As previously mentioned, chemical reactions in achiral media always give racemic mixtures of enantiomers. Therefore, the first step in obtaining homochiral metal complexes is determining a strategy for separating the two enantiomers. We decided to use chiral anions to selectively crystallize one or the other enantiomer, leaving the undesired enantiomer in solution. The first family of anions is based on enantiopure arsenic tartrate ions (AsT), while the second family is formed by helicochiral arsenic tris-catecholate ions (TRISCAS). Both of the selected salts were synthesized in multigram scale.
Chiral AsT ions were used to separate racemic mixtures of helico-chiral tricobalt complexes (Fig. 2). These compounds were selected particularly for their strong optical chirality and stability in solution. We were able to separately crystallize both enantiomers and study their optical properties and absorption of left and right-handed circularly polarized light. As expected for two enantiomeric species, we obtained mirror-image circular dichroism spectra, not only using standard UV/visible circular dichroism techniques, but also using infrared and X-rays (Fig. 3). Although the signals were very intense, X-ray magnetochiral dichroism signals could not be detected, most likely because the spin state was too low, with only one unpaired electron shared over the three cobalt ions. We then decided to work with trinickel analogues, which were likewise separated by crystallization withchiral AsT, and are currently under study.
We then turned our attention to building chiral extended structures from these trimetal units. We had previously found a way to synthesize one-dimensional polymers using fluorine-based linkers, and thought that we could use a similar strategy to synthesize chiral polymers, using the homochiral trimetal building units. We were interested in these complexes because some of them showed a high spin state at low temperature, suggesting they would be promising candidates for magnetochiral studies. Surprisingly, we found that samples of homochiral tricobalt complexes converted into racemic mixtures upon polymer formation. Moreover, this unexpected isomerization occurred whenever we attempted to replace the axial ligands, limiting the chemistry that can be done withthem . However, we found that using chiral linkers prevented racemization and yielded chiral polymers. For example, we discovered two different chiral one-dimensional polymers made from alternating tricobalt or trinickel compounds bridged by chiral AsT ions (Fig. 4).
We also investigated a variety of mononuclear chiral complexes (Fig. 5). For example, we synthesized homochiral iron complexes with a propeller-like structure using chiral AsT and TRISCAS anions. In this case, spectroscopic methods evidenced the racemization of these compounds in solution. Nonetheless, in the solid state, the complexes demonstrated strong non-linear optical properties. A series of similar compounds with a variety of metals was prepared in order to study the influence of the spin state on the chiroptical properties. We were excited to observe magnetochiral effects in the UV-visible range for some of these complexes (unpublished work).
Some of these results have been already published in the following per-reviewed articles: 1) Eur. J. Inorg. Chem. 2018, 320. 2) Chem. Sci., 2018, 9, 1136. 3) Polymers 2018, 10, 311. These results have been presented in the following conferences: 42nd ICCC (Brest, 2016), 2nd and 3rd BOOK-D (Pessac, 2016 and 2017), 2016 and 2017 GdR MCM-2 (Dourdan), 6th ECMM (Bucharest, 2018) and 2018 JCC (Brest). The results have been presented in two formal group meetings to the teams involved in the project and several visiting researchers. The fellow attended to 2016 AIC International Crystallography School (Rimini) and a workshop on chirality intended for young students was developed in the 2017 Fête de la science (Pessac).
