Synthesis and structural studies of two pyridine-armed reinforced cyclen chelators and their transition metal complexes

Abstract

Two novel pyridine pendant-armed macrocycles structurally reinforced by an ethyl bridge, either between adjacent nitrogens (for side-bridged) or non-adjacent nitrogens (for cross-bridged), have been synthesized and complexed with a range of transition metal ions (Co²⁺, Ni²⁺, Cu²⁺ and Zn²⁺). X-ray crystal structures of selected cross-bridged complexes were obtained which showed the characteristic cis-V configuration with potential labile cis binding sites. The complexes have been characterized by their electronic spectra and magnetic moments, which show the expected high spin divalent metal complex in most cases. Exceptions are the nickel side-bridged complex, which shows a mixture of high-spin and low spin, and the cobalt cross-bridged complex which has oxidized to cobalt(III). Cyclic voltammetry in acetonitrile was carried out to assess the potential future use of these complexes in oxidation catalysis. Selected complexes offer significant catalytic potential enhanced by the addition of the pyridyl arm to a reinforced cyclen backbone.

Synopsis

Mono-pyridyl pendant armed side-bridged and cross-bridged cyclen ligands have been synthesized and complexed to Co, Ni, Cu, and Zn. Their electronic, electrochemical, and structural characterization suggests that they may be useful as catalysts.

1. Introduction

Tetraazamacrocycles form complexes of high thermodynamic and kinetic stability with transition metals. [1] Depending on the chelated metal, tetraazamacrocyclic complexes have found use in a range of applications, including medical imaging, [2] [3] [4] [5] protein binding agents, [6] [7] and as anti-malarial drugs. [8] [9] Recently we have been interested in the use of this complex type as oxidation catalysts. [10] [11] [12] A recent papers by Que, et al, has shown that addition of a pyridyl pendant arm to a tetraazamacrocycle can result in new and improved oxidation catalysis. [13] The literature contains numerous other pyridyl pendant armed tetraazamacrocycles, indicating a general interest in the community and known synthetic techniques for adding pendant pyridyl donors. [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24]

Our specific exploration is aimed at forming catalysts which can perform under a range of aqueous conditions. Transition metal complexes of tetraazamacrocycles reinforced with additional ethylene bridges have produced oxidation catalysts with high kinetic stability under such harsh conditions. [25] [26] [27] [28] [29] Transition metal complexes of macrocycles of this type are known to be stable to a range of pH’s and temperatures. A cross-bridged cyclen ligand with two pyridyl pendant arms (CB-Py₂Cyclen, Figure 1) has been reported [30], but its hexadentate nature coordinatively saturates its complexed metal ions, while we wished to leave at least one coordination site available for interaction with oxidant and/or substrate. Our first communications regarding complexes incorporating a single pyridyl labile coordinating group have described reinforced cyclam derivatives with modified steric and electronic properties to cause improved properties as oxidation catalysts, [10] [12] see Figure 1. Herein, we report the synthesis and characterization of analogous smaller cyclen derivatives as potential future oxidation catalysts. Comparisons between our cyclam and cyclen derivatives, as well as with other closely related cyclen-pyridyl ligands (Figure 1) will be presented.

Figure 1.

Ligands discussed within this work; SB-PyCyclam [10] PyMeEBC [12] Me₂BCyclen, [31] PyCyclen [21] [22] CB-Py₂Cyclen [30]; alongside SB-PyCyclen and CB-MePyCyclen first described in this study.

2. Results and discussion

2.1. Synthesis

We have recently demonstrated that alkylation of tetraazamacrocycle-glyoxal condensates using typical acetonitrile conditions described by Weisman et al. [32] [33] [34] and Handel et al. [35] are not ideal when using pyridyl derived alkyl halides and nonpolar chlorinated solvents are preferred. [10] [12] Analogous conditions were used for the formation of 1. Using dichloromethane gave the desired product in relatively high yields (~80%), see Scheme 1. Methylation of the non-adjacent nitrogen atom with methyl iodide gives a bis-quaternary ammonium salt 2 by what are now routine methods for cyclam, [12] [36] although this is the first reported use of methylation to form unsymmetrical ethyl bridged cyclen derivatives. Reduction of mono (1) and bis (2) quaternary ammonium salts is carried out with sodium borohydride to form ethyl side-bridged (SB-PyCyclen) 3 and cross-bridged (CB-MePyCyclen) 4 derivatives respectively using well described analogous reactions. [32] [33] [34]

Scheme 1.

Ligand synthesis. (a) (i) CH₂Cl₂, 2-picolyl chloride hydrochloride, NaHCO₃ (ii) Decahydro-2a,4a,6a,8a-tetraazacyclopenta[fg]acenaphthylene, KI. (b) MeCN, CH₃I. (c) EtOH, NaBH₄. (d) EtOH, NaBH₄.

Metal complexes have been formed with both SB 3 and CB 4 ligands with a range of divalent transition metals (Co²⁺, Ni²⁺, Cu²⁺ and Zn²⁺). Due to the proton sponge nature of cross-bridged tetraazamacrocyclic ligands, complexation reactions were carried out in an inert atmosphere glovebox using anhydrous metal salts in acetonitrile (with additional DMF for acetonitrile-insoluble NiCl₂). In a similar manner to our cyclam analogues, [10] [12] the reactions did not require heating and the complexes did not precipitate from the reaction solution. After drying the crude solutions the complexes were purified by changing the counter ion from chloride to hexafluorophosphate in methanol, with the conversion causing a reduction in solubility and precipitation of the pure complex salt.

