Synthesis, structural studies, and oxidation catalysis of the manganese(II), iron(II), and copper(II) complexes of a 2-pyridylmethyl pendant armed side-bridged cyclam

Abstract

The first 2-pyridylmethyl pendant armed structurally reinforced cyclam ligand has been synthesized and successfully complexed to Mn²⁺, Fe²⁺, and Cu²⁺ cations. X-ray crystal structures were obtained for the diprotonated ligand and its Cu²⁺ complex demonstrating pentadentate binding of the ligand with trans-II configuration of the side-bridged cyclam ring, leaving a potential labile binding site cis to the pyridine donor for interaction of the complex with oxidants and/or substrates. The electronic properties of these complexes were determined by means of solid state magnetic moment, with a low value of μ = 3.10 μ_B for the Fe²⁺ complex suggesting it has a trigonal bipyramidal coordination geometry, matching the crystal structure of the Cu²⁺ complex, while the μ = 5.52 μ_B value for the Mn²⁺ complex suggests it is high spin octahedral. Cyclic voltammetry in acetonitrile revealed reversible redox processes in all three complexes, suggesting catalytic reactivity involving electron transfer processes are possible for these complexes. Screening for oxidation catalysis using hydrogen peroxide as the terminal oxidant identified the Fe²⁺ complex as the oxidation catalysts most worthy of continued development.

We have studied cross-bridged cyclam complexes of manganese and iron for nearly a decade and a half as oxidation catalysts. [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] The manganese complex of 4,11-dimethyl-1,4,8,11-tetraazabicyclo[6.6.2]hexadecane [15] (Me₂EBC, Scheme 1) in particular, has a rich oxidation chemistry. [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] This compound, which we propose to call “the Busch catalyst”, was initially synthesized as an oxidation catalyst because the cross-bridged ligand could rigidly bind the oxygen-reactive manganese metal and stop it from being lost in the form of MnO₂. [1] [2] [3] [4] Que has determined that the iron complex of Me₂EBC is an efficient olefin epoxidation catalyst with H₂O₂ oxidant under appropriate conditions as well. [16]

Scheme 1.

Ligands discussed in this paper.

Cyclam ligands structurally reinforced with an ethylene bridge between nitrogens 1 and 4 (i.e. “side-bridged”) have been known since 1980. [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] The additional bridge provides rigidity, configurational selectivity, and kinetic stability and has been synthetically put in place by several different methods. [17] [18] [19] [20] [21] Like other cyclam ligands, additional utility can be added by addition of pendant arms, which have included in Scheme 1, structure 1: R and/or R′ = H, [17] [21] Me, [19] Et, [20] Bz, [19] [23] [22]Bz-p-NO₂, [22] [25]Bz-p-NH₂, [22] CH₂CO₂H, [24] [26]CH₂PO₃H [25] [26] and others.

An interesting 2-pyridylmethyltrimethylcyclam ligand (Scheme 1, structure 2) was reported by Que and coworkers [27] and stimulated our present work. Its iron complex activated dioxogen and formed an oxoiron(IV)intermediate that was crystallographically characterized, but has not been pursued further as a catalyst. [27] Although 2-pyridylmethyl pendant armed unbridged cyclams are ubiquitous, [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] no 2-pyridylmethyl pendant armed side-bridged or cross-bridged cyclams had been reported prior to our work. Due to our experience with bridged cyclam oxidation catalysts, we set out to synthesize and characterize side- and cross-bridged cyclam ligands containing a 2-pyridylmethyl pendant arm, along with their most biologically relevant oxidation active metal ion complexes: manganese, iron, and copper. Our cross-bridged work has been reported elsewhere. [39] Here, we report on the results of our efforts with the side-bridged cyclam ligand, (Scheme 1, structure 3) and the screening of its complexes as potential oxidation catalysts.

