Transition-state lattice modes and the breakdown of adiabatic tunneling for hydrogen and deuterium in bcc Nb

This study demonstrates that hydrogen and deuterium tunneling in body-centered-cubic niobium is a fundamentally nonadiabatic, collective process mediated by anharmonic lattice couplings, which can only be accurately described by a five-dimensional lattice-renormalized framework that treats interstitial and transition-state lattice modes on equal quantum footing.

The Gist

The Big Picture: A Quantum Dance in a Metal

Imagine a piece of metal (Niobium) as a giant, crowded dance floor made of heavy atoms. Sometimes, tiny, light particles like Hydrogen or Deuterium get stuck in the gaps between these heavy dancers. Because they are so light, they don't just sit still; they act like ghosts, able to "tunnel" (teleport) from one gap to another without climbing over the walls.

Scientists have long believed that the heavy dance floor stays perfectly still while the tiny ghost particle does its teleporting dance. They thought the floor was just a static stage. This paper says: That assumption is wrong for Hydrogen and Deuterium.

The authors show that for these specific particles, the dance floor doesn't just sit there; it actually moves and wiggles in sync with the particle. The particle and the floor are dancing together as a team, not as separate entities.

The Main Characters

1. The Heavy Dancers (The Lattice): The Niobium atoms. They are heavy and usually move slowly.
2. The Light Ghosts (The Interstitials): Hydrogen (H), Deuterium (D), and a special particle called a positive muon ($\mu^+$).
   • Hydrogen & Deuterium: These are the main stars of this study. They are light, but not too light.
   • The Positive Muon ($\mu^+$): This is a particle that is about 9 times lighter than a proton (Hydrogen nucleus). It's the "ultra-light" version.

The Old Theory vs. The New Discovery

The Old Theory (The "Static Stage" View):
Previously, scientists used a model called "Adiabatic Separation." Imagine a heavy stage and a light acrobat. The theory assumed the stage is so heavy and slow that it doesn't notice the acrobat jumping. The acrobat jumps, and the stage just sits there. This works well for the Positive Muon ($\mu^+$), which is so light it barely disturbs the stage.

The New Discovery (The "Collective Dance" View):
The authors found that for Hydrogen and Deuterium, the stage does move.

• The Analogy: Imagine a trampoline. If a heavy person stands on it, the trampoline sags. If a tiny mouse runs across, the trampoline barely moves. But if a medium-sized cat runs across, the trampoline bounces and warps with the cat.
• The Finding: Hydrogen and Deuterium are like that cat. When they try to tunnel from one spot to another, they pull the surrounding metal atoms with them. The metal atoms distort to help the particle cross the barrier.
• The Result: You cannot calculate how fast they tunnel by looking at the particle alone. You have to calculate the movement of the particle and the specific wiggles of the metal atoms at the same time.

The "Five-Dimensional" Solution

To get the math right, the authors had to stop looking at the problem in 3D (just the particle moving in space). They had to add two extra dimensions representing the specific way the metal atoms wiggle.

• Dimension 1-3: Where the Hydrogen is.
• Dimension 4: How the metal atoms shift to make the two spots look the same (symmetry).
• Dimension 5: How the metal atoms shift to create the "bridge" or "hill" the particle has to cross (the transition state).

By using this 5D model, they were able to predict the exact speed of the tunneling, matching real-world experiments perfectly. The old 3D models failed to get the numbers right.

Why Does Mass Matter?

The paper explains that the "Static Stage" theory only works if the particle is incredibly light (like the Muon).

• Muon ($\mu^+$): It's so light that the metal atoms don't really care. The stage stays still. The old theory works here.
• Hydrogen & Deuterium: They are heavy enough that the metal atoms must move to help them tunnel. If you ignore the metal's movement, your math is wrong.

Why Should We Care? (The "Superconducting Qubit" Connection)

The paper mentions that these tunneling particles are a problem for superconducting qubits (the tiny computers used in quantum computing).

• The Problem: These "ghost" particles in the metal can cause "decoherence," which is like static noise that ruins the computer's memory.
• The Insight: Because the tunneling is a collective dance (particle + metal moving together), the energy levels are different than we thought. This means we might have been looking for the "noise" in the wrong places or with the wrong assumptions. To fix the noise in quantum computers, we need to understand that the metal and the hydrogen are dancing together, not separately.

Summary

• Old Idea: The metal stays still; the particle jumps alone. (True for Muons, False for Hydrogen).
• New Idea: For Hydrogen and Deuterium, the metal moves with the particle. They are a team.
• Proof: Only a complex 5D model that includes the metal's movement can predict the real experimental results.
• Takeaway: To understand how these tiny particles move in metals, you can't treat the metal as a static background. You have to treat the whole system as a single, moving, quantum unit.

Technical Summary

Technical Summary: Transition-state lattice modes and the breakdown of adiabatic tunneling for hydrogen and deuterium in bcc Nb

Problem Statement
Interstitial hydrogen (H) and deuterium (D) in body-centered-cubic (bcc) niobium (Nb) serve as archetypal quantum tunneling systems, exhibiting low-temperature anomalies in thermodynamic and dynamical responses. These systems are increasingly relevant to superconducting technologies, as baths of such tunneling systems (TS) are known to generate subgap quasiparticle states in BCS superconductors like Nb, contributing to excess dissipation and decoherence in superconducting qubits.