According to elemental analysis, the copper complexes of both ligands, as well as the Ni²⁺, and Zn²⁺ complex of 3, replace both chloride counter ions with hexaflurophosphates, indicating no bound chloro ligands and five-coordinate geometries. In contrast, the other metal ion complexes show replacement of only one chloro counter ion, indicating a six-coordinate complexes with a coordinated chloride anion (vide infra). The five-coordinate behavior of the side-bridged ligand is consistent with the complexation of its cyclam analogue SB-PyCyclam, which coordinates Fe²⁺ and Cu²⁺ in a 5-coordinate manner. [10] Similarly, the 6-coordination (in combination with a chloro ligand) of the cross-bridged 4 is consistent with its cyclam analogue PyMeEBC with Mn²⁺, Fe²⁺, Co²⁺, Ni²⁺, and Zn²⁺. PyMeEBC is only 5-coordinate, as seen for 4 here, with Cu²⁺. [12] Uniquely, the cobalt complex of 4 has oxidized to Co³⁺ upon its workup in air. Its elemental analysis shows one chloro ligand completing its six-coordination, but the presence of two additional hexafluorophosphate counter ions. This behavior has been observed for other cross-bridged cyclen ligand analogues with cobalt. [30] [37]

2.2. X-ray crystal structures

X-ray crystal structures could be determined for the Ni²⁺, Cu²⁺ and Zn²⁺ complexes of CB-MePyCyclen (4), see Figure 2. These solid state structures are consistent with the elemental analyses in which Ni(4)Cl⁺ and Zn(4)Cl⁺ retain a chloro ligand, but Cu(4)²⁺ displaces both. All three complexes crystalized in the monoclinic P21 space group. Ni(4)Cl⁺ and Zn(4)Cl⁺ are isostructural and will be discussed collectively, with Cu(4)²⁺ crystalizing differently.

Figure 2.

X-ray crystal structures of Zn(4)Cl⁺ (left) Ni(4)Cl⁺ (middle) and Cu(4)²⁺ (right) showing 50% ellipsoids.

Ni(4)Cl⁺ and Zn(4)Cl⁺ have the metal ion in a slightly distorted octahedral environment, coordinating with the four nitrogen atoms of the folded macrocycle along with the pyridyl arm and a chloride counter ion located cis to each other. Cu(4)²⁺ has displaced the chloride ion with a non-coordinating hexafluorophosphate and is subsequently square based pyramidal. The Addison Parameter [38] (τ) has been used to quantify the geometry of 5-coordinate structures, with τ = 0 for a perfect square pyramidal complex, and τ = 1 for a perfect trigonal bipyramidal one. For Cu(4)²⁺ τ = 0.09, which confirms our square pyramidal designation. Of note for Cu(4)²⁺, The “empty” coordination position on the square pyramidal Cu²⁺ ion is occupied in the solid state structure by one of the PF₆⁻ anions, with the Cu(1)-F(1) distance at 2.787 Å (see Figure 2c). This Cu-F distance is rather typical for similar interactions where copper is coordinated by four or five nitrogen atoms. [39]

The cross-bridged tetraazamacrocycle complexes show the well-recognized “cis-V” configuration, [40] with the metal ion being slightly pushed out of the macrocyclic plane. The degree of distortion from octahedral can be quantified by looking at N_ax-M-N_ax and N_eq-M-N_eq bond angles, where N_ax are the non-cross-bridged nitrogen atoms of the cyclen ring, while N_eq are the cross-bridged nitrogen atoms. For comparison (Table 1), the metrical parameters of the complexes of the two most relevant ligand analogues, Me₂Bcyclen and PyMeEBC, with the same metal ions and in the same coordination geometries are provided.

Table 1.

Structural data for transition metal complexes

Metal ion	Ionic radiusa [41]	Me2BCyclen		Ref	PyMeEBC		Ref	CB-MePyCyclen
		Nax-M-Nax	Neq-M-Neq		Nax-M-Nax	Neq-M-Neq		Nax-M-Nax	Neq-M-Neq
Zn2+	88	150.75	78.32	[31]	168.66	81.84	[12]	153.69(10)	81.26(11)
Ni2+	83	161.58	85.46	[42]	173.56	84.85	[12]	160.8(2)	84.46(19)
Cu2+	79	164.06	85.95	[37]	175.66	85.24	[12]	170.6(2)	83.0(2)

^aHigh spin 6-coordinate (Zn²⁺/Ni²⁺) or 5-coordinate (Cu²⁺)

The first trend that is apparent from this data is that the N_ax-M-N_ax and N_eq-M-N_eq bond angles increase (becoming less distorted) as the ionic radius of the metal ion decreases. For example, the smaller Cu²⁺ ion fits into the CB-MePyCyclen cavity (N_ax-M-N_ax = 170.6° and N_eq-M-N_eq = 83.0°) with less distortion of the ideal 180° and 90° expected bond angles than the larger Zn²⁺ ion (N_ax-M-N_ax = 153.69° and N_eq-M-N_eq = 81.26°). This trend is also followed by analogous bridged ligands Me₂Bcyclen and PyMeEBC.

