Discriminating between carbohydrates

In recent years, scientists have created chemosensor systems inspired by the human tongue's combinatorial sensing approach. These lab sensors, though, still have a difficult time discriminating between molecules that are structurally very similar, such as carbohydrates containing only hydroxyl groups. Now, Young-Tae Chang and coworkers at New York University have used a new strategy that allows them to distinguish between 23 carbohydrates with 100% accuracy (Angew. Chem. Int. Ed., DOI: 10.1002/anie.200602055). The chemists take advantage of the fact that carbohydrates bind to boric acid or phenylboronic acid with unique strengths, a process that leads to diverse changes in pH. The pH changes resulting from each acid-carbohydrate reaction were determined by using a variety of organic pH indicator dyes and UV-visible spectroscopy. The data were used to construct a matrix of the colorimetric pH changes. Surprisingly, the researchers find that, using only six acid-indicator pairs, they can discriminate between the matrix patterns produced by all 23 carbohydrates, including mono-, di-, and trisaccharides. Practical application of this approach is under study.