Edinburgh Research Explorer Synthesis and complexes of a constrained-cavity Schiff-base dipyrrin macrocycle

A new constrained-cavity [1 + 1] Schi ﬀ -base dipyrrin macrocycle comprising an N 4 donor-pocket has been synthesised by spontaneous oxidation and in situ crystallisation. Access to Fe( II ) and Zn ( II ) complexes is achieved by salt elimination reactions of the lithium salt. All compounds have been characterised by NMR and UV-vis spectroscopy, X-ray crystallography, and DFT analysis.

Dipyrrins, also known as dipyrromethenes, represent a fascinating class of organic chromophores and are readily synthesised via oxidation from their dipyrromethane analogues.][16][17][18] Due to their planar N 4 donor-pockets, Schiff-base dipyrrins are similar to porphyrins and their expanded porphyrin cousins and display a rich coordination chemistry along with unique redox-activity and photo-physical properties.0][21] Sircoglou, Aukauloo and co-workers investigated the photo-physical properties of an acyclic zinc(II) and nickel(II) dipyrrin-pyridine complex. 22In contrast to acyclic ligands, non-porphyrinic, macrocyclic dipyrrins are rare.Naruta and co-workers reported the phenanthroline-dipyrrin macrocycle 2 (Fig. 1) and described the photo-physical properties of its magnesium(II) complex. 23,24Sessler, Katayev, and co-workers studied competitive binding of palladium(II) with the expanded Schiff-base dipyrrin-diamide macrocycle 3 (Fig. 1). 25ccordion-like dimanganese(II) Schiff-base dipyrrin macrocycles have been investigated by Bowman-James and coworkers as a functional model for catalases. 26,27][30][31] The chemical and photochemical properties of the dipyrrin framework combined with the unique control of coordination and organometallic chemistry afforded by a macrocyclic environment led us to investigate the formation of macrocyclic analogues of 1.We report here the straightforward formation and isolation of a unique constrained-cavity [1 + 1] Schiff-base dipyrrin macrocycle HL that comprises an N 4 dipyrrin-imine donor set and describe the formation and properties of its lithium, iron, and zinc complexes.
The condensation of equimolar quantities of the monomeso-substituted dipyrromethane dialdehyde 4 and bis(m-aminoisopropyl)benzene 32 in acetonitrile at room temperature causes the macrocyclic product HL to crystallise directly from the reaction mixture as greenish-red crystals in 80% yield (Scheme 1 and ESI †).The 1 H NMR spectrum of HL shows an imine proton resonance at 8.47 ppm and additional resonances indicative of C 2 -symmetry in solution.In addition, the disappearance of the meso-proton resonance from 4 reveals that spontaneous oxidation of the dipyrromethane to the dipyrrin has occurred.
The MALDI-TOF mass spectrum of HL shows the expected molecular ion at m/z 524.5 ([M + H] + ) with no higher-order cyclisation products observed.Attempts to form HL using the dipyrrin analogue of 4 (4 ox , ESI †) in a condensation with the diamine under similar conditions results in a mixture of the starting materials and a variety of imine-containing products.][35] The X-ray crystal structure of HL (Fig. 2) reveals a fully planer conformation of the N 4 donor set with the meso-C 6 F 5 substituent approximately orthogonal (82°) and the paraproton (H1) of the bridging phenyl group pointing into the macrocyclic cavity.The size of the cavity is defined by the pyrrole (N2⋯N3) and imine (N1⋯N4) separations of 2.672(2) Å and 4.391(2) Å, respectively, so resulting in a small and rigid trapezoid-shaped binding pocket.This is significantly different to porphyrins such as tetraphenylporphyrin (TPP) in which an approximately square cavity of 2.921 Å by 2.914 Å is seen. 36eaction between HL and one equivalent of LiN(SiMe 3 ) 2 in toluene cleanly generates the lithium complex Li(L) which was isolated as a dark purple solid in 91% yield.Its formation is apparent by the disappearance of the pyrrole proton in the 1 H NMR spectrum and the appearance of a broad singlet at 1.57 ppm in the 7 Li NMR spectrum.The X-ray crystal structure of Li(L) shows that the lithium coordinates to all four N-donor atoms within the macrocyclic cavity with an axial Li⋯F3′ interaction from an adjacent molecule (Li⋯F3′ 2.354(3) Å), forming an infinite chain structural motif (Fig. 2 and ESI †).The presence of the Li cation causes the bridging phenyl ring to tilt significantly, from 176.56°in HL to 112.79°in Li(L), resulting in a C1⋯F3′ distance of 3.570(2) Å. Attempts to doubly deprotonate HL (e.g., at C1-H1) using two equivalents of LiN(SiMe 3 ) 2 were unsuccessful.The twist of the bridging aryl group such that the unique H-atom H1 avoids interaction with the Li cation within the N 4 -doner set shows that, unlike the porphyrin analogues, the macrocyclic cavity is size-compromised.
