Computational Linker Design for Highly Crystalline Metal–Organic Framework NU-1000

NU-1000 is a metal–organic framework (MOF) with Zr6(μ3–OH)4(μ3–O)4(OH)4(OH2)4 nodes (called Zr6 nodes) and tetratopic 1,3,6,8-tetrakis(p-benzoate)pyrene (TBAPy) linkers. NU-1000 has pores of 31 Å diameter, and it has smaller pores connected perpendicularly to these large pores; this nanoporous structure makes it suitable as a catalyst or a support for catalysis. However, as ordinarily synthesized, NU-1000 has disordered Zr6 nodes populated in its large pores, and this introduces nonuniformity into the crystal and complicates structural characterization. We have examined possible MOF morphologies composed of Zr6 nodes and TBAPy linkers, and we propose that the extra Zr6 nodes in the large pores of NU-1000 are structurally similar to those of an NU-901 phase, which differs from NU-1000 in the relative orientation of Zr6 nodes and the linker conformations. This polymorphism comes from two rotamers of the TBAPy linkers; in NU-1000 the benzoates at the 1 and 3 positions and the 6 and 8 positions are disrotatorily situated, whereas in NU-901 these benzoates are conrotatorily situated. Substituting carbon-2 and carbon-7 of pyrene with groups that are unevenly disposed with respect to the pyrene plane should stabilize the rotamer that can form NU-1000 and therefore stabilize NU-1000 itself. We confirmed this by applying the PBE-D2 and M06-L exchange-correlation functionals in periodic electronic structure calculations to compare the energies of NU-1000 and NU-901 with different substituents on the linkers. In NU-1000, the benzene rings of the benzoates coordinating to the Zr6 nodes form pocketlike structures. In unsubstituted NU-1000 the pocket in the small pore is smaller and attracts precursors of the atomic layer deposition (ALD) step more strongly because of shorter benzene–precursor distances and more favorable dispersion-like interactions, leading to selective metal deposition in the small pores. We examined the pocket sizes and the adsorption energy of Zn(Et)2 (the ALD precursor) on the Zr6 node in large and small pores of substituted NU-1000, and found that bulkier substituents not only compress the pocket in the large pore, but also have shorter distances to the ALD precursor in the large pore. Both factors contribute to the more favorable dispersion interaction and lead to less tendency to deposit metals in the small pore.