A Four-Coordinate Fe(III) Porphyrin Cation

Few families of compounds have been as intensely studied as iron porphyrin complexes. In spite of this, a truly four-coordinate Fe(III) porphyrin cation has never before been reported.¹ The exceptional electrophilicity of the four-coordinate [Fe^III(Porph)]⁺ cation will even coordinate arene solvent upon crystallization if the counteranion is extremely non-coordinating, as in the work of Reed with silver dihexabromocarborane,^2,3 and in solution, a Br atom of the hexabromocarborane anion binds to the Fe.⁴

Here we present the first purely four-coordinate Fe(III) porphyrin both in the solid state and in solution. To do so, we have combined both steric and electronic factors by creating a very sterically hindered bis-pocket siloxyl porphyrin in conjunction with a bulky and very weakly coordinating anion.

We have synthesized an extremely hindered bis-pocket siloxyl porphyrin, 5,10,15,20-tetrakis(2′,6′-bis(triisopropylsiloxy)phenyl)-porphyrin (H₂TipsiPP) by the reaction of the octahydroxyl iron porphyrin⁵ (5,10,15,20-tetrakis(2′,6′-dihydroxyphenyl)porphyrin with triisopropyl chlorosilane using imidazole as a catalyst (Scheme 1). In spite of the driving force of strong Si–O bond formation, high temperatures (250 °C) are required to overcome steric hindrance and fully silylate all eight sites. The resulting porphyrin is extremely sterically hindered with a pocket opening of only 2 Å. Figure 1 compares the space-filling models of the simple FeTPP(Cl) and the bis-pocket siloxylporphyrin Fe^III(TipsiPP)(Cl). The steric hindrance prevents anion coordination by reducing the thermodynamic binding constant.

Scheme 1.

Synthesis of a Bis-pocket Siloxylporphyrin, Fe^III(TipsiPP)(Cl)

Figure 1.

Space-filling model of Fe^IIITPP(Cl) (left) and Fe^III(TipsiPP)-(Cl) (right); side view, same scale.

To replace the sterically undemanding chloro ligand, a dichlo-romethane solution of Fe^III(TipsiPP)(Cl) was mixed with 1 equiv of AgCB₁₁H₆Br₆ or AgCF₃SO₃, which produces a color change from yellow to red (Figure 2). NMR, MALDI-TOF, and elemental analysis were used to confirm the purity of the final products. Both complexes react with water readily to form a five-coordinate spin-admixed Fe(III) complex, which has diagnostic UV–vis and NMR features (Figures S1 and S2 in Supporting Information (SI)).

Figure 2.

UV–visible spectrum of five-coordinate starting material Fe^III(TipsiPP)(Cl) (dotted line) and four-coordinate [Fe^III(TipsiPP)]⁺[CB₁₁H₆Br₆]⁻ (solid line). Inset shows an expanded absorbance scale.

Proton NMR is a very sensitive probe for the spin state of iron.^4,6,7 On the basis of the β-pyrrole proton NMR isomer shift, Reed has established a magnetochemical series for different anions.⁸ Surprisingly, solutions of [Fe^III(TipsiPP)]⁺ with a variety of weakly coordinating anions (i.e., CB₁₁H₆Br₆⁻, SbF₆⁻, ClO₄⁻, and CF₃SO₃⁻) all have exactly the same porphyrin ¹H NMR chemical shifts with β-pyrrole proton chemical shift at −81 ppm in CD₂Cl₂ at 290 K (Figure S3 in SI), which corresponds to the Fe(III) intermediate spin state (S = ³/₂).⁸ Because these anions are very different both in nucleophilicity and size, the porphyrin NMR spectra would be different if the anions were bound to the iron. This is the case for Fe^IIITPP⁺: β-pyrrole proton chemical shifts for FeTPP(X) are −60 ppm for CB₁₁H₆Br₆⁻, −31.5 ppm for SbF₆⁻, 13 ppm for ClO₄⁻, and 39.3 ppm for CF₃SO₃⁻.^8,9 Possible assignments to either Fe^III(TipsiPP^•+) porphyrin cation radicals or Fe^II(TipsiPP) were excluded by independent synthesis of the pure complexes, which show very different chemical shifts and have different spin states.

The solvent plays a central role in stabilization of the four-coordinate Fe(III) cation. The four-coordinate species is only stable in halogenated solvents (i.e., CH₂Cl2, CHCl₃, CH₂Br₂). In aromatic solvents, such as toluene and benzene, it converts to an admixed spin state, probably due to arene coordination to the iron center.³ The β-pyrrole chemical shift of [Fe^III(TipsiPP)]⁺ is the same in either CH₂Cl₂ or CH₂Br₂, which shows that the halocarbons do not perturb the iron center. The compound is not soluble in aliphatic solvents.

