Quantum effect enables more information to be encoded on microchip

Phenomenon can be used in technology for processors and various types of sensors.

By José Tadeu Arantes  |  Agência FAPESP – The idea of using the effects of quantum mechanics in information technology began to be taken seriously in the 1970s. It gained momentum in the early 1980s when charismatic US physicist Richard Feynman (1918-88) pronounced himself in favor of the possibility of a quantum computer. Then, in January 2017, an article published in Nature speculated that with the participation of giant corporations such as Google and Microsoft, “Quantum computers [are] ready to leap out of the lab in 2017”.

In the interval between Feynman’s verdict and the promise raised in Nature, much progress has been made in science and technology. A study published in January of this year makes a new contribution by proposing to take advantage of a typical quantum mechanical phenomenon of light interference to encode more information on silicon microchips. The research was conducted by Professor Michal Lipson and her group at Cornell University, with the collaboration of Paulo Nussenzveig, Full Professor at the University of São Paulo’s Physics Institute (IF-USP) in Brazil.

The results were published in Nature Communications in a paper entitled “Quantum interference between transverse spatial waveguide modes”. Nussenzveig participated in the study under the aegis of the Thematic Project “Exploring quantum information with atoms, crystals and chips”, with Marcelo Martinelli as the principal investigator and with funding from FAPESP.

“My interaction with the group focuses on exploring the quantum effects that result from the integration of electronic processing and photonic communication in silicon microchips,” Nussenzveig told Agência FAPESP.

The effect explored in the study was Hong-Ou-Mandel (HOM) interference. Discovered in 1987 by Chung Ki Hong, Zhe Yu Ou and Leonard Mandel, HOM interference occurs when two indistinguishable photons interfere in a 50:50 glass beam splitter (50% reflection and 50% transmission), one on each side of the glass.

In this case, four situations are ideally possible, as shown in the figure: (1) the photon that comes in from above is reflected, and the photon that comes in from below is transmitted; (2) both photons are transmitted; (3) both photons are reflected; or (4) the photon that comes in from above is transmitted, and the photon that comes in from below is reflected.

The quantum effect, described by what are known as the Feynman rules, means that situations (2) and (3) cancel each other out owing to destructive quantum interference. The output is either both photons exiting from the top of the beam splitter (1) or both photons exiting from the bottom (4), but never one photon in each direction. 

“This is a typical quantum effect, because the same phenomenon behaves as both waves (interference) and particles (two discrete photons),” Nussenzveig said. “The deviation of the pair of photons to one side or the other can be considered one bit of information.”

Experiments with HOM interferometry became almost commonplace in quantum optics during the 1990s and 2000s. The main novelty of this latest study is the confinement of the phenomenon in a microchip, replacing the beam splitter with a painstakingly fabricated microscopic waveguide.

“Confinement of light in a very small region imposes restrictions on its propagation,” Nussenzveig explained. “It propagates freely in the longitudinal direction, along the waveguide axis, but propagation in directions transverse to the axis is hindered by reflection from the walls of the waveguide. Depending on the wavelength of the light and the dimensions of the waveguide, the intensity of the light wave in both transverse directions may also increase or decrease, also owing to the interference effect. This defines different possible spatial modes for the wave.”

By varying the internal geometry of the waveguide, it is possible to obtain several spatial modes for the wave and numerous transitions from one mode to another. Thus, the waveguide performs a function equivalent to that of the HOM interferometer but enables a far larger amount of information to be encoded for only one pair of photons. This is a crucial goal for any type of processor or sensor – encoding the maximum amount of information in the minimum space. 

“The device is made of silicon nitride and targeted by a special light source that produces twin photons. This is obtained by means of a non-linear crystal that converts the individual photons in the beam that lights up the crystal into pairs of photons. Given the energy conservation principle and because electromagnetic radiation frequencies depend on energy, the sum of the frequencies of the two photons produced is equal to the frequency of the original photon. This ensures that two indistinguishable photons reach the waveguide simultaneously, one in each of the transverse modes. On the way out, two different detectors ensure that the observed effect has really been produced by the two photons,” Nussenzveig said.

The point of such a structure is to encode more information per photon than a single bit. Several other research groups, some in Brazil, are also working toward this goal. The key is to encode information via the orbital angular momentum of light. However, this is done in free space, on the laboratory bench. The technology will be of practical use in many more applications if it can be achieved compactly on a microchip.

“The first part of the article shows that this interference effect occurs efficiently and can be controlled. The second presents proof of the principle that multiple applications of interferometry are possible based on the states of two photons,” Nussenzveig said. “More sensitive measurements in metrology, heat sensors and devices to gauge the purity of materials are some of the potential technological developments. But there’s a considerable distance between proof of concept and the production of a device.”

Efforts to integrate quantum optics with electronics have been underway for some time. The originality of the study lies in the demonstration that transverse waveguide modes can be used to encode information and that increasing amounts of information can be stored at the microchip scale by multiplying these transverse modes.

Another unique feature of the study relates to the research group’s composition. It includes younger as well as older scientists, women as well as men, and researchers of Indian, Chinese and European origin. It is led by an experienced Israeli scientist, with the collaboration of an experienced Brazilian scientist.

The team’s multi-national, multi-gender and multi-age makeup is characteristic of a great deal of contemporary science. Not long ago, it would not have been necessary to highlight this composition, but in the present context, it is relevant to know who produces science in US centers of excellence.

The Israeli scientist is Michal Lipson, daughter of US scientist Reuven Opher. Opher is a Full Professor at the University of São Paulo’s Institute of Astronomy, Geophysics & Atmospheric Sciences (IAG-USP). Lipson collaborates closely with researchers at USP, where she did part of her undergraduate coursework. Her group, initially based at Cornell and now at Columbia, is considered the leader in silicon nanophotonics. 

Paulo Nussenzveig also comes from a family with several renowned scientists, including his father, physicist Herch Moysés Nussenzveig. Paulo graduated in physics from the Pontifical Catholic University of Rio de Janeiro (PUC-RJ), earned a PhD in quantum physics from Paris University VI (Pierre et Marie Curie), and became a professor at IF-USP’s Department of Experimental Physics in 1996.

More information: youtube.com/watch?v=M0XsmNGAa98.