Access Official Technical PDF

Share Summary:LinkedInEmail

CMOS image sensors are widely used, but conventional designs can have limited sensitivity in low-light and near-infrared conditions. A sensor architecture combining CMOS integration with photodiodes designed for internal gain aims to address that limitation. The approach is being developed for applications such as solid-state LiDAR and passive imaging, where detecting weak optical signals can help a system capture useful information from its surroundings.

The central idea is to pair CMOS manufacturing advantages with light-absorbing materials and device structures that can detect very low light levels. For industrial automation engineers and system integrators, this points to a potential path toward imaging arrays that combine high sensitivity with compact electronics. The technology remains under development, so its practical value will depend on how designs progress from prototype arrays to application-ready sensors.

Why increase sensitivity in CMOS image sensors?

Image sensors convert incoming light into electrical signals that can be processed into an image or used to detect objects. In low-light scenes, weak signals are harder to distinguish, which can constrain imaging performance. Near-infrared (NIR) sensing presents an additional challenge: conventional CMOS image sensors may have low sensitivity in this part of the spectrum.

Improving a pixel’s ability to detect light can support earlier or more reliable detection in imaging and ranging systems. For LiDAR, which uses light to sense distance, sensitivity is relevant to receiving reflected signals. For passive imaging, it can help capture information without relying on an active illumination source. The source technology describes a goal of single-photon detection or near-photon counting, alongside high dynamic range.

These terms describe different aspects of sensing. Photon counting concerns detecting individual light particles, while dynamic range describes the span of signal levels a sensor can handle. In combination, they are intended to support imaging across varied light conditions. The stated application targets include visible and NIR imaging, as well as short-wave infrared (SWIR) sensing.

Internal-gain photodiodes and CMOS integration

The proposed sensor uses silicon-germanium-carbon (SiGeC) photodiodes integrated with CMOS circuitry. According to the technology description, the photodiodes provide large internal gain at low voltage, using a mechanism described as noiseless and different from conventional avalanche gain. Internal gain amplifies the signal generated by absorbed light before it is read by the surrounding circuitry.

That distinction matters because sensitivity is not only about producing a larger signal. The device must also manage noise, power requirements and integration with the readout electronics. The stated design objective is to achieve high gain at CMOS voltages without relying on a separate readout integrated circuit (ROIC). Monolithic integration—placing the light-sensing elements and associated CMOS circuitry together—can also allow pixel and sensor circuitry to be designed as part of one platform.

The technology materials report gain above 100,000 at voltages below 3.5 V for the described SiGeC photodiode. These are development specifications, not a guarantee of performance across every sensor design or operating condition. Engineers evaluating the approach would need to consider measured sensitivity, noise, operating range, array uniformity and interface requirements for their specific use case.

Group-IV materials extend the spectral opportunity

A second part of the approach is the use of Si-Ge-C superlattices as light-absorbing regions within photodiodes. A superlattice is a deliberately structured arrangement of very thin material layers. By engineering the material structure, developers aim to tailor optoelectronic properties and extend the wavelengths a silicon-based sensor can detect.

The stated development direction includes superlattice films for SWIR sensing, with a target absorption coefficient comparable to InGaAs around wavelengths up to 1.6 micrometers. The broader technology vision describes the possibility of covering multiple spectral regions, from ultraviolet to long-wave infrared, through different material structures. Those are development goals; the specific spectral range and performance of a sensor depend on the material and device implementation.

Using Group-IV materials is intended to bring some of the light-sensing capabilities associated with compound semiconductors into a CMOS-compatible manufacturing context. The potential benefit is not simply a different detector material: it is the prospect of integrating sensitive pixels, circuitry and larger arrays within a CMOS-based process.

What this could mean for LiDAR and machine vision

For automation applications, a more sensitive imaging array could be relevant where systems must detect objects or features in challenging lighting or use infrared wavelengths. The proposed uses include solid-state LiDAR, visible and NIR imaging, and SWIR imaging. In principle, CMOS integration may support smaller, lighter and lower-power camera designs, while commercial CMOS manufacturing is presented as a route to high-volume production and lower manufacturing cost.

These potential advantages should be assessed against system-level requirements. A plant or equipment OEM would need to establish whether the sensor’s spectral response, dynamic range, frame rate, resolution and environmental suitability fit the application. Integration with optics, illumination, processing hardware and control systems also affects the final system. The source describes improved frame rate and imaging capability as intended benefits, but does not provide comparative frame-rate measurements.

One development roadmap described a first-generation 128-by-128 array for visible and NIR imaging, followed by a planned one-megapixel SWIR array using Group-IV superlattices. These roadmap details illustrate the progression from device development to larger imaging arrays; they should not be interpreted as confirmation that the planned generations are commercially available.

Manufacturing and engineering considerations

CMOS compatibility is a key part of the proposition. The development organization describes a fabless model and experience with U.S.-based foundry manufacturing, including test-chip work. Its technical activities include photodiode and pixel design, sensor circuitry, simulation of materials and devices, and fabrication flows. This range of work reflects the engineering challenge: material properties, device structure, pixel layout and readout design must work together.

For technical decision-makers, the next step is to evaluate evidence at the component and system levels. Relevant questions include:

  • What sensitivity and noise performance have been demonstrated across the intended wavelength range?
  • How do array size, pixel design and dynamic range match the sensing task?
  • What are the power, voltage and readout requirements in the target system?
  • What manufacturing and reliability data are available for the intended production scale?

Practical implications

CMOS image sensors with internal-gain photodiodes and engineered Group-IV materials represent a development path toward more sensitive visible and infrared imaging. If the approach achieves its stated objectives, it could combine low-light detection and broader spectral coverage with CMOS integration for LiDAR and passive imaging applications. For automation teams, the practical priority is to track validated device and array performance, then assess how it fits the full sensing and control architecture—not to assume that a promising material concept alone guarantees a production-ready system.

Access Official Technical PDF

About The Author