Red-brown needle-shaped crystals of [Fe^III(TipsiPP)]⁺[CB₁₁H₆Br₆]⁻ were grown by slow evaporation of a 1:1 mixture of dichloro-methane/heptane solution inside an inert atmosphere box. The single crystals were redissolved in dry CD₂Cl₂ and tested by NMR to confirm no contamination with water during the crystal growing process. The crystal structure (Figure 3) confirms the absence of axial ligation to the iron porphyrin. The average Fe–N bond length is 1.94 ± 0.01 Å with the iron sitting in the porphyrin plane. This very short bond length is consistent with the removal of the unpaired electron from the d_x²_−y² orbital.^1,10 The porphyrin is slightly ruffled, with meso-carbons above and below the average nitrogen plane by 0.23 Å. The carborane anion is definitively not coordinated to the iron and is fully outside of the porphyrin pocket. The resulting salt has a layered structure with alternating planes of positive and negative charges (a crystal packing diagram is given as Figure S9 in SI). A similar packing structure has been reported for [FeTPP]-[Ag(CB₁₁H₆Br₆)₂]·4(p-xylene) with the exception of the coordination of the xylene to the iron center.^2,3

Figure 3.

X-ray single-crystal structure of (a) [Fe^III(TipsiPP)]⁺[CB₁₁H₆Br₆]⁻ and (b) Fe^III(TipsiPP)(CF₃SO₃). Figure S9 in SI provides a crystal packing diagram for[Fe^III(TipsiPP)]⁺[CB₁₁H₆Br₆]⁻ showing a layered salt structure with alternating planes of positive and negative charges.

In the solid state, the four-coordinate iron porphyrin cation can only be isolated with hexabromocarborane as the counterion. In crystals of solids with other anions (e.g., CF₃SO₃⁻), the anion enters the porphyrin pocket and coordinates to the iron, presumably driven by the loss of solvation of the anion. The powder X-band EPR of Fe^III(TipsiPP)(CF₃SO₃) shows a g_⊥=5.6 signal, consistent with an admixed spin assignment (Figure S4 in SI). The average Fe–N bond length in Fe^III(TipsiPP)(CF₃SO₃) is 2.053(2) Å (which is 0.1 Å longer than in the true four-coordinate structure), and the iron atom is out of the plane of the four nitrogens by 0.406 Å. To accommodate the triflate ion, the phenyl groups rotate ~12° about the plane normal to porphyrin, which further demonstrates the interplay of electronic and steric factors in these highly hindered systems.

Further characterization data on the four-coordinate [Fe^III-(TipsiPP)]⁺[CB₁₁H₆Br₆]⁻ are all consistent with a ³/₂ intermediate spin ground state. The solution magnetic susceptibility of the four-coordinate complex is 4.1 μ_B in a dichloromethane solution measured by Evan’s method from room temperature to 190 K (Figure S5 in SI); SQUID data show no change in the magnetic moment between 300 and 10 K. EPR data of both frozen halocarbon solutions and the crystalline [Fe^III(TipsiPP)]⁺[CB₁₁H₆Br₆]^minus; have a g_⊥ = 4.2 signal (Figure S4 in SI). Mössbauer spectra were obtained with the ⁵⁷Fe-enriched sample at 6 K, with an isomer shift of 0.33 mm/s (a typical value for Fe(III)), but with a very large quadrupole splitting value of 5.16 mm/s (Figure S6 in SI). Solid-state NMR also gives an upfield signal at −80 ppm for the β-pyrrole hydrogens, confirming that the four-coordinate Fe(III) porphyrin is stable both in solid and solution phases.

Although four-coordinate Fe(III) heme is unlikely in any heme protein, the intermediate spin state of Fe(III) porphyrins has an interesting counterpart in nature.^11–17 In cytochrome c′, the heme is believed to be a quantum admixed S = ⁵/₂ and S = ³/₂ spin state.^18–20 Depending on the field strength of the ligand coordinated to the iron, the ratio of S = ⁵/₂ to S = ³/₂ character varies: as the ligand strength becomes weaker, the spin state of iron approaches the pure intermediate spin state of ³/₂.

In summary, the first four-coordinate Fe(III) porphyrin complex has been isolated and fully characterized in both solution and solid state and shows a pure S = ³/₂ intermediate spin state. The combination of steric hindrance of a bis-pocket porphyrin with weakly coordinating anions has proved essential in isolation of this highly elusive species, whose reactivity we are still exploring.