Another interesting comparison is between the larger PyMeEBC complexes and the small CB-MePyCyclen complexes. From the bond angles it is obvious the 14-membered cyclam ring of PyMeEBC better accommodates these metal ions with less distortion from the expected bond angles than does the 12-membered cyclen ring of CB-MePyCyclen. For example, the Zn²⁺ ion is difficult for CB-MePyCyclen to hold in an undistorted octahedral geometry (N_ax-M-N_ax = 153.69° and N_eq-M-N_eq = 81.26°) while the larger PyMeEBC accommodates Zn²⁺ with much less distortion of the octahedral geometry (N_ax-M-N_ax = 168.66° and N_eq-M-N_eq = 81.84°).

Finally, the effect of the pyridine pendant arm on the coordination geometry can be analyzed by comparing the coordination geometries of the CB-MePyCyclen complexes with the pendant-arm-lacking Me₂BCyclen complexes. In each ion’s case, addition of the pendant arm allows the metal ion to be better engulfed by the bridged macrocycle with less geometric distortion than for Me₂Bcyclen, at least judged by the N_ax-M-N_ax angles. For example Cu(Me₂Bcyclen)Cl⁺ has N_ax-M-N_ax = 164.06°, while Cu(CB-MePyCyclen)²⁺ has N_ax-M-N_ax = 170.6°. Interestingly, the trend towards larger bond angles for CB-MePyCyclen complexes does not always extend to the N_eq-M-N_eq angles, which are smaller for Cu²⁺ and Ni²⁺ (see Table 1), but larger for Zn²⁺.

Structural comparisons with CB-Py₂Cyclen [30] and PyCyclen [21] [22] (see Figure 1) are also desirable. Unfortunately, although Co³⁺, Cu²⁺, and Zn²⁺ complexes with CB-Py₂Cyclen, the most direct analogue including an ethylene cross-bridge, have been synthesized and characterized, no X-ray crystal structures have been published. However, [Ni(PyCyclen)(CH₃CN)]²⁺ [22] and [Cu(PyCyclen)]²⁺ [21] have published crystal structures.

Comparison between octahedral Ni(4)Cl⁺ and octahedral [Ni(PyCyclen)(CH₃CN)]²⁺ [22] begins by recognition that the pyridine donor of Ni(4)Cl⁺ is forced by the short cross-bridge to be located on an axially positioned nitrogen of the coordinated folded macrocycle, positioning the pyridine donor on the equator of the distorted octahedral complex with the pyridine ring perpendicular to the equator. However, in [Ni(PyCyclen)(CH₃CN)]²⁺, although the unbridged cyclen is similarly folded, the pyridyl donor is attached to an equatorial positioned cyclen nitrogen with the pyridine ring parallel to the equator. For comparison to Table 1, the N_ax-Ni-N_ax and N_eq-Ni-N_eq bond angles for in [Ni(PyCyclen)(CH₃CN)]²⁺ are 156.00° and 104.01° respectively. These values are 160.8° and 84.46° for Ni(4)Cl⁺. Clearly, the most significant difference is the ~20 degree larger N_eq-Ni-N_eq bond angle for the unbridged PyCyclen complex, demonstrating the severe effect of the short ethylene cross-bridge on complex coordination geometry. Very little difference in Ni-N bond lengths are noted, with the Ni-N_py bond lengths nearly identical at 2.085 and 2.083 Å, respectively.

Comparison of Cu(4)²⁺ and Cu(PyCyclen)²⁺ can begin with their Addison Parameters [38] which are t = 0.09 for both compounds, indicating similar square pyramidal coordination geometries. However, the axial position is occupied by a macrocycle nitrogen in cross-bridged Cu(4)²⁺, while the axial position is occupied by the pyridine nitrogen in unbridged Cu(PyCyclen)²⁺. Due to the Jahn-Teller distortion of d⁹ Cu²⁺, the axial bond is typically lengthened and weakened. Here, that results in a long Cu-N_py bond of 2.164 Å for Cu(PyCyclen)²⁺, while the equatorial pyridine of Cu(4)²⁺ has a Cu-N_py bond length of 2.012 Å. The short, strong, equatorial coordination of the non-macrocycle donor has been credited with the more effective CXCR4 antagonism and longer residence time of Cu²⁺ cross-bridged tetraazamacrocyle CXCR4 antagonists. [43] The N_eq-Ni-N_eq bond angle for Cu(4)²⁺ is 83.0° and is this small because of the short ethylene cross-bridge. Cu(PyCyclen)²⁺ does not have a completely analogous bond angle, because the macrocycle occupies the base of the square pyramid, rather than having on macrocycle nitrogen at the axial position. The smallest nonadjacent N_cyclen-Cu-N_cyclen bond angle in Cu(PyCyclen)²⁺ is 146.94°, reflecting the very large structural difference with Cu(4)²⁺.

