Three years after a team of Caltech scientists showed that pairs of entangled photons could double the resolution of a light microscope, the same lab has figured out a way to double down on that improvement. They have now achieved a fourfold resolution boost compared to a classical microscope, using a new optical design that sends one of the entangled photons through the microscope's optics three times rather than just once.
The work, led by Lihong Wang, the Bren Professor of Medical Engineering and Electrical Engineering and the Andrew and Peggy Cherng Medical Engineering Leadership Chair at Caltech, builds on the lab's 2023 demonstration of quantum microscopy by coincidence (QMC). The approach relies on one of those bizarre quantum-mechanical phenomena called entanglement, in which two particles are linked such that the state of one particle is intimately tied to the state of the other no matter how far apart they might be.
In QMC, entangled pairs of photons, called biphotons, are split so that one photon, called the signal photon, passes through the sample while its entangled partner, called the idler photon, travels a separate parallel path. In some ways, the pair behaves as a single particle that has twice the momentum of an individual photon. According to the laws of quantum mechanics, a particle's wavelength is inversely related to the momentum of the particle. That means that either one of the photons in a biphoton pair effectively images with a wavelength that is half that of the original light. And because microscope resolution improves as wavelength shrinks, this creates a twofold improvement in resolution.
In the new setup, the team still passes a signal photon through the object to be imaged just once. However, the idler photon is routed back through the same pair of lenses three times before it reaches the detector. This is accomplished by applying a magnetic field and using optical tools, including special beam splitters that help the scientists control a property of light called polarization, which is roughly the direction in which the electric field of a light wave travels.
The team describes the setup and experimental validation in a paper published in the journal Science Advances.
"Now we have entered a new physical regime," says Wang, who is also the executive officer for medical engineering at Caltech. "In general, people think that with a single photon pair, you can, at most, increase the resolution from the classical diffraction limit by two times. But we've gone beyond that. This is the most exciting aspect of our new paper, because it points us in a direction where we can improve even further—10 times or maybe even 100 times—down the road."
Wang and his team investigated the power of the new system by imaging a standard test target with the classical imaging setup, their previous twofold quantum configuration, and the new triple-pass configuration. When they measured the sharpness of edges and fine features of the target, they found that the twofold configuration improved resolution by about 1.8 times over classical imaging, while the new setup achieved an enhancement of roughly four times.
The work has clear biomedical applications. Standard high-resolution retinal imaging sometimes requires light that is so bright that it can temporarily affect a patient's vision. The light that reaches the imaging target in the new quantum imaging technique is much less intense and will not damage living tissue. The scientists say higher-resolution low-light imaging could eventually make it possible to see finer structures, such as cell nuclei and mitochondria, without causing damage.
It could also be useful in the inspection of semiconductor devices, where quality-control imaging sometimes entails using longer wavelengths that lead to poor resolution with classical optics. Wang says a fourfold resolution improvement at the same wavelengths could substantially enhance the detection of defects in chips.
Wang notes that while the experimental work validates the new approach, his team is still developing a theoretical model that explains the technique's ability to enhance resolution. In fact, Wang says, he originally had a hard time convincing his lab to try the approach. "Based on our previous theory, I thought this might be possible, but it was just a hunch," he says. "People felt it was risky, but eventually they started working on it. One night, one of my postdocs emailed me and said, 'It worked!' It was such a great moment. And now that we've observed this, we definitely want to push it even further."
The lead authors of the paper, titled "Above twofold quantum super-resolution microscopy enabled by multiple idler passes with entangled biphotons," are Wang lab alumnus Xin Tong (PhD '26), and former postdocs Zhe He, now at Shandong Institute of Advanced Technology in China, and Yide Zhang, now at the University of Colorado Boulder. Caltech graduate students Wenyu Liu and Chien-Ying Huang (MS '23) are also authors of the paper. The work was supported by Caltech's Center for Sensing to Intelligence, the Chan Zuckerberg Initiative DAF, the Silicon Valley Community Foundation, and the National Institutes of Health.
Comparing images of a resolution target using classical microscopy (left), twofold super-resolution imaging (SR2) where the idler beam only passes once, and fourfold super-resolution imaging (SR4) where the idler beam passes back and forth three times. Notice how much more detail in the image is provided by the SR4 quantum imaging technique. The scale bars represent 10 µm.
Credit: Caltech
The basic concept behind the different imaging setups. Classical imaging uses a single count of individual photons. Twofold super-resolution imaging (SR2), the technique introduced by Wang and colleagues in 2023, uses an entangled photon pair but the idler beam passes through just once. The new setup, fourfold super-resolution imaging (SR4) involves an entangled photon pair with an optical setup that causes the idler beam to be routed back through the same pair of lenses three times before it reaches the detector.
Credit: Caltech