Let’s talk about the popular gallium oxide

Efficient ultra-high voltage power conversion devices (voltage >20kV) require semiconductors with a much larger energy gap than silicon. The wide-bandgap (WBG) semiconductor silicon carbide (SiC) has matured as a commercial technology platform for power electronics, but ultra-wide-bandgap (UWBG) (bandgap >4.5eV) semiconductor devices have the potential to enable higher voltage electronics. Candidate UWBG semiconductors include aluminum nitride (AlN), cubic boron nitride and diamond, but it is probably gallium oxide (Ga 2 O 3 ) that has seen the greatest increase in research activity over the past decade . This interest is due in part to its large bandgap of 4.85 eV and breakthroughs in crystal growth that led to the demonstration of the first Ga2O3 transistor in 2012 . Ga2O3 holds promise as a power electronics platform, but there are challenges in bringing this UWBG semiconductor to commercial use in the next decade .

The electrification process that has captured the attention of many industries could be disrupted if ultra-high voltage electronics penetrate applications such as next-generation grid control and protection, ultra-fast electric vehicle chargers, or efficient point-of-load converters with size, weight and power advantages. Disruptively accelerated. While silicon carbide devices cost more than traditional silicon power electronics, at the system level these costs are expected to be offset by savings due to simpler circuit requirements.

If a viable UWBG technology platform emerges, power conversion at very high voltages in excess of 20kV and high switching speeds can be achieved. Even at 10kv, it is difficult to increase the switching frequency of a power converter above 10khz without sacrificing circuit efficiency. UWBG semiconductors inherently require thinner device layers, thereby reducing conduction losses (proportional to channel resistance). The reduced carrier transit time through smaller UWBG devices will also reduce switching losses (proportional to capacitance) and provide a platform for high-speed electronics without sacrificing output power. Such high-speed power transistors will be disruptive in the power electronics industry because system volume is inversely proportional to frequency.

Among the six crystalline Ga 2 O 3 phases, low-symmetry monoclinic β-Ga 2 O 3 is furthest along in its development cycle due to its thermal stability at high temperatures (>650°C), discussed below Relates to this phase. Unlike other WBG or UWBG semiconductors, the melt growth method originally developed for silicon substrates has been used for commercial Ga2O3 substrates . Beta- Ga2O3 wafers have reached the 4-inch (100mm) commercial milestone and are on track to reach the 6-inch (150mm) size in 2027. At the same time , the infrastructure for high-quality epitaxy is expanding to keep up with growing Ga2O3 substrate sizes. Ga 2 O 3 epitaxial growth methods, such as chemical vapor deposition (CVD), molecular beam epitaxy and halide vapor phase epitaxy, are being extensively studied with the goal of producing the highest quality materials.

Although the basic infrastructure building blocks of UWBG technology have entered the development cycle, researchers are still actively exploring UWBG device architecture. Vertical field-effect transistors (FETs), such as FinFETs (see figure left), can theoretically block very high fields, but are more susceptible to extended defects in the epitaxial layer. Lateral transistors, such as heterojunction FETs (see figure right), have the potential to switch faster and more efficiently due to their smaller capacitance and shorter transfer times, and it can also use Ga2O3 ternary alloys , in In this case it is β-(AlxGa1 - x ) 2O3 to further improve power performance .

Gallium oxide (β-Ga 2 O 3 ) device

The presence of shallow energy donors and acceptors (charged impurities) plagues all UWBG semiconductors, as increasingly wider energy gaps typically cause external impurities to reside further away from the conduction band (or valence band). However, silicon is an excellent external shallow donor for Ga2O3 , enabling a wide range of controllable conductivities from ^{below 10} cm to ^{above 10} cm . Controllable n-type conductivity even extends to the ternary alloy (Al x Ga 1-x ) 2 O 3 , which has a wider bandgap that can also be tuned based on phase and Al concentration. Furthermore, the purity of CVD-grown homoepitaxial β-Ga2O3 is surpassed only by silicon. Recently, homoepitaxial CVD ^Ga 2 O 3 has extremely ^high low-temperature mobility (23000 cm ² V ^{- 1} s ^-1 ), which may originate from point defects formed unintentionally in the crystal lattice.

However, growing very thick (>30 μm) epitaxial β - Ga2O3 at this purity level is very challenging and its development needs to compete with SiC for ultra-high power switching applications. The understanding of Ga2O3 epitaxial defects must advance over the next few years before high-voltage Ga2O3 devices can be commercialized . Point defects, such as vacancies and their associated complexes (e.g., vacancy-interstitial defects), as well as extended defects in thick epitaxial layers, currently inhibit Ga2O3 device size . Overall, defect characterization in Ga2O3 promises to be a rich area of ​​research that will also enable any Ga2O3 power electronics commercial enterprise looking to break the 20kv barrier with useful device sizes .