2.3. Electronic structure

The magnetic moments and UV-vis spectra were recorded on all paramagnetic ions in an attempt to glean structural information about the complexes, see Table 4. The elemental analysis of Ni(3)²⁺ shows no bound chloride, making it at most 5-coordinate. However, the electronic spectrum appears to indicate a square-planar 4-coordinate structure. Square planar Ni²⁺ usually has only one absorption near 460 nm, [44] in this case, it is at 441 nm with an extinction coefficient of 234 M⁻¹cm⁻¹, indicating that the pyridine pendant arm does not bind to the metal center in this complex. The magnetic moment of this complex is 1.13 BM, which is well below the expected value for a high spin Ni²⁺ ion of 2.80–3.50. [45] This low value indicates that there may be a mixture of high and low spin Ni²⁺ complexes in the solid state, which is in agreement with the low-spin square planar electronic spectrum in an acetonitrile solution.

Table 4.

Electronic structural data for transition metal complexes

	Electronic Spectra		Magnetic moment
Compound	λmax (nm)	ε (M−1cm−1)	Literature range (BM)	Experimental value (BM)
Co(3)Cl+	219	35100	4.30–5.20	4.10
	521	100.3
	263(sh)	5760
Ni(3)2+	240	4820	2.80–3.50	1.13
	441	230
Cu(3)2+	612	210	1.70–2.20	1.92
	289	6140
	263.5	8520
	216	34460
Co(4)Cl2+	227	44500	0	0
	499	280
	245(sh)	2310
Ni(4)Cl+	847	31	2.80–3.50	3.00
	538	26
	233	440
	332(sh)	150
Cu(4)2+	216	33600	1.70–2.20	1.91
	262	5700
	295	3220
	559	390

The elemental analysis of the cross bridged Ni(4)Cl⁺ points to a bound chloride ligand, while its electronic spectrum supports an octahedral Ni²⁺ ion. Such an ion generally exhibits three weak absorptions from 300–1100 nm. [44] In this case, these absorptions are at 332, 538, and 847 nm. The electronic spectra of octahedral Ni²⁺ complexes can be used to determine the ligand field strength of the ligands, [46] equivalent to the energy of the lowest energy absorption. In this case Δ_o = 11,800 cm⁻¹. This can be compared to the bridged cyclen ligand having two methyl groups and no pyridine donor, [Ni(Me₂BCyclen)Cl₂] which has Δ_o = 9,840 cm⁻¹. [42] The pyridine donor and 5-coordinate ligand in the present case greatly increases the ligand field strength. Pyridine is a strong field ligand in the spectrochemical series [45] as its empty π orbitals allow for back bonding, while Cl⁻ is a weak field ligand because its filled p-orbitals reduce Δ_o. This trend is clearly reflected in the Δ_o values observed. Also, comparing with ligand field strength of the very similar [Ni(PyMeEBC)Cl₂] complex (Δ_o = 11,223 cm⁻¹) [12] shows a slight increase for Ni(4)Cl⁺. Additionally, the magnetic moment for Ni(4)Cl⁺ (3.00 BM) is consistent with high-spin Ni²⁺, which generally ranges from 2.80–3.50 BM. [45]The elemental analysis, electronic spectrum, and magnetic moment all point to an octahedral, high spin, Ni²⁺ complex in this case, in contrast to the low spin, square planar complex of Ni(3)²⁺. This result is expected, as the short 2-carbon cross bridge allows only for a folded, cis configuration of the cyclen ring, which leads only to the distorted octahedral structure and a high spin Ni²⁺ cation. The electronic data for Ni(4)Cl⁺ is in general very similar to the slightly larger cyclam analogue. [12]

The side-bridged complex Co(3)Cl⁺ has an elemental analysis matching a Co²⁺ ion with one bound chloride and a PF₆⁻ anion. Its electronic spectra has a weak absorbance at 521 nm (ε = 100 M⁻¹cm⁻¹). This is consistent with high spin cobalt(II), which generally has a single band centered between 500–600 nm. [44] This observation is in agreement with the magnetic moment of the compound, 4.10 BM, which is also consistent with high spin cobalt(II). [45]

However, the cross-bridged complex Co(4)Cl²⁺ has an elemental analysis that indicates an oxidized Co³⁺, having one bound chloride ligand and two PF₆⁻ anions. This behavior was not noted with the cross-bridged cyclam analogue. [12] This compound also gave a diamagnetic magnetic moment, consistent with low spin d⁶ Co³⁺ in an octahedral environment. Its electronic spectrum is similar to Co(3)Cl⁺, yet the absorbances are stronger and there are two peaks plus a shoulder: 499 (281), 387 (281), and 252 nm (2307 M⁻¹cm⁻¹). These bands are consistent with the di-methyl ligand complex [47] [Co(Me₂EBCyclen)Cl₂]⁺: 577 nm (250), 415(210), and 346(950). This cobalt(III) complex had three bands with similar intensities and included a shoulder for the middle absorbance. The major difference between the pyridine-armed and unarmed complexes is the increase in energy of each of the bands for the pyridine armed complex. This can be quantified by calculating Δ_o for the pyridine-armed complex, which is accomplished by adding the Racah parameter, 3800cm⁻¹ to the energy of the lowest energy band for an octahedral Co³⁺ complex. [48] In this case Δ_o = 23,840 cm⁻¹. This compares to the Δ_o = 21,130 cm⁻¹ value for the [Co(Me₂EBCyclen)Cl₂]⁺ complex. As in the case with nickel(II), a clear increase in ligand field strength is observed, which makes sense as we have added a strong field pyridine ligand in place of a weak field chloride anion.

