Vacuum Wannier Functions for First-Principles Scattering and Photoemission

This paper establishes a first-principles theory of vacuum Wannier functions that unifies tight-binding and nearly-free-electron models to enable predictive photoemission calculations without semiempirical potentials by constructing dense k-space scattering states at solid-vacuum interfaces.

The Gist

The Big Picture: The "Empty Room" Problem

Imagine you are trying to describe how a person (an electron) jumps out of a crowded dance floor (a solid material like graphene) and runs into an empty, vast hallway (the vacuum).

In the world of quantum physics, scientists use a special tool called Wannier functions to map out where electrons are likely to be. Think of these functions as "flashlights" that shine on specific spots to show where an electron is hiding.

The Problem:
These flashlights work perfectly on the dance floor. But when you try to shine them into the empty hallway (the vacuum), they break.

• The Old Way: If you try to map the empty hallway using standard methods, the "flashlight" beam gets so wide and blurry that it never really settles down. It's like trying to measure the exact position of a ghost; the math says the beam is infinitely wide, making it impossible to calculate how the electron moves into the air.
• The Consequence: Because the math breaks in the empty space, scientists have to use "guesswork" (semi-empirical models) to predict things like photoemission (when light kicks electrons out of a material). They can't calculate it perfectly from first principles.

The Solution: The "Perfect Packing" Trick

The authors, Tyler Wu and Tom´as Arias, discovered a new way to arrange these flashlights in the empty hallway so they work perfectly.

The Analogy: The Egg Carton vs. The Pile of Oranges

• The Old Way: Imagine trying to pack oranges into a box. If you just throw them in randomly or stack them in a square grid, there are huge gaps. The oranges (electrons) don't fit well, and the structure is unstable.
• The New Discovery: The authors realized that in the vacuum, the most efficient, stable way to arrange these "flashlights" is to pack them like oranges in a crate (a "close-packed" lattice).
  • They proved mathematically that if you arrange these vacuum flashlights in a tight, honeycomb-like or face-centered cubic pattern, they stop being blurry and infinite. They become sharp, compact, and stable.
  • It's like realizing that to fill a room with sound evenly, you don't just shout from the corners; you arrange speakers in a specific, dense grid that covers every inch perfectly.

Why This Matters: The "Photoemission" Test

To prove their new method works, they tested it on two materials: Graphene (a single layer of carbon) and Hexagonal Boron Nitride (h-BN). They wanted to predict what happens when you shine light on them to knock electrons out (photoemission).

Think of this like a slingshot contest:

1. The Goal: You want to know exactly how fast and in what direction the "rock" (electron) flies when you pull the slingshot (light).
2. The Old Prediction: Previous methods were like guessing the wind speed. They often got the direction wrong because they ignored how the "rock" interacts with the air (the vacuum) right after leaving the slingshot.
3. The New Prediction: Using their new "perfectly packed" flashlights, the authors could simulate the entire flight path of the electron, from the solid material all the way into the air, without guessing.

The Surprising Results:

• Graphene: It behaved exactly as the simple models predicted. It's symmetrical, so the physics is straightforward.
• Boron Nitride (h-BN): This was the shocker. The simple models failed completely. Because h-BN is not symmetrical (it's like a lopsided coin), the electron interacts with the vacuum in a complex way that only the new "full scattering" method could catch. The new method showed that the electrons fly out much more efficiently (with less energy spread) than anyone thought.

The Takeaway: Why Should You Care?

This paper is like upgrading the GPS for electrons.

1. No More Guessing: Before, scientists had to use "rules of thumb" to predict how electrons leave materials. Now, they can calculate it exactly from the laws of physics.
2. Better Tech: This is huge for developing ultrafast electron microscopes (which take movies of atoms moving) and better photocathodes (the parts of cameras or particle accelerators that generate electron beams).
3. The "Close-Packing" Secret: The most important discovery is the geometric rule: To describe empty space in quantum physics, you must arrange your mathematical tools in the tightest possible packing pattern.

In a nutshell: The authors found the secret recipe for organizing mathematical tools in empty space. This allows us to predict exactly how electrons jump from solids into the air, leading to better, faster, and more precise technologies in the future.

Technical Summary

1. Problem Statement

The paper addresses a critical gap in modern electronic structure theory: the inability to efficiently and accurately describe vacuum–bulk coupling using Wannier functions (WFs).

• Limitations of Current Methods: Standard Wannier function techniques are optimized for bulk materials. When applied to systems with large vacuum regions (e.g., slabs, surfaces), the standard Marzari–Vanderbilt (MV) localization procedure fails.
• The Vacuum Pathology: In the vacuum limit, Bloch states become plane waves. The resulting Wannier functions exhibit $1/x$ algebraic decay, leading to divergent second moments (variance). This renders the localization functional ill-defined.
• Numerical Instability: While the Souza–Marzari–Vanderbilt (SMV) disentanglement procedure can remove formal divergences, it often results in Wannier functions with large residual spreads and irregular center positions. This "ill-conditioning" prevents the construction of dense, regular lattices of vacuum states, which are necessary for modeling scattering and photoemission without prohibitively large Density Functional Theory (DFT) supercells.
• Application Gap: Existing photoemission models often rely on phenomenological "three-step" models or simplified one-step approaches that neglect the material-specific Born-series structure of continuum states near the surface, leading to inaccurate predictions for observables like Mean Transverse Energy (MTE).

