Silicon CCD Camera Peers into Mid-Infrared

Researchers at the University of California, Irvine used the nonlinear optical properties of the silicon (Si) chip in a Si-based camera to enable a mid-infrared (MIR)-specific response in the camera. Their method for detecting MIR images with a Si-based camera does not rely on phase matching, is alignment free, and does not require complex post-processing of the images.

Unlike MIR cameras, Si-based cameras exhibit low noise characteristics and have high pixel densities, making them more attractive candidates for high-performance imaging applications. Currently, the most direct way for achieving the MIR-to-visible conversion required by an Si-based camera is through the use of a nonlinear optical crystal. When the MIR light and an additional near-infrared (NIR) light beam are coincident in the crystal, a visible light beam is generated through the process of sum-frequency generation. Although this up-conversion trick works well, it is sensitive to alignment and it requires numerous orientations of the crystal to produce a single MIR-derived image on the Si camera.

To simplify MIR imaging with Si-based cameras, the Irvine researchers used the process of nondegenerate two-photon absorption (NTA) in a standard Si-based charge-coupled device (CCD).

Artistic rendering of the principle of nondegenerate two-photon absorption for the detection of mid-infrared by a silicon-based camera. In this detection technique, the sensor is illuminated directly by the mid-infrared light beam, while a second, near-infrared beam is also incident on the sensor. The energies of the mid-infrared and near-infrared photons combine to excite charge carriers in the silicon material, inducing a response in the camera. This method enables fast mid-infrared imaging with regular silicon-based cameras. Courtesy of David Knez, Adam Hanninen, Richard Prince, Eric Potma, and Dmitry Fishman.

First the team confirmed that Si was a suitable material for MIR detection through NTA. Using MIR light with pulse energies in the femtojoule range, the researchers found that NTA in Si could efficiently detect MIR wavelengths, and they performed vibrational spectroscopy measurements of organic liquids using a Si photodiode as the detector.

The researchers then replaced the photodiode with a CDD camera. Using an MIR beam in combination with an NIR pump beam, they induced a two-photon absorption process in Si. They sent the pump beam to the camera, and they sent the MIR beam through the sample to be imaged before it reached the camera. An image was produced through two-photon absorption of the pump and MIR light in the camera’s CCD sensor.

Through NTA, the researchers were able to capture MIR-derived images on a 1392- × 1040-pixel sensor at 100-millisecond exposure times, yielding chemically selective images of several polymer and biological materials as well as living nematodes. Despite using technology not specifically optimized for NTA, the team was able to detect small changes in optical density in the images.

“We have high hopes that the simplicity and versatility of this approach allows for the broad adoption and development of the technology,” researcher David Knez said. The use of NTA to detect MIR light for imaging could speed up analysis in pharmaceutical quality assurance, geologic mineral sampling, and microscopic inspection of biological samples.

The Irvine team was led by Dmitry Fishman and Eric Potma.

The research was published in Light: Science & Applications (www.doi.org/10.1038/s41377-020-00369-6).