The electronic spectra of the Cu²⁺ complexes in acetonitrile demonstrate the expected ligand field transitions [45] for d⁹ Cu²⁺. A number of high-intensity bands are present in the UV, which are likely ligand to metal charge transfer bands. Each complex has one d-d absorption: at 559 nm (390 M⁻¹cm⁻¹) for Cu(3)²⁺ and at 612 nm (210 M⁻¹cm⁻¹) for Cu(4)²⁺. Cu²⁺ electronic spectra generally exhibit one such d-d band, which has an energy equivalent to Δ_o. Thus, Δ_o = 17,900 cm⁻¹ for Cu(3)²⁺ and Δ_o =16,300 cm⁻¹ for Cu(4)²⁺. The oxidation state is confirmed by the magnetic moments, Cu(3)²⁺ μ = 1.91 BM and Cu(4)²⁺ μ = 1.92, both consistent with the usual Cu²⁺ range of 1.70–2.20 BM, [31] and similar to that of the cross-bridged cyclam analogue. [12]

2.4. Electrochemical Studies

After our encouraging results using transition metal complexes of SB-PyCyclam [10] and PyMeEBC [12] in oxidation catalysis, we wished to determine if a reduction in macrocyclic size would result in beneficial shifts in the red-ox potentials of their transition metal complexes and also determine the degree of influence achieved by addition of the pyridyl arm compared to the analogous non-pyridyl complexes [M(Me₂BCyclen)Cl₂]. [31] [42] [42] The electrochemical properties were assessed via cyclic voltammetry in acetonitrile; all redox potentials given are vs. SHE.

[Ni(Me₂BCyclen)Cl₂] in acetonitrile has an irreversible reduction to Ni⁺ with E_red = −2.036 V, a reversible Ni^2+/3+ couple at E_1/2 = +0.863 V (ΔE = 68 mV), and an additional unassigned irreversible oxidation at E_ox = +1.450 V that is most likely due to oxidation of a bound chloride ligand. [42] Ni(3)²⁺ has only an irreversible reduction at E_red = −1.287 V, likely from Ni²⁺ to Ni⁺, and an irreversible oxidation at E_ox = +1.375 V, likely from Ni²⁺ to Ni³⁺. In comparison to [Ni(Me₂EBCyclen)Cl₂], the reduction to Ni⁺ is substantially (749 mV) easier for this side-bridged complex with a pyridine arm. Recall that this complex appears to be low spin, d⁸, square planar Ni²⁺ in its electronic structure, with no bound chloride. Reduction would result in a d⁹ Ni⁺ ion, which has a preference for 5-coordinate structures. [48] Perhaps the easier reduction is due to lack of a negatively charged chloro ligand combined with the ready presence of a fifth donor in the form of the pyridine pendant arm. The irreversibility of this reduction would be explained by the change in structure if the pyridine donor does indeed bind upon reduction to Ni⁺. The lack of a chloro donor would also explain the missing second oxidation, which was assigned to oxidation of a bound chloro ligand in [Ni(Me₂EBCyclen)Cl₂]. Finally, the oxidation that is present at +1.375 V (likely Ni²⁺ to Ni³⁺) is at a much higher potential compared to this reversible oxidation in [Ni(Me₂EBCyclen)Cl₂], which is at +0.863 V. Again, lack of negatively charged chloro ligands may explain this difference, as the Ni³⁺ ion would not be as stabilized in Ni(3)²⁺. The lack of reversibility of this oxidation may also be explained by the Ni³⁺ ion binding to the pyridine pendant arm as it would require more electron donation to be stabilized.

Ni(4)Cl⁺ is quite different from either Ni(3)²⁺ or [Ni(Me₂EBCyclen)Cl₂. This complex gave no reduction wave in acetonitrile out to approximately −2.0 V which is very different from the reduction of the analogous Ni(PyMeEBC)Cl⁺ (−1.026 V). Ni(4)Cl⁺ does demonstrate a reversible oxidation at E_1/2 = +1.034 V (ΔE = 76 mV) and another quasi-reversible oxidation at E_1/2 = +1.313 V (ΔE = 87 mV). The first oxidation is likely from Ni²⁺ to Ni³⁺. Recall that the electronic structure of this complex, from its magnetic moment and electronic spectrum in acetonitrile appears to be octahedral Ni²⁺, which along with its elemental analysis implies binding of the pyridine pendant arm and one chloro ligand. This structure is apparently maintained upon oxidation, with a very reversible reduction wave present. This is similar to the [Ni(Me₂BCyclen)Cl₂] complex, where this Ni^2+/3+ couples occurs at +0.863 V. The increase in oxidation potential for the pendant arm complex could be explained by the presence of only one negatively charged chloro ligand, which makes it more challenging to reach Ni³⁺ than in [Ni(Me₂EBCyclen)Cl₂], which has two negatively charged chloro ligands and would make it easier to oxidize. When compared with the analogous Ni(PyMeEBC)Cl⁺ complex, we see a similar (+1.290 V) oxidation potential to Ni³⁺ which is slightly (0.171 V) stabilized compared with Ni(4)Cl⁺. The identity of the second quasi-reversible wave, which is less intense than the first one, could be assigned, as in [Ni(Me₂BCyclen)Cl₂] to oxidation of a chloro ligand. Although the quasi-reversible nature of this oxidation would not be expected for oxidation of a chloro ligand, which would not likely stay bound upon oxidation. An alternative explanation could be the presence of a second form of the complex in which the pyridine arm is not bound to the nickel ion. This form would have less electron density to stabilize Ni³⁺ and would therefore likely require a larger voltage to reach this oxidation state.