2. Methodology

The authors develop a unified first-principles framework based on Vacuum Wannier Functions (VWFs).

• Analytic Derivations:

  • They derived analytic solutions for the global minimizers of the MV localization functional in arbitrary dimensions ($d$).
  • They provided a fully analytic solution for the SMV disentanglement procedure in 1D ($d=1$) for an empty lattice. This involves constructing a smooth gauge (using a mixing function $\theta(k)$) to cure the divergence of the gauge-invariant spread functional ($\Omega_I$).
  • Result: The analytic 1D solution yields Wannier functions with super-algebraic decay ($\sim x^{-2}$), finite variance, and centers arranged on a regular grid.

• Numerical Verification and the "Close-Packing" Principle:

  • Extending the 1D analytic results to 3D, the authors performed numerical calculations on cubic and hexagonal supercells.
  • Key Finding: They discovered that when disentanglement is applied to vacuum states, the resulting Maximally Localized Wannier Functions (MLWFs) spontaneously distort their centers to form regular close-packed lattices (e.g., Face-Centered Cubic for cubic cells, Hexagonal Close-Packed for hexagonal cells).
  • Stability Tests: Random perturbations of the initial WF centers (up to 30% of nearest-neighbor spacing) consistently relaxed back to the regular close-packed grid, confirming this arrangement as the global minimum for the localization functional.

• Application to Photoemission:

  • The authors constructed a unified basis set containing both material-bound states and vacuum scattering states using the VWFs.
  • They applied this to monolayer graphene (Gr) and hexagonal boron nitride (h-BN).
  • They calculated photoemission observables, specifically the Longitudinal Energy Spread ($\sigma_{EL}$) and Mean Transverse Energy (MTE), using a full Born-series scattering approach rather than the First Born Approximation (plane-wave final states).

3. Key Contributions

1. Theory of Vacuum Wannier Functions: Established a rigorous first-principles theory showing that optimal vacuum localization naturally selects close-packed lattices as stable solutions. This resolves the numerical pathologies associated with vacuum padding in DFT.
2. Scalable Vacuum Extension: The framework allows for the arbitrary extension of vacuum regions in supercell calculations without increasing the computational cost of the underlying DFT calculation. The vacuum states are constructed via Wannierization rather than explicit DFT plane-wave expansion.
3. Unified Scattering Framework: The method unifies tight-binding (bulk) and nearly-free-electron (vacuum) descriptions, enabling the construction of full Born-series scattering states at interfaces.
4. Analytic 1D Disentanglement: Provided a closed-form analytic solution for the disentanglement of vacuum bands, proving that the resulting functions are orthonormal, maximally localized, and possess finite variance.

4. Key Results

• Structural Stability: Numerical tests confirmed that VWFs on FCC and HCP lattices are stable against significant initial perturbations, converging to symmetry-equivalent positions with high precision ($\sim 10^{-5}$ bohr).
• Photoemission in Graphene vs. h-BN:
  • Longitudinal Spread ($\sigma_{EL}$): For ultrafast electron diffraction (UED), graphene showed a lower energy spread coefficient ($\alpha_{Gr} \approx 0.235$) compared to h-BN ($\alpha_{h-BN} \approx 0.273$) and the simple effective-mass model ($0.298$), identifying graphene as a superior candidate for UED.
  • Mean Transverse Energy (MTE):
    • Graphene: Due to inversion symmetry, the optical matrix elements vanish at the zone center ($k_\parallel=0$). Consequently, the MTE follows simple theoretical expectations, and higher-order scattering effects do not qualitatively change the result.
    • h-BN: Lacking inversion symmetry, higher-order scattering processes (beyond the First Born Approximation) generate substantial spectral weight at the zone center. This leads to a dramatic suppression of MTE (slope $\approx 0.160$) compared to the effective-mass prediction ($0.25$) and the First Born Approximation (which incorrectly predicts a "hollow" zone center and inflated MTE).
• Beyond First Born Approximation: The study demonstrates that capturing the correct physics for non-centrosymmetric materials (like h-BN) requires the full Born series. Plane-wave final state approximations fail to capture the zone-center intensity required for accurate MTE prediction.

5. Significance

• Predictive Photoemission: The work enables predictive, parameter-free calculations of photoemission observables (MTE, energy spread) for photocathodes and ultrafast electron sources, removing the need for semi-empirical vacuum potentials.
• Material Design: By accurately modeling the interplay of symmetry, matrix elements, and scattering, the framework provides a tool to screen materials for specific applications (e.g., identifying graphene's superiority for low-emittance beams).
• Methodological Advancement: The "Vacuum Close-Packing Principle" offers a general solution for handling open boundary conditions in electronic structure theory, bridging the gap between localized bulk states and delocalized continuum states. This is crucial for advancing fields ranging from field emission and tunneling to plasmonics and photonics.