Qubits enter the mechanical world

The basic unit of data storage for classical computing—the bit—saves information as either a 0 or a 1. For quantum computing, the basic unit of storage is the qubit, which can save not just one of two values but also a superposition of both. That fundamental difference is expected to one day open a realm of unprecedented computing capabilities (see, for example, the article by Klaus Liegener, Oliver Morsch, and Guido Pupillo, Physics Today, September 2024, page 34 ). Most qubits are built from systems that can inherently possess superposed states, such as trapped ions, atoms, or superconducting loops.

Now Yu Yang, Igor Kladarić, and colleagues in ETH Zürich’s Hybrid Quantum Systems Group, led by Yiwen Chu, have demonstrated a method that turns a much larger system—a mechanical resonator containing roughly 10¹⁷ atoms—into a qubit. A mechanical qubit offers advantages in computational design and sensing that existing qubit platforms lack. To induce quantum superposition in the larger device, the researchers coupled the resonator with superconducting qubits. An antenna connects the qubits to a piezoelectric dome of aluminum nitride that converts the electrical signals from the qubits to vibrations in a sapphire chip.

The research group has previously used the setup to generate superposed states—sometimes called “cat states,” after Erwin Schrödinger’s famous thought experiment—in a similar acoustic-wave resonator (see Physics Today, July 2023, page 16 ). In that work, the researchers had demonstrated the superposition of two opposite-phase oscillations in a sapphire chip. The collective oscillations, known as phonon modes, aren’t suitable for use as a qubit because of their harmonicity, in which the energy gaps between modes are all equal. The resonator’s state, therefore, could easily jump to more than two modes. But by tuning the superconducting qubit’s current to be slightly offset from the resonator’s frequency, the researchers were able to induce variable energy spacing between different states and restrict the phonon modes to just two states. And that two-level quantum system makes it a qubit.

The mechanical qubit has a lifetime of about 100 microseconds, which isn’t bad, but it is not competitive with existing state-of-the-art qubit technologies. Its potential advantages come from its differences. Because the chip is much more massive than typical quantum systems, it could be used for gravitational-wave measurements that other quantum devices are too small to perform. The mechanical resonator can also be used to translate optical information, which the superconducting qubits cannot. That means the resonator could be controlled not only by the microwave photons used to manipulate information in superconducting qubits but also by optical photons.

The mechanical resonator can also support more than one phonon mode at a time. According to Yang, a PhD student in the research group and first author of the paper, the group can manipulate about 10 phonon modes before the system decays. But he sees room for improvements that could make the system uniquely useful. “If we want to build a quantum circuit with 100 qubits, we don’t necessarily need to make 100 chips,” says Yang. “We can fabricate one chip with hundreds of phonon modes, and each individual one can be used as a qubit.” (Y. Yang et al., Science 386, 783, 2024 .)