The VCSEL industry has witnessed strong growth in the past few years, since the emergence of 3D sensing in smartphones. According to Yole Développement’s latest report “VCSELs – Market and Technology Trends 2020“, the VCSEL market for mobile and consumer applications is expected to grow from $844M in 2020 to $2.1B in 2025, representing a 20% compound annual growth rate (CAGR). Since their integration in smartphones, VCSELs are progressing, with higher optical output power and a reduced chip size.
Pierrick Boulay, Market and Technology analyst at Yole Développement, had the opportunity to talk about VCSELs with Mary Hibbs-Brenner, CEO of Vixar, an Osram Opto Semiconductor company. Discover the detail of their discussion below.
Yole Développement (YD): Can you please introduce yourself, your activities and the activities of Vixar?
Mary Hibbs-Brenner (MHB): I got started working on VCSELs in the mid-1990s at Honeywell. Honeywell became the leading supplier of VCSELs in the data communication market until the business was sold to Finisar in 2004. Klein Johnson and I then started Vixar to develop VCSELs for the sensor market. We initially developed red VCSELs and moved on to a range of wavelengths up to 1000nm, as well as a range of output power. Since 2018 Vixar has become part of Osram and I am responsible for their VCSEL group. We are now supplying VCSELs for 3D sensing in consumer, industrial, medical and automotive applications, as well as other VCSELs for things such as atomic sensors, and medical treatment applications.
YD: Vixar recently showed progress with multi-junction VCSELs. What are your expectations for this technology? What would be the targeted applications for such components?
MHB: The technical advantages of multi-junction VCSELs include the ability to produce a chosen amount of optical output power at a much reduced current, which in turn allows for a fast-pulsed rise time to reach that peak power. In addition, the overall power conversion efficiency is improved. This is beneficial for 3D sensing and imaging applications in consumer, industrial and automotive industries. It also enables a reduction in chip size, and therefore cost. We expect multi-junction VCSELs will become the standard for 3D sensing and will open up a wider range of VCSEL-based LiDAR applications.
YD: The number of VCSEL manufacturers for smartphone is still limited to 6-7 main players. Why are so few players involved? Do you see newcomers emerging? How can they enter the market?
MHB: The very large consumer companies want suppliers who have the capacity, quality systems, R&D, and financial resources to ramp up quickly and weather the inevitable bumps in the road. There are a limited number of players that can do that. Furthermore, the R&D projects in the mobile area require a high degree of flexibility for a smooth and timely product launch.
It is hard for a new entrant, and particularly a start-up company, to gain a foothold in an already established market. A more promising approach is to look for a new application or niche market that you think could develop into a larger market in the future, or possibly a new solution for an existing market that leapfrogs the existing solutions. Start-up companies, especially, need to demonstrate a technology which is both unique and has a clear value proposition. Then you need to establish partnerships that will help you get across the finish line.
YD: Most LiDAR manufacturers using VCSELs are using addressable VCSEL arrays. What are the challenges associated with it manufacture such VCSELs? Are they related to the high number of emitters? Or the definition of addressable zones?
MHB: A small number of addressable zones is quite easy to implement – simply a mask change of the metal contact layer. Matrix addressing of a larger number of segments implies a two-level metal, which requires some process development, but is not too challenging. A higher degree of addressability of a larger number of segments is probably best addressed by a flip chip design, with the routing accomplished in the submount. If the submount is not transparent, then back-side emission is necessary. A flip-chip, back-side emitting VCSEL design requires changes to the epitaxial layers, process flow, mask layout and packaging, so is a much bigger investment.
YD: Red VCSELs have a slightly different materials structure than infrared devices, making them more temperature sensitive. Are they still limited to niche applications like printing, industrial sensing and medical, or do you see some killer applications coming?
MHB: It remains to be seen whether a killer application will develop. Some candidates include wireless battery powered consumer health and medical sensors like pulse oximetry, near-to-eye displays, probably also requiring blue and green VCSELs, or personalized medicine such as fluorescence-based gene sequencing and molecular diagnostics – particularly point of care.