[Co(Me₂BCyclen)Cl₂] in acetonitrile has an irreversible reduction to Co⁺ with E_red = −2.202 V, a quasi-reversible Co^2+/3+ couple at E_1/2 = −0.157 V (ΔE = 288 mV), and an additional unassigned irreversible oxidation at E_ox = +0.983 V that is most likely due to oxidation of a bound chloride ligand. Co(3)Cl⁺, has a voltammogram much like [Co(Me₂EBCyclen)Cl₂]. Similarly, there is an irreversible reduction to Co⁺ with E_red = −1.884 V, compared to −2.202 volts for [Co(Me₂BCyclen)Cl₂]. Since there is only one negatively charged chloride ligand in Co(3)Cl⁺, it makes sense that it is easier to reduce this complex to Co⁺ than for [Co(Me₂EBCyclen)Cl₂]. The irreversible nature of both complexes upon reduction is probably due to loss of a ligand, either chloride or the pendant arm pyridine. The oxidation behavior of both complexes differs more than the reduction. [Co(Me₂EBCyclen)Cl₂] has a reversible oxidation to Co³⁺ at a mild potential of −0.157 V, whereas Co(3)Cl⁺ doesn’t oxidize to Co³⁺ until a potential of +1.121 V is reached. Again, having only one bound chloro ligand may make it less favorable to oxidize to Co³⁺ in this case. [Co(Me₂BCyclen)Cl₂] shows reversible behavior for this oxidation, while Co(3)Cl⁺ does not. Perhaps the pendant arm pyridine is dissociated or modified causing the irreversible behavior, where this pathway is not available to [Co(Me₂BCyclen)Cl₂]. Another similarity is that [Co(Me₂BCyclen)Cl₂] has a second oxidation, probably of bound chloro ligand at +0.983 V while Co(3)Cl⁺ also has a second irreversible oxidation at +1.411 V.

The cross-bridged pyridine pendant arm complex Co(4)Cl²⁺ differs significantly from [Co(Me₂BCyclen)Cl₂] or Co(PyMeEBC)Cl⁺, as might be expected since it appears to be a diamagnetic, octahedral Co³⁺ complex, where the others are paramagnetic Co²⁺ complexes. Starting as Co³⁺, there is no observed oxidation for this complex, even of the irreversible variety of a bound chloro ligand. Reduction to Co²⁺, however is observed, and is quasi-reversible: E_1/2 = −0.127 V (ΔE = 107 mV). Apparently both 3+ and 2+ oxidation states are stable with the bound chloride and, presumably, the bound pyridine pendant arm. Unlike the other two complexes, no further reduction to Co⁺ is observed. Perhaps the bound pyridine and chloro ligands particularly stabilize Co²⁺ in this case, or provide too much electron density to allow for reduction to Co⁺ in the observed region (out to ~ −2.0 V). The difference in the Co^2+/3+ reversible couple for Co(4)Cl²⁺ and Co(PyMeEBC)Cl⁺ is instructive. The smaller cyclen-based ligand 4 clearly selects for the smaller Co³⁺ ion with an easily obtained oxidation to Co³⁺ at −0.127 V. This is likely the reason this complex oxidized in air to Co³⁺ during workup. In contrast, Co(PyMeEBC)Cl⁺ is much more difficult to oxidized (+0.657 V), likely because the larger cyclam-based ligand prefers the larger Co²⁺ ion.

A copper complex serving as the basis for comparison is [Cu(Me₂BCyclen)(CH₃CN)₂](PF₆)₂. [31] Interestingly, this complex in acetonitrile does not bind chloride, but prefers acetonitrile, as evidenced by a crystal structure grown from acetonitrile and molar conductance data. [31] The explanation for this behavior is the strained, distorted octahedral structure of this complex due to the small cavity of the bridged cyclen ligand. Cu(3)²⁺, like [Cu(Me₂BCyclen)(CH₃CN)₂](PF₆)₂ has no bound chloride, according to its elemental analysis. This is to be expected if the pyridine pendant arm is bound, as Cu²⁺ prefers 5-coordinate geometries. [48] Cu(3)²⁺ has an irreversible reduction to Cu⁺ at E_red = −0.406 V, which occurs about 250 mV less negative than the same reduction in [Cu(Me₂EBCyclen)(CH₃CN)₂](PF₆)₂. Perhaps the bound pendant arm helps to stabilize the large Cu⁺ ion, making reduction easier. Cu(3)²⁺ also has an irreversible oxidation at E_ox ~ +1.75 V (solvent oxidation prevents exact measurement of the peak), this oxidation is not seen in [Cu(Me₂BCyclen)(CH₃CN)₂](PF₆)₂. The explanation for the lack of observation of Cu³⁺ in the Me₂BCyclen complex is that strong bonds are required to stabilize the reactive Cu³⁺ ion, which are not possible because of the small cavity size and distorted geometry of the complex. [31] Perhaps the pyridine pendant arm provides enough additional bonding strength to allow the formation of Cu³⁺ in this case.

