Study on the Effect of Annealing on Ga$_2$O$_3$ Thin Films Deposited on Silicon by RF Sputtering

This study demonstrates that thermal annealing of RF-sputtered $\beta$-Ga$_2$O$_3$ thin films on silicon substrates significantly enhances their crystalline structure and refractive index, with the most pronounced improvements observed at 1000 $^\circ$C.

The Gist

Imagine you have a very special, ultra-thin sheet of glass made of Gallium Oxide (Ga₂O₃). This isn't just any glass; it's a "super-material" for the future of electronics. It's incredibly strong, can handle massive amounts of electricity without breaking, and is transparent to light that our eyes can't even see (like deep ultraviolet). Scientists want to use this material to build faster computers, better solar-blind cameras, and advanced sensors.

However, there's a catch. When the researchers first made this material using a technique called RF Sputtering (think of it like blasting tiny particles of the material onto a silicon chip, similar to spray-painting a wall), the resulting film was messy. It was amorphous, meaning the atoms were jumbled up like a bowl of spaghetti rather than neatly stacked like a brick wall. This "spaghetti" state made the material less efficient for its intended high-tech jobs.

The big question this paper answers is: What happens if we bake this messy film in an oven?

Here is the story of what they found, explained simply:

1. The Experiment: The "Thermal Spa"

The researchers took their jumbled, "spaghetti-like" films and put them in a furnace (an oven) at different temperatures, ranging from a warm 550°C up to a scorching 1000°C. They wanted to see if heat could organize the atoms into a neat, orderly structure (crystallization).

2. What Happened to the Surface? (The "Bumpy Road" Analogy)

Before baking, the surface of the film was as smooth as a polished marble floor.

• At lower temperatures: It stayed mostly smooth.
• At 1000°C: The surface got significantly rougher. Imagine the smooth floor turning into a cobblestone street.
• Why? As the heat got intense, the atoms started moving around and clumping together to form larger, organized "grains" (like grains of sand). This process made the surface bumpy, but it also made the material much stronger and more organized underneath.

3. The Chemical Change: The "Oxygen Sandwich"

While the film was baking, something interesting happened at the bottom, where the film touched the silicon chip.

• The silicon chip is like a piece of bread that loves to get toasted. As the heat increased, the silicon reacted with the air, growing a thicker layer of Silicon Dioxide (SiO₂)—basically, a layer of rust or glass—between the film and the chip.
• Also, the film itself got a little "healthier." It started with a slight shortage of oxygen (like a cake with too little flour). The heat allowed more oxygen to sneak in, fixing the recipe so the chemical balance became perfect. This is crucial because a perfect recipe means the material works better as a sensor or light detector.

4. The Optical Magic: The "Denser Glass" Effect

This is the most exciting part for engineers. The researchers measured how light travels through the film.

• The Refractive Index: Think of this as how much the material "bends" or slows down light.
• The Result: After baking at 1000°C, the film became denser. Because it was denser, it bent light much more strongly.
• The Analogy: Imagine a sponge. When it's dry and full of air holes, light passes through easily. If you squeeze the sponge tight (making it denser), light has a harder time getting through and changes direction more. The 1000°C film was like that squeezed sponge—it was packed so tightly with atoms that its optical properties changed dramatically.

5. The Crystal Structure: From "Spaghetti" to "Bricks"

Using X-rays (like an X-ray machine for crystals), they looked at the internal structure.

• Before baking: The atoms were a chaotic mess (amorphous).
• After baking: The atoms lined up perfectly in a specific pattern called β-Ga₂O₃.
• The Grain Size: The "grains" (the little organized blocks of atoms) grew much larger.
• The Strain: Before, the atoms were stretched and stressed (like a rubber band pulled too tight). After baking, they relaxed. The material became "calm" and stable.

The Big Takeaway

This paper tells us that heat is the key to unlocking the potential of Gallium Oxide films made on silicon.

By simply baking the film at high temperatures (specifically 1000°C), the researchers turned a messy, low-quality coating into a high-performance, crystalline material.

• It got denser.
• It got more transparent to certain lights.
• It got a better chemical recipe.
• It bent light more effectively.

Why does this matter?
Because this material is now ready to be used in the next generation of electronics. It means we can build better devices that can see in the dark (UV light), handle high-voltage power without exploding, and fit onto the same silicon chips we use in our phones today. The "baking" process was the secret ingredient that turned a good idea into a great material.

Technical Summary

Here is a detailed technical summary of the paper "Study on the Effect of Annealing on Ga2O3 Thin Films Deposited on Silicon by RF Sputtering" by Sousa et al.

1. Problem Statement

Gallium oxide ($\beta$-Ga$_2$O$_3$) is an ultra-wide bandgap semiconductor ($\sim$4.85 eV) with exceptional optoelectronic properties, making it ideal for deep-UV photodetectors, power electronics, and photonic integrated circuits. While Radio-Frequency (RF) sputtering is a cost-effective and scalable technique for depositing these films on various substrates (including silicon), it typically produces amorphous films when performed at room temperature.

To achieve the desired crystalline structure and optical performance, post-deposition thermal annealing is required. However, there is a lack of systematic understanding regarding how thermal annealing specifically correlates the optical properties (such as refractive index and bandgap) with the structural, morphological, and compositional changes in RF-sputtered Ga$_2$O$_3$ films on silicon. This study aims to bridge that gap by investigating the effects of annealing temperatures ranging from 550°C to 1000°C.