Alkylation of the cyclam-glyoxal bisaminal 4 in typical S_N2 solvents like acetonitrile was unsuccessful due to self-reaction of the picolyl chloride. So, we explored the chlorinated solvents [40] to stabilize this reagent in the presence of the bisaminal and found successful, although low- yielding, monoalykylation with picolyl chloride in chloroform at room temperature. [41] We raised the yields of the monoalkyl salt 5 to 66% by addition of KI and increasing the temperature to reflux over 6 days. Typical reduction conditions [23] gave 3 in good yield. Metal complexation with anhydrous metal chlorides in dry solvents under nitrogen gave adequate yields of the desired complexes. [41]

The X-ray crystal structures of the H₂3Cl₂ • 2 H₂O and [Cu(3)][PF₆]₂ • C₃H₆O were obtained and are depicted in Figure 1. [42] The ligand is protonated at one of the piperazine tertiary nitrogens, and at the secondary nitrogen of the cyclam ring. There is an extensive hydrogen bonding network between the chloride anions, the two waters of crystallization, the cyclam nitrogens, and their protons. The pyridine nitrogen is oriented away from the center of the cyclam ring and is not involved in the hydrogen bonding network.

Figure 1.

(a) Structure of H₂3Cl₂ • 2 H₂O (b) Structure of Cu(3)²⁺ depicted as an approximate trigonal bipyramidal geometry (c) Structure of Cu(3)²⁺ depicted as an approximate square-based pyramidal geometry and showing that the two slightly different ligand orientations are approximately enantiomers.

Upon coordination of Cu²⁺, the pyridine nitrogen is oriented into the center of the cyclam ring and coordinates the Cu²⁺ ion located there. This compound contains Cu²⁺ in two different coordination geometries. The ligand coordinates through each of the nitrogen atoms but in a slightly different orientation. The two copper complexes are approximately enantiomers, as shown in Figure 1(c). The coordination geometry around the copper(II) ion is somewhat between trigonal bipyramidal (as shown in Figure 1(b)) and square pyramidal (as shown in Figure 1(c)). To illustrate the slight differences between the two different metal ion coordination geometries of this structure, square pyramidal type bond angles for the two different copper(II) ions are given: Cu(1) N–Cu–N angles / °: (in plane) 100.9(2), 74.0(2), 94.75(19), 87.79(19); (to N5) 82.91(17), 118.02(18), 101.94(17), 97.03(18). Cu(2) N-Cu-N angles / °: (in plane) 102.10(19), 73.4(2), 91.7(2), 86.2(2); (to N25) 80.5(2), 107.85(18), 107.37(19), 120.5(2). Complete metrical parameters for both structures are given in the Supplementary Information.

Unfortunately, crystal structures of the Mn²⁺ and Fe²⁺ complexes were not obtained. We sought other hints at their structures and turned to solid state magnetic moments to help determine them. First, the magnetic moment of [Cu(3)][PF₆]₂ was determined to be 2.32 μ_B, which is consistent with n = 1 for a d⁹ Cu²⁺ ion. [43] [Mn(3)Cl]PF₆ was determined to have μ = 5.52 μ_B, which is consistent with n = 5 for a high spin octahedral d⁵ Mn²⁺ metal ion. [43] High spin n = 5 is not diagnostic of a particular coordination geometry for Mn²⁺, but the consistent presence of one chloride in the elemental analysis of this complex led us to conclude that the most likely structure is a 6-coordinate pseudo-octahedral one. If the chloride is not coordinated and the complex is only 5-coordinate, precipitation by PF₆^- would have led to the dihexafluorophosphate salt.

Perhaps the most interesting contribution to structural information came from the magnetic moment of [Fe(3)]Cl₂. μ = 3.10 μ_B, which is most consistent with n = 2 for a low spin 5-coordinate d⁶ Fe²⁺ metal ion. [44] In our previous experience with Mn²⁺ and Fe²⁺ complexes of bridged tetraazamacrocycles, [1] [3] [45] both ions agree in that they were high spin and 6-coordinate. However, in this case, Mn(3)Cl⁺ appears 6-coordinate and high spin, while Fe(3)²⁺ appears 5-coordinate and low spin. Two supporting pieces of evidence for this perhaps unexpected geometry for the Fe²⁺ complex are (1) the 5-coordinate Cu(3)²⁺ crystal structure presented above for this ligand system; and (2) an unpublished crystal structure in which Fe²⁺ is coordinated to the four nitrogens of a similar ethylene side-bridged ligand having a methyl group in place of the pyridylmethyl group of 3, and a chloro ligand. [39] Interestingly, a six-coordinate 1,8-bispyridylmethylcyclam Fe²⁺ complex with spin-crossover properties and a transition temperature to low spin of 150 K is known. [46] Of course, low spin d⁶ Fe²⁺ complexes are ubiquitous in organometallic chemistry. A number of 5-coordinate low-spin Fe²⁺ examples have been published. [47] [48] [49] [50]

