Capitalizing on growth: the compound semiconductor epiwafer industry eyes opportunities in RF, power, and photonics – An interview with IQE

Compound semiconductors are surging in popularity in power electronics, radio frequency (RF), photonics, and microLED display applications, thus eclipsing silicon. At the core of these applications lie various compound semiconductor materials:

To achieve the expected performance from compound semiconductor-based devices, high-quality epitaxy is essential. This epiwafer market can be categorized as either captive, involving vertically integrated players (IDMs) with internal epitaxy capabilities, or open, involving epi-players offering epitaxy services with significant epitaxy capabilities.

As outlined in Yole Group’s recent report, Status of the Compound Semiconductor Industry, its analysts forecast the open compound semiconductor epiwafer market to reach $2.5 billion in 2029 with a 19% CAGR between 2023 and 2029.

To gain deeper insights into the dynamics of the compound semiconductor open epiwafer market, Yole Group’s analysts Dr. Ali Jaffal, Dr. Taha Ayari, and Mr. Aymen Ghorbel had the opportunity to engage in discussion with Dr. Rodney Pelzel, IQE’s CTO.

Discover today the conversation about current trends, opportunities, and challenges facing the open epiwafer business in photonics, power electronics, and RF applications.

Yole Group (YG): Please introduce yourself and your company.

Rodney Pelzel (RP): I am the CTO, managing a group of around 30-40 engineers/scientists (various levels), and am responsible for all the Group’s product and process technology. Stated simply, I make sure that our products and processes are best in class, giving our customers what they need at the price point required by us and them. I also ensure that we are positioned with next-gen technologies for the future.

Our products are the base materials for devices that CONNECT, SENSE, POWER, and DISPLAY. The reason IQE has limited its business to only epiwafers is that the value is in the epi (up to 80%, depending on the application). Epi is the lithography for CS materials.

YG: What are IQE’s activities in compound semiconductor epitaxy as well as your differentiators?

RP: IQE has a broad and deep product/technology portfolio. We manufacture epi wafers for all four of the III-V families (GaAs, InP, GaN, and GaSb). We have both metal-organic chemical vapor deposition (MOCVD) technology and molecular beam epitaxy (MBE) capabilities. We also have GaSb substrate manufacturing capability in the US (Spokane, Washington) and the UK (Milton Keynes).

We are differentiated from our competitors since we are the only pure-play epi-wafer producer that has (1) the entire portfolio of materials (competitors typically have only one or two) and (2) is global with manufacturing and development capabilities in the three key regions of the world (US, EU-UK, and Asia). We also span the full product life cycle from R&D to mass production.

IQE has dominated the GaAs VCSEL epitaxy market for several years, being a key supplier for major OEMs in the consumer space.

YG: How do you see the demand from the consumer evolving in the coming years?

RP: For 3DS, we continue to see evolution in design for the current applications, which will continue to be released in line with the market predictions. Ultimately, the same technology has applications in other markets, such as automotive LiDAR, and we expect development in these areas to continue. Although the technology progression is solid, the timing for market insertion remains uncertain.

YG: What are the current trends in this market and your technological roadmap that could help you maintain your current position?

RP: For 3DS VCSELs, the trends are to higher power and longer wavelength, an active area of development for IQE, where we are working with leading OEMs.

After a difficult 2023 for the optical communication segment, which strongly impacted not only the InP epiwafer business but also the InP substrate and InP bare die business, this year, the market seems to be recovering, driven by Artificial Intelligence.

YG: How big is IQE’s opportunity in the AI market? What are the main laser technologies based on InP and GaAs driving your AI photonics business?

RP: For InP, we are seeing an uptick in requests for InP edge-emitting lasers and DFBs. There is also activity in 850 nm GaAs VCSELs (for which IQE has a reference capability). Regarding the size of the opportunity, it is important to realize that the device sizes are much smaller than required by other applications (e.g., 3DS), and therefore, the associated wafer quantities are modest (albeit increasing).