The reactions between Li(L) and either ZnBr 2 or FeBr 2 cleanly generate the metal bromide complexes ZnBr(L) and FeBr(L), respectively (Scheme 2); the isolated yield for FeBr(L) is low due to its poor solubility.HRMS of ZnBr(L) and FeBr(L) show molecular ion peaks at m/z 667.025720 ([M + H] + ) and 577.110887 ([M − Br] + ), respectively.The 1 H NMR spectrum of ZnBr(L) shows mirror symmetry with the geminal methyl groups appearing as two singlets at 1.80 ppm and 1.34 ppm, indicative of top/bottom asymmetry; this asymmetry is also seen in the 19 F NMR spectrum which displays five fluorine resonances.
The X-ray crystal structures of ZnBr(L) and FeBr(L) were determined (Fig. 3) and both exhibit square pyramidal geometries with axially coordinated bromide ligands.There is a noticeable difference in distance between the metal and the plane of the N 4 donor set; for zinc this is 0.593 Å above the plane while for iron it is 0.304 Å. Similarities can be found in previously reported structures of iron chloride octaethylpophyrin 37 and zinc chloride m-benziporphodimethene 28,38 in which the Fe and Zn are ca.0.47 Å and 0.51 Å, respectively, out  of plane.In contrast, for iron chloride and zinc chloride TPP complexes the metal ion is coordinated on the macrocyclic N 4 donor plane. 39,40This difference could be explained by the additional steric effect of the para-proton (H1) of the bridging phenyl group that points towards the cavity.
The UV-vis spectrum of HL shows two absorptions at 277 and 512 nm (Fig. 4) that are similar to those seen for other expanded dipyrrins and substituted BODIPY compounds. 20,41,42Upon deprotonation to form Li(L) and subsequent transmetallation to form ZnBr(L) and FeBr(L), significant red shifts occur in the UV-vis spectra, with the lower wavelength adsorption splitting into two.The experimental UV-Vis spectra (with the exception of FeBr(L)) were reproduced by TD-DFT calculations using B3LYP functional and 6-311G(d,p) basis set (Fig. 4, 5 and ESI †).The energies of the transition states are over estimated, as seen with previously reported hetero-diarylmethene ligands and their complexes. 43The electronic density of both the HOMO and the LUMO of HL and ZnBr(L) comprise π orbitals that are centred on the dipyrrin unit and the imines, highlighting the extended conjugation in these molecules.The dominant absorption band in HL at 277 nm is assigned as a mixture of HOMO to LUMO+3 (41%) and HOMO to LUMO+1 (13%) transitions, with the HOMO-LUMO transition appearing at 512 nm.This latter transition is comparable with (Fig. 1) the acyclic dipyrrin ligand 1, which has the HOMO-LUMO transition at 540 nm. 20The red-shifted dominant absorption bands at 595 and 557 nm for Li(L) and ZnBr(L), respectively, correspond to the HOMO-LUMO transition with the HOMO-LUMO+3 transition shifting to 303 and 300 nm, respectively.The UV-vis spectrum of the ZnBr(L) resembles that of the deprotonated ligand Li(L), suggesting the absence of any metal-centred transitions.
We have synthesised a new constrained-cavity [1 + 1] Schiffbase dipyrrin macrocycle that is spontaneously oxidised and crystallised during its preparation.The straightforward deprotonation of the macrocycle unveils a uninegative, N 4 -donor set that allows the formation of iron(II) and Zn(II) complexes; unlike porphyrinic analogues, these complexes retain the ancillary bromide ligands.The observation of ligand-centred electronic transitions suggests a potentially rich ligand-centred redox chemistry that, combined with the Lewis acidity of the   metal, could lead to new chemical transformations.Our current investigations focus on the catalytic redox behaviour of this ligand and its complexes as well as exploring the possibility of using the NC(H)N pocket in the formation of dinuclear complexes.

Fig. 2
Fig. 2 Solid-state structure of HL and Li(L) viewed from side and top.For clarity, all hydrogen atoms (except H1) and solvent of crystallisation are omitted (displacement ellipsoids are drawn at 50% probability).Li(L) top view is depicted including close-contact atoms of adjacent molecule.Crystals of Li(L) were grown by slow evaporation of a concentrated n-hexane solution.

Fig. 3
Fig.3Solid-state structures of ZnBr(L) and FeBr(L), respectively, viewed from side and top.For clarity, all hydrogen atoms (except H1) and solvent of crystallisation are omitted (displacement ellipsoids drawn at 50% probability).Crystals were grown by slow evaporation of dilute, saturated n-hexane solutions.