal proxy for 5$f$4 itinerancy in uranium systems, in strong contrast to the rare-earth series or transition metals. Theoretical treatment using DFT+DMFT, in conjunction with rigorous experimental cross-validation, sets a new standard for the electronic structure determination in actinide intermetallics.

From a practical perspective, these insights refine the criteria for material selection in heavy-fermion, multipolar, or exotic superconducting system searches, where true 5$f$5 localization may be required or expressly avoided. The theoretical framework can be extended toward the description of excited-state processes, ultrafast dynamics, and the role of crystal field splitting in CEF- and hybridization-driven transitions.

Anticipated future developments include further systematic benchmarking of DFT+DMFT parameter choices across the actinide series, explorations of pressure/chemical tuning-induced delocalization transitions, and refinement of many-body core-level solvers to include full multiplet and lifetime effects. This will enable more nuanced interpretations of subtle variations observed in PES and other core-level sensitive techniques.

Conclusion

A combination of advanced x-ray spectroscopy and DFT+DMFT methodologies has unambiguously established UCd$f$6 as a strongly localized uranium $f$7 system. The analysis discredits the simplified correlation between U\,4$f$8 satellite strength and itinerancy, instead highlighting the necessity of full many-body modeling to interpret core-level spectra in actinides. These results solidify the 5$f$9 assignment in UCd$^{3+}$0 and clarify its local-moment antiferromagnetic properties, while providing a reproducible protocol for the future characterization of complex heavy-fermion uranium compounds.