Fluctuating Pair Density Wave in Finite-temperature Phase Diagram of the $t$-$t^\prime$ Hubbard Model

Using state-of-the-art thermal tensor network methods, this study reveals that while the electron-doped $t$-$t^\prime$ Hubbard model supports a $d$-wave superconducting regime, the hole-doped side is characterized by a finite-temperature phase dominated by fluctuating pair-density-wave states with inter-arc pairing rather than robust superconductivity.

The Gist

Imagine you are trying to understand how a group of people (electrons) behave when they are packed tightly together in a crowded room (a material called a cuprate superconductor). Sometimes, these people can pair up and dance in perfect unison without bumping into each other, creating a state called superconductivity (where electricity flows with zero resistance).

For decades, scientists have been trying to figure out the exact rules of this dance. A famous mathematical model called the Hubbard Model is like the "rulebook" for this crowded room. However, there's a big mystery: the rulebook seems to work perfectly when the room is slightly empty (electron-doped), but it gets confusing when the room is slightly crowded (hole-doped). In the crowded version, experiments show strong dancing (superconductivity), but the math says the dancing should be weak or non-existent.

This paper by Qiaoyi Li, Yang Qi, and Wei Li uses a super-powerful new computer simulation method to finally look at the "rulebook" in action, not just at the very end of the night (absolute zero temperature), but throughout the whole party (finite temperatures).

Here is what they found, explained simply:

1. The Two Different Parties

The researchers simulated the party under two different conditions:

• The "Electron-Doped" Party (Emptying the room): When they removed a few people, the electrons paired up perfectly. They formed a classic, smooth dance called d-wave superconductivity. This matched what we expect and confirmed the rulebook works well here.
• The "Hole-Doped" Party (Crowding the room): When they added a few extra people, things got weird. Instead of a smooth, unified dance, the electrons started doing something chaotic. They didn't form a single, stable dance floor. Instead, they formed fluctuating waves of pairs.

2. The "Pair Density Wave" (PDW) Analogy

To understand the weird behavior on the crowded side, imagine a dance floor where couples are trying to hold hands.

• Normal Superconductivity (dSC): Everyone holds hands with a partner directly across the room. The whole room moves as one giant, synchronized wave.
• The Fluctuating PDW: Imagine the couples are still holding hands, but they are constantly shifting their positions. One moment, a couple is here; the next, they are a few steps away. It's like a "wave" of dancing couples moving back and forth across the floor, but the wave never settles down into a single, stable pattern.

The paper calls this a Pair Density Wave (PDW). It's a state where pairs exist, but they are "wiggling" with a specific rhythm (momentum) rather than sitting still. The researchers found that on the hole-doped side, this "wiggling" is actually stronger than the attempt to form a stable dance.

3. The "Fermi Arc" Mystery

Why does this happen? The authors discovered a difference in the "dance floor" itself.

• On the empty side: The dance floor is a complete circle. Everyone can find a partner opposite them easily.
• On the crowded side: The dance floor is broken. It looks like a C-shape or an arc rather than a full circle.
  • Because the floor is broken, electrons can't find partners directly opposite them. Instead, they try to pair up with partners on the ends of the arc.
  • This "arc-to-arc" pairing forces the pairs to move with a net momentum (the PDW wave), rather than staying still. It's like trying to dance in a hallway instead of a ballroom; you have to keep moving to stay in sync.

4. The Temperature Twist

The most exciting part of this paper is that they looked at the party at different temperatures.

• At the very bottom (Ground State): If you cool the crowded room down enough, the "wiggling" PDW eventually freezes into a static pattern called a Charge Density Wave (CDW). This is like the dancers getting tired and just standing in a grid pattern, no longer dancing.
• At "Warm" Temperatures: But right before they freeze, in the "pseudogap" region (a mysterious state where the material acts like a metal but isn't quite a superconductor yet), the fluctuating PDW is the dominant force.

Why This Matters

For a long time, scientists were confused because the math (Hubbard Model) predicted weak superconductivity for the crowded side, while real experiments showed strong superconductivity.

This paper suggests a new explanation: The math isn't wrong; we just weren't looking at the right thing.
The electrons are pairing up strongly, but they aren't forming the "standard" superconducting dance. They are forming these fluctuating PDW waves. The "standard" superconductivity is weak because the electrons are too busy doing this weird, wiggling PDW dance.

The Takeaway:
The single-band Hubbard model (the rulebook) might be missing a few ingredients to fully explain high-temperature superconductors. It seems that to get the "super" in superconductivity to work perfectly in the crowded room, we might need to add more complex rules (like interactions with the lattice or other bands). However, this study provides a crucial map of the "temperature-doping" landscape, showing us exactly where these strange, wiggling pair waves live, and helping us understand why high-temperature superconductors are so tricky to crack.

In short: The electrons aren't failing to dance; they are just doing a very complicated, wiggly dance that we didn't expect!

Technical Summary

1. Problem Statement

The paper addresses the long-standing discrepancy in understanding high-temperature superconductivity (HTSC) in cuprates, specifically regarding the particle-hole asymmetry between electron-doped and hole-doped regimes.

• The Conflict: While experiments show robust $d$-wave superconductivity ($d$SC) in both regimes, numerical simulations of the single-band $t$-$t'$ Hubbard model have yielded conflicting results. Some methods (e.g., Constrained-Path AFQMC) predict strong $d$SC in hole-doped systems, while others (e.g., DMRG, iPEPS) suggest weak or absent $d$SC in hole-doped systems, often finding competing Charge Density Wave (CDW) orders instead.
• The Gap: Most theoretical studies focus on ground-state ($T=0$) properties. The finite-temperature pairing behavior, particularly the nature of fluctuations in the pseudogap regime and the transition to the ground state, remains largely unexplored due to the computational difficulty of simating fermionic systems at low temperatures without the "sign problem."

