Squinting to See a Single Photon

By Emily Conover

What’s the dimmest flash of light the human eye can perceive? A handful of photons? What about a single photon? It’s a basic biological question that has yet to be conclusively answered, but a group of quantum optics researchers could fill in the blanks soon. And if humans can see a single photon, those researchers may be able to test for quantum effects on human vision.

According to Rebecca Holmes, a graduate student at the University of Illinois at Urbana-Champaign, her group’s research has already provided evidence that humans are capable of seeing a burst of several photons. She presented the results this October at the 2015 Frontiers in Optics / Laser Science meeting, a joint meeting of the APS Division of Laser Science and The Optical Society, in San Jose, CA. In the study, Holmes and her collaborators presented human subjects in a dark room with faint flashes of light and asked them to record what they observed.

When light enters your eye through your pupil, it is focused on the retina, in the back of the eye. This area is chock-full of photoreceptor cells: cones, which operate best in bright light and provide our color vision, and rods, which are important for night vision, and are highly sensitive to small amounts of light.

In a laboratory dish, a rod that absorbs just a single photon springs to life, producing an electrical signal in response. But the question of whether a single-photon signal can make it all the way through the visual pathway to the brain is still unanswered — the visual system may filter out such tiny signals to avoid unwanted noise. And testing whether humans can observe a lone photon isn’t straightforward: “It turns out it’s really hard to answer that question if you can’t actually make precise numbers of photons,” Holmes said.

This is where the quantum optics expertise of her group, which is led by Paul Kwiat of the University of Illinois, comes into play. Previous experiments have used classical light sources that produce a small but indeterminate number of photons, making the result less clear-cut. Kwiat’s team will utilize a quantum source — a device that produces individual photons and is standard in quantum optics labs — but turn it on human subjects.  They shine an ultraviolet laser on a nonlinear crystal, in which photons produce pairs of lower-energy photons. The researchers detect one photon in a pair, using it to herald the other, which is sent to the subject. The laser is tuned so that the output photons will be at a wavelength of 500 nanometers, where sensitivity of the rods is highest.

Holmes’ apparatus randomly sends each photon to one of two optical fibers, directed at the left and right sides of the subject’s retina. The subject then must say what side it was on. The design is an improvement on previous studies — which simply asked subjects if they saw a flash of light — because subjects may hesitate to give a response that could be a false positive, the researchers say.

In preparation for their experiment with single photons, the team first measured the result with an average of 3 or 4 photons on the retina. The subjects picked the correct side 54 ± 1% of the time — only slightly better than chance, implying that humans can just barely see these dim flashes.

To get those 3 or 4 photons to the retina, the researchers send in 100 photons — 70% of these are lost in the optical apparatus, and only about 10% of the photons that go on to hit the eye actually make it to the retina. But inefficiency in the visual pathway or lapses in attention could be just as important: “Maybe a bigger problem is that the observers don’t always seem to notice the stimulus, even though it was probably technically bright enough for them to see it,” Holmes said.

In the final version of their experiment, the researchers will allow only one photon through at a time, making it crucial to increase the efficiency of the system so that as many photons make it to the retina as possible. “Every trial where that doesn’t happen [the subjects] really are just randomly guessing — that’s just noise,” Holmes said.  The researchers are now optimizing their setup to improve optical efficiency before photons reach the eye.

And the researchers have ideas on how to improve the human part of the equation as well. Adaptive optics could compensate for aberrations or eye motion, and an EEG might help monitor subjects’ attention, allowing the researchers to send photons only when the observer is most likely to see them. And, if researchers’ anecdotes are to be believed, subjects watching for single photons may get better with practice. “I’ve done it a lot of times and I’m definitely better than anyone else,” Holmes said.

If humans can serve as single-photon detectors, the researchers have more ideas in store. “Since we’re quantum physicists, we’d love to try to test quantum effects on the visual system,” Holmes said. For example, rather than sending the photon to the left or the right, they could send the subject a superposition, and study what the subject observes. Quantum mechanics predicts a superposition would look no different. But that’s not something that’s been tried before,” Holmes says.