Intertwined fluctuations and isotope effects in the Hubbard-Holstein model on the square lattice from functional renormalization

Using an advanced functional renormalization group approach that resolves full frequency dependence and self-energy feedback, this study systematically analyzes the Hubbard-Holstein model to reveal how self-energy effects enhance $d$-wave superconductivity at high phonon frequencies while causing a breakdown of Migdal-Eliashberg theory and a reduction of $s$-wave superconductivity at low frequencies, thereby clarifying the complex interplay of intertwined fluctuations and isotope effects.

The Gist

The Big Picture: A Dance Floor with Heavy Music

Imagine a crowded dance floor where people (electrons) are trying to move around. In a normal room, people just bump into each other and push away (repulsion). But in this specific quantum dance hall, there is also a DJ (the crystal lattice/atoms) playing music.

The music isn't just background noise; it physically shakes the floor. When a dancer steps on a spot, the floor vibrates. If another dancer steps there a moment later, they feel that vibration. This creates a weird, delayed attraction between the dancers.

The scientists in this paper are trying to figure out: When do these dancers pair up and waltz together (superconductivity), and when do they just stand in angry groups (magnetism) or form rigid lines (charge order)?

They are studying a model called the Hubbard-Holstein model, which combines two forces:

1. The Push: Electrons hate being in the same spot (Coulomb repulsion).
2. The Pull: Electrons are attracted to each other because of the vibrating floor (electron-phonon interaction).

The Main Characters

1. The Electrons (The Dancers): They want to pair up to become superconductors (zero resistance), but they are also stubborn and repel each other.
2. The Phonons (The Floor Vibrations): Think of these as the "beat" of the music. The heavier the atoms in the floor (the "mass"), the slower the beat.
   • Fast Beat (Light atoms): The floor vibrates quickly.
   • Slow Beat (Heavy atoms): The floor vibrates slowly.
3. The "Isotope Effect": This is the paper's title concept. In the old days, scientists thought that if you made the atoms heavier (slower beat), the dancers would pair up better and the music would last longer (higher temperature for superconductivity). It was like saying, "If we slow the music down, everyone can dance better."

The Big Surprise: The "Heavy" Beat Breaks the Dance

The authors used a powerful new mathematical tool (Functional Renormalization Group) to simulate this dance floor. They didn't just look at the average; they looked at the timing of every single step and vibration.

Here is what they found, which breaks the old rules:

1. The "Slow Motion" Trap (Adiabatic Limit)

When the atoms are very heavy, the floor vibrates very slowly.

• Old Theory: Slower vibrations should help electrons pair up (s-wave superconductivity).
• New Finding: Actually, when the vibrations get too slow, they start acting like a sticky trap. The electrons get "distracted" by the heavy, slow vibrations. Instead of pairing up to dance, they get stuck in a rigid formation (charge density waves).
• The Analogy: Imagine trying to dance a fast waltz, but the floor is made of thick, slow-moving jelly. You don't glide; you sink. The "slow beat" actually kills the dance.

2. The "Heavy" Dancer Effect (d-wave Superconductivity)

In high-temperature superconductors (like cuprates), electrons pair up in a specific shape called "d-wave" (like a four-leaf clover).

• Old Theory: Making the atoms heavier (slowing the beat) should make this d-wave dancing stronger.
• New Finding: The authors found the exact opposite! When they included the "self-energy" (a fancy way of saying "how much the dancer gets tired or slowed down by the floor"), they discovered that heavier atoms actually make the d-wave pairing worse.
• The Analogy: It's like a marathon runner. If the track is soft and heavy (heavy atoms), the runner gets tired faster and slows down. The "tiredness" (self-energy) destroys the runner's ability to sprint (superconduct).

3. The "Feedback Loop" (Intertwined Fluctuations)

The paper shows that the different types of dancing are all connected.

• If the electrons start trying to form rigid lines (charge order), it kills the superconducting dance.
• If the electrons start spinning in a magnetic pattern, it also changes how they pair up.
• The Metaphor: It's like a group of friends trying to decide on a game. If half the group wants to play soccer (magnetism) and the other half wants to play chess (charge order), nobody ends up playing the game they actually wanted (superconductivity). The vibrations of the floor (phonons) tip the scales, often making the "wrong" game win.

