Unlocking Deep UV: A Photoluminescence Journey into ZnGaO Thin Films

Hey there! Let me tell you about something pretty cool we’ve been digging into – the world of zinc gallium oxide (ZnGaO) thin films. Think of it as exploring a material that could be a big deal for future tech, especially stuff that works with deep ultraviolet (UV) light. We used a technique called photoluminescence, or PL for short, to peek inside these films and see how light interacts with them. It’s like shining a special flashlight and seeing what colors the material glows back, which tells you a lot about its inner workings.

Why ZnGaO? It’s an Ultra-Wide Bandgap Star!

So, why bother with ZnGaO? Well, it belongs to a family of materials known as ultra-wide bandgap (UWBG) semiconductors. These aren’t your everyday silicon chips; they have this super-wide energy gap that makes them perfect for demanding applications like high-power electronics and, you guessed it, deep UV photonics. Gallium oxide (Ga₂O₃) is a well-known player here, with a bandgap around 4.9 eV. It comes in a few different forms, but the beta phase (β-Ga₂O₃) is usually the most stable.

Now, while Ga₂O₃ is great, it has some challenges. One big one is finding a good way to make it p-type, which is crucial for many electronic and optical devices, especially things like lasers. Adding other elements can change its properties drastically. For instance, adding magnesium can actually *increase* the bandgap and change its crystal structure. But adding zinc? That’s where things get interesting.

Adding zinc to Ga₂O₃ to form ZnGaO allows us to potentially tune the bandgap, bringing it down from around 4.9 eV towards something closer to 3.3 eV, depending on how much zinc you add. And here’s the neat part: changing the zinc composition also changes the material’s crystal structure. It can go from that stable beta phase to a spinel structure, and maybe even Wurtzite. While other phases are cool, the spinel structure ZnGa₂O₄, with a bandgap around 5.2 eV, hasn’t been studied as much, and it holds a lot of promise for those deep-UV and high-power applications. To really make the most of it, we need to understand its band structure – essentially, where the electrons and holes hang out and how they move.

Our Approach: Growing Films and Shining Light

To get our hands on these ZnGaO samples, we used a fancy technique called plasma-assisted molecular beam epitaxy (MBE). It’s a bit like atomic-level spray painting in a super clean vacuum chamber. MBE is known for growing really high-quality films, which is exactly what we needed. We grew nine different samples, carefully changing the amount of zinc in each one, ranging from basically none (pure Ga₂O₃) up to almost 50%.

After growing the films, we put them through the wringer with various tests to see what we made. We used things like EDX and XPS to figure out the exact composition, SEM to look at the surface (though you don’t see those pics here, they’re in the original paper!), and UV/VIS/NIR spectroscopy to check how transparent they were and estimate their bandgap.

But the real star of our show was the photoluminescence study. Most previous PL studies on similar materials used lasers with lower energy, which means they couldn’t really excite electrons all the way across the wide bandgap from the valence band to the conduction band. We decided to hit our samples with a powerful 193 nm ArF laser. Why 193 nm? Because its photons have an energy of 6.42 eV – way higher than the bandgap of our materials. This high energy is key because it *can* pump electrons across the entire gap, giving us a more complete picture of what’s happening inside.

Structure Changes: Beta, Mixture, and Spinel!

One of the first things we noticed was how the crystal structure changed as we added more zinc. We used X-ray diffraction (XRD) to figure this out. For the sample with no zinc (pure Ga₂O₃), we saw clear peaks indicating the pure beta phase structure. Pretty standard stuff for Ga₂O₃.

As we added just a little bit of zinc (0.9 at%), things got interesting. The XRD showed a *mixture* of both beta and spinel phases. It’s like the material couldn’t quite decide which structure it preferred. With a bit more zinc (3.4 at%), the spinel phase started to show up more, but it was still kind of weak. But push the zinc composition above 7.3 at%, and bam! All the samples showed a strong spinel structure.

We used a technique called reciprocal space mapping (RSM) with XRD to confirm these structural changes. It’s a bit technical, but think of it as getting a 3D map of the crystal structure. This confirmed our findings: pure beta at 0% Zn, a mix at 0.9%, weak spinel at 3.4%, and strong spinel for everything above 7.3%. This change in structure is important because it affects all the material’s properties, including how it interacts with light.

We also calculated the lattice parameters – basically, the size and shape of the tiny repeating units in the crystal structure – from our XRD data. For the spinel samples, we saw the lattice parameter increase slightly as the zinc composition went up. This makes sense because zinc ions are a bit larger than gallium ions, so sticking more of them into the structure makes it expand a little. These values matched up nicely with what other folks have reported for beta Ga₂O₃ and spinel ZnGa₂O₄, which was a good sign we were on the right track.