Similar behavior is seen in Cu(4)²⁺. One difference is that the reduction to Cu⁺ is now quasi-reversible with E_1/2 = −0.341 V (ΔE = 86 mV). The 5-coordinate ligand, including the pyridine pendant arm can obviously stabilize the Cu⁺ oxidation state, which is reached at a milder potential than either of the other two compared complexes. Finally, an oxidation to Cu³⁺ is also observed for Cu(4)²⁺ with E_ox ~ 1.86 V (again solvent oxidation prevents exact measurement of the peak). This is about 100mV more positive than for Cu(3)²⁺. Due to their expected structural similarities as distorted 5-coordinate complexes, there is very little difference in reduction potentials between all four analogous compounds (Cu(3)²⁺, Cu(4)²⁺, Cu(SB-PyCyclam)²⁺ and Cu(PyMeEBC)²⁺ being in the range −0.3 to −0.65 V.

3. Conclusion

Two novel topologically constrained cyclen chelators bearing pyridyl arms were described which were synthesized in a simple and scalable fashion. Access to unsymmetrically substituted tetraazamacrocycles is valuable and was achieved here by controlling the reactivity of the pyridyl alkyl halide by selection of an appropriate solvent. Transition metal complexes were formed with both ligands using Zn²⁺, Ni²⁺, Co²⁺ and Cu²⁺ which were characterized using UV-vis spectroscopy and magnetic moments to gain insights into their electronic structure along with X-ray crystal structure determination for selected cross-bridge complexes. The magnetic and electronic spectral studies revealed both five- and six-coordinate structures controlled by selection of the appropriate side- or cross-bridged ligand. The X-ray crystal structures were limited to the cross-bridged ligand complexes only, yet revealed interesting trends in geometric parameters based on metal ionic radius, macrocyclic ring size, and the effect of the additional pendant pyridine donor, especially in comparison to other known structural analogues. Electrochemical studies were carried out using cyclic voltammetry to determine oxidation and reduction potentials and give indications for their future potential as oxidation catalysis. Our current model relies on forming simple to work with divalent metal complexes (such as those used in this study) as a more rapid pre-screening of ligands which could be useful for oxidation catalysis. Indications from this study, particularly with respect to reversible access to multiple oxidation states stabilized by ligands 3 and 4, encourage us to subsequently form the Mn²⁺ and Fe²⁺ complexes which have shown oxidation catalytic promise previously and assess their potential for this application.

4. Experimental

Materials and methods

1,4,7,10-tetraazatetradodecane (Cyclen), 98%, was purchased from CheMatech. Glyoxal (40% wt in water), methyl iodide (99%), and sodium borohydride (98%) were purchased from Sigma Aldrich. Elemental analyses were performed by Quantitative Technologies Inc. Electrospray Mass spectra were collected on a Shimadzu LCMS 2020 Electrospray Mass Spectrometer. NMR spectra were obtained on a Varian Bruker AVANCE II 300 MHz NMR Spectrometer instrument. Electronic spectra were recorded using a Shimadzu UV-240 UV-Vis Spectrometer. Conductance measurements were obtained with an Oakton CON510 Bench Conductivity/TDS Meter on 0.001 M solutions at room temperature. Magnetic moments were obtained on finely ground solid samples at ambient temperatures using a Johnson Matthey MSB Auto magnetic susceptibility balance. Electrochemical experiments were performed on a BAS100B Electrochemical Analyzer. A button Pt electrode was used as the working electrode with a Pt-wire counter electrode and a Ag-wire pseudo-reference electrode. Scans were taken at 200 mV/s. Acetonitrile solutions of the complexes (1 mM) with tetrabutylammonium hexafluorophosphate (0.1 M) as a supporting electrolyte were used. The measured potentials were referenced to SHE using ferrocene (+0.400 V versus SHE) as an internal standard. All electrochemical measurements were carried out under N₂. The Co²⁺ complexes were analyzed using tetrabutylammonium chloride (0.1 M) as supporting electrolytes due to the simplification of the voltammograms in the latter. Partial replacement of chloro ligands by solvent in the former complicated the voltammograms. Synthesis of decahydro-2a,4a,6a,8a-tetraazacyclopenta[fg]acenaphthylene was carried out via literature methods. [11]

Crystal Structure Analysis

Single crystal X-ray diffraction data were collected in series of ω-scans using a Stoe IPSD2 image plate diffractometer utilizing monochromated Mo radiation (λ = 0.71073 Å). Standard procedures were employed for the integration and processing of the data using X-RED. [49] Samples were coated in a thin film of perfluoropolyether oil and mounted at the tip of a glass fibre located on a goniometer. Data were collected from crystals held at 150 K in an Oxford Cryosystems nitrogen gas cryostream.

Crystal structures were solved using routine automatic direct methods implemented within SHELXS-97. [50] Completion of structures was achieved by performing least squares refinement against all unique F2 values using SHELXL-97. [50] All non-H atoms were refined with anisotropic displacement parameters. Hydrogen atoms were placed using a riding model. Where the location of hydrogen atoms was obvious from difference Fourier maps, C-H bond lengths were refined subject to chemically sensible restraints.

