While traditional computers save data in the form of bits, essential units of logic that represent a value of either 0 or 1, quantum computers rely on qubits. These can exist in a state that is simultaneously both 0 and 1. This peculiar characteristic, a feature of quantum physics called superposition, is central to the potential of quantum computing to eventually resolve issues that are unsolvable by classical computers.

Numerous current quantum computers utilize superconducting electronic systems where electrons move without resistance at very low temperatures. In these frameworks, the quantum mechanical characteristics of electrons flowing through precisely engineered resonators give rise to superconducting qubits. These qubits excel at rapidly executing the logical tasks required for computation. However, retaining information—in this context, quantum states, which are mathematical representations of specific quantum systems—is not their forte. Quantum engineers have been striving to enhance the retention times of quantum states by developing so-called “quantum memories” for superconducting qubits.

Now, a group of Caltech researchers has adopted a hybrid technique for quantum memories, effectively converting electrical data into sound so that quantum states from superconducting qubits can be preserved for a duration up to 30 times longer than in prior methods.

The new research, spearheaded by Caltech graduate students Alkim Bozkurt and Omid Golami, under the guidance of Mohammad Mirhosseini, assistant professor of electrical engineering and applied physics, is featured in a publication in the journal Nature Physics.

“Once you possess a quantum state, you might not wish to immediately manipulate it,” Mirhosseini explains. “You need a method to return to it when you do want to perform a logical operation. For that purpose, a quantum memory is essential.”

Previously, Mirhosseini’s team demonstrated that sound, specifically phonons—individual particles of vibration (similar to how photons are single particles of light)—could serve as an effective means for preserving quantum information. The devices evaluated in classical experiments appeared optimal for integration with superconducting qubits since they functioned at the same high gigahertz frequencies (humans perceive sound at hertz and kilohertz frequencies, which are at least a million times slower). They also operated effectively at the low temperatures required to sustain quantum states with superconducting qubits and exhibited long lifespans.

Now, Mirhosseini and his team have constructed a superconducting qubit on a chip and linked it to a tiny apparatus referred to as a mechanical oscillator. Essentially a small tuning fork, the oscillator comprises flexible plates that vibrate due to sound waves at gigahertz frequencies. When an electric charge is applied to those plates, they can engage with electrical signals that carry quantum information. This enables data to be introduced into the device for storage as “memory” and retrieved later, or “remembered.”

The researchers meticulously assessed how long it took for the oscillator to lose its valuable quantum content once information entered the device. “It turns out these oscillators have a lifespan approximately 30 times longer than the best superconducting qubits currently available,” Mirhosseini notes.

This technique for creating a quantum memory presents several benefits over earlier methods. Acoustic waves travel significantly slower than electromagnetic waves, permitting more compact devices. Furthermore, mechanical vibrations, unlike electromagnetic waves, do not spread in free space, meaning that energy remains contained within the system. This allows for prolonged storage durations and reduces undesirable energy interactions between adjacent devices. These benefits hint at the potential for many such tuning forks to be integrated into a single chip, offering a possibly scalable solution for constructing quantum memories.

Mirhosseini states that this research has shown the minimal level of interaction between electromagnetic and acoustic waves required to evaluate the value of this hybrid system as a memory component. “For this platform to be genuinely beneficial for quantum computing, the ability to input quantum data into the system and extract it much more rapidly is essential. This implies that we need to discover methods for increasing the interaction rate by a factor of three to ten beyond our current system’s capabilities,” Mirhosseini explains. Fortunately, his group has concepts regarding how this can be achieved.

Co-authors of the paper, “A mechanical quantum memory for microwave photons,” include Yue Yu, a former visiting undergraduate student in the Mirhosseini lab; and Hao Tian, a postdoctoral research associate in electrical engineering at Caltech, affiliated with the Institute for Quantum Information and Matter. The research received support from the Air Force Office of Scientific Research and the National Science Foundation. Bozkurt was funded by an Eddleman Graduate Fellowship.