Enhanced chemical detection for manufacturing

Augmenting spectral databases for performant chemical detection.

Near-infrared spectroscopy is a key method for real-time chemical detection. It promises cheap and reliable quality assurance, such as in pharmaceuticals.

Problem

Extracting chemical data from near-infrared spectra is challenging. Large datasets are required to train statistical models, a data-intensive process specific to each chemical and setup, limiting its use. High-quality simulations could augment these datasets, but strong vibrational anharmonicities make these spectra prohibitively challenging to obtain on classical computers.

Solution

Our breakthrough is a quantum algorithm for efficient simulation of high-fidelity near-infrared spectra. It computes experimentally-accurate spectra for classically intractable molecules using only about a hundred logical qubits, running in just hours or days.

Efficient computation of high-quality near-infrared spectra for chemometric modeling can unlock cheap, reliable, non-invasive chemical characterization, enabling broad industrial process control.