A two-qubit gate depth of 1442 two-qubit gates has been achieved by simulating the liquid-state proton NMR spectrum of 1,2-di-tert-butyl-diphosphane using Quantinuum’s System Model H2 trapped-ion computer. Reproducing key spectroscopic features in this challenging molecule was previously unattainable on existing quantum hardware, but the reconstructed spectrum now agrees with classical reference calculations. This is currently the most accurate large-scale nuclear magnetic resonance simulation result obtained to date on a quantum computer.

Quantum simulation mirroring how molecules interact with magnetic fields, the process underpinning nuclear magnetic resonance (NMR) spectroscopy, has been accurately modelled for 1,2-di-tert-butyl-diphosphane, a complex molecule difficult to simulate on conventional computers. The achievement reproduces key features of its NMR spectrum and indicates improved computational capabilities for modelling molecular properties. Accurate simulation of complex molecular behaviour has been achieved using Quantinuum’s trapped-ion technology by modelling interactions between molecules and magnetic fields, central to nuclear magnetic resonance (NMR) spectroscopy.

A trapped-ion computer utilises individual charged atoms as qubits held in place by electromagnetic fields, visualising them like tiny balls suspended on invisible springs manipulated for computation. HQS Quantum Simulations GmbH successfully simulated the NMR spectrum of 1,2-di-tert-butyl-diphosphane; this molecule is notoriously difficult to analyse classically because its internal interactions are so intricate and require calculations that grow exponentially more demanding with size. Reaching a depth of 1442 two-qubit gates in their simulation has produced results agreeing with conventional methods.

Demonstration of extended quantum simulations reproducing detailed NMR spectra of diphosphanes

A two-qubit gate depth of 1442 two-qubit gates now surpasses previous efforts in quantum simulation, as earlier experiments struggled with circuits exceeding several hundred gates limiting their ability to accurately represent molecular interactions. The advance unlocks access to previously unattainable levels of detail within NMR spectra, enabling reproduction of key spectroscopic features in the 3.9, 4.3 ppm region that were absent from prior demonstrations on other platforms. Accurate simulation of complex molecules like 1,2-di-tert-butyl-diphosphane requires modelling intricate relationships between atomic nuclei which grows exponentially more difficult for classical computers as molecule size increases.

Twenty-one system qubits alongside twenty-one ancilla qubits allowed for modelling interactions within the diphosphane molecule using up to seventy steps in their calculations. The team employed a technique called ‘Trotterization’, breaking down complex quantum movements into smaller, manageable stages; this enabled them to simulate how nuclear spins change over time.

Tailored error suppression techniques designed specifically to minimise noise during these extended simulations were implemented and are key because signal quality degrades rapidly with increasing circuit complexity. While representing strong progress, current simulations still require substantial computational resources and do not yet demonstrate any practical speed advantage over optimised classical methods for simulating similar molecules.

Diphosphane nuclear spin Hamiltonian reduction and Trotterised dynamics modelling

Careful reduction of the complexity of the nuclear-spin Hamiltonian tackled a core challenge in quantum NMR simulation; simplifying interactions allows for more manageable calculations akin to understanding how gravity governs planetary motion. This wasn’t about ignoring physics but rather identifying key elements within diphosphane’s structure that allowed them to focus on the most impactful magnetic relationships between atoms. The technique then broke down an otherwise impossibly complex calculation into smaller sequential steps, visualising it like creating a flipbook animation where each frame differs slightly from its predecessor gradually building up the complete picture over time.

Quantinuum’s System Model H2 trapped-ion quantum computer enabled simulation of a 21-spin system representing diphosphane using twenty-one qubits plus an additional twenty-one ancilla qubits for calculations. Reaching circuits with a depth of 1442 two-qubit gates, representing significant complexity, the team accurately reproduced key spectroscopic features previously missed by other demonstrations on quantum hardware.

Quantum computation successfully models challenging diphosphane nuclear magnetic resonance spectra

Accurate modelling of how molecules respond to magnetic fields is vital for nuclear magnetic resonance (NMR) analysis, promising to revolutionise fields reliant on spectroscopy such as drug discovery and materials science. Nuclear magnetic resonance (NMR) spectroscopy relies on precisely determining how atomic nuclei interact within a magnetic field to reveal molecular structure, yet conventional computational techniques struggle with molecules possessing many interacting parts.

This demonstration signifies an important step beyond simply replicating classical results establishing a pathway toward utilising quantum computers where conventional methods are limited and shows improved capabilities in simulating complex molecular systems using trapped ions. The achievement opens questions regarding extending these simulations to even larger molecules and more varied chemical environments demanding further development of both the underlying hardware and algorithmic approaches.

Researchers successfully simulated the nuclear magnetic resonance spectrum of the challenging molecule 1,2-di-tert-butyl-diphosphane on Quantinuum’s System Model H2 trapped-ion computer. This simulation involved modelling interactions within a 21-spin system with circuits reaching 1442 two-qubit gates, accurately reproducing key features previously difficult to capture. The work demonstrates quantum computation can extend beyond replicating classical results in NMR spectroscopy, offering potential for analysing molecules where conventional methods struggle. Authors suggest further development is needed to expand simulations to larger and more complex molecular systems.

👉 More information
🗞 Large-scale NMR simulation on a trapped-ion quantum computer
✍️ Pascal Stadler, Alec Owens, Etienne Granet, David Muñoz Ramo and Michael Marthaler
🧠 ArXiv: https://arxiv.org/abs/2609.17102

More like this