The Glow Show: Deconvoluting the PL Peaks

Now for the fun part: the photoluminescence! When we hit our samples with that powerful 193 nm laser, they glowed back with different colors of light. By analyzing the spectrum of this emitted light, we could learn about the energy levels within the material’s forbidden gap – the region where electrons usually aren’t supposed to be, but where defects can create stepping stones.

For the samples that had the pure beta or strong spinel structure, we could break down the PL spectrum into five distinct peaks. These peaks corresponded to different colors: UV, violet, blue, green, and red emissions. Each color tells us about a specific type of optical transition happening inside the material.

Think of it like this: the material has a main energy gap (the bandgap). Electrons normally live in the lower energy levels (valence band) and can jump to higher energy levels (conduction band) if you give them enough energy (like from our laser). When they fall back down, they release energy as light. But defects in the crystal can create intermediate energy levels *within* that gap. The light we see comes from electrons falling from the conduction band to one of these intermediate levels, or from one intermediate level to another.

Understanding the Energy Levels

Let’s break down what those five peaks revealed:

• UV Emission: In the pure beta Ga₂O₃ sample (0% Zn), this peak seems to be from electrons falling from the conduction band to an acceptor level related to gallium-oxygen complex vacancies (V_Ga-O). We estimated this level to be about 0.91 eV above the valence band edge.
• Violet Emission: This one is super interesting! It’s linked to transitions involving self-trapped holes (STHs). Imagine a “hole” (the absence of an electron) getting stuck or “trapped” in a specific spot in the crystal structure. This violet light corresponds to electrons from the conduction band recombining with these trapped holes. We could even estimate the binding energy of these STHs – how tightly they’re held in place. For our spinel samples with higher zinc content, the calculated STH binding energies (around 0.15 to 0.24 eV) were actually quite similar to what theoretical calculations had predicted! This was a nice experimental validation.
• Blue, Green, and Red Emissions: These visible light peaks are likely due to transitions between different types of oxygen vacancies (V_O) and acceptor levels. Oxygen vacancies are basically missing oxygen atoms in the crystal structure, and they can create their own energy levels. We estimated the positions of three different oxygen vacancy levels (V_OI, V_OIII, and V_OII) relative to the conduction band edge. The acceptor levels involved could be those V_Ga-O vacancies we saw in beta phase, or potentially zinc vacancies (V_Zn) or gallium vacancies (V_Ga) in the spinel samples, or even zinc atoms sitting where gallium should be (Zn_Ga). It depends a bit on the exact zinc composition.

Interestingly, for the samples with the mixture phase or weak spinel structure (0.9 at% and 3.4 at% Zn), we didn’t see the violet emission clearly. This might be because the crystal quality wasn’t as good, and other transitions (like those causing the blue, green, and red light) were stronger and masked the STH signal.

We calculated the energies of these oxygen vacancy levels (relative to the conduction band) and the acceptor levels (relative to the valence band) for all the samples where we could deconvolve the peaks. This gives us a detailed map of the electronic landscape within these materials, which is super valuable for designing devices.

Temperature’s Role: Hot and Cold Effects

We also looked at how the PL changed when we heated or cooled the samples, ranging from a chilly 14 K up to room temperature. Generally, as we increased the temperature, the PL intensity went down. This is a common phenomenon called positive thermal quenching. Basically, at higher temperatures, other ways for electrons and holes to recombine that *don’t* produce light become more likely, so you get less glow.

However, we also saw something called negative thermal quenching in some samples, especially at lower temperatures. This is where the PL intensity *increases* as the temperature goes up slightly. It sounds counter-intuitive, but it happens when the radiative transitions (the ones that make light) involving those intermediate energy levels in the gap become *more* efficient as you add a little thermal energy. It’s like needing a little nudge to get the light-producing process going more strongly in that temperature range.

We even used an equation to try and model this temperature behavior and extract some activation energies related to these quenching processes. It showed that the energy needed for non-radiative recombination (the light-killing kind) varied a bit depending on the crystal structure, being slightly higher in the mixture phase compared to the pure beta or strong spinel phases.

Wrapping It Up: What We Learned

So, what’s the takeaway from all this? We successfully grew ZnGaO thin films with varying zinc compositions using MBE. We clearly showed how increasing the zinc content drives a transition in the crystal structure from beta phase Ga₂O₃ to spinel phase ZnGa₂O₄.

Crucially, by using high-energy PL excitation, we were able to systematically study the optical transitions happening within these films. We identified and mapped out the energy levels associated with different types of defects, including three distinct oxygen vacancy levels, acceptor levels, and importantly, we experimentally validated the binding energy of self-trapped holes in the spinel structure.

This detailed understanding of the energy landscape and defect states is a big step forward. It gives us valuable insights into how these materials behave optically and electronically. This knowledge is absolutely essential for unlocking the full potential of spinel ZnGa₂O₄ for practical applications, particularly in the exciting field of deep ultraviolet photodetectors and other advanced optoelectronic devices. It feels good to contribute a piece to that puzzle!

Source: Springer