Emerging Uses of Photonic Device Technologies-EEWORLD

Collect

Photonic device technology has long existed in applications such as laser scanning and printing, telecommunications, and industrial material processing. In recent years, light-emitting diode (LED) lighting has been widely used. Photonic devices such as lasers, photodetectors, microLEDs, and photonic integrated circuits (PICs) have become building blocks for a range of new technologies, including facial recognition, 3D sensing and laser imaging, detection and ranging (lidar), etc. To meet the needs of today's applications, these technologies require innovative device architectures, new material development, monolithic and heterogeneous integration of materials, larger wafer sizes, and single-wafer processing.

introduction

Silicon has been the backbone of all semiconductor IC technologies, making the development of electronic technology possible, from computers, the Internet, smartphones, and now artificial intelligence and 5G. However, for some applications, photonic device technology better meets the technical and environmental requirements.

Compound semiconductors such as gallium arsenide (GaAs), indium phosphide (InP) and gallium nitride (GaN) have direct energy band gaps to support photonic device technologies such as lasers and LEDs. 1.3μm and 1.5μm single-mode lasers using indium gallium arsenide phosphide (InGaAsP) materials can build extremely efficient fiber-optic communication systems and are already in use today.

With the advancement of visible light LED technology based on gallium arsenide and gallium nitride, the lighting industry has produced efficient and high-brightness LED products for indoor and outdoor lighting, automotive lighting and displays. In addition to energy efficiency, LEDs also provide lighting designers with greater freedom, which can be seen from the latest automotive headlight design (Figure 1).

Figure 1. High-brightness LEDs are used in recent automotive headlight designs.

Emerging Photonic Applications

Photons are considered an important enabler of emerging technologies such as 3D sensing, autonomous vehicles, and optical interconnects. Just as electronics have been the backbone of designing the "brain" of machines, photons will give "vision" to future machines, and lasers will be the source of these photons.

3D Sensing

As smartphones are increasingly used for computing, more personal information is retained on them, requiring a rigorous security setup that goes beyond fingerprint recognition and two-dimensional iris scanning-based authentication. Vertical-cavity surface-emitting lasers (VCSELs) have attracted considerable attention in the consumer market in recent years, following Apple's introduction of facial recognition in the iPhone X in 2017. VCSELs shine tens of thousands of laser beams onto a user's face, which are then collected to generate a 3D depth map of the face, creating a unique identification image for that user (Figure 2).

Figure 2. VCSEL is the basis for facial recognition technology used in devices

The latest product from a leading consumer product manufacturer extends this technology to use a time-of-flight laser sensor that uses a VCSEL to flash a scene several meters away, creating a 3D image of the space with depth information. For example, furniture or artwork can now be placed virtually in a space to see how it will look before purchasing. Today’s technology is limited in wavelength range for eye safety, but we can expect future developments to longer wavelengths and more devices, including smartphones.

Optical Interconnect

Traditional data centers consume more than 2% of the world's electricity today, while global data traffic is expected to double every four years. In the future, using electronic packet switches to transmit data between racks will not be able to meet both bandwidth and energy requirements. The shift in data center business models to cloud computing will involve even greater amounts of data processing and transmission in the coming years (Figure 3).

Figure 3. Cloud computing will intensify the challenge of data center energy consumption

Optical interconnect technologies based on silicon photonics and indium phosphide photonic integrated circuits (PICs) are currently being developed to address these challenges facing data centers. 100GbE transceiver modules are already on the market and are steadily advancing toward 400GbE and beyond. Silicon photonics enables faster and longer-distance data transmission than conventional electronics, while also taking advantage of the efficiencies of semiconductor lasers and high-volume silicon manufacturing.

LiDAR

In addition to electrification, the next big paradigm shift in the automotive industry is autonomous driving. Today's Level 3 autonomous driving requires highly sophisticated lighting, detection, perception, and decision-making systems to work seamlessly together (Figure 4). LiDAR's high resolution, 3D imaging capabilities, and a detectable range of more than 200 meters are in stark contrast to radar or camera-based solutions, and it has been widely recognized as the best solution for autonomous driving.

