The limits matter. The device does not identify and recover every original source; it separates incoming light into coherence modes.
Researchers at Stanford University, working with collaborators in Austria and Italy, have built a photonic chip roughly the size of a fingernail that can take a mixture of light and separate it into individual, clean beams, according to a report in the Times of India. The problem the device addresses has been discussed by scientists since the 1980s, the report says, but had not previously been demonstrated in this physical form.
The difficulty lies in the nature of most real-world light. A laser produces coherent light, whose waves move in step and can reinforce or cancel one another; a light bulb produces light whose waves are largely out of sync. Most light falls between these extremes and is described as partially coherent, meaning some of its waves are organised and others are not. When light from several sources overlaps, the contributions become hard to tell apart, and the resulting interference can degrade the performance of optical instruments.
The Stanford device is a photonic chip, guiding photons through extremely small channels on a silicon surface rather than moving electrons. In the reported experiment, two infrared lasers were directed into the chip, producing a partially coherent mixture. Inside, a network of tiny interferometers, supported by algorithms, examined the statistical properties of the incoming light and separated it into its components, known as coherence modes, directing them to different outputs. The report is explicit that the chip does not simply measure the mixture and reconstruct the original beams computationally: it physically separates the modes into individual channels.
To verify the separation, the team recombined the two output beams and found their waves did not synchronise, confirming they had been rendered independent. A third laser was then introduced, and the device was still able to sort the resulting mixture, suggesting the approach can handle more than two sources. According to the report, the researchers separated the two strongest coherence modes with a fidelity of more than 98 per cent against a benchmark method.
The potential applications are in imaging. Cameras record mainly the intensity of light and lose much of its phase information, which describes the relative timing and alignment of waves. A chip able to separate coherence modes might help distinguish two fluorescent particles that appear as one under a microscope, or tease apart contributions from several objects in what a telescope records as a single blurry source. The researchers are exploring a collaboration with NASA on the latter, the report says, and see possibilities in optical communications, sensing and computing, since the chip can adjust to changing light conditions.
The limits matter. The device does not identify and recover every original source; it separates incoming light into coherence modes, which correspond to the original sources only under certain conditions, such as when the beams are spatially orthogonal. The work is a laboratory demonstration, and the report gives no timetable for practical instruments, no cost figures and no detail on how the chip would perform outside controlled conditions. Whether the approach survives contact with real telescopes and microscopes is not yet known.