YD: Currently, bulky secondary optics are implemented on top of VCSELs in 3D sensing modules. Wafer level optics and micro lens arrays could also be used. What would be the drivers of such optics? What are the impacts on the VCSEL structure? Is there use of back-side emitting VCSELs? Or more complex manufacturing/packaging?
MHB: The drivers for optics integration are the usual ones of size and cost. It is hard to make a general statement about the benefits of the optical integration, as there are performance, process, test, size and cost trade-offs. Let’s take the case of the integration of collimation lenses with VCSELs. To get the optical standoff distance required for the focal length of a lens, you would most likely integrate on the back side of the VCSEL wafer. Theoretically, you could replace an expensive chip by chip alignment of a lens or lens array to the VCSELs with a wafer scale alignment of lenses in one or more photolithographic steps. But there are other things to consider. Is the integrated lens quality sufficient? What is the cost increment of the more complex back side emission process? What are the implications for test if the electrical contacts are on the opposite side from the light emission? So, given these challenges, we need to take it on a case by case basis to determine if optics integration is the right answer for an application. And, very importantly, the manufacturing infrastructure must exist in order to support implementing these new technologies in high volume.
YD: It seems that there are no standards regarding VCSEL drivers and each smartphone OEM uses custom-made drivers. What are the challenges associated with VCSEL drivers? Why are these components still customized? Do you see some standardization coming?
MHB: In the data communication market, interoperability drove standardization, but that is not really an issue in this market. I think we are currently at a stage where every solution provider is trying to develop a different offering. There are enough variables that it is difficult to come up with a driver chip that would satisfy the requirements of the wide range of solutions. Segmented chips, approaches taken for feedback of eye safety, the optical power level, the use of multi-junction vs. single junction chips are all variables that affect driver design. Perhaps as the market matures, a few solutions will become the winners, and standardization can start to occur.
YD: Is there an advantage to place the VCSEL driver close to the VCSEL? What are the advantages to use GaN-based VCSEL drivers? Is it only limited to LiDAR applications or can it also be used in smartphones?
MHB: In time of flight (ToF) or LiDAR applications, rise time of the pulse is important to distance resolution, and so the inductance between the driver and VCSEL is important. The problem becomes even more challenging when pulsing with high current, making the location of the driver relative to the laser key for LiDAR, and direct ToF. While it still matters for indirect ToF (iToF) in mobile phones, it is somewhat less critical. For LiDAR, where one of the key challenges is driving enough current into the VCSEL, the GaN based drivers appear to be the preferred solution today. At the other end of the power range, the drivers for iToF have incorporated other functionality, like processing the signal from an eye safety mechanism. This causes low-cost CMOS drivers to be attractive for consumer applications where required power is limited and cost is a primary consideration.
YD: Do you have a final word for our readers?
MHB: I’ve long been a believer in the benefits and potential of VCSEL technology and continue to be optimistic about the long-term prospects for the industry.
Dr. Mary Hibbs-Brenner received a B.A. in Physics from Carleton College, a Ph.D. in Materials Science from Stanford University, and an MBA from the University of Minnesota. She is currently the CEO of Vixar, a subsidiary of Osram Opto Semiconductors. Mary co-founded Vixar in 2005 to commercialize VCSELs for sensor applications. Previously, Mary held business development, technical management and research positions at Honeywell. She was responsible for initiating Honeywell’s VCSEL activity and directed R&D within that business unit. She spent two years doing post-doctoral research at the University of Linköping in Sweden.
As part of the Photonics, Sensing & Display division at Yole Développement (Yole), Pierrick Boulay works as Market and Technology Analyst in the fields of Solid State Lighting and Lighting Systems to carry out technical, economic and marketing analysis. Pierrick has authored several reports and custom analysis dedicated to topics such as general lighting, automotive lighting, LiDAR, IR LEDs, UV LEDs and VCSELs.
Prior to Yole, Pierrick has worked in several companies where he developed his knowledge on general lighting and on automotive lighting. In the past, he has mostly worked in R&D department for LED lighting applications. Pierrick holds a master degree in Electronics (ESEO – Angers, France).
VCSELs – Market and Technology Trends 2020
VCSEL market growth is triggered now but still under evolution. Changes are happening at design, manufacturing, supply chain and application levels.