Staying in the AI market, we are witnessing a growing interest in transitioning to 6” InP substrates to reduce costs and support the increasing demand for AI applications.

YG: How big is the 6” InP project at IQE? What are the bottlenecks that you are facing in this transition? Do you see concerns with using 6” InP in front-end manufacturing lines?

RP: We have a leadership position in this area through our government-funded projects, have demonstrated the epi capability at 6”, and are confident in our ability to supply into this market. One challenge for 6” InP pertains to substrate supply, more specifically the cost. The current price for 6” InP is a challenge. Certainly, the price will come down as volumes increase, but InP substrate manufacture is more challenging than GaAs, so yield is likely to be a factor, particularly at the larger wafer diameter. A second consideration is on the fabrication side. As InP is more brittle than GaAs, it is more prone to breakage and yield loss. This is a challenge that the 6” CS fabs will need to address.

In power electronics applications, we see strong demand for both SiC- and GaN-based devices. However, in Yole Group’s understanding, IQE mainly focuses on GaN epiwafer for power applications.

YG: What is the reason behind this choice? How confident is IQE about the GaN epiwafer business in power applications?

RP: It is true that IQE’s current focus is on GaN, and we are confident in the GaN market and our unique position in it. For GaN, the epitaxy defines the device; it is where the value is. This is the primary driver behind our GaN focus, and our position has intentionally leveraged our 20+ year history in GaN, which dates back to GaN/SiC for military and RF infrastructure applications, and we have a successful business in GaN RF today. 

IQE consistently reviews opportunities for SiC, and if/when there is a compelling business case where we can capture value, we will consider entering.

In the RF GaN business, GaN-on-SiC is still the mainstream wafer. However, we are witnessing initial shipments of GaN-on-Si for 5G base stations and growing interest from the mobile market.

YG: From the epitaxy point of view, what are the challenges of GaN-on-Si epitaxy for RF applications on 8”, and is it comparable to power GaN-on-Si? Are you planning to enter this business?

RP: We already have GaN on Si for RF and have been a leader in this area with work that originated on 4”. See public information: more info. As well as IQE Chosen as Key Supply Chain Partner for MACOM’s Global Gallium Nitride on Silicon Program. Today, we are working with leaders in GaN on Si for RF on 8” wafers. For example: the partnership with GlobalFoundries in 2021.

As GaN-on-Si for RF layer stacks are thinner, the challenges are different than for power. Wafer flatness, morphology, and defectivity are less challenging for thinner epi. However, RF has unique challenges. For example, the initiation of growth is critical for RF to avoid the creation of parasitics at the epitaxy/silicon interface that degrades performance.

RF GaAs technology is well established as a standard for applications such as mobile power amplifiers. The market is expected to remain steady. However, geographic restrictions and the development of a local ecosystem in China are impacting the market.

YG: How does IQE plan to position itself in this changing RF GaAs market in the coming years?

RP: This is one area where IQE’s global footprint is a critical asset. We are able to partition our technology development and create purely “local” technologies in different regions of the world. We have already had success with this, working closely with Asian supply chains.

YG: Would you like to add any closing thoughts for our readers?

RP: CS materials content in semiconductor devices is on the rise because they are essential for next-generation devices. This requirement has fundamentally shifted the innovation landscape for the semiconductor industry. For over 50 years, most semiconductor innovations have focused on CMOS silicon, where device design and fabrication have been the key enablers. In these instances, the starting materials, silicon wafers, have been a commodity and not an area of innovation and development. For compound semiconductors, this is no longer the case; the key enabler and differentiator is at the materials/epiwafer level. Advancing compound semiconductors requires fundamental materials engineering. Specifically, the enabling innovation is in the epitaxial growth process, thus overcoming technological challenges at the substrate level. As such, IQE is ideally positioned to enable the markets and capture value.

This interview has been developed in collaboration with Ezgi Dogmus, Activity Leader, Compound Semiconductors and her team at Yole Group.

Source: www.yolegroup.com