Atomistic Simulations of H-Cu Vacancy Cosegregation and H Diffusion in Cu Grain Boundary

This study combines density functional theory and bond-order potential simulations to reveal that hydrogen and copper vacancies cosegregate at copper grain boundaries, forming stable complexes that significantly lower hydrogen diffusion barriers and facilitate rapid accumulation, thereby elucidating the atomistic mechanism of hydrogen embrittlement.

The Gist

The Big Picture: Why Copper Gets Brittle

Imagine copper as a super-reliable highway system for electricity and heat. It's strong, flexible, and used everywhere from your phone to power grids. But sometimes, this highway develops cracks and potholes, a problem known as hydrogen embrittlement.

Think of hydrogen not as a gas, but as tiny, invisible "gremlins" that sneak into the metal. The big mystery scientists have been trying to solve is: How do these gremlins get in, where do they hide, and how do they cause the metal to break?

This paper acts like a high-tech detective story, using computer simulations to follow the path of these hydrogen atoms from the air, onto the copper surface, and deep into the metal's internal structure.

---

The Story of the Hydrogen Gremlins

1. The Entry Point: The "Open Door" Policy

Copper is usually very picky about who it lets inside. If a hydrogen atom tries to squeeze into the middle of a perfect copper block (the "bulk"), the door is locked tight. It's hard work, and the hydrogen doesn't want to go there.

However, the edges of the copper grains (the boundaries where different crystals meet) are like open doors with a welcome mat.

• The Analogy: Imagine a crowded concert. The middle of the crowd is packed tight; you can't move. But the aisles and the exits (the grain boundaries) are wide open. The hydrogen atoms prefer to hang out in these "aisles" because there's more room to breathe.
• The Twist: If there is a missing copper atom (a "vacancy" or a hole) near the surface, it's like a VIP lounge. The hydrogen gremlins love these spots even more. They stick there with a strong grip.

2. The Team-Up: The "Gremlin and the Hole" Alliance

Once the hydrogen gets into the grain boundary (the "aisle"), something interesting happens. It finds a missing copper atom (a vacancy).

• The Analogy: Imagine a hole in a fence (the vacancy). A stray dog (the hydrogen) wanders in. The hole makes the fence weak, but the dog makes the hole even more attractive to other dogs. They form a team.
• The Science: The paper shows that hydrogen and the missing copper atom stick together to form a stable "complex." This team-up is so strong that it lowers the energy of the system by a significant amount (up to -0.8 eV). It's like they found a perfect hiding spot that neither could enjoy alone.

3. The Fast Lane: The "Bullet Train" Effect

This is the most critical part of the discovery.

• In the Bulk (The Middle): If a hydrogen atom tries to move through the solid middle of the copper, it has to push through a wall of atoms. It's like trying to run through a dense forest. It's slow and requires a lot of energy (a "barrier" of 0.42 eV).
• In the Grain Boundary (The Aisle): Inside the grain boundary, the atoms are less organized. It's like a wide, empty highway. The hydrogen can zip along this path with very little effort (a barrier of only 0.2 eV).
• The Result: Hydrogen doesn't just sit still; it rushes along the grain boundaries. It acts like a bullet train, quickly transporting itself to weak spots in the metal.

4. The Disaster: How the Metal Breaks

Here is the chain reaction the paper describes:

1. Hydrogen gas hits the copper surface and splits into atoms.
2. These atoms stick to the surface, especially near holes or cracks.
3. They slide down into the grain boundaries (the "aisles").
4. They race along these boundaries because it's the "fast lane."
5. They find missing copper atoms (vacancies) and team up with them.
6. This team-up creates a cluster of empty space. Over time, these clusters grow into voids (tiny bubbles or holes).
7. These voids merge, creating cracks. The metal, which used to be tough, suddenly snaps like a dry twig.

Why This Matters

Before this study, scientists knew hydrogen was bad for copper, but they didn't have the full map. They knew hydrogen got stuck in grain boundaries, but they didn't know exactly how it got there or how fast it moved.

This paper connects the dots:

• It shows that grain boundaries are not just traps; they are also highways for hydrogen.
• It proves that hydrogen and vacancies are best friends that team up to weaken the metal.
• It explains that the process starts at the surface and moves inward, creating a continuous path of destruction.

The Takeaway

Think of the copper metal as a city. The grain boundaries are the streets. The hydrogen is a flood of water.

• In the past, we thought the water just seeped slowly into the buildings (the bulk).
• This paper shows that the water actually rushes down the streets (grain boundaries) at high speed, finds the empty lots (vacancies), and fills them up until the whole neighborhood collapses.

By understanding this "street map" of how hydrogen moves, engineers can design better copper materials that block these highways or fill the empty lots before the water arrives, preventing the metal from breaking.

Technical Summary

1. Problem Statement

Hydrogen embrittlement (HE) is a critical failure mechanism in structural and electronic copper (Cu) applications, yet the atomistic mechanisms driving it remain poorly understood. While Cu has very low bulk hydrogen solubility, experimental evidence suggests that voids and cracks form preferentially at grain boundaries (GBs) under stress, potentially initiated by hydrogen accumulation.
Key knowledge gaps addressed in this study include:

• The specific atomistic pathway connecting $H_2$ gas exposure to hydrogen accumulation at GBs.
• The interaction between hydrogen and copper vacancies ($V_{Cu}$) at interfaces.
• The limitations of previous studies that treated surfaces, bulk, and GBs as isolated systems or relied on small supercells that introduced artificial periodic image interactions.

