New Developments in Photonics

An article based on an interview with Dr. Emre Sari, Professor of Photonics at Izmir Institute of Technology, Izmir, Turkey

By Jacob Doyle

Dr. Emre Sari is a professor of photonics, an emerging field that studies the application of light for industrial purposes. These include computing, health care and energy.

In computing, Sari described how photonics is being applied in such a way that the microchips used in certain high-computation-power computers communicate with each other via fiber-optic cables. The next step, he explained, could be optic connections within chips themselves. He described the challenges facing the development of an “all-optic” or “all-photonic” computer wherein computer components that use electricity as a medium of communication and processing would be replaced with components that use light. He described his own thesis work, which involved the use of photonics as applied to a specific aspect of computing.

He described how light-sensitive genetic material has been applied to neurons in the brain of a mouse, which then could be activated by a fibre optic cable inserted into the mouse’s brain. This resulted in experimenters being able to control parts of the mouse’s body and “thoughts” with light. Such research, Sari hopes, will prove useful in seeking treatments for neural diseases such as Parkinson’s and others involving neurological degeneration.

Sari was recently awarded an entrepreneurship grant to develop a new kind of solar cell that uses “perovskites” which are potentially cheaper, lighter, cooler and more efficient in its conversion of light to electricity than solar cells currently in use.

The concept of building a new kind of computer that uses exclusively light rather than electricity has been discussed for at least two decades. Such promises as dramatic increases in processing speed, reduced energy consumption and material cost have all been made by advocates of the hypothetical machines. These “all-photonic” computers would use light as their processing medium in much the same way that fiber-optic cables use light to carry voice and data traffic over distances, as opposed to old-fashioned copper telecom cables that used slower, less efficient electricity. How close are we to such a photonic computer today?

“We are not quite there yet,” said Dr. Sari. “An ‘all photonic computer’ would be a radical change and the industry is not willing to do that. This also has to do with capabilities and limitations of photonics in such applications as computing.”

Sari explained that entirely new components would need to be invented for the photonic computer to become a reality, and there are a number of technical hurdles that need to be jumped first.

“For example, you have to guide your light along the circuits within your hypothetical photonic chip.” he said. Unlike electricity, which travels through an electronic circuit along conductive metal wire, light would travel along a “waveguide” such as fiber-optic cable made from glass fiber. This brings with it problems of scale.

“In order to guide the light through the waveguide or other means around the circuits of the chip, it has to be at a size that is comparable to the wavelength of the light that you are using. Most commonly, this would be infrared light, which has around 1.5 µm and it has to be of the same order of magnitude as far as the size of the waveguide is concerned. And if you compare these dimensions with what is currently used in electronic chips, in interconnects of electrical lines that go through a chip, the width of these lines is like three orders of magnitude less, or more than 100 times smaller than the channel required for infrared light to pass through a circuit.”

Another hurdle is dealing with heat. The most compact way to generate light, Sari added is with electricity. “And when a light source and electricity are together, that almost always brings with it a heating problem.”

Heat can damage circuits and cause other problems that could interfere with functionality. Sari added that there are ways to mitigate problems associated with heat, such as distributing it to avoid “hot spots” within the circuits.

Applying photonics to the health care and energy sectors are two areas that have Sari excited and have become the focus of his latest research.

In health care, photonic devices called optical neural actuators are being used to study neural diseases such as Parkinson’s, Alzheimer’s, and ALS, he said, in an emerging field known as optogenetics.

“Normally, some of our neural cells are sensitive to light and also, in nature, there are some organisms that are inherently sensitive to light,” said Sari, “a sensitivity that’s achieved through some membrane proteins, that these organisms synthesize, and which can also synthesized in the lab. By genetically isolating the DNA code that makes it possible to synthesize those membranes proteins and injecting them to any neuron that we like, we can make those neurons light-sensitive. We can control the operation of these neurons, in principle, either in an inhibitory way or an actuating way, by sending them like pulses. This enables us to study neural damage diseases.”

Such neural diseases, he added, utilize pathways from one functional part of the brain to another in order to transport information or even disease.

“To better understand how these diseases operate, one demonstrated way is to shut down or actuate some parts of the brain and study the effects on these pathways,” said Sari. “Optogenetics has proven itself to be very useful in this area. An early demonstration had a fiber-optic cable injected inside a mouse’s brain which served to actuate a neuron in the mouse’s brain that had been treated with light-sensitive membrane protein. This is an example of what’s called a tethered solution, which means that one end of the fiber is connected to the external world and the other end is in the mouse’s head”.

