Molecular Structure Analysis Technology for High-Quality and High-Performance

Trace components contained in raw materials at parts-per-billion (ppb) or parts-per-million (ppm) levels at times affect product quality or are sometimes involved in the development of prominent performance. For these reasons, analyzing the molecular structures of such components precisely is indispensable for selecting raw materials in a timely and proper way and for optimizing manufacturing technology.
In general, the trace components are isolated after complicated separation and purification processes, and their molecular structures are then analyzed with a nuclear magnetic resonance (NMR) spectrometer or a mass spectrometer (MS). However, analyzing the molecular structures of unknown components precisely is very difficult and sometimes takes a long time.

Kao is working on the development of precise and effective molecular structure analysis systems that use spectrum data from a gas chromatograph–mass spectrometer (GC-MS) or a liquid chromatograph–tandem mass spectrometer (LC-MS/MS). We have previously created a database by melding spectral information from public database and our in-house original database. Furthermore, we have developed a spectral similarity search technology to automatically search for compounds similar to target compounds. We then analyze the substructures of such compounds by introducing machine learning and a unique search algorithm to which natural language processing is applied.

We are currently aiming to develop a technique for automatically analyzing molecular structures on the basis of mass spectra. Furthermore, we are trying to establish a method for examining the structures of a wide variety of molecules more precisely. This is being done by incorporating the most advanced molecular structure analysis technology that we have been introducing in joint research with universities.