2. Methodology

The researchers employed a multi-faceted experimental approach combining thin-film deposition with four complementary characterization techniques:

• Deposition: $\beta$-Ga$_2$O$_3$ thin films were deposited on silicon substrates via RF sputtering at room temperature (60 W, 6 mTorr, 1 hour).
• Thermal Treatment: The as-deposited films were annealed in air using a tube furnace at temperatures of 550°C, 700°C, 850°C, and 1000°C for 1 hour each.
• Characterization Techniques:
  1. Variable Angle Spectroscopic Ellipsometry (VASE): Used to extract optical functions (refractive index $n$, extinction coefficient $k$, bandgap) and layer thicknesses using Tauc-Lorentz and Wemple-DiDomenico (WDD) models.
  2. Atomic Force Microscopy (AFM): Used to measure surface roughness (RMS) and morphological changes.
  3. Rutherford Backscattering Spectrometry (RBS): Used to determine elemental composition, stoichiometry (Ga/O ratio), and depth profiling of layers (including interfacial SiO$_2$).
  4. X-ray Diffraction (XRD): Used to analyze crystalline structure, phase identification, grain size, and microstrain (via Scherrer and Wilson equations).

3. Key Contributions

• Comprehensive Correlation: The study provides a direct correlation between thermal annealing and the simultaneous evolution of optical, structural, and compositional properties in RF-sputtered Ga$_2$O$_3$ on Si.
• Interfacial Dynamics: It highlights the formation and thickening of an SiO$_2$ interfacial layer during air annealing, a critical factor often overlooked in rapid thermal annealing (RTA) studies but significant for long-duration furnace annealing.
• Optical Modeling: The application of the Wemple-DiDomenico (WDD) single-effective oscillator model to extract dispersion energy ($E_d$) and static refractive index ($n_0$), linking these parameters to structural disorder reduction.
• Density Estimation: By combining RBS and ellipsometry data, the authors successfully estimated film density, demonstrating its convergence toward the theoretical density of $\beta$-Ga$_2$O$_3$ with higher annealing temperatures.

4. Key Results

A. Structural and Morphological Evolution

• Crystallization: As-deposited films were amorphous. XRD revealed the emergence of the $\beta$-Ga$_2$O$_3$ (400) reflection starting at higher temperatures. At 1000°C, the films showed significant crystallinity with increased grain size and a dramatic reduction in microstrain.
• Surface Roughness: AFM results showed a marked increase in RMS roughness. While roughness remained low (<1 nm) up to 850°C, it jumped to 5.4 nm at 1000°C. This is attributed to grain growth and structural reorganization.
• Interfacial Layer (RBS): RBS analysis confirmed the presence of an SiO$_2$ layer at the Si/Ga$_2$O$_3$ interface. Its thickness increased significantly with temperature, reaching ~202 $\times$ 10$^{15}$ at/cm$^2$ at 1000°C, due to the oxidation of the silicon substrate.
• Stoichiometry: The Ga/O stoichiometry shifted from 44/56% (as-grown) to 40/60% after annealing, indicating a reduction in oxygen vacancies and a move toward stoichiometric balance.

B. Optical Properties

• Refractive Index ($n$): A significant increase in the refractive index was observed, rising from 1.851 (as-grown) to 1.915 (at 1000°C) at 632.8 nm. This correlates with the increased film density and improved crystallinity.
• Bandgap ($E_g$): The optical bandgap initially decreased slightly at 550°C but increased to 4.28 eV at 1000°C. This increase is linked to the reduction of oxygen vacancies and improved stoichiometry.
• Transparency: The films exhibited excellent transparency (extinction coefficient $k \approx 0$) in the UVA–visible–NIR range, with absorption onset aligning with the bandgap.
• Dispersion Parameters: The WDD analysis showed an increase in dispersion energy ($E_d$) and static refractive index ($n_0$) with annealing, confirming a reduction in structural disorder.

5. Significance and Implications

This study is critical for the development of Silicon-based Ga$_2$O$_3$ optoelectronic devices.

• Device Integration: The findings demonstrate that while high-temperature annealing (1000°C) is necessary to achieve high-quality crystalline films with optimal optical constants (high $n$, low loss), it inevitably leads to the thickening of the SiO$_2$ interface and increased surface roughness. Engineers must balance these factors when designing waveguides or photodetectors.
• Optical Tuning: The ability to tune the refractive index and bandgap via annealing allows for the customization of Ga$_2$O$_3$ films for specific photonic applications, such as low-loss waveguides and integrated photonic circuits (PICs).
• Process Optimization: The results suggest that for applications requiring smooth interfaces, lower annealing temperatures or alternative annealing atmospheres (to suppress SiO$_2$ growth) might be necessary, whereas high-performance detectors requiring high crystallinity may benefit from the 1000°C treatment despite the interface changes.

In conclusion, the paper establishes that thermal annealing is a powerful tool for transforming amorphous RF-sputtered Ga$_2$O$_3$ into high-quality crystalline films with tunable optical properties, provided that the trade-offs regarding interfacial oxidation and surface roughness are managed.