Researchers at Google Quantum AI have demonstrated how their Willow quantum computer can augment Nuclear Magnetic Resonance spectroscopy to reveal molecular structures. The technique, called Quantum Echoes, uses qubit perturbations to emulate molecular analysis. While promising, it has not yet shown a clear advantage over classical methods.
Google Quantum AI's Willow quantum computer, featuring 103 qubits, has been employed to interpret data from Nuclear Magnetic Resonance (NMR) spectroscopy, a key tool in chemistry and biology for determining molecular structures. The team, led by Hartmut Neven, developed a protocol named Quantum Echoes, which draws on a quantum analog of the butterfly effect. In the process, researchers apply a sequence of operations to the qubits, perturb one specific qubit as a 'quantum butterfly,' reverse the sequence, and then measure the system's quantum properties to extract information about the whole.
This approach mimics NMR's use of electromagnetic perturbations on molecules to map atomic positions, potentially creating a 'longer molecular ruler' for observing distant atoms. As team member Tom O’Brien explained, “We’re building a longer molecular ruler.” The quantum method proved reproducible across two quantum computers, aided by Willow's improved hardware with lower error rates. However, for two organic molecules, only up to 15 qubits were used, and results were reproducible by classical computers. The team estimates that a supercomputer would take 13,000 times longer for similar computations, though the demonstration remains preliminary and unpublished in peer-reviewed form.
Experts offered mixed views. Keith Fratus of HQS Quantum Simulations called it an important link between NMR and quantum computing but limited to specialized biological studies. Dries Sels of New York University noted it advances quantum simulation of complex NMR protocols, providing motivation despite few industrial examples. Curt von Keyserlingk of King’s College London praised the experimental feat but questioned its broad utility, suggesting classical methods could compete and its appeal lies mainly in fundamental quantum physics research. As qubit performance improves, O’Brien anticipates broader applications for larger molecules. The work appears in Nature (DOI: 10.1038/s41586-025-09526-6).