Self-trapped holes and acceptor impurities in orthorhombic Ga2O3

This study utilizes hybrid density functional theory to demonstrate that self-trapped holes in orthorhombic $\kappa$-Ga$_2$O$_3$ are energetically favorable and stabilized by acceptor dopants, creating localized gap states that induce red-shifted optical absorption and suggest a potential pathway for achieving p-type conductivity through isoelectronic doping if self-compensation can be mitigated.

The Gist

Imagine a crystal made of Gallium Oxide ($\kappa-Ga_2O_3$) as a bustling city built of tiny atoms. In this city, the "streets" are made of Oxygen atoms, and the "buildings" are Gallium atoms. Usually, this city is very stable, but sometimes, a "hole" appears. In physics, a hole isn't an empty space; it's a missing piece of electricity (a positive charge) that acts like a restless traveler looking for a place to sit down.

This paper is a study of where that restless traveler decides to sit, and what happens when we swap some of the city's buildings with different types of materials.

The Natural Habit: The "Self-Trapped" Hole

In the pure, unaltered city, the hole doesn't like to wander around aimlessly. It behaves like a person who gets tired and immediately sits on a specific bench (an Oxygen atom). When it sits, it pulls the bench slightly closer to itself, making the bench wobble a bit. This wobble actually helps the hole feel more comfortable and stay put. Scientists call this a "self-trapped hole" or a "polaron."

The paper confirms that in this specific version of the crystal (the $\kappa$ phase), the hole loves to sit on an Oxygen atom and stay there. It's a very stable arrangement, much like a magnet sticking firmly to a fridge.

The Experiment: Changing the Neighborhood

The researchers asked: "What happens if we swap some of the Gallium buildings with different materials?" They tested four new "neighbors":

1. Aluminum (Al) and Indium (In): These are "isoelectronic" neighbors. Think of them as twins of the original Gallium building. They have the same electrical "personality" but are slightly different sizes.
2. Magnesium (Mg) and Zinc (Zn): These are "acceptor" neighbors. They are like new tenants who bring their own electrical baggage, potentially changing the rules of the neighborhood.

The Results: How the Neighbors Changed the Game

1. The Twins (Aluminum and Indium): The "Disruptors"
When the researchers swapped in Aluminum or Indium, the hole got confused. Instead of sitting comfortably on one specific bench, the hole became restless and spread out across the whole neighborhood.

• The Analogy: Imagine the hole was a cat that usually likes to nap on one specific chair. When you put a twin of the chair next to it, the cat gets nervous and starts pacing around the whole room, refusing to settle down.
• The Result: These neighbors made the hole delocalize (spread out). They actually made it harder for the hole to trap itself in one spot.

2. The New Tenants (Magnesium and Zinc): The "Partners"

• Magnesium: This neighbor was a bit like a quiet roommate. The hole still liked to sit on the Oxygen bench, and Magnesium didn't really interfere. The hole stayed put, just as it did in the original city.
• Zinc: This neighbor was very interactive. When Zinc moved in, the hole didn't just sit on the bench; it started "holding hands" with the Zinc atom. The hole's energy mixed with the Zinc's energy, creating a special bond.
• The Result: Zinc actually made the hole more stable and even more likely to stay in that specific spot, but now it's a "team effort" between the hole and the Zinc atom.

The Hidden Catch: The "Vacancy" Problem

The paper also looked at the "thermodynamics" of the city—basically, how easy it is to build these new neighbors or create empty spots (vacancies) in the crystal.

They found that Oxygen Vacancies (empty spots where an Oxygen atom is missing) are the easiest defects to create, especially when the environment is "Oxygen-poor" (like a dry season).

• The Analogy: Imagine you are trying to build a specific type of house (the impurity) in a neighborhood. However, the local construction rules make it incredibly cheap and easy to just knock down a wall and leave a hole (an Oxygen vacancy) instead.
• The Consequence: These empty holes act like "donors" that cancel out the effects of the new tenants (the impurities). If you try to use Magnesium or Zinc to change the electrical properties of the crystal, these empty spots might show up and neutralize your efforts, acting like a counter-force.

Summary

In simple terms, this paper tells us that:

• In pure $\kappa-Ga_2O_3$, holes naturally get stuck on Oxygen atoms.
• If you swap Gallium with Aluminum or Indium, the holes get scared and spread out, losing their "trap."
• If you swap with Magnesium, the holes stay put as usual.
• If you swap with Zinc, the holes get even more attached, bonding with the Zinc.
• However, nature has a tendency to create missing Oxygen spots (vacancies) easily, which can mess up the electrical balance and cancel out the effects of the new materials you add.

The study helps scientists understand the "personality" of holes in this material so they can better predict how to control electricity in future electronic devices, without accidentally letting the "vacancies" ruin the plan.

Technical Summary

Technical Summary: Self-trapped holes and acceptor impurities in orthorhombic $\kappa$-$\text{Ga}_2\text{O}_3$

Problem Statement
Gallium oxide ($\text{Ga}_2\text{O}_3$) is a wide-bandgap semiconductor with multiple polymorphs. While the monoclinic $\beta$ phase is the most thermodynamically stable and extensively studied, the metastable orthorhombic $\kappa$ phase has attracted interest due to its polar crystal structure and anisotropic optoelectronic properties. A critical phenomenon in these materials is hole self-trapping, where holes localize as small polarons on oxygen atoms, accompanied by local lattice distortions. Although hole self-trapping has been established in several $\text{Ga}_2\text{O}_3$ polymorphs, the specific behavior of holes in $\kappa$-$\text{Ga}_2\text{O}_3$ and their interaction with substitutional impurities remain less understood. Specifically, it is unclear how isoelectronic and acceptor impurities influence the stability, character, and formation energy of these localized hole states, which is essential for understanding charge transport and defect compensation mechanisms in this material.