Why This Matters

For decades, physicists have been trying to understand how to make better superconductors (materials that conduct electricity with zero loss). The old rule of thumb was: "Make the atoms heavier to slow down the vibrations, and you'll get better superconductivity."

This paper says: "Not so fast."

• The "Self-Energy" is Key: The authors showed that you can't ignore how the electrons themselves get "bruised" or slowed down by the vibrations. When you account for this, the old rules flip upside down.
• No More "Lattice Instability": They also proved that the floor doesn't actually collapse (a lattice instability) just because the vibrations get strong. Instead, the floor just gets softer (phonon softening), which changes the dance dynamics without breaking the building.

The Takeaway

Think of this paper as a new instruction manual for a complex dance. The old manual said, "Slow the music down, and the dancers will pair up."

The new manual says: "If you slow the music down too much, the dancers get tired and sticky, and they stop dancing. Also, the way they get tired depends on the specific shape of their dance steps. To get the best dance, you need to find the perfect speed, not just the slowest one."

This helps scientists understand why some materials are great at superconducting and others aren't, paving the way for designing new materials that can carry electricity without any loss, even at higher temperatures.

Technical Summary

1. Problem Statement

The paper addresses the complex interplay between electron-electron (Coulomb) and electron-phonon interactions in layered quantum materials, specifically modeled by the two-dimensional Hubbard-Holstein model on a square lattice.

• Context: While electron-phonon coupling drives conventional superconductivity (BCS theory), and Coulomb repulsion drives magnetic instabilities and unconventional superconductivity (Hubbard model), their combined effect is non-trivial.
• Gap in Literature: Previous studies often neglected the full frequency dependence of the interaction vertex or the feedback from the electronic self-energy. Many approaches relied on the Migdal-Eliashberg (ME) theory, which assumes weak coupling and neglects vertex corrections and self-energy feedback in certain regimes. There is a need to understand how these interactions compete or cooperate, particularly regarding isotope effects (dependence of physical properties on ion mass $M_{ion}$) and the stability of superconducting phases against charge density wave (CDW) instabilities.
• Specific Challenges: The model exhibits a "fermion sign problem" in Quantum Monte Carlo (QMC) simulations away from half-filling, making numerical exploration of doped regimes difficult.

2. Methodology

The authors employ the Functional Renormalization Group (fRG) method, enhanced by several recent methodological developments to achieve quantitative accuracy beyond standard approximations:

• Single-Boson Exchange (SBE) Formulation: Instead of treating the full two-particle vertex directly, the authors use the SBE decomposition. This splits the vertex into:
  • $\nabla^X$ (Single-boson exchange): Represents the exchange of a collective bosonic mode (spin, charge, or pairing).
  • $M^X$ (Rest functions): Represents multi-boson processes.
  • This decomposition significantly reduces computational complexity by separating the frequency dependence of the bosonic propagators ($w^X$) from the fermion-boson couplings ($\lambda^X$).
• Full Frequency Dependence: Unlike previous fRG studies on this model, this work resolves the full frequency dependence of the two-particle vertex and the self-energy.
• Self-Energy Flow & Katanin Substitution: The flow of the electronic self-energy ($\Sigma$) is explicitly included. The authors utilize the Katanin substitution, which replaces the single-scale propagator with the full derivative of the Green's function. This ensures the fulfillment of Ward identities and allows the method to recover ME theory in the appropriate limit while going beyond it in others.
• Fluctuation Diagnostics: To disentangle the intertwined fluctuations, the authors perform a systematic analysis of the contributions of different channels (spin, charge, pairing) to the physical susceptibilities using diagnostic matrices and bubble decompositions.
• Parameter Space: The study scans a broad range of parameters:
  • Hubbard repulsion ($U$) and electron-phonon coupling ($V_H$).
  • Phonon frequency ($\omega_0$), varying from the adiabatic limit (low $\omega_0$, heavy ions) to the anti-adiabatic limit (high $\omega_0$, light ions).
  • Doping levels: Half-filling (nesting present) and finite doping (Van Hove filling).

