Researchers have designed a phototransistor that can detect extremely weak infrared light and directly power a visible LED with a voltage of just 1.4 volts. In this way, the prototype turns infrared into an amber glow without the need for a camera or screen, a specific path to ultimately illuminate night vision systems worn in front of the eyes.
A detector that amplifies the signal twice
The conversion seems simple on paper: a detector absorbs infrared photons, produces a current, which then feeds a visible light source. In practice, the light available at night is so weak that the detector must provide a high gain. Previous solutions often imposed high voltage or did not combine sensitivity, amplification and efficiency in a single compact component.
The team at the Universities of Washington, Illinois at Urbana-Champaign and Yale fabricated a heterojunction phototransistor made of indium gallium arsenide and indium phosphide, two common materials in infrared optoelectronics. Its architecture combines the amplification of a transistor with an avalanche effect: the accelerated charges of the material release others, which multiplies the current.
The device adds a positive feedback loop between these two mechanisms. This is what makes it possible to obtain an avalanche regime with a breakdown voltage announced in 1.4Vmuch easier to deliver via a handheld device than the high voltages traditionally associated with this type of gain.
In the experiment described on October 30 in Nature photonicsthe phototransistor directly powered an amber LED. With an incident infrared intensity of 6.4 nanowatts per square centimeter, close to the low natural radiation of the night sky, the visible light produced was more than 17,000 times brighter than input infrared signal according to the measurement reported by the authors.
Diagram of the operation of the infrared phototransistor developed by the researchers. The component amplifies extremely weak infrared radiation thanks to an avalanche effect, obtained from 1.4 V, and then directly powers an LED that emits visible light. This technology could ultimately simplify night vision goggles.
Direct night vision as a goal
Current night vision goggles are generally based on an image intensifier tube or a digital sensor associated with a display. These architectures work, but they add power, electronics, optics, and sometimes display lag. A robust infrared-to-visible converter could simplify this chain by allowing the detector to directly drive a visible emitter.
The published result does not yet constitute a pair of glasses. The researchers demonstrated a component connected to an LED, not a high-definition matrix capable of reconstructing an entire scene. To form an image it would be necessary to manufacture and connect a large number of pixels, control their uniformity, manage noise and integrate everything with the appropriate optics. The amber color of the LED is not a final choice of the interface, but rather that of the transmitter used to validate the conversion.
The field of photonics specifically seeks to translate more optical functions into compact components. A photonic quantum processor has recently been tested in orbit, in a very different context but with the same limitation of operating optical elements outside of a protected laboratory.
Promising figures, an integration yet to be invented
The phototransistor solves three obstacles at once: it detects infrared at low intensity, provides very high gain, and operates at low voltage. The authors believe that the mechanism can be adapted to a broader infrared range by modifying the materials and joints, which also opens up uses for optical sensors, imaging or communications.
Several limitations are outside the scope of the article. The listing does not demonstrate field of view, resolution, or battery life on a handheld device. Nor does it compare a complete prototype of glasses with commercial intensifiers in terms of latency, weight or robustness. The results refer to components manufactured and tested in the laboratory, with data available in the article and its complementary information.
The work was peer-reviewed and funded primarily through DARPA’s ENVISION program. Its immediate contribution is a very low voltage amplification mechanism, validated in an actual conversion to an LED. Moving to a display surface composed of thousands or millions of pixels will require a new stage of manufacturing, addressing and optical integration.