For power electronics, the development of p - type (hole carrier) materials is necessary because holes in Ga2O3 form localized polarons, leading to self-trapping phenomena that limit their conduction. Regardless of device geometry, the lack of p-type conductivity in Ga2O3 poses challenges for high electric field management, and any practical solution will require innovations in heterogeneous integration not faced by previously developed semiconductors.

Unlike p-type semiconductors, such as SiC, gallium nitride (GaN) or diamond, WBG p-type nickel oxide (NiO) can be sputtered at room temperature, thus facilitating integration with Ga 2 O 3 devices. Recent studies, such as the 8 kv NiO/Ga2O3 pn diodes demonstrated by Zhang et al. , have shown that Ga2O3 can potentially be managed by combining heterojunctions with field management and charge balancing in these devices Lack of medium p-type conductivity. The prospects of Ga2O3 as a power electronics material will be greatly enhanced if robust heterogeneous integration with p-type WBG semiconductors such as GaN or AlN is developed . Such a development could lead to the commercialization of reliable junction barrier Schottky rectifiers, as is the case with SiC.

A key requirement for using UWBG materials in practical high-voltage electronic devices is effective electric field termination at the surface. Nitrogen deep domination is effective in making Ga2O3 nearly insulating and creating an effective dielectric layer that reduces electric fields. Selective ion implantation can create conductive and insulating surface areas during device fabrication. Dry etching, a common processing step to create such patterns, can introduce surface defects that impact device reliability. If patterning could be achieved entirely by ion implantation, dry etching might be eliminated entirely. Unlike other UWBG materials, Ga2O3 can even be wet etched in phosphoric acid and etched using vapor phase Ga, both of which eliminate chemical and mechanical damage from plasma etching, which will always Etching the surface introduces defects. Developing high-quality , thick epitaxial layers while developing Ga2O3 - specific manufacturing processes could accelerate the commercialization of Ga2O3 devices over the next decade , at least to the scale of two-terminal devices such as diodes.

The extremely low thermal conductivity of Ga 2 O 3 (11 to 27 W m ^-1 K ^-1 ) must be carefully considered . Cooling of Ga2O3 transistors is even more critical than GaN transistors, which also have self-heating effects . Although the power output by Ga2O3 devices during operation is still an order of magnitude lower than that of GaN devices, the top and bottom side cooling methods developed for GaN can be applied to Ga2O3 . In fact, coating lateral transistors with AlN or nanocrystalline diamond can achieve DC output powers of 5-6w mm for Ga2O3 , similar to early results with GaN high electron mobility transistors ⁱⁿ the 1990s. Heterogeneous integrated WBG p-type semiconductors with high thermal conductivity, such as SiC, GaN, or even diamond, are particularly suitable for pn and junction barrier Schottky rectifiers.

Looking back at WBG Semiconductor’s early commercialization efforts, SiC’s success was driven in part by substantial government investment and continued innovative scientific research efforts. Solving the problem of silicon carbide microtubule and basal plane dislocation defects relies on advanced characterization techniques such as UV photoluminescence imaging and spectroscopy. Materials scientists continue to develop their understanding of defects in SiC wafers with larger diameters.

Similar efforts are required to understand and control point and extended defects in thick (>30 μm) Ga2O3 epitaxial layers . Government funding is critical to support these efforts early on. The U.S. Office of Naval Research launched a Small Business Technology Transfer program in 2017 to initiate the development of beta-Ga2O3 CVD , which commercialized the capability before the end of the program, highlighting the importance of this new technology . . The recently enacted American Chips and Science Act will not only provide funding for chip manufacturing facilities, but will also provide $13 billion to the U.S. Department of Commerce and the U.S. Department of Defense for semiconductor and microelectronics research and development. These investments will spur additional funding for UWBG semiconductor and related materials research in the coming years, with the expectation that a diverse portfolio of heterogeneous integrated semiconductor modules will overcome the shortcomings of chips fabricated using specific semiconductors. Furthermore, higher frequency devices will only become useful at the system level if passive devices can keep up. Advances in magnetic materials also help prevent components such as inductors and transformers from becoming too lossy at higher frequencies.

*Disclaimer: This article is original by the author. The content of the article is the personal opinion of the author. The reprinting by Semiconductor Industry Watch is only to convey a different point of view. It does not mean that Semiconductor Industry Watch agrees or supports the view. If you have any objections, please contact Semiconductor Industry Watch.

Today is the 3232nd content shared by "Semiconductor Industry Observation" with you. Welcome to pay attention.