Our group is developing chip-scale systems and devices that process electromagnetic, optical, and mechanical fields in qualitatively new ways. Silicon chips form the material basis of our current information technologies and comprise the vast majority of electronic circuits. However, electronic signals in transistors are hardly the only physical degrees of freedom available to us. Such electronic circuits are especially ill-suited for dealing with quantum information – states of a physical system that possess nonclassical correlations and enable vastly more powerful paradigms of computation, communications, and sensing (Preskill2012). Moreover, by developing spatio-termporal control over optical fields we can revolutionize classical sensing, imaging, and communications systems. Our goal is to create ground-breaking new functionality by developing on-chip systems that operate on photons (light), phonons (mechanical motion), as well as the quantum electrical signals from emerging superconducting quantum computers. We are developing the fabrication, design, theory, and characterization methods to realize new classes of systems, some which are described below.
Different physical forms of information are best suited for different tasks. The simple act of sending a message from a mobile phone requires conversion of classical information from electronic signals to radio frequency signals and then onto light that is transmitted over optical fibers across long distances. All computation, sensing, and communication systems depend on devices that convert information from one form to another. The same capability, currently lacking in the quantum domain, is crucial for creating truly compelling quantum technologies. We are addressing this basic challenge by developing the physical devices that convert quantum information.
Quantum computers promise to harness the principles of quantum mechanics to perform certain tasks extremely quickly. In the near future, they will enable us to solve problems that are intractable even for the largest supercomputers today, such as simulating the behavior of new drug molecules, predicting the properties of novel materials, and factoring large numbers for cryptography. In order to get the maximum performance out of these new computers, it will be useful to connect them into networks much like today’s internet. However, because quantum signals are extremely fragile, quantum computers must be connected using optical fibers rather than normal wires which are inherently noisy. The goal of our project is to create a microchip which will convert the microwave frequency electrical signals from a quantum processor into light signals which can be sent via an optical fiber. This chip will function as a dual purpose transmitter and receiver, allowing quantum computers to be linked into a quantum network. Our device works by combining electrical circuits and optical waveguides together on a single microchip and using an electro-optic material to exchange energy between electricity and light.
Jeremy D. Witmer, Joseph A. Valery, Patricio Arrangoiz-Arriola, Christopher J. Sarabalis, Jeff T. Hill, Amir H. Safavi-Naeini, High-Q photonic resonators and electro-optic coupling using silicon-on-lithium-niobate, Scientific Reports 7, 46313 (April 13, 2017), arXiv:1612.02421.
Jeremy D. Witmer, Jeff T. Hill, and Amir H. Safavi-Naeini, Design of nanobeam photonic crystal resonators for a silicon-on-lithium-niobate platform, Optics Express Vol. 24, Issue 6, pp. 5876-5885 (March 10, 2016), arXiv:1604.05647.
Marek Pechal describes our motivation and effort in to creating high bandwidth quantum networks:
Microwave-to-optical conversion (as described above) requires a relatively high amount of energy per converted bit of information, which makes its operation difficult at the extremely low temperatures required by quantum computing hardware specially if we’re interested in very rates of information transfer. A direct connection using microwave waveguides without the need for frequency conversion, on the other hand, suffers from higher losses than in optical fibers and is complicated by the need to cool the whole waveguide to the same ultra-low temperatures as the quantum computers themselves.
We are studying an alternative to these two approaches - a way to exchange quantum information between microwave-frequency quantum machines by converting it to millimeter waves. These are electromagnetic waves whose frequency is higher than microwaves but lower than optics. We consider an implementation using so-called kinetic inductors - superconducting circuit elements that possess a nonlinearity which enables frequency conversion. By carefully analyzing the details of the conversion process, we find that it achieves an appealing compromise between the optical and microwave approach: A millimeter waveguide does not need to be cooled to extremely low temperatures and while frequency conversion is still necessary, it requires much less energy per converted bit of information than conversion to optics.
Rishi Patel describes his work, on using light to control circuits for sound below:
Our research lately deals with manipulating the flow of sound on a chip. In particular, we are trying to make waveguides for phonons (the particles of sound, much like photons are the particles of light) to get them from one place to another, without getting lost or distorted. Our devices are fabricated out of silicon, the same material that is used in much of the electronics industry. They are about a thousand times thinner than a human hair, and extend up to several millimeters. For an analogy, imagine making a one kilometer long flute. How can we learn about the properties of this flute by placing an ear on one end?
Making such an “ear” is itself an interesting problem. The frequencies (“pitches”) that we deal with are about a million times higher than what we hear on a day-to-day basis. To build a sensitive ear for these pitches, we trap light and sound in a small volume on our chip. By analyzing the color change of light reflected from the device, we learn about the sound. Interestingly, we can also use light to “push back” on the mechanical structure. Finally, these systems will let us explore quantum effects, where the physics of single phonons and photons is important, on a surface a chip.
Rishi N. Patel, Christopher J. Sarabalis, Wentao Jiang, Jeff T. Hill, and Amir H. Safavi-Naeini, Engineering Phonon Leakage in Nanomechanical Resonators, Phys. Rev. Applied 8, 041001(October 16, 2017), arXiv:1705.07869. more…
Photonic circuits — circuits that control light in the way electronic circuits control electrical currents — are an ever more powerful way of detecting and controlling light. There are good ways of coupling these circuits to the fiber optic cables that backbone the internet and so they’re being used by companies like IBM and Luxtera to send and receive more and more data. Can we use photonic circuits to control, detect, and analyze light in our environment in the same way they’re being used in fiber for telecommunications? To do so we’ll need an “on-chip” means of forming and directing a beam of light : a microscopic lighthouse of sorts that can direct a laser around a room.
Many of the current approaches to on-chip beam steering use optical gratings. Gratings scatter light by color like the reflections of a white light off a CD : red is reflected at one angle and blue at another resulting in a rainbow. If you shine a laser beam at a CD and change its color, you can steer that beam around the room. What if you want to steer a red, green, and blue beam independently? For example, you might want to use a photonic circuit to form a color image of a room, or you might want to project a color image on a surface. With a grating, you’d only be able to see a red ball at a particular angle (or project a red ball at that angle).
There is a link between the color of incident light, the angle of scattered light, and the grating’s period. Gratings are a series of scatterers and light hitting them is like waves crashing on a chain of rocks on a shore. The period is the spacing between the rocks. If you look at a CD under a scanning electron microscope, you’ll see exactly what I’m talking about. If the period, the spacing, was somehow tunable, the scattering angle could be varied independently from the color of the light (which is analogous to the wavelength of the ocean wave). It turns out that sound is a kind of naturally tunable grating.
It shouldn’t be surprising that sound, which is nothing more than vibrations of a material, scatters light. If you toss a stone into a pond, you can watch the ripples as they wander off because those ripples scatter light. Similarly if you make a photonic wire — a wire for light which is called a waveguide — on a chip and you send sound waves down that wire, you can scatter guided light out from the wire and off the chip. By changing the pitch of the sound, that is whether it’s a high or a low note, you can change the angle at which light is scattered out.
We’re currently investigating whether we can use sound to make an on-chip beam steering system, a microscopic lighthouse. We call our device an optomechanical antenna array. Such a system may prove useful in making photonic circuits practicable across a wide range of optical systems, potentially enabling cheap, manufacturable 3D color scanning, holographic display, optical wireless communication, and optical interconnects.