2. Methodology

The authors employed a hybrid computational framework combining Density Functional Theory (DFT) and Bond-Order Potentials (BOP) to overcome the scale limitations of each method individually.

• Density Functional Theory (DFT):

  • Software: VASP with PBE-GGA functional.
  • System: Used smaller periodic cells (76 atoms for $\Sigma5(210)[100]$ GB; 108 atoms for bulk/surface).
  • Role: Calculated electronic structure changes, charge density redistribution, and accurate adsorption/incorporation energies. Validated BOP results for diffusion barriers in bulk Cu.
  • Specifics: Used dispersion corrections (D2) for adsorption energies.

• Bond-Order Potentials (BOP):

  • Software: LAMMPS.
  • System: Large-scale supercells (>900 atoms, specifically 932 atoms) containing free surfaces, the GB, and bulk regions.
  • Role: Enabled long-range structural relaxations and the study of continuous diffusion pathways (Surface $\to$ GB $\to$ Bulk) impossible in DFT due to cell size constraints.
  • Techniques:
    • Molecular Dynamics (MD): Performed in the NPT ensemble (up to 700 K) to identify preferential diffusion trajectories.
    • Nudged Elastic Band (NEB): Used to calculate precise energy barriers along the pathways identified by MD.

• Key Models:

  • Focused on the $\Sigma5(210)[100]$ grain boundary, a common and well-studied symmetry in Cu.
  • Investigated $H_2$ adsorption, dissociation, atomic H incorporation, and $H$-$V_{Cu}$ cosegregation.

3. Key Contributions

• Unified Framework: First study to model the continuous pathway from gas-phase $H_2$ adsorption on a Cu surface, through dissociation, into the GB, and subsequent diffusion within a single, consistent model.
• Mechanism Elucidation: Proposed a complete atomistic mechanism for the early stages of hydrogen-induced degradation in Cu, linking surface defects to void nucleation.
• Cosegregation Quantification: Quantified the thermodynamic driving force for the formation of stable $H$-$V_{Cu}$ complexes at GBs.
• Diffusion Pathway Mapping: Mapped specific low-energy diffusion channels within the GB network that are distinct from bulk diffusion.

4. Key Results

A. Adsorption and Incorporation Energetics

• Surface Adsorption: $H_2$ dissociates on the Cu(100) surface with a low barrier (0.12 eV). Atomic H adsorption is favorable ($E_{ads} = -0.24$ eV) and becomes even more favorable ($E_{ads} = -0.30$ eV) in the presence of a surface vacancy.
• Incorporation: Incorporating H into bulk Cu is thermodynamically unfavorable ($E_{inc} = 0.68$ eV). However, incorporation at the $\Sigma5$ GB is significantly more favorable ($E_{inc} = 0.35$ eV), confirming GBs as preferential sinks.

B. Vacancy Segregation and Cosegregation

• Vacancy Segregation: Cu vacancies naturally segregate to the GB ($E_{seg} = -0.72$ eV) more strongly than to the surface ($-0.59$ eV).
• Synergistic Effect: The presence of hydrogen further stabilizes vacancies at the GB, lowering the segregation energy to $-0.83$ eV.
• Cosegregation Energy: The simultaneous presence of H and a vacancy at the GB yields a cosegregation energy gain of up to $-0.8$ eV. This indicates a strong thermodynamic drive to form stable $H$-$V_{Cu}$ complexes, which can act as seeds for void nucleation.

C. Diffusion Barriers

The study revealed distinct diffusion barriers for different pathways (Table 1 summary):

• Surface Diffusion: Low barrier ($0.29$ eV).
• Surface $\to$ Bulk: High barrier ($0.60$ eV), indicating resistance to bulk penetration.
• Bulk Diffusion: $0.42$ eV.
• Surface $\to$ GB: $0.20$ eV.
• GB $\to$ GB (along the boundary): $0.20$ eV.

Critical Finding: Hydrogen diffuses along the grain boundary network with a barrier of 0.20 eV, which is significantly lower than in the bulk (0.42 eV). This establishes GBs as fast diffusion highways for hydrogen.

5. Significance and Implications

• Mechanism of Failure: The paper proposes that hydrogen embrittlement in Cu is driven by a specific sequence:
  1. $H_2$ adsorption and dissociation at surface defects (steps/vacancies).
  2. Rapid diffusion into GBs via low-barrier pathways.
  3. Accumulation at GBs where it stabilizes vacancies, forming $H$-$V_{Cu}$ complexes.
  4. These complexes act as nuclei for voids, leading to decohesion and cracking.
• Differentiation from Other Alloys: Unlike steels (where HE is linked to phase transformations) or Ti/Zr (hydride precipitation), HE in FCC Cu is driven by fast GB diffusion and vacancy stabilization.
• Multiscale Modeling: The calculated diffusion barriers and trapping energies provide essential input parameters for Kinetic Monte Carlo (KMC) and Phase Field models, enabling the prediction of long-term hydrogen degradation in Cu interconnects and structural components.
• Experimental Guidance: The findings suggest that experimental validation should focus on detecting voids and vacancy clusters at GBs using high-resolution microscopy (SEM/TEM) and FIB cross-sectioning, particularly in regions of stress concentration.

In summary, this work bridges the gap between surface chemistry and bulk microstructural failure, demonstrating that grain boundaries in copper act not only as traps but as rapid transport channels that facilitate the accumulation of hydrogen and vacancies, ultimately driving material degradation.