Turning charge-density waves into Cooper pairs

Superconductivity and charge-density waves (CDWs) are symmetry-breaking phases characterized by energy gaps and complex order parameters involving electron pairing (electron-electron for superconductivity, electron-hole for CDWs). Both typically require strong coupling to another degree of freedom, such as phonons, and often display precursor phenomena above their critical temperatures (e.g., pseudogaps and Fermi surface arcs). Conventional BCS theory successfully explains basic superconductors and 1D Peierls CDWs, where phonons and the details of the Fermi surface (fermiology) are central. However, in higher-dimensional and high-Tc systems, the mechanisms remain debated. In cuprates, CDWs are ubiquitous across hole-doped compounds, and in quasi-2D transition metal dichalcogenides (TMDs) such as 2H-NbSe2 and 2H-TaSe2, CDWs coexist with superconductivity. The role of Fermi surface nesting in 2D CDWs remains controversial. Notably, in 2H-NbSe2 nesting is weaker and a saddle-point singularity lies below EF, while in 2H-TaSe2 it lies above EF, suggesting that fermiology is crucial for both CDW and superconductivity. TMDs offer tunable platforms: 2H-NbSe2 exhibits TICDW = 33 K and Tc = 7.3 K, while 2H-TaSe2 shows TICDW = 120 K, a commensurate transition at TCCDW = 90 K, and Tc = 0.133 K. External perturbations (intercalation, pressure, irradiation) strongly modify these transitions. Since Pd intercalation enhances Tc in 2H-PdxTaSe2 by more than an order of magnitude and the pristine electronic structure is well characterized, the present ARPES study targets the intercalated compound. The key questions are how Pd intercalation modifies fermiology, whether CDW is suppressed, and whether a van Hove singularity (vHs) can be tuned to EF to enhance superconductivity.

The study builds on extensive prior work on CDWs and superconductivity in layered materials. Foundational BCS theory and Peierls physics frame the role of phonons and Fermi surface topology. In TMDs, ARPES and neutron scattering studies have characterized CDW transitions (incommensurate and commensurate) and Fermi surface reconstructions in 2H-TaSe2 and 2H-NbSe2, including pseudogap formation and 3×3 folding in the commensurate phase. Prior ARPES identified weaker nesting and a vHs below EF in NbSe2 compared to TaSe2, implicating fermiology in stabilizing respective orders. External tuning (intercalation, pressure, disorder) was shown to alter CDW and superconductivity, with Pd intercalation in TaSe2 yielding a superconducting dome. Broader literature links enhanced superconductivity to strong Fermi-surface singularities (extended saddle points in cuprates, flat bands in iron-based superconductors) and discusses possible roles of CDW fluctuations (competing or aiding pairing) as well as disorder effects in suppressing long-range CDW order.

Angle-resolved photoemission spectroscopy (ARPES) measurements were carried out at the 13-ARPES endstation at BESSY. Samples were cleaved in situ, measurements performed at T = 3 K with photon energies 50–80 eV. Overall energy and momentum resolutions were ~8 meV and ~0.013 Å^-1, respectively. ARPES mapped Fermi surfaces and dispersions along principal Brillouin zone directions to detect band topology, pseudogaps, folding, and the position/nature of van Hove singularities. Sample synthesis: Polycrystalline PdxTaSe2 was synthesized via solid-state reaction. Single crystals were grown by chemical vapor transport using SeCl4 as a transport agent; millimeter-sized hexagonal crystals were collected from the cold end of the quartz tube. Phase purity was confirmed by powder X-ray diffraction with Cu Kα radiation at room temperature. Transport and magnetization: In-plane resistivity ρab(T) and magnetic susceptibility (ZFC/FC, H = 10 Oe) were measured to identify superconducting transitions and check for CDW signatures. Data near optimal Pd content (x ≈ 0.08–0.09) were emphasized and compared to the known phase diagram.

