Anyone whose rearview camera has saved them from a broken bumper knows the value of eyes in the back of your head. That’s why new cars come complete with rear-facing visible cameras, RADAR and near-infrared LiDAR sensors that will tell you when something is getting too close. LiDAR is currently the workhorse of automated driving; its ability to sense the distance of objects is why it can perform the perfect parallel park.

But LiDAR has its limitations. It can sense the distance of an object, but not what an object is. It can tell that something is standing by the side of the road but it doesn’t know if that thing is a kangaroo. If LiDAR could make that distinction, the car’s automation system could prepare for a sudden impact.

Infrared (IR) detectors are able to determine what an object is based on its unique thermal signatures. It can tell a rubbish bin from a child. The military uses it to determine if a floating structure is a ship or organic material. The most expensive consumer car brands, such as Audi, have included IR detectors in their vehicles but even they are limited to IR detectors made from Silicone-based CMOS* technology and/or III-V Indium Gallium Arsenide (InGaAs), which only detect wavelengths in the range of near-infrared to short-wave infrared, require a light source, frequent replacement of IR illuminators, and have limited sensitivity.

IR detectors made from Mercury Cadmium Telluride (MCT) are far superior, detecting a much wider range of wavelengths with greater sensitivity and less noise in a faster timeframe. They are the detector of choice for the military. They are also incredibly expensive, far beyond what commercial car manufacturers can include in their products, and in order to provide the necessary 360 degree vision, they require bulky, complex optics or multiple sensors to achieve a wide field of view.

Conventional flat IR detectors distort the edges of an image due to the way light passes through a lens and lands on the sensor. To correct this problem, high-end IR cameras use multiple lens elements to flatten the focusing plane, but these systems can be large, bulky, and expensive. Using a curved sensor instead of a flat one reduces edge distortion. With a curved imaging plane, you only need one lens, which makes this technology smaller, lighter and cheaper.

Researchers at TMOS are working on curved/flexible MCT detectors, as published in Advanced Materials Interfaces. These detectors are orders of magnitude smaller, decreasing the cost of production significantly. Because they are so thin and are grown on a 2D material with a weak bonding between the sensor and the substrate, they can be peeled off, bent, and placed on a curved plane that mimics the retina of our eyes. This increases the sensor’s field of view without the need for additional optics.

Lead author Wenwu Pan says, “We have discovered a way to grow high-quality MCT thin films IR detectors using a method called van der Waals epitaxy on 2D layered substrates. This created weak bonds, which allows us to lift the thin film off the substrate so that we can shape it into a curved IR imaging array. This could lead to simpler lenses, wider field of view, and better imaging quality in comparison to conventional flat IR detectors.”

TMOS Chief Investigator Lorenzo Faraone says, “The integration of MCT detectors with optical metasurfaces has the potential to improve IR detection and provide new functionality. In addition to the application in curved imaging sensors, a free-standing MCT thin film offers even more design flexibility and fabrication options for realizing these integrated structures. For instance, these MCT device layers can be transferred or integrated onto various optical platforms using metamaterial layers, creating more complex and versatile optical systems, including higher performance imaging devices for long-range imaging and advanced sensors with polarization sensitivity for enhanced target identification.”

For more information about this research, please contact connect@tmos.org.au

* CMOS stands for Complementary Metal–Oxide–Semiconductor

Van der Waals Epitaxy of HgCdTe Thin Films for Flexible Infrared Optoelectronics

Wenwu Pan, Zekai Zhang, Renjie Gu, Shuo Ma, Lorenzo Faraone, Wen Lei
Advanced Materials Interfaces 4th December 2022

Van der Waals epitaxial (vdW) growth of semiconductor thin films on 2D layered substrates has recently attracted considerable attention since it provides a potential pathway for realizing monolithically integrated devices and flexible devices. In this work, direct growth of epitaxial HgCdTe (111) thin films on 2D layered transparent mica substrates is achieved via molecular beam epitaxy. The full width at half maximum of the ω-mode X-ray diffraction peak is measured to be around 306 arc sec. Mid-wave infrared photoconductors based on the as-grown HgCdTe thin films have been demonstrated and the self-heating effect has been evaluated. A peak responsivity at the wavelength of around 3500 nm is measured to be about 110 V W⁻¹ at 80 K and 8 V W⁻¹ at room temperature under a bias of 25 V cm⁻¹. Twinning defects are observed, limiting the crystallinity and mobility-lifetime product in HgCdTe/mica. Benefiting from the vdW epitaxial growth, an etch-free layer transfer process for lifting off the HgCdTe from the mica substrate has been demonstrated, leading to large area free-standing HgCdTe thin films.