Figure 4. Safe operation of autonomous vehicles requires the seamless collaboration of many different systems.

LiDAR is available in two frequency options: 905nm and 1550nm. 905nm is the preferred choice because it has a well-established ecosystem of lasers and photodetectors. However, the industry is actively investigating 1550nm because it has a wider range and an eye safety limit that is 40 times that of 905nm. Beam steering technologies currently being evaluated include mechanical rotation, MEMS, and optical phased arrays. Mechanical rotation has significant reliability issues, while MEMS-based beam steering technology has recently appeared in many cars as an option for Level 3 advanced driver assistance systems (ADAS), but has limited range and field of view. Solid-state optical phased arrays for beam steering are in the early stages of development and have good performance, cost, and form factor prospects, and can be used in more areas besides autonomous driving. In order to meet the cost and performance requirements of LiDAR systems, heterogeneous integration or co-packaging of lasers, detectors, and beam steering chips in high-volume manufacturing is required. Today, MEMS-based LiDAR technology shows promising prospects in meeting these industrial requirements.

MicroLED

In addition to enabling higher resolution in existing devices such as TVs, smartphones, and smartwatches, microLED technology has the potential to enable exciting new products, such as augmented reality/virtual reality (AR/VR) products, as shown in Figure 5. These new applications require self-emissive red, green, and blue (RGB) displays, without color conversion or filtering. The challenges involved are achieving the required quantum efficiency of RGB microLED die, cost-effectively transferring microLEDs to backplanes in large quantities, and testing each individual microLED. Innovative device designs, epitaxial growth optimization, substrate engineering, die transfer methods, and new backplane architectures are being researched and developed to make microLED technology competitive with existing liquid crystal display (LCD) and organic light-emitting diode (OLED) technologies.

Figure 5. AR/VR applications are among the consumer products that benefit from MicroLED technology.

Device Technology

The key device technologies that enable these emerging photonic applications are gallium arsenide and indium phosphide-based lasers, silicon and indium gallium arsenide (InGaAs) photodetectors, MEMS devices, gallium nitride and gallium arsenide LEDs, silicon and silicon nitride (SiN) waveguides, and optical modulators. For 3D sensing applications, gallium arsenide laser devices are moving from 100mm substrates to 150mm substrates. Gallium arsenide and gallium nitride LEDs for high-brightness applications are produced on 150mm gallium arsenide substrates and sapphire substrates, respectively. However, in some applications, the use of microLEDs is driving the demand for RGB LEDs on silicon substrates. Indium phosphide laser diodes are produced on 75mm and 100mm indium phosphide substrates. Compound semiconductor devices are typically processed in batch reactors, but the manufacturing focus is increasingly on improving yield and intra-wafer uniformity and enhancing process control, which in turn drives the transition to single-wafer processing equipment.

Currently, MEMS devices for beam steering technology rely on 200mm silicon MEMS production lines. Silicon photonics technology mainly operates on 200mm silicon-on-insulator (SOI) platforms and continues to push the transition to 300mm wafers to address the technical limitations of 200mm lithography and etching equipment. Thin-film technologies with high electro-optic coefficients have been under research to expand the speed and bandwidth envelope of optical interconnects.

The above photonic applications are expected to achieve huge growth in the next 5-10 years. The market size of the four key applications of 3D sensing, LiDAR, optical interconnection and AR/VR display is expected to grow at a compound annual growth rate of 31%, from US$8 billion in 2020 to US$23.3 billion in 2025 (Figure 6). 3D sensing technology is seeking new applications, while LiDAR and AR/VR displays are still in the early stages of development and are expected to grow at a higher compound annual growth rate. The growth of optoelectronic applications will require solving the challenges of device technology in performance, manufacturing and system integration. Today, various forces are driving the demand for new process equipment that can not only solve the difficulties in device performance, but also achieve excellent process control and improve overall manufacturing yield.