• Pd intercalation suppresses CDW order in 2H-TaSe2 near optimal doping (x ≈ 0.08–0.09). ARPES Fermi surface maps show no signs of either incommensurate or commensurate CDW: no 3×3 reconstruction, no replica bands, and no pseudogap-induced spectral-weight suppression.
• Superconductivity is strongly enhanced: the phase diagram exhibits a dome with optimal Tc at x ≈ 0.08–0.09. For x = 0.08, resistivity shows onset Ton ≈ 3.29 K and zero resistance at Tc ≈ 3.0 K; diamagnetic onset ~3 K with 4πχ reaching ~93% at 1.9 K, indicating nearly full Meissner shielding.
• A Lifshitz transition is observed. The M-centered electron-like "dog-bone" pockets expand and touch along Γ–K, transforming into two hole-like sheets around Γ and K due to movement of a van Hove singularity.
• ARPES reveals the van Hove singularity of saddle-point character tuned to EF (at or slightly below) in 2H-Pd0.08TaSe2. Dispersions perpendicular to Γ–K are hole-like, confirming the saddle nature. In pristine 2H-TaSe2 the vHs lies above EF; in 2H-NbSe2 it lies below EF.
• No CDW signatures appear despite occupation of the saddle points, arguing against a saddle-point nesting mechanism for CDW in this system.
• Comparison among 2H-TaSe2, 2H-Pd0.08TaSe2, and 2H-NbSe2 shows similar qualitative low-energy structures but different bandwidth renormalizations; stronger renormalization and/or higher DOS near EF in NbSe2 may underlie its higher Tc.
• The propensity for an extended van Hove singularity is indicated, suggesting that further electronic-structure tuning could raise Tc even more.

The results directly address the interplay between CDW and superconductivity by showing that electron doping via Pd intercalation shifts the chemical potential, alters Fermi surface geometry, and suppresses nesting that stabilizes CDW in pristine 2H-TaSe2. Increased separation between nesting-prone Fermi surface sections (Γ/K hole barrels and M "dog-bones") destroys CDW order, while the simultaneous tuning of a saddle-point vHs to EF enhances the density of states and promotes superconductivity, consistent with BCS considerations. The absence of CDW despite occupied saddle points disfavors a saddle-point nesting driven CDW in this material. In 2H-NbSe2, where the vHs is below EF without added carriers, CDW persists at lower temperatures and remains incommensurate, highlighting the sensitivity to fermiology rather than doping per se. Disorder introduced by intercalated Pd could additionally suppress long-range CDW, but the dominant, coherent fermiology changes provide a natural mechanism. The findings suggest that small shifts in vHs energy can rapidly modify electronic susceptibility (via low Fermi velocities near the saddle) and tip the balance between CDW and superconductivity. This framework aligns with observations of stress-driven Lifshitz transitions enhancing Tc in Sr2RuO4 and with pressure-enhanced Tc in TaSe2, emphasizing vHs proximity to EF as a key tuning knob. The broader implication is that controlling DOS peaks relative to bosonic pairing spectra (e.g., phonons) can optimize Tc while suppressing competing CDW order; the role of CDW fluctuations may be dual, with long-range order competing but quantum fluctuations potentially aiding pairing and suppressing static CDW.

Pd intercalation in 2H-TaSe2 drives a Lifshitz transition that shifts a saddle-point van Hove singularity to the Fermi level, simultaneously suppressing CDW order and boosting superconductivity by over an order of magnitude (to ~3 K at optimal x ≈ 0.08–0.09). ARPES shows no CDW-related reconstruction or pseudogap at optimal doping and confirms the saddle nature of the vHs at EF. Comparison with 2H-NbSe2 suggests that bandwidth renormalization and DOS near EF further influence Tc. These results highlight fermiology—and specifically the energetic tuning of van Hove singularities—as a crucial parameter governing the competition and proximity of CDW and superconductivity in TMDs. Future work should aim to: (i) systematically tune the vHs (via pressure, strain, fine-tuned intercalation) to map Tc optimization and potential further increases; (ii) directly probe the relevant bosonic spectra (phonons, collective modes) and their evolution with tuning; (iii) quantify the role of disorder and short-range/quantum CDW fluctuations; and (iv) extend similar strategies to related layered superconductors.

The study primarily relies on ARPES, transport, and magnetization near optimal Pd content; a full doping- and temperature-dependent ARPES map would further quantify the evolution of nesting and vHs position. Potential disorder from Pd intercalation could contribute to CDW suppression and is not independently disentangled. The bosonic pairing spectrum (e.g., phonons) and CDW fluctuation dynamics were not measured directly, limiting conclusions about their quantitative roles. Absolute bandwidth renormalization is inferred by comparison to related materials rather than from ab initio–experiment benchmarks for the same samples.