Evidence for Umklapp electron scattering emission from metal photocathodes

This paper presents evidence for a new one-photon emission process in single-crystal Cu(001) and W(111) photocathodes near the photoemission threshold, attributing the phenomenon to momentum-resonant Franck-Condon mechanisms mediated by inelastic Umklapp electron scattering and validating the theory through a combined first-principles simulation and direct one-step band emission model.

The Gist

The Big Picture: A Mystery at the Exit Door

Imagine a metal photocathode (like a piece of Copper or Tungsten) as a crowded nightclub. The electrons are the partygoers, and the "work function" is the bouncer at the door. To get out of the club (into the vacuum), an electron needs enough energy to jump over the bouncer's rope.

Usually, scientists have a perfect rulebook for how these electrons get out. They say: "If you hit the metal with a light beam, the electrons absorb the energy. If they have enough, they jump out. If they don't, they stay inside."

The Problem:
The researchers in this paper looked at two specific metals (Copper and Tungsten) and found a glitch in the rulebook.

• Copper behaved exactly as the rulebook predicted.
• Tungsten, however, was acting weird. Even when the light didn't have quite enough energy to kick the electrons out, they were still escaping. Furthermore, the electrons that escaped were "wobbly" (moving in a messy, spread-out way) rather than a clean, straight line.

The existing rulebooks couldn't explain why Tungsten was sneaking electrons out the back door when the front door was locked.

The New Theory: The "Umklapp" Shuffle

The authors propose a new mechanism to explain this. They call it Umklapp electron scattering, mediated by a Franck-Condon mechanism. That sounds complicated, so let's break it down with an analogy.

The Analogy: The Trampoline and the Wall

1. The Standard Way (Direct Emission):
Imagine an electron is a ball bouncing on a trampoline (the metal surface). If you hit the ball hard enough (with a photon of light), it flies straight over the wall. This is what we expect to happen.

2. The New Way (Umklapp Scattering):
Now, imagine the ball hits a wall inside the room before it can leave.

• The Collision: The excited electron (the ball) crashes into another electron (a friend) inside the metal.
• The "Umklapp" Twist: In physics, when particles collide in a crystal, they can bounce off in a way that changes their "direction code" completely. It's like hitting a wall that doesn't just stop you, but teleports you to a different spot on the map. This is the "Umklapp" process.
• The Resonance (Franck-Condon): Because of this specific type of collision, the electron suddenly finds itself in a "sweet spot" where it can easily slip through the bouncer's rope, even if it didn't have enough energy to jump it on its own. It's like the collision gave it a hidden boost or a "key" to the door.

Why Tungsten vs. Copper?

Why did this only happen with Tungsten and not Copper?

• Copper is like a simple, round room. The electrons move in smooth, predictable circles. When they collide, they just bounce around normally. The "hidden boost" doesn't happen often.
• Tungsten is like a complex, maze-like room with many different paths and walls (multiple "Fermi surfaces"). When an electron collides here, it's much more likely to hit a wall that triggers the "teleportation" (Umklapp) effect. This allows many more electrons to sneak out, even when the light is weak.

What Did They Measure?

The scientists measured two things:

1. QE (Quantum Efficiency): How many electrons got out?
   • Result: Tungsten let out way more electrons than the old rules predicted when the light was weak.
2. MTE (Mean Transverse Energy): How "messy" was the beam of electrons?
   • Result: The electrons coming out of Tungsten were much "messier" (higher energy sideways) than expected. This is because the "collision" process adds a bit of chaos to their movement.

Why Does This Matter?

You might ask, "Who cares if Tungsten acts weird?"

These photocathodes are used to create super-bright electron beams for machines like:

• X-ray Free Electron Lasers (XFELs): These take movies of atoms moving.
• Electron Microscopes: These let us see tiny details of materials.