Methodology
The authors employed hybrid density functional theory (DFT) using the HSE06 functional to investigate hole self-trapping and the effects of cation substitution in $\kappa$-$\text{Ga}_2\text{O}_3$.

• Computational Setup: Calculations were performed using the VASP code with projector augmented wave (PAW) potentials. A plane-wave cutoff of 500 eV was used, and the Hartree–Fock exact exchange fraction was set to 0.32 to reproduce the experimental bandgap of 4.9 eV.
• Supercells: 80-atom supercells were utilized for all systems, with spin polarization explicitly included and symmetry switched off to allow for polaron localization. The adequacy of the 80-atom cell was verified against 160-atom cells, showing negligible differences in energy (0.04 eV) and structural distortion (<0.02 Å).
• Defect Modeling: The study examined intrinsic hole localization and substitutional impurities at the Ga site: isoelectronic substitutions (Al, In) and acceptor impurities (Mg, Zn). Defect formation energies were calculated as a function of oxygen chemical potential and Fermi level, utilizing the anisotropic correction method of Kumagai and Oba for charged defects.
• Search Strategy: To ensure global ground state identification, the authors used the doped and shakenbreak Python libraries to sample the potential energy surface, employing bond distortion and rattling ranges before final HSE06 relaxation.

Key Contributions and Results

1. Intrinsic Hole Self-Trapping:
   In pure $\kappa$-$\text{Ga}_2\text{O}_3$, the hole localizes on a single oxygen atom, forming a small polaron stabilized by outward relaxation of neighboring Ga–O bonds. The calculated self-trapping energy ($E_{ST}$) is 0.173 eV, with a maximum structural distortion ($\Delta R_{max}$) of 0.044 Å. This confirms that $\kappa$-$\text{Ga}_2\text{O}_3$ exhibits hole localization behavior comparable to the $\beta$ phase, driven by the O 2p character of the valence band.

2. Impact of Isoelectronic Substitution (Al and In):
   Substitution of Ga with Al or In suppresses hole localization.

   • Energetics: Both systems exhibit negative self-trapping energies ($E_{ST} = -0.215$ eV for $\text{Al}_{\text{Ga}}$ and $-0.200$ eV for $\text{In}_{\text{Ga}}$), indicating that the hole favors a delocalized state over a localized one.
   • Structure: The lattice distortions associated with the hole are significantly reduced ($\Delta R_{max} = 0.014$ Å for Al and 0.020 Å for In) compared to the intrinsic case.
   • Mechanism: The impurities modify the local bonding environment, destabilizing the polaron state and spreading the spin density across the cell.

3. Impact of Acceptor Impurities (Mg and Zn):
   The behavior of acceptor impurities differs based on the specific element:

   • Mg: Acts as a weak perturbation. The hole remains localized on a nearest-neighbor oxygen atom (retaining >70% of spin density on a single O), with a self-trapping energy of 0.131 eV. The character of the polaron remains oxygen-centered, similar to the intrinsic system.
   • Zn: Induces a significant modification of the localized state. The hole remains primarily on oxygen atoms coordinated to the Zn impurity but exhibits hybridization between O 2p and Zn 4s orbitals. This results in an impurity-bound polaron complex with a much higher self-trapping energy (0.453 eV) and larger structural distortion (0.221 Å), indicating that Zn stabilizes the localized hole state more strongly than Mg or the intrinsic lattice.

4. Defect Thermodynamics and Compensation:
   Formation energy calculations reveal that oxygen vacancies ($V_O$) are the most favorable defect species, particularly under oxygen-poor conditions.

   • Compensation: $V_O$ acts as a donor defect that competes with acceptor impurities (Mg, Zn). The study finds that $V_O$ formation is highly favorable near the valence band maximum, suggesting that oxygen vacancies will likely compensate for the effects of extrinsic acceptor defects, limiting p-type conductivity.
   • Deep Levels: Both isoelectronic and acceptor impurities introduce deep levels. The ionization energies for Mg and Zn are calculated to be over 3 eV, indicating they form deep acceptor levels within the bandgap.

Significance
The paper establishes that $\kappa$-$\text{Ga}_2\text{O}_3$ intrinsically supports stable self-trapped holes, a feature shared with the $\beta$ phase. The primary significance of this work lies in demonstrating that substitutional impurities can fundamentally alter this intrinsic behavior. Isoelectronic doping (Al, In) effectively delocalizes the hole, while acceptor doping (Mg, Zn) preserves or modifies the localized state depending on the impurity's electronic structure (with Zn showing strong hybridization). Furthermore, the thermodynamic analysis highlights that oxygen vacancies are the dominant compensating defects, a critical constraint for achieving p-type conductivity in $\kappa$-$\text{Ga}_2\text{O}_3$. These findings provide a reference point for understanding how the interplay between polaron formation and defect thermodynamics dictates the electronic behavior of wide-bandgap oxide polymorphs.