In the lab, a catalyst sorts mirror-image molecules apart

Hold a left glove and a right glove side by side. Same materials, same size, and yet one does not fit the other hand. Molecules can behave the same way, and for chemists that is a real headache. Mirror-image molecules, called enantiomers, are so similar that standard separation methods struggle to pull them apart. A team in Japan now reports, in the journal Science Advances, that it has separated an equal mixture of such mirror-image molecules, made by replacing hydrogen with deuterium in one form.

Deuterium is a heavier cousin of hydrogen: its nucleus holds a proton and a neutron. Swapping hydrogen for deuterium can change a molecule's properties slightly, which makes these labelled molecules useful in drug development and materials engineering. The difficulty is that the two mirror forms stay almost identical. The available tools were weak. Spectroscopy, which uses nuclear magnetic resonance (NMR), can tell such molecules apart, but the authors note these methods are mainly analytical and offer no practical route to separating them. Specialized chromatography gave poor separation, and one attempt that worked better needed 65 passes through a recycling chromatography system. So researchers have mostly built these molecules through carefully controlled synthesis, rather than sorting existing mixtures.

The team's idea was to use the one difference that deuterium introduces. A carbon–deuterium bond is harder to break than a carbon–hydrogen bond, and the authors call this difference “the only energetic handle available” for telling the two forms apart. They combined it with a catalyst's preference for one molecular orientation. The result was a ruthenium-based catalyst that reacts with one mirror form faster than the other. In tests on 2-ethylnaphthalene, using a reaction that adds a nitrogen-containing group, the recovered material was estimated to contain more than 99.9% of one form. In a test with another molecule, one form reacted 83 times faster than the other.

Those numbers come with real limits. Most enrichment values were inferred from the reaction products, not measured directly in the recovered molecules. The one independent check, which converted a sample and then ran NMR, gave a measured 93:7 ratio against a predicted 94:6, a close match that supports the estimates. Performance also varied considerably with molecular structure, and some compounds separated less effectively. The method is a starting point, not a finished production process.

For pharmaceutical research, the practical gain would be a faster supply of isotope-labelled building blocks, the starting pieces that chemists assemble into drugs. One of the molecules tested is used in making tofogliflozin, a diabetes drug, and it was successfully separated in these tests. The authors say the method still needs further testing, and the article gives no timeline for when labs could adopt it. For now, the change is concrete for chemists: a route that could replace some of the slow, controlled synthesis they rely on today, once it has been proven beyond these laboratory examples.