Existing theoretical treatments of H and D tunneling in Nb typically rely on an adiabatic separation between the light interstitial and the host lattice. In this framework, the light particle is assumed to respond instantaneously to lattice motion, allowing lattice forces to be computed using the Hellman-Feynman theorem (e.g., the model by Sugimoto et al.). While this approximation may be formally justified for extremely light particles like positive muons ($\mu^+$), its validity for hydrogenic species in real, anharmonic host lattices has not been rigorously established. Previous work by the authors indicated that lattice participation plays a non-perturbative role, but the specific level of coupling required for quantitative agreement with experiment and the validity of effective two-level descriptions remained unresolved.

Methodology
The authors employ a Lattice-Renormalized Born-Oppenheimer (LRBO) framework to treat the interstitial particle and selected lattice modes on equal quantum footing.

• Hamiltonian Construction: The study constructs a five-dimensional (5D) subspace of the nuclear Hamiltonian. This includes three interstitial modes ($q$) describing motion between adjacent tetrahedral sites and two specific lattice modes:
  1. $Q$ mode: Transforms the lattice between configurations with degenerate hydrogen tetrahedral sites (connecting the two self-trapped minima).
  2. $T$ mode: Parametrizes distortions toward the transition-state configuration (the barrier maximum along the minimum-energy path).
• Potential Energy Surface (PES): The potential $V(q, Q, T)$ is derived from density functional theory (DFT) simulations using the Vienna Ab initio Simulation Package (VASP) with the PBE functional. Three fully relaxed atomic configurations are used to define the lattice coordinates: two degenerate sites and one unstable equilibrium (transition state).
• Numerical Solution: The Schrödinger equation is solved numerically on a grid spanning these 5D coordinates using the Implicitly Restarted Lanczos Method.
• Comparative Analysis: The authors systematically compare the LRBO results against the Light-Particle Approximation (LPA), a nested Born-Oppenheimer hierarchy where the light particle evolves in a potential defined by the average positions of heavier lattice atoms. This comparison is performed across a range of particle masses, from positive muons ($\mu^+$) to hydrogen (H) and deuterium (D).
• Energetic Criteria: The study utilizes simple energy estimates involving ground-state light-particle energies at fixed lattice configurations to assess the validity of adiabatic separation.

Key Results

1. Quantitative Reproduction of Tunnel Splittings: The experimentally measured tunnel splittings for O-trapped H ($J_H \approx 0.19$ meV) and D ($J_D \approx 0.021$ meV) are quantitatively reproduced only within the 5D LRBO framework. The inclusion of the transition-state lattice mode ($T$) is critical; excluding it yields significantly lower bounds ($0.064$ meV for H) that disagree with experiment. The ratio of splittings $J_D/J_H$ increases from 0.075 to 0.092 when the $T$ mode is included, indicating that lattice distortions toward the transition state become increasingly significant for heavier species.
2. Breakdown of Adiabatic Separation: The study demonstrates that the adiabatic separation of the light particle from lattice dynamics is satisfied only in the positive-muon ($\mu^+$) mass limit.
   • For $\mu^+$, the ground-state wavefunction is strongly biased toward the symmetric lattice configuration and the transition-state distortion, and the LPA yields results nearly identical to the fully coupled LRBO model.
   • For H and D, the adiabatic approximation fails. The ground-state probability density for H remains localized in the self-trapped configurations with suppressed population in the symmetric configuration. The LPA incorrectly predicts enhanced occupation of the symmetric configuration, leading to a significant overestimation of tunnel splittings.
3. Collective Nature of Tunneling: Tunneling for H and D is shown to be a fundamentally collective, nonadiabatic process mediated by anharmonic lattice couplings. The excited-state structure for H cannot be described by a simple adiabatic hierarchy of lattice modes; instead, it exhibits hybrid character combining tunneling bias toward the transition state with harmonic lattice vibrations.
4. Energetic Criteria for Adiabaticity: The breakdown of adiabaticity can be anticipated by comparing the tunneling-induced delocalization energy against the elastic energy cost of distorting the lattice to the transition state. For $\mu^+$, the symmetric configuration is energetically favorable due to the large tunnel splitting; for H, the elastic cost prevents this stabilization unless anharmonic relaxation is explicitly included.

Significance and Claims
The paper establishes that hydrogen tunneling in superconducting Nb is a lattice-mediated, multilevel quantum phenomenon.

• Theoretical Impact: The work challenges the prevailing adiabatic light-interstitial paradigm for hydrogenic species in real lattices, showing that effective two-level descriptions emerge from an underlying multilevel process involving coupled lattice modes. It provides a controlled, converged description of lattice-renormalized tunneling that achieves accuracy comparable to state-of-the-art instanton approaches.
• Implications for Superconducting Qubits: The findings suggest that the classification of defect systems based solely on resonance conditions (matching the qubit frequency) is insufficient. Even if a defect's tunnel splitting exceeds the qubit frequency (deactivating resonant coupling), it can still contribute significantly to decoherence by generating subgap quasiparticle states. The impact of defects should instead be understood through their contribution to the quasiparticle density of states, which depends on the full tunneling spectrum rather than isolated resonant transitions.
• Practical Criterion: The authors provide a practical criterion for assessing the validity of adiabatic tunneling theories in other systems: simple energy estimates involving the ground-state light-particle energy evaluated at a small number of fixed lattice configurations can predict whether adiabatic separation holds.

In summary, the paper argues that accurate predictive microscopic theories for hydrogenic tunneling in Nb must treat the light particle and lattice degrees of freedom on equal footing, moving beyond the static or quasi-static lattice potentials assumed in previous models.