In anticipation of carrying out oxidation studies, we obtained cyclic voltammograms (Figure 2) on these complexes in acetonitrile looking for multiple stabilized oxidations states if catalytic processes were to be likely. Unfortunately, only one reversible redox wave was observed for each complex. Relative to SHE, Cu(3)²⁺ gave a reversible Cu^1+/2+ redox wave at E_1/2 = -0.586 V (ΔE = 77 mV). No oxidation to Cu³⁺ was observed, which is perhaps not surprising, as there are no negatively charged ligands to help stabilize the Cu³⁺ cation. Cu(Me₂EBC)Cl⁺ has an irreversible Cu²⁺ to Cu⁺ reduction wave at E_red = -0.544 V and an irreversible Cu²⁺ oxidation to Cu³⁺ at E_ox = +1.530 V. [51] In comparison, Cu(3)²⁺ is more difficult to oxidize (due to lack of negatively charged ligands) and more reversibly reduced.

Figure 2.

Cyclic voltammograms in acetonitrile vs. SHE for (a) Cu(3)²⁺, (b) Fe(3)²⁺, and (c) Mn(3)Cl⁺.

Mn(3)Cl⁺ gave a reversible Mn^2+/3+ redox wave at E_1/2 = +0.856 V (ΔE = 93 mV) and an irreversible reduction at E_red = -0.685 V. This can be compared to the well-known Mn(Me₂EBC)Cl₂ catalyst [1] which has reversible Mn^2+/3+ and Mn^3+/4+ waves at E_1/2 = +0.585 V (ΔE = 61 mV) and E_1/2 = +1.343 (ΔE = 65 mV), respectively. The single chloro ligand allows only Mn³⁺ to be accessed for Mn(3)Cl⁺, and at a higher potential, since only one negatively charged chloro ligand is present to stabilize the growing positive charge. Reduction is possible here, when not observed for Mn(Me₂EBC)Cl₂, but not reversible, likely due to the loss of the chloro ligand upon reduction to Mn⁺.

5-coordinate Fe(3)²⁺ exhibits a complex cyclic voltammogram with a reversible Fe^2+/3+ wave at E_1/2 = +0.456 V (ΔE = 78 mV), two irreversible reductions at E_red = -0.802 V and -1.671 V, respectively, and a large return oxidation at E_ox = -0.252 V. A much simpler behavior is observed for Fe(Me₂EBC)Cl₂, with only E_1/2 = 0.110 (ΔE = 63 mV) observed for Fe^2+/3+. [1] Oxidation is obviously much easier for the latter complex, where two negatively charged chloro ligands stabilize the positive charge. Reduction is observed only for Fe(3)²⁺, where no negatively charge ligands inhibit it. The complex behavior of Fe(3)²⁺ after the initial reduction is the subject of ongoing study.

Finally, we present initial oxidation screening data on these complexes. As shown in Table 1, compared with its analog, Mn(Me₂EBC)Cl₂ complex, the redox activities of Cu(3)²⁺, and Mn(3)Cl⁺ are quite poor. In sulfide oxidation using H₂O₂ as oxidant, while the Mn(Me₂EBC)Cl₂ catalyst provided nearly complete conversion of thioanisole (99.8%) with 44.3% yield of sulfoxide and 46.5% of sulfone, the Cu(3)²⁺ and Mn(3)Cl⁺ complexes are almost inactive for sulfide oxidation. However, Fe(3)²⁺ demonstrated some activity, providing 9.7% yield of sulfoxide, and 2.9% of sulfone with 15.6% conversion. Similarly, in hydrogen abstraction from 1,4-cyclohexadiene, Mn(Me₂EBC)Cl₂ gave 71.4% yield of benzene with 86.2% conversion, and Cu(3)²⁺ and Mn(3)Cl⁺ complexes were still inactive for benzene formation (∼3% yields represent natural benzene content in commercial 1,4-cyclohexadiene). Again, Fe(3)²⁺ provides more encouraging results, generating 23.6% yield of benzene with 44.5% conversion.

Table 1.

Oxidation catalysis screening results for manganese(II), iron(II), and copper(II) complexes of 3, in comparison to Mn(Me₂EBC)Cl₂.

