There are few technologies more fundamental to modern life than the ability to control light with precision. From fiber-optic communications to quantum sensors, the manipulation of photons underpins much of our digital infrastructure. Yet one capability has remained frustratingly out of reach: controlling light with light itself at the most fundamental level using single photons to switch or modulate powerful optical beams.
Now, researchers at Purdue University have achieved this long-sought milestone, demonstrating what they call a "photonic transistor" that operates at single-photon intensities. Their findings, published in the journal Nature Nanotechnology , report a nonlinear refractive index several orders of magnitude higher than the best-known materials, a leap that could finally make photonic computing practical.
"We demonstrated a way to realize a photonic transistor working at single-photon intensities," said Vladimir Shalaev , Purdue's Bob and Anne Burnett Distinguished Professor in Electrical and Computer Engineering. "This was a long-standing problem, and we found a potential way of solving it."
The achievement addresses a fundamental challenge in photonics: Traditional optical nonlinearity, where one beam of light affects another, requires enormous power levels.
"Usually there is optical nonlinearity, which allows two beams to interact with each other," said Demid Sychev, a postdoctoral researcher in Shalaev's group in the Elmore Family School of Electrical and Computer Engineering and the first author of this paper. "But typically, this interaction works only for macroscopic beams, for classical light, because the nonlinear refractive index is very small. This is a problem because this method cannot be used for single photons."
Amplifying the quantum world
The solution came from an unexpected source: the avalanche multiplication process used in commercial single-photon detectors. When a single photon strikes silicon and creates a single electron, that electron can trigger an avalanche that generates up to 1 million new electrons, a cascade that bridges the microscopic quantum world with macroscopic, measurable effects.
"This multiplication is a very powerful tool for connecting the microscopic quantum world with the macroscopic world," Sychev said. "This principle was often used for single-photon detection, but what we did was apply this process to create a huge nonlinearity for optical beams, where one single-photon beam can control a huge macroscopic beam."
Peigang Chen, a fourth-year PhD student in Shalaev's group, noted the elegance of the approach. "When I first came to the group, I just thought this was a genius idea from Demid," Chen said. "In the future, we're going to fabricate our own single-photon avalanche diodes (SPAD) for this specific design. But the easiest way for us to get this first result was to use a commercial SPAD."
The device functions as an optical switch: A single photon in the control beam can modulate the properties of a much more powerful probe beam, effectively switching it on or off.
Three critical advantages
The Purdue team's approach offers three key advantages over alternative methods that have been explored for single-photon nonlinearity.
First, it operates at room temperature. "Typically, what people use for single-photon nonlinearity these days are quantum systems where they use two-level systems, like a single-photon emitter coupled to a cavity," Sychev said. "But this method is very sensitive to temperature. It cannot be applied at room temperature."
Second, the technology is compatible with complementary metal-oxide-semiconductor, meaning it can be integrated into existing semiconductor manufacturing processes. "This is seamless and compact," Chen said. "For the others, it's very different and complicated physics systems. This one is semiconductor, and it can always be fabricated on chip."
Third, and perhaps most importantly, it operates at gigahertz speeds and could potentially reach hundreds of gigahertz, dramatically faster than existing approaches. "Clock rates of such systems may go up to gigahertz, but with the methods we developed, in principle, it can be extended to hundreds of gigahertz," Sychev said.
Applications: From quantum to classical
While the research has obvious applications in quantum computing, where it could increase the efficiency of generating single photons and enable faster quantum teleportation protocols, Sychev believes classical computing applications may be even more transformative.
"The reason why a photonic computer is not realized is because the current approaches using photons are supposed to be much better. Photons consume less energy; they are faster," he said. "Ideally, from photons, you can get terahertz clock rates of CPUs, compared to currently existing 5 gigahertz in the best cases. But the problem is that there are no photonic switches like this. The needed interaction between photons typically requires high powers of optical light. With our method, in principle, you can do it with single photons."
The implications extend beyond computing to data centers, optical communications and data transfer systems — anywhere that the speed and energy efficiency of photons could replace slower, more power-hungry electronics.
"I feel like here I'm starting to change the world," Chen said. "This work really means a lot, not only for my career, but also because this device can make a difference in the industry and the science community."
A four-year journey into unknown territory
The path from concept to publication was neither quick nor straightforward. Sychev had been working on the idea for about a year when Chen joined the group three years ago. Over four years, the team tried several different experiments before achieving success.
"It was an iterative process," Sychev said. "It took a lot of experimental efforts because it's completely unknown territory."
The work was conducted at Purdue's Birck Nanotechnology Center under the guidance of professors Shalaev and Alexandra Boltasseva, the Ron and Dotty Garvin Tonjes Distinguished Professor in Electrical and Computer Engineering. Both Sychev and Chen are affiliated with the Purdue Quantum Science and Engineering Institute .
Next steps and broader impact
The team is now focused on optimizing the technology. "Previously, all commercially available SPADs we used were not designed for this purpose," Sychev said. "Now our goal is to make a device which will be optimized to work as a single-photon switch." They plan to explore different device geometries and materials to further enhance performance.
Sychev emphasized that while the demonstration is significant, substantial work remains. "This work indeed can bring more results in the future for industry and for academia, for science and technology," he said. "It's a long-standing problem, and we found some potential way of solving that problem. It still requires a lot of work toward this goal, but at least some interesting direction was found, and we are very happy about this."
The research team also includes Yuheng Chen, Morris Yang and Colton Fruhling, all from Purdue's Elmore Family School of Electrical and Computer Engineering, who performed optical measurements alongside Sychev and Chen. Alexei Lagutchev, a senior research scientist at the Birck Nanotechnology Center, conceived and planned the experiments and analyzed the experimental data. Alexander Kildishev, a professor of electrical and computer engineering at Purdue, performed finite-element simulations to support the work.
The researchers believe their breakthrough has created what Shalaev and Boltasseva have described in related work as a new "playground" for physics and engineering, a platform where fundamental discoveries in controlling light at its most basic level can lead to transformative applications across quantum and classical technologies. As the demand for faster, more efficient computing and communication systems continues to grow, the ability to manipulate photons at the single-photon level represents a critical step toward realizing the full potential of light-based technologies.
Quantum research is a pillar of Purdue Computes , a strategic university initiative to further scale Purdue's research and educational excellence in computing.
The paper acknowledges cooperative research with DEVCOM Army Research Laboratory supported by ACC-APG-RTP Agency. A provisional patent application related to this work has been filed by the Purdue Innovates Office of Technology Commercialization .
