Field-tuned quasiparticles and electronic structure in heavy-fermion YbNi4P2

The gist: The study investigates Fermi surface topology and quasiparticle properties in heavy-fermion YbNi4P2 at high magnetic fields using quantum oscillation measurements, revealing a large decrease of the quasiparticle mass with increasing field and an abrupt change of electronic structure at 17 T

Introduction to Heavy Fermions and Tuning Parameters

Heavy-fermion compounds respond strongly to external stimuli because of the reduced Fermi energy scale ≈ 1 meV to 10 meV of the composite quasiparticles forming from the hybridisation of local moments and conduction electrons Electronic quasiparticles with enhanced mass and strong renormalisation of other properties are formed in heavy-fermion compounds The volume enclosed by the Fermi-surface increases in heavy-fermion compounds forming the so-called “large” Fermi surface when the degrees of freedom of the local moments contribute to the Fermi volume Prominent frequencies include the highest fundamental frequencies F+α = 5.5 kT and F+α = 4.1 kT at intermediate polar angles 20◦ ≤ θ ≤ 60◦ <ref:2608.281194, Fβ = 1.6 kT for fields along [100] extending to polar angles θ ≥ 50◦ (orange traces in Figure 1(a) and (b)) and Fγ = 0.5 kT and Fδ = 0.4 kT for fields along [001] extending to polar angles θ ≤ 65◦ <ref:2608.28119#pg5> The frequency of the oscillations is proportional to the cross-sectional area of extremal orbits perpendicular to the magnetic field

De-renormalisation and Mass Evolution

With increasing magnetic field, a continuous decrease of the quasiparticle effective mass, m⋆ in YbNi4P2 is observed Below B9, the continuous decrease of m⋆ shows directly that the quasiparticles become continuously de-renormalised as the Kondo singlet formation is suppressed in high magnetic fields up to 17.5 T, i.e. up to the LT at B9 <ref:2608.281195, Beyond 17 T, i.e. beyond the LT at B9, we find different masses for the spin-split branches of F±γ at m⋆γ+ = 4.9(5) me and m⋆γ− = 3.5(6) me <ref:2608.28119#pg6> The decrease of m⋆γ below B9 reflects the bulk behaviour of all electronic states as it matches the evolution of the Sommerfeld coefficient, γS

Abrupt Change at B9 and Frequency Splitting

At B9, a drastic change of quantum oscillation frequencies is observed Instead, new frequencies F+γ = 0.7 kT, F−γ = 0.6 kT, F+δ = 0.34 kT, and F−δ = 0.23 kT are detected in the FFT for fields B > B9 <ref:2608.28119#pg6> This abrupt change of quantum oscillation frequencies suggests a sudden change of the quasiparticle properties and/or Fermi surface as discussed below> The splitting of the quantum-oscillation frequencies is linked to the magnetic anisotropy of YbNi4P2

Models for Frequency Splitting

The observed change of frequencies at B9 can be modeled by two scenarios: A) sudden jumps of FA(B) due to a reconstruction of the Fermi surface and B) a sudden change of slope dFA/dB ∝ m⋆g linked to the product of effective mass and g-factor Scenario B suggests that an abrupt change in the electronic g factor from the magnitude of the frequency splitting <ref:2608.281195, The change in slope at B9 is set by the change of the product of effective mass and g-factor, i.e. ∆F = F+γ −F−γ ∝ ∆(m⋆g) <ref:2608.28119#pg6>

Fermi Surface Topology

The resulting Fermi surface (Figure 8) comprises three major sheets: Two flat sheets (yellow) perpendicular to the [001] direction are located half way between Γ and Z in agreement with earlier DFT [13] and ARPES studies [43] The fact that Fβ is observed over the full rotation between [100] and [110] proves that the doughnut has no necks along the X point

Conclusion

Our combined quantum oscillation and DFT study provides evidence that the Fermi surface as shown in Figure 8 is indeed realised in YbNi4P2 in the heavy-fermion state with small modifications including a decreased inner tube inside the doughnut

Experimental Details and Analysis

Samples were synthesised using the Czochralski method as described previously [10] <ref:2608.281193, Measurements were conducted at the high-field laboratory, HFML, at fields up to 35 T in a 3He cryostat with a base temperature of ≈ 0.3 K and at the University of Bristol in a 3He/4He dilution refrigerator in fields up to 12 T <ref:2608.28119#pg10> Effective masses, m⋆ have been determined from Lifshitz-Kosevich fits to the temperature dependence of the amplitude determined from Fourier transforms (cf. Figure S2)

Additional Information on Models

The difference between the back-projected frequencies is given by ∆Fbp = F+bp − F−bp = 2B9(a1 − a2) <ref:2608.281195, We estimate the change in g-factor to be ∆g ≈ 0.5(2) <ref:2608.281195, This suggests that the frequency splitting is linked to the magnetic anisotropy of YbNi4P2 <ref:2608.28119#pg6>

Computational Details

We approximate the band structure of YbNi4P2 by confining the f electron to the core states <ref:2608.

1. Bold header: High-Field Electronic Structure Modeling

AI systems can now accurately model the evolution of the Fermi surface and quasiparticle properties in high magnetic fields by incorporating field-dependent changes in effective mass and Fermi surface topology, as demonstrated by the study's ability to compare experimental results with theoretical models like the single-ion Kondo model.

2. Bold header: Lifshitz Transition Prediction

Improved AI can predict abrupt changes in electronic structure by analyzing a sudden change of quantum oscillation frequencies suggesting an abrupt change of the electronic structure and/or quasiparticle characteristics at putative Lifshitz transitions, specifically identifying the field dependence where the LTs shift to larger fields as the magnetic field is rotated away from the magnetically easy direction [001].

3. Bold header: Quasiparticle Mass Renormalization Tracking

AI can track a large decrease of the quasiparticle mass with increasing field by using quantum oscillation measurements as a direct probe of specific orbits on the Fermi surface and their effective masses, providing quantitative agreement with theoretical predictions like those from Renormalised Perturbation Theory (RPT) for the evolution of the renormalised effective mass.

4. Bold header: Spin-Splitting Mechanism Identification

The system can differentiate between competing physical mechanisms, such as A sudden jump of FA(B) due to a reconstruction of the Fermi surface versus an Abrupt change of g-factor, by analyzing the asymmetry in split frequencies and relating it to whether the change is proportional to the product of effective mass and g-factor.

5. Bold header: Fermi Surface Topology Mapping

AI can construct a detailed, high-resolution map of the Fermi surface topology (e.g., doughnut-shaped Fermi surface, jungle-gym Fermi surface) by correlating the angular dependence of observed quantum oscillation frequencies with DFT calculations, such as those shown in Figure 8.

6. Bold header: Anisotropy Analysis for Transition Drivers

The system can analyze the complex relationship between magnetic anisotropy and frequency splitting, noting that the change in quantum oscillation frequencies is tied to the transition at B9 and follows the anisotropy of B9, thereby identifying that Large spin splitting is thus linked to the strongly correlated state of YbNi4P2.

We study the Fermi surface topology and quasiparticle properties in the heavy fermion compound YbNi4P2 at high magnetic fields using quantum oscillation measurements. We observe a large decrease of the quasiparticle mass with increasing field and demonstrate good qualitative agreement with the single-ion Kondo model. At the putative Lifshitz transition at 17 T, we observe a sudden change of quantum oscillation frequencies suggesting an abrupt change of the electronic structure and/or quasiparticle characteristics. Our results demonstrate the ability to tune the electronic structure and provide input for theoretical models of YbNi4P2 and correlated electron systems in high magnetic fields.