To get a clear picture, you need a beam of electrons that is:

1. Bright (lots of them).
2. Tight (they all travel in a straight line, not wobbling).

If we understand this "Umklapp" trick, we can:

• Design better metals: We can choose metals that don't do this messy scattering if we want a tight beam, or use it if we need more electrons.
• Fix the math: We can update the rulebooks so engineers can design better machines without guessing.

The Conclusion

The paper proves that for certain metals (like Tungsten), electrons don't just jump out when hit by light. Sometimes, they take a "shortcut" by colliding with other electrons and using the crystal structure of the metal to boost themselves out.

The authors created a new mathematical model that includes this "collision shortcut." When they ran the numbers, their new model matched the messy experimental data perfectly, while the old models failed. This is a big step forward in understanding how to make the world's most powerful electron microscopes and lasers work better.

Technical Summary

1. Problem Statement

Current theoretical models for photoemission from single-crystal metal photocathodes fail to accurately predict experimental data near and below the photoemission threshold. Specifically:

• Existing Models: The temperature-extended Dowell-Schmerge (DS) formalism and the one-step direct band emission model (incorporating Density of States and band dispersion) work reasonably well for energies well above the threshold.
• The Discrepancy: Experimental measurements of Quantum Efficiency (QE) and Mean Transverse Energy (MTE) for Tungsten (W(111)) and Copper (Cu(001)) show significant deviations from these theories in the sub-threshold region ($\Delta E < 0$).
  • For W(111), the measured QE is 4–5 orders of magnitude higher than predicted below threshold, and the MTE is ~70 meV (more than double the predicted ~30 meV).
  • For Cu(001), the deviation is less severe but still present, with MTE slightly higher than the DS prediction.
• The Gap: These models neglect the role of inelastic electron-electron scattering, specifically Umklapp scattering (UES), which involves momentum transfer via a reciprocal lattice vector ($G$). The authors hypothesize that an additional emission mechanism, mediated by UES, is responsible for the excess emission and higher transverse energy observed near the threshold.

2. Methodology

Experimental Approach

• Samples: High-purity (99.999%), single-crystal Cu(001) and W(111) photocathodes with sub-10 nm surface roughness to minimize roughness-induced effects.
• Setup: A 300 K vacuum system with a DC electron gun. Samples were illuminated by tunable, sub-picosecond UV radiation (3–5.3 eV) generated by an optical parametric amplifier (OPA).
• Measurements:
  • QE: Measured via calibrated micro-channel plate (MCP) detection.
  • MTE: Determined by analyzing the transverse momentum spread of the electron beam after a 42 cm drift, imaged onto a phosphor screen and CMOS camera.
  • Cleaning: Samples underwent chemical etching and "laser cleaning" to stabilize the work function ($\phi$).

Theoretical Framework

The authors developed a hybrid model combining two emission processes:

1. Direct One-Step Band Emission: Based on established theory, incorporating DFT-calculated band dispersions (longitudinal and transverse) and vacuum DOS.
2. Proposed FC-UES Mechanism: A momentum-resonant Franck-Condon (FC) emission process mediated by inelastic Umklapp Electron Scattering (UES).
   • Mechanism: A photoexcited electron ($p_1^*$) collides inelastically with a bulk electron ($p_2$), transferring momentum via a reciprocal lattice vector $G$ ($p_1^* + p_2 = p_0 + p \pm G$). This allows an electron to reach the vacuum state ($p_0$) even if direct excitation is forbidden or inefficient.
   • Key Parameters: The model utilizes the thermal electron effective mass ($M_{th}$), the number of Fermi surfaces ($N_{fs}$), and the inelastic mean free path ($\lambda$) at the vacuum level.
   • Resonance: The model includes a resonance effect near zero vacuum energy ($E_{vac} \approx 0$), attributed to the penetration of low-energy vacuum wavefunctions into the bulk, described by a Lorentzian profile dependent on $\lambda$.

