Voxtel claims huge gains with new detectors
Oregon-based Voxtel has launched a new class of photodiodes that could open up new commercial and military applications for InGaAs detectors.
The company has developed “carrier multiplication devices”, or CMDs, that it claims overcome many of the problems associated with conventional avalanche photodiodes (APDs).
“These new photodetectors exceed the capabilities of APDs in both gain and noise performance,” said Voxtel.
According to company CEO and founder George Williams, the improved performance results from the very different epilayer structure and doping strategy employed in the InGaAs/InAlAs structures.
“This is a new type of device,” Williams said. “The way that it operates is fundamentally different [to APDs].”
The CEO explains that the noisiness of an APD relates directly to the ratio between electron and hole ionization coefficients. For a conventional InGaAs APD, this ratio is approximately 0.4.
As a result, near-infrared APDs can only reach a maximum gain of about 15. And even at a relatively modest gain of about seven, the noise of the avalanche process begins to impact detector performance, he told compoundsemiconductor.net.
In the CMDs, however, the epiwafer structure means that detector noise is not directly dependent on the ionization properties of the materials. As a result, they can produce gains of the order of 10,000.
“This is achieved by modulating the electric field and doping the device in a specific pattern, such that only one carrier gets amplified,” Williams said.
Multi-layered strategy
APDs typically employ a monolithically composed, intrinsic InP multiplication layer for amplification. Instead of this, the CMDs feature 20 or more layers of different quaternary alloy compositions such as InAlGaAs or InGaAsP.
Doped in a specific pattern and with varying layer thicknesses, this structure modulates the electrical field potential to preferentially allow only one charge carrier to multiply.
Voxtel, which Williams says has been granted a patent protecting the CMD concept, has out-sourced device manufacture to three different foundries. Either MOCVD or MBE can be used for CMD epiwafer production.
And although the cost of that epiwafer production might be higher for CMDs than that for conventional APDs, this is more than off-set by the reduced overall system costs that result.
“The usefulness of an APD is severely restricted "“ sensitivity is low,” said Williams. “This means that for telecommunications applications, higher-power lasers are required and repeaters have to be used.”
“If a CMD is used [instead], sensitivity can be increased by several decibels, which allows for lower laser power and fewer repeaters.”
Those advantages should find the CMDs employed in what Williams describes as "signal-starved applications", where very low light intensities need to be detected. These include laser range finding, laser radar and free-space optical links.
But initial demand is being driven by the military community, where system integrators have been able to incorporate lower-power lasers into their products at a lower overall cost, improved reliability and with a longer range.
Williams also thinks that laser-based collision-avoidance technology for cars could represent a considerable market in the future, as OEMs seek to develop eye-safe systems operating at 1550 nm.
“Currently laser rangefinders and laser designators operate at an unsafe 1064 nm wavelength,” he explained. “The CMD can operate at 1550 nm, which is eye-safe. If the military and even auto manufacturers move to the eye-safe wavelength, there is a considerable market.”