Untethered Optogenetics

By using photonic technologies, specifically very small LEDs, Sari explained, one can create implants which function as untethered actuators, or neural drivers in a practice now called “wireless optogenetics.”

“Those are very interesting to me,” said Sari. “In the future I’m aiming to study such novel devices for such applications and moving my interests to neurotechnology, the exploration of the ways of using technology to control the brain.”

Sari elaborated on neurotechnology, saying that adding light-sensitive DNA to regions of the brain that control movement and implanting untethered micro LED actuators, then it could be possible to further study or even overcome the loss of motor control caused by such diseases as Parkinson’s. Moreover, such photonics-based solutions would not have unintended effects on untargeted parts of the brain, as do older devices developed for this purpose that use electricity instead of light.

The Brain Initiative launched by President Obama, which pledged US$110m for brain research lines up with Sari’s interest in using photonics to study the brain.

“Each part of the brain has its own detail, which needs to be understood on the cellular and connectedness level,” Sari said. “All these pathways of the brain must be understood. If one pathway is blocked, how does the brain find another pathway to achieve the same thing? This also needs to be understood.”

He mentioned that in addition to the US, a number of other countries are funding similar research including a €100 million EU initiative to develop a computer chip whose function would mimic that of the human brain, both for the benefit of brain disease research and to enhance chip innovation. Other such efforts are underway in Japan.

“This is a wave I am aiming to catch,” said Sari.

Photonics in solar energy

Generating electricity from sunlight can be enhanced with help of a recent development in photonics, said Sari with much enthusiasm. Solar cells that employ synthetic perovskites, he explained, are taking photovoltaics to new levels of efficiency, flexibility, as well as cost.

A perovskite is a naturally occurring mineral, which is synthesized for use in solar cells, thanks to its special properties. They are much easier to synthesize compared to traditional silicon, used in conventional solar cells. Perovskites are cheap and easy to manufacture, according to Sari, and people have recently demonstrated more than 20% of power conversion efficiency, albeit in a very small scale experiment. It is however still very promising, he added.

“The scaling issue can be solved by industrial optimization work,” said Sari. “Also, there is room for improvement in the achievable power conversion efficiency, meaning it could go even higher than 20%. In the thermodynamic limit it can reach up to 30%.”

Sari spoke of an entrepreneurship grant from The Scientific and Technological Research Council of Turkey (TÜBİTAK), for developing perovskite solar cells in Turkey. The modest sum, along with the facilities of the University should help him to develop the new cells and bring them to market, he said. “Now that I founded my company Esinotek, named after my 2-year-old daughter Esin, I will be able to apply for industrial grants in Europe. If I make a difference in this big market, it could be helpful for the future of our planet, for our kids. I’m pretty excited about it.”

Sari went into some detail about what parts of the solar spectrum are converted more effectively using perovskite cells.

“The absorption edge of a perovskite solar cell where the absorption starts is just in the near infrared region, 800 nm wavelength, so they can successfully convert photons with larger energies, meaning they can convert the whole visible spectrum, but they are not responsive to other parts of the infrared spectrum,” he said. “So their are thermal electric properties are not viable for electricity generation from heat but their optical properties are very much suitable for converting visible light to electricity.”

The photovoltaic heat problem

When sunlight beams onto a solar cell, Sari eplained, the ratio of generated electrons per incoming photon determines how inefficient the solar cell is and how large a heating problem it has.

“The amount of energy that you cannot convert into your electricity turns into heat, because energy is conserved, he said. “But if you have an efficient material and you are able to successfully collect your electrons in external circuits, then those electrons do not contribute to heat. Those electrons contribute to the sum of the electricity that you generate.”

Inherently in silicon, he added, there is a certain amount of electrons that are converted to heat because of their inefficiency. This is the case with both conventional silicon solar cells and more recent “thin film” solar cells.

“But for perovskites, the quantum efficiencies are high, as is the optical absorption and the ratio of photons converted to electricity,” he said. “But the problem that we are trying to solve is the so-called ‘traps’ when you are collecting the electrons generated in the external circuits. Before you can collect them, they can be trapped between the ‘grains’ of the perovskite semiconductors, as perovskites are polycrystals rather than single crystals. These lost electrons that cannot be collected causes the solar cell to be dumped with extra-unutilized energy and that extra energy eventually converts into heat. That becomes a problem. People are working to mitigate this problem by way of materials science and engineering.”