(682a) Insights and Rational Design of Metal-Organic Frameworks for Enantiomers Separations

Recent advances in the design and synthesis of metal-organic frameworks have demonstrated the potential of this material class. This research paper analyzes the existing literature on chiral metal-organic frameworks as gathered in the CoRE MOF database [2]. The CoRE MOF database contains most of the reported MOF structures from literature. The crystal structure and synthesis condition are known for each represented network. In many cases, only specific properties in the domains of adsorption, catalysis, electrochemistry… are reported. The extend of reported properties is not only limited by the scope of the original author(s) but in some cases also by the technological advance at the time of reporting. A combination proper chemical and topological analysis with selective computational screening of adsorptive properties was performed.

Chiral frameworks are characterized by the presence of either chiral topology (e.g. srs, unc, lcy…) or enantiomers as linkers, or both. Chiral topologies are identified using structural analysis of their 3D net. Interestingly, only 19 out of the 362 chiral nets in the Reticular Chemistry Structure Resource (RCSR) database are found within reported MOF structures. MOF structures with chiral linkers (or introduced by post synthetic modification) are identified by data mining of article contents. Analysis of the reported frameworks gives insight in the organic and inorganic building blocks needed to form a chiral frameworks. There are 2 topological factors mentioned in literature that contribute to chirality: asymmetric metal node or helix formation. Interestingly, our analysis shows that in many cases chirality is the result of symmetry distortion at the metal node by chelation or mixed O- and N-terminated organic linkers. This essentially means that no chiral building blocks are required to result in some degree of chirality in the framework. Several asymmetric metal clusters are identified for Zn/Cu, In, Co, Ni and Cd that are of interest in the design of novel frameworks (no need for chiral linkers). The formation of helixes or warped pore architecture is more likely if a three-connecting organic building block is present rather than 2 or 4 connecting linkers. On the basis thereof, several new hypothetical structures are proposed and evaluated for chiral separations. The analysis of structural components is complemented with molecular simulations on the adsorption for R- and S-isomers (organic acid, alcohol, amine, amino acids) to establish relationships and compare impact of key features (linker, node, helix) on the separation potential. A one-to-one comparison between framework types is not straightforward as pore size, topological net type and chemical composition are typically linked but select cases where only one aspect is different indicate synergetic effects.

Particular attention is given to cyclodextrin Metal-Organic structures as a subset to illustrated the effect of pore size and topology towards chiral separation potential. The three different ring sizes in alpha, beta and gamma cyclodextrin lead to different pore dimensions irrespective of the packing in a crystal structure. Our results on the CD-MONT metal-organic cage structures [2] indicate that a more confined pore enhances the chiral separation potential. This finding is in line with observations for stereoisomers in shape selective nanoporous materials. The differences in free energy between R- and S- isomers are partly enthalpic and partly entropic. The effect of topology is illustrated by another subset of gamma cyclodextrin [3] metal-organic frameworks with different chiral net. Both the number of cyclodextrin rings that a probe enantiomer simultaneously interacts with as the relative pore dimension (resulting from tubular or cage type pore formation) have significant effects.

[1] Chung et al. Chem. Mater., 2014, 26 (21), pp 6185–6192.

[2] Wei et al. Chem. Sci., 2012, 3, pp 2282-2287.

[3] Forgan et al. JACS, 2012, 134, 406, Hartlieb et al. J. Am. Chem. Soc., 2016, 138 (7), pp 2292–2301.