Using phonons, Tsinghua team introduces scalable, programmable ion trap quantum processor In conversation with Wentao Chen, Yisaku
Boson sampling is a benchmark problem for optical quantum computers and a potential way to realize the superiority of quantum computing. Now, the Tsinghua team has demonstrated a scheme to implement a boson sampler based on capturing vibrational modes in a chain of ions.
Trapped ions are a versatile experimental platform at the forefront of quantum technology and are often used in quantum computing and quantum simulations. In the confines of an electric field, individual atomic ions are typically arranged linearly and possess vibrational modes - phonons - a combination much like the overtones on a guitar string.
In most cases, the system is encoded in the internal electronic states of the ions, for example, as quantum bits in a digital quantum circuit or as interacting spins in a quantum simulation. In a recent paper published in Nature Physics ("Scalable and Programmable Phononic Network with Trapped Ions"), Qi Huan Jin's team in the Department of Physics at Tsinghua University Wentao Chen et al. put the phonon modes "center stage" and used the internal ion states as auxiliary degrees of freedom.
In conventional ion trap experiments, phonon modes are only used briefly to generate spin-spin interactions before being decoupled. However, these modes represent a large Hilbert space that previous researchers have barely begun to exploit, and this opens up opportunities for different applications.
One such application is the boson sampling problem. The goal is to predict the output probability of a network of interacting boson modes - a formidable mathematical computational task. In this work by the Tsinghua ion trap team, phonons take the place of photons and act as beam splitters by interacting with the internal ion states.
A phonon trapping ion boson sampler. The different transverse vibrational modes of the trapped ion chain are initialized and then coupled in a fully reconfigurable network using beam splitter interactions. The beam splitter is implemented by coupling the two modes to the internal spin state of an ion, using a pair of laser beams to generate the coupling between spin state and motion. The system is fully connected, which means that any pair of modes can be interfered with at any stage At any stage. Finally, each mode is mapped to the spin state of one ion for readout.
This approach has multiple advantages, as individual phonons are easier to generate than individual photons and have longer lifetimes. Now, Chen and colleagues have successfully assembled an ion-based boson sampler.
"have shown that phonons at one harmonic potential can be coherently transferred to another harmonic potential and that these phonons can interfere with each other, and when we learned that modified boson sampling (Gaussian boson sampling) can also be applied to chemical problems (i.e., vibrational sampling), we demonstrated the sampling of SO2 molecules and developed a method to create highly entangled phonon states; However this was limited to a single ion. In this work, we finally implemented phonon networks in a scalable way, overcoming the limitations of single ions."
(a) Schematic diagram of the structure of the boson system. (b) Schematic diagram of the four-mode phonon network; (c) phonon state preparation scheme; (d) phonon beam splitter implementation scheme; (e) phonon state detection scheme
This phonon-based scheme constructs a completely new boson sampling platform thanks to its flexibility and the absence of phonon loss mechanism in the system. In contrast to a fixed optical network, it can be reconfigured at will, allowing all modes to be interconnected, since many ions are involved in each mode. In addition, the ion heating rate (the source of spurious phonons that cause errors) is small for all modes, except for the center-of-mass mode, which the team avoided in their experiments.
Ultimately, the results demonstrate a clear pathway to extend phonon networks for quantum information processing and to break through some of the technical limitations of classical and photonic systems. Compared to other previously proposed optical quantum processors, it is easier to scale up and can eventually achieve better performance on complex problems - the research opens up new possibilities for advances in quantum computing and quantum technology.
This time, the experimental system demonstrated by the Tsinghua team offers another way of looking at trapped ion systems, and the boson control it provides presents a new experimental platform that will inspire the development of additional ion-phonon control schemes. Potential applications range from the realization of long term theoretically exotic systems, such as the study of condensates and other many-body phenomena of phonons, to advances in quantum simulations of strongly interacting systems, and some frequently encountered problems in nuclear physics.
This research was supported by the Quantum Science and Technology Innovation Program and the National Natural Science Foundation of China. Corresponding authors of the paper are Wentao Chen, a 2017 PhD student in the Department of Physics, Tsinghua University, Prof. M.S. Kim, Imperial College of Science and Technology, and Prof. Qi Huan Kim, Department of Physics, Tsinghua University; among them, Dr. Wentao Chen is the first author of the paper. Recently, PhotonBox had the honor to invite Dr. Wentao Chen to give a more detailed and in-depth interpretation of the related results.
