Interview with PolyU's Lao Zihuan: Taking Atomic-Level CT of Molecular Sieves, Giving Tens of Thousands of Tons of Industrial Catalysts a 'Design Blueprint'

Molecular sieve catalysts have been industrialized since the 1960s-70s, widely used in petroleum cracking and chemical production, but the spatial distribution of their active sites—aluminum atoms—has long been a 'black box'. Lo points out that in the past, we could only know the performance of a catalyst, not why it had that performance.

The acidity of molecular sieves arises from aluminum substituting silicon in the framework: aluminum is trivalent, silicon is tetravalent, and to balance the charge, a nearby hydrogen ion is needed, forming a Brønsted acid site. The position of aluminum determines the location of acid sites, but aluminum and silicon are nearly indistinguishable under traditional X-ray diffraction: aluminum has 13 electrons, silicon has 14, and after removing outer electrons, their electron clouds are similar, making X-ray differentiation difficult.

Although electron microscopy has high resolution, it has significant errors at the atomic level, and aluminum and silicon are 99% similar, so existing techniques struggle to separate them. During his DPhil at Oxford, Lo used probe molecules (such as pyridine, methanol, ammonia) to indirectly infer acid site positions, but could not directly see aluminum atoms.