Using the same technology at the center of facial recognition, engineers and chemists at the Hiroshima University have successfully developed chiral crystals. This is the first research reporting the use of this technology, known as logistic regression analysis, to predict which chemical groups are ideal for creating chiral molecules. The research findings have been published in Chemistry Letters.
In order to build a chiral magnet (pictured), it is necessary to first design a chiral crystal. (Image credit: Julien Zaccaro / Center for Chiral Science)
Chirality defines the quality of having a mirror image to something else, but without the ability to superimpose it. The left foot, for instance, is a mirror of the right. They look similar, but they are not identical. This is why a left shoe cannot be worn on the right foot.
The idea is similar in chemistry. Two molecules can have the same composition of elements, but their geometry can vary. A left-handed chiral helix can have an equivalent right-handed helix.
However, making a mirror image of a chiral molecule is more difficult than just rearranging a few bonds. An additional layer of complexity comes up when making a crystal, an extremely ordered series of molecules or atoms that can spread in three dimensions.
“The most difficult part of making a chiral crystal,” main author Katsuya Inoue said, “is knowing how to design them.” Inoue is a researcher at the Graduate School of Engineering at Hiroshima University (HU).
According to Inoue, it is tough to blend different atoms so that their chiral geometry coexists in one crystal. Independently, two types of atoms will want to develop bonds with the same angles. When integrated, though, they may not.
The team examined 686 chiral crystals and 1000 achiral crystals from the Inorganic Crystal Structure Database. By using logistic regression, Inoue’s team built a model to demonstrate the ideal way to engineer chiral crystals.
They deliberated which chemical groups of the periodic table have elements that are more probable to coexist in a chiral crystal. The groups that correspond to carbon, nitrogen, and oxygen were most suitable – or group numbers 14, 15, and 16, respectively.
Logistic regression is a statistical technique that can distinguish two objects. Eri Shimono, co-author and research assistant in the Department of Chemistry and Chirality Research Center at HU, related it to use in smartphones.
In face recognition, smartphones use machine learning to classify faces and things that are not faces. We can train our model to detect chiral and non-chiral possibilities. In this case, though, the input is not an image. It is information.
In the days to come, the team will be tweaking the prediction model in dual ways. First, they are keen to account for more atoms in a crystal.
“We started from two atoms. In reality, though, many crystals are made with three or four,” Inoue said. “ We have to extend this model to fit for these cases.”
Second, they will be applying deep learning. The present model, which uses standard machine learning, is made from prevalent data. Deep learning would allow researchers categorize new data as chiral or not. Based on these results, the team plans to begin developing some predicted crystals and explore how to produce a chiral magnet from them.