2. Methodology

The authors employ state-of-the-art Thermal Tensor Network (ThermoTN) simulations to overcome the limitations of previous methods.

• Model: The single-band $t$-$t'$ Hubbard model on a square lattice with parameters relevant to cuprates: nearest-neighbor hopping $t=1$, next-nearest-neighbor hopping $t' = -0.2$, and on-site repulsion $U=8$.
• Algorithm: They utilize the tangent-space tensor renormalization group (tanTRG) method. This involves representing the thermal density operator $\rho(\beta) = e^{-\beta H}$ as a Matrix Product Operator (MPO) and evolving it in imaginary time using the Time-Dependent Variational Principle (TDVP).
• Geometry & Scaling: Simulations are performed on cylindrical geometries (width $W=4, 6, 8$; length $L=18$) with open boundaries in the $x$-direction and a combination of Periodic (PBC) and Anti-Periodic (APBC) boundary conditions in the $y$-direction to improve momentum resolution.
• Key Technical Achievement: The method allows for bond dimensions up to $D=32,768$, enabling simulations deep into the low-temperature regime ($T/t \approx 0.02$) where Determinant Quantum Monte Carlo (DQMC) fails due to the sign problem. The space-time volume reached ($W \times L \times \beta \approx 4600$) significantly exceeds previous DQMC limits.
• Observables:
  • Imaginary-Time Proxy (ITP): To probe spectral weight near the Fermi level ($\omega=0$), they compute $G(k, \beta/2)$ and pairing correlators $\Phi(q, \beta/2)$. The kernel functions in these imaginary-time quantities act as "low-pass" filters, isolating low-frequency fluctuations that are physically relevant for pairing instabilities, unlike equal-time structure factors which are dominated by high-frequency noise.

3. Key Contributions & Results

A. Asymmetric Phase Diagrams

The study maps out the temperature-doping phase diagram, revealing a stark contrast between electron and hole doping:

• Electron-Doped Side ($\delta < 0$):
  • A robust $d$-wave superconducting ($d$SC) dome emerges near optimal doping ($\delta \approx -1/9$) once Antiferromagnetic (AFM) correlations are suppressed.
  • The pairing is dominated by zero-momentum ($q=0$) fluctuations.
  • The Fermi surface (FS) shows spectral weight concentrated in antinodal regions.
• Hole-Doped Side ($\delta > 0$):
  • No robust $d$SC phase is detected at finite temperatures or in the ground state on the cylinders studied.
  • Instead, a regime dominated by fluctuating Pair Density Wave (PDW) states emerges at intermediate temperatures.
  • Upon further cooling, these PDW fluctuations give way to a Charge Density Wave (CDW) order in the ground state.

B. Discovery of Fluctuating PDW

• Nature of PDW: The hole-doped side exhibits strong pairing fluctuations with a finite momentum $Q_{PDW} \approx (0, \pi)$.
• Strength: The strength of these finite-momentum PDW fluctuations substantially exceeds that of conventional zero-momentum $d$-wave pairing in the hole-doped regime.
• Mechanism: The PDW arises from inter-arc pairing. Unlike the electron-doped side, the hole-doped Fermi surface consists of disconnected Fermi arcs concentrated near the nodal points ($k = (\pm \pi/2, \pm \pi/2)$). The PDW state pairs electrons across these arcs, resulting in a net momentum of $(0, \pi)$.
• Temperature Window: This fluctuating PDW phase occupies the lower portion of the pseudogap regime. It is a distinct intermediate-temperature state that precedes the CDW ground state.

C. Node-Antinode Dichotomy

The paper provides a microscopic explanation for the asymmetry based on the evolution of the Fermi surface:

• Electron Doping: Spectral weight is suppressed at nodes and enhanced at antinodes. This favors conventional opposite-momentum pairing ($k, -k$) at the antinodes, leading to uniform $d$SC.
• Hole Doping: Spectral weight is suppressed at antinodes, leaving only nodal Fermi arcs. This geometry prevents standard $d$SC condensation but facilitates the inter-arc pairing characteristic of the PDW.

4. Significance

• Resolution of Numerical Discrepancies: The results reconcile previous conflicting numerical studies. The "absence" of $d$SC in hole-doped DMRG/iPEPS studies is confirmed, but the paper adds the crucial insight that this is due to the dominance of fluctuating PDW states at finite temperatures, which eventually freeze into CDW order.
• Pseudogap Connection: The study links the pseudogap phase directly to the PDW instability. The PDW fluctuations are not just a competing order but are intrinsic to the nodal-antinodal dichotomy of the hole-doped Fermi surface.
• Limitations of Single-Band Models: The authors conclude that the single-band $t$-$t'$ Hubbard model alone is insufficient to capture the full phenomenology of hole-doped cuprates (which exhibit high-$T_c$ superconductivity). The model naturally favors CDW/PDW over $d$SC in the hole-doped regime, suggesting that extensions (e.g., multi-band models, electron-phonon coupling, or density-assisted hopping) are necessary to stabilize the observed high-$T_c$ superconductivity.
• Methodological Benchmark: The work demonstrates the power of thermal tensor networks in accessing low-temperature fermionic physics, providing a benchmark for future studies of correlated electron systems where the sign problem is prohibitive.

In summary, the paper establishes that in the $t$-$t'$ Hubbard model, hole-doping leads to a fluctuating Pair Density Wave state driven by Fermi arc topology, which competes with and ultimately suppresses $d$-wave superconductivity, offering a new perspective on the microscopic origins of the pseudogap and the challenges in modeling high-$T_c$ superconductors.