3. Key Contributions

1. Extension of SBE to Retarded Interactions: The authors demonstrate that the SBE formalism, previously applied to extended interactions, can be successfully adapted to retarded (frequency-dependent) electron-phonon interactions, providing a computationally efficient yet accurate framework.
2. Resolution of Isotope Effect Contradictions: The study resolves a contradiction between previous fRG results and expectations regarding the isotope effect in $d$-wave superconductivity. They show that self-energy effects are the critical missing ingredient that reverses the sign of the isotope effect.
3. Breakdown of Migdal-Eliashberg Theory: The work identifies a regime where ME theory fails. Specifically, at small phonon frequencies, strong charge fluctuations suppress $s$-wave superconductivity, leading to a non-monotonic behavior of the superconducting susceptibility with respect to coupling strength.
4. Exact Phonon Renormalization: The authors derive an exact expression for the phonon self-energy in terms of the charge density susceptibility. This proves that lattice instabilities (imaginary phonon frequencies) only occur concomitantly with electronic charge-density wave instabilities, correcting previous RPA-based predictions of unphysical lattice melting.

4. Key Results

A. Isotope Effects and Self-Energy

• $d$-wave Superconductivity: In the anti-adiabatic limit, $d$-wave susceptibility increases as phonon frequency decreases (standard expectation). However, as the system approaches the adiabatic limit, the electronic self-energy causes a sign change in the isotope effect ($d\chi_{dSC}/dM_{ion}$).
  • Mechanism: The self-energy dampens coherent quasiparticles. As phonons soften (adiabatic limit), this damping increases, reducing the size of the $d$-wave bubble (the coherent part of the Green's function), thereby suppressing $d$-wave pairing despite enhanced magnetic fluctuations.
• $s$-wave Superconductivity: At small phonon frequencies, increasing electron-phonon coupling initially enhances $s$-wave pairing but eventually leads to a suppression due to the growth of charge-density wave (CDW) fluctuations. This non-monotonic behavior signals the breakdown of ME theory.

B. Interplay of Fluctuations (Fluctuation Diagnostics)

• Adiabatic vs. Anti-Adiabatic:
  • In the anti-adiabatic limit, the system behaves like an effective Hubbard model with a renormalized interaction ($U_{eff} = U - V_H$).
  • In the adiabatic limit, retardation effects become dominant. The analysis reveals that low-frequency phonons invert the role of charge fluctuations:
    • In the pure Hubbard model, charge fluctuations often compete with spin fluctuations.
    • With Holstein phonons, charge fluctuations suppress $s$-wave superconductivity and enhance spin fluctuations (magnetism).
• Diagnostic Matrices: The authors derive analytic diagnostic matrices that predict the signs of contributions from different channels. These matrices successfully explain the numerical observation that charge fluctuations switch from being supportive to destructive for $s$-wave pairing as the system moves toward the adiabatic regime.

C. Phonon Softening and Lattice Instability

• The authors derive an exact relation between the phonon self-energy ($\Xi$) and the charge susceptibility ($\chi_D$).
• Result: The renormalized phonon frequency $\omega_{eff}$ softens (decreases) as coupling increases but never becomes imaginary unless the charge susceptibility diverges (i.e., a CDW instability occurs).
• Significance: This refutes earlier RPA predictions suggesting a lattice instability could occur independently of electronic ordering. There is no "lattice melting" without an accompanying electronic charge order.

5. Significance

• Theoretical Advancement: This work establishes the SBE-fRG approach as a powerful tool for studying retarded interactions in correlated systems, capable of capturing physics beyond Migdal-Eliashberg theory.
• Material Relevance: The findings are crucial for understanding high-temperature superconductors (cuprates, iron-based) and moiré materials where both strong correlations and electron-phonon coupling are present.
• Isotope Effect Prediction: The prediction of a sign change in the isotope effect for $d$-wave superconductivity due to self-energy feedback provides a specific, testable signature for experimental verification in doped systems.
• Methodological Robustness: By proving that lattice instabilities are strictly tied to electronic instabilities in this model, the paper clarifies the conditions under which polaronic or bipolaronic states might form versus charge-ordered states.

In summary, the paper provides a comprehensive, frequency-resolved view of the Hubbard-Holstein model, demonstrating that self-energy feedback and retardation effects fundamentally alter the nature of superconducting instabilities and isotope effects, often in ways that contradict simpler mean-field or static approximations.