Simulation Strategy

• The total QE and MTE were calculated as a weighted sum of the direct band and FC-UES contributions.
• The relative weight was determined by the product of the UV absorption coefficient ($\alpha$) and the IMFP ($\lambda$).
• The model was fitted to experimental data by adjusting the relative strength parameter ($f$) of the UES contribution, while keeping material-specific parameters ($M_{th}, N_{fs}, \lambda$) fixed based on literature and DFT calculations.

3. Key Contributions

• Identification of a New Mechanism: The paper provides strong evidence that inelastic Umklapp scattering is a dominant photoemission mechanism near and below the threshold for metals with complex electronic structures (like Tungsten).
• Unified Theoretical Model: The authors successfully combine the one-step direct band model with a simplified FC-UES formalism. This combined model resolves the long-standing discrepancies between theory and experiment for both QE and MTE in the sub-threshold regime.
• Material-Specific Insights: The study demonstrates that the magnitude of the UES effect is material-dependent, driven by the ratio of thermal effective mass to free electron mass ($M_{th}/m_0$) and the number of Fermi surfaces.
  • W(111): High $M_{th}$ ($3.5m_0$) and complex band structure lead to a strong UES contribution, explaining the high sub-threshold QE and MTE.
  • Cu(001): Near-free electron behavior ($M_{th} \approx 1.36m_0$) results in a weaker UES contribution, making direct band emission dominant, consistent with the smaller observed deviations.
• Analytical Approximation: The authors derived an approximate expression for the sub-threshold MTE (Equation 12), showing it scales with the thermal effective mass and the number of Fermi surfaces: $MTE_{UES} \approx (\frac{6}{5} + \frac{M_{th}}{2m_0 N_{fs}}) k_B T_e$.

4. Results

• W(111) Photocathode:
  • QE: The combined model (Direct + FC-UES) accurately reproduces the experimental QE, including the 4–5 order of magnitude excess below threshold, whereas the direct-only model fails completely.
  • MTE: The model correctly predicts the measured ~70 meV MTE below threshold. The direct-only model predicts only ~30 meV.
  • Fit Parameters: A UES strength factor ($f$) of 4 was required, consistent with the high thermal mass and low absorption depth of Tungsten.
• Cu(001) Photocathode:
  • QE & MTE: The combined model fits the data well with a UES strength factor ($f$) close to 1.
  • Comparison: The direct band model alone is nearly sufficient for Cu, as the material behaves more like a "Sommerfeld" metal with a spherical Fermi surface. The UES contribution is present but less dominant than in Tungsten.
• Work Function Extraction: Using the $QE^{1/n}$ vs. photon energy method, the authors extracted work functions of 4.2 eV for W(111) and 4.1 eV for Cu(001), which were used to define the excess energy ($\Delta E$) for all comparisons.

5. Significance

• Fundamental Physics: The work challenges the assumption that photoemission from metals is purely a direct one-step process or a simple thermal tail effect. It establishes inelastic Umklapp scattering as a critical, momentum-resonant pathway for electron emission.
• Impact on Advanced Instruments: Understanding and minimizing MTE is crucial for the performance of X-ray Free Electron Lasers (XFELs), Ultrafast Electron Diffraction (UED), and Dynamic Transmission Electron Microscopes.
  • The findings suggest that for materials with high $M_{th}$ (like W), the intrinsic emittance (MTE) cannot be reduced simply by cooling, as the UES mechanism introduces a "thermal" tail dependent on the effective mass.
  • Conversely, materials with low $M_{th}$ and simple band structures (like Cu) are better candidates for low-emittance applications.
• Future Directions: The paper proposes Beryllium (Be(0001)) as an ideal candidate for future study. Be has no direct emitting bands near the Fermi level in the relevant direction, potentially allowing the isolation and study of pure FC-UES emission. The authors predict Be could yield sub-thermal MTE (< 5 meV), offering a pathway to ultra-high brightness electron sources.

In conclusion, this paper provides a robust theoretical and experimental framework that attributes sub-threshold photoemission anomalies to Umklapp scattering, offering a new paradigm for designing and optimizing high-brightness metal photocathodes.