How AI is transforming the world of analytical chemistry

Artificial intelligence is reshaping analytical chemistry from an empirical model based on "experiments plus manual interpretation" into an intelligent paradigm characterized by data-driven processes, automated closed-loop workflows, and high-precision prediction. This core transformation is manifested in the automation of spectral analysis, the autonomy of experimental workflows, real-time quality control, and a significant leap in capabilities for the reverse identification of unknown substances.

Analytical chemists have been using machine learning long before ChatGPT made headlines. Rasmus Bro, who researches machine learning in analytical chemistry at the University of Copenhagen in Denmark, stresses that generative methodologies will not necessarily help chemists do what they do much better, but they will broaden the kind of problems that can be solved. "We think it will be a revolutionary change, not an incremental one when it comes, but we're not there yet," he explains.

Long before generative AI emerged, established forms of AI, particularly machine learning, had revolutionized how scientists approach data analysis. The power of machine learning to classify and discern patterns in large datasets makes analysis simultaneously more rigorous and less time-consuming. Any chemist who uses large quantities of data will be using machine learning, whether they realize it or not.

Jerome Workman Jr., a former instrument and software development scientist from California, says generative models will "augment, simulate, and better characterise spectral data." This provides the logical link between scientific uses of generative AI and the ubiquitous chatbots.

A recent feature in Spectroscopy magazine described spectroscopy as "at a crossroads." The authors list three unrelated trends as contributing to the challenges facing spectroscopists: artificial intelligence, automation, and miniaturisation. Workman believes that generative AI became particularly compelling for spectroscopists when it was able to offer the possibility of mapping data space itself, not just mapping inputs to outputs.

Researchers caution against overreliance on AI outputs, noting that both chatbots and scientific AI models can produce convincing but incorrect results or "hallucinations," making validation against established physical and chemical principles essential. Farooq Wahab, an analytical chemist at the University of Texas at Arlington, warns: "We should beware in particular of a 'beautifully correct' answer from AI, because it may still be based on incorrect reasoning."

Donatella Puglisi at Linkoping University in Sweden and her group have developed an artificial olfactory system, called an "e-Nose," in which volatiles bind to a sensor array, generating signals that machine learning algorithms can classify quickly, precisely and from small samples. The e-Nose is being tested in oncology to distinguish between blood plasma from ovarian cancer patients and healthy controls. "We aim to produce a machine that can screen for ovarian cancer in minutes using a simple blood sample, with more accuracy than any other technique," says Puglisi.

Key dimensions of transformation include intelligent spectral analysis and quantitation using deep learning algorithms, autonomous closed-loop experimental processes integrating robot chemists, reverse identification and structural deduction of unknowns using large-scale models, and intelligent quality control with multi-source data fusion.

Will AI replace analytical chemists? Workman suggests the answer will be "more, not less." "The more routine work will be automated," he explains. "But the analytical chemist's role will switch to quality control and interpretation; demand will grow for scientists who understand chemometrics and can work with AI."