Complex	Conversion %	Yield % Sulfoxide	Yield % Sulfone
Mn(Me2EBC)Cl2	99.8	44.3	46.5
[Mn(3)Cl]+	8.3	0.2	0.02
[Fe(3)]2+	15.6	9.7	2.9
[Cu(3)]2+	6.1	3.4	1.9
Complex	Conversion %	Yield % Benzene
Mn(Me2EBC)Cl2	86.2	71.4
[Mn(3)Cl]+	23.1	3.5
[Fe(3)]2+	44.5	23.6
[Cu(3)]2+	29.6	4

These results are perhaps not surprising, when reconciled with the electrochemical studies (Figure 2). Mn(3)Cl⁺ has only a redox couple of Mn^2+/3+ at E_1/2 = +0.856 V, which is much higher than that of the corresponding Mn^2+/3+ couple of Mn(Me₂EBC)Cl₂ (E_1/2 = +0.585 V), and no access to Mn(IV). Therefore, the catalytic cycle of the Mn(3)Cl⁺ complex is very sluggish in oxidations. Similar sluggish redox behavior is observed in Cu(3)²⁺ since its oxidation to Cu(3)³⁺ is not observed. For the Fe(3)²⁺ complex, although its catalytic activity in sulfide oxidation, an oxygen transfer process, is poor, it still demonstrates a relatively good activity in hydrogen abstraction because its redox couple for Fe^2+/3+ is modest (E_1/2 = +0.456 V) and even lower than that of Mn(Me₂EBC)Cl₂. Consistent with the hydrogen abstraction activity of the Fe(3)²⁺ complex, Stack has reported[55] that Fe(PY5)(OH)³⁺, which has a potential of E_pc = + 0.555 V, is capable of stoichiometric hydrogen abstraction from 9,10-dihydroanthracene (PY5 = 2,6-bis-(bis(2-pyridyl)methoxymethane)pyridine). This result encourages us to continue the exploration of the Fe(3)²⁺ complex and its biological relevance to lipoxygenase[56] in the future.

In conclusion, a new pyridylmethyl N-pendant arm, side-bridged cyclam ligand, 3, has been synthesized, and its diprotonated salt structurally characterized, with a key synthetic step the use of non-polar chloroform to decrease the self-reactivity of picolyl chloride in the presence of bisaminal 4. Divalent Mn, Fe, and Cu complexes were synthesized and the copper(II) complex structurally characterized by X-ray crystallography as 5-coordinate Cu(3)²⁺ showing the cyclam ring of 3 in a trans-II configuration and the chelated pyridine nitrogen bound to the Cu²⁺ ion. Solid state magnetic moment determination and elemental analysis revealed a high spin, octahedral Mn(3)Cl⁺ cation, but a 5-coordinate, low-spin Fe(3)²⁺ cation. Electrochemical studies in acetonitrile revealed reversible access to only two oxidation states for each metal ion, whereas optimal behavior for successful oxidation catalysis is likely a greater redox range. Preliminary screens for oxidation catalysis using H₂O₂ as the oxidant carried out on all three complexes showed promising results only in the hydrogen atom abstraction of 1,4-cyclohexadiene by Fe(3)²⁺, which is consistent with another iron complex from the literature having a similar redox potential. [55] Future work will include expanding the range of oxidation reactions possible with this catalyst and determination of its oxidation catalysis mechanisms.

Scheme 2.

Synthesis of 3. a) i. CHCl₃, 2 eq. picolyl chloride hydrochloride, 4 eq. NaHCO₃, 1 h, filter to remove solids; ii. 1 eq. of (1), 2 eq. KI, reflux 6 d; b) i. 95% EtOH,5 eq. NaBH₄, N₂, reflux, 1.5 h; ii. 12 M HCl(aq), 30% KOH(aq), benzene extraction.

Research Highlights.

• -The first 2-pyridylmethyl side-bridge cyclam ligand and its X-ray crystal structure

• -The first transition metal complex of a 2-pyridylmethyl side-bridge cyclam ligand and its X-ray crystal structure

• -Cyclic voltammetry and oxidation screening of Mn²⁺, Fe²⁺, and Cu²⁺ complexes of this new ligand showing promising hydrogen atom abstraction for the Fe²⁺ complex