Synthesis of 6a-(pyridin-2-ylmethyl)decahydro-5H-2a,4a,6a,8a-tetraazacyclopenta[fg]acenaphthylen-6a-ium chloride (1)

Sodium bicarbonate (8.7 g, 0.103 ml) was added to 2-picolyl chloride hydrochloride (8.5 g, 0.052 mol) in dichloromethane (500 mL) and stirred at room temperature for 1 h. The tan, solid precipitate was filtered and discarded. Decahydro-2a,4a,6a,8a-tetraazacyclopenta[fg]acenaphthylene (5.2 g, 0.027 mol) was added to the light yellow liquid filtrate along with potassium iodide (0.854 g, 0.00514 mol). The reaction was heated at reflux under nitrogen for 24 hours. The solution was allowed to cool to room temperature and filtered. The filtrate was also precipitated with ether (1.5 L) and more product was collected via filtration. Yield 6.818 g (79.43%) of 1. ¹H-NMR (300 MHz, D₂O): δ 2.62 (m, 2H), 2.89 (m, 3H), 3.03 (m, 1H), 3.23 (m, 2H), 3.39 (m, 3H), 3.57 (m, 1H), 3.72 (m, 2H), 3.85 (m, 2H), 4.17 (m, 1H), 4.45 (td, 1H), 4.87 (m, 1H), 5.06 (m, 1H), 7.68 (m, 1H), 7.81 (d, 1H), 8.09 (m, 1H), 8.76 (d, 1H). ¹³C{¹H} NMR (75.6 MHz, D₂O): δ 43.91, 47.64, 47.74, 48.20, 48.33, 51.32, 58.06, 61.81, 62.08, 83.11, 125.92, 128.21, 138.98, 147.23, 150.33. MS (ESI) 286.2 (M-Cl⁻)⁺. Anal calc. for C₁₅H₂₄N₅Cl•0.38CH₂Cl₂ calc: C 55.56%, H 7.05%, N 19.78%; Found: C 55.36%, H 6.95%, N 20.02%.

Synthesis of mono(2a-methyl-6a-(pyridin-2-ylmethyl)dodecahydro-2a,4a,6a,8a-tetraazacyclopenta[fg]acenaphthylene-2a,6a-diium) diiodide (2)

Methyl iodide (0.306 moles, 19 mL) was added to 6a-(pyridin-2-ylmethyl)decahydro-5H-2a,4a,6a,8a-tetraazacyclopenta[fg]acenaphthylen-6a-ium chloride (1, 6.56 g, 0.02 moles) in dry acetonitrile (250 mL) and stirred at room temperature for 3 days. After which, ether (500 mL) was added and the precipitate was filtered and washed with acetonitrile (200 mL) and ether (200 mL). Yield 9.41 g (84.70%) of 2. ¹H-NMR (300 MHz, D₂O): δ 2.49 (m, 1H), 2.88 (m, 1H), 3.00 (m, 2H), 3.34 (m, 8H), 3.81 (m, 7H), 4.32 (m, 1H), 4.68 (m, 1H), 4.84 (m, 1H), 5.08 (m, 1H), 7.56 (m, 1H), 7.76 (d, 1H), 8.02 (m, 1H), 8.73 (d, 1H). ¹³C{¹H} NMR (75.6 MHz, D₂O): δ 42.36, 42.74, 46.31, 46.50, 46.72, 56.43, 58.33, 60.38, 61.84, 64.62, 76.94, 77.19, 125.14, 127.33, 138.15, 148.68, 150.58. MS (ESI) 428.1 (M-I⁻)⁺. Anal calc. for C₁₇H₂₇N₅I₂ calc: C 36.77%, H 4.90%, N 12.61%; Found: C 36.79%, H 4.69%, N 12.69%.

Synthesis of 4-(pyridin-2-ylmethyl)-1,4,7,10-tetraazabicyclo[8.2.2]tetradecane (SB-PyCyclen, 3)

Sodium borohydride (1.30 g, 0.0345 mol) was added slowly to a solution of 1 (2.22 g, 0.0069 mol) in ethanol (100 mL) and the reaction was heated under reflux for 1 hour. After cooling to room temperature, 6 M hydrochloric acid was added drop wise until a pH of 1–2 was reached. The ethanol was removed via rotary evaporation. Water (50 mL) and 30% w/w potassium hydroxide solution was added until a pH of 14 was reached. The product was extracted into benzene, dried (NaSO₄) and reduced under vacuum and subsequently purified via short-path distillation. Yield 1.89 g (94.49%) of 3. ¹H-NMR (300 MHz, CDCl₃): δ 2.42 (m, 4H), 2.59 (m, 2H), 2.72 (m, 5H), 2.90 (m, 4H), 3.11 (m, 4H), 3.23 (m, 2H), 3.83 (m, 2H), 7.15 (m, 1H), 7.56 (d, 1H), 7.67 (m, 1H), 8.51 (d, 1H). ¹³C{¹H} NMR (75.6 MHz, CDCl₃): δ 45.72, 47.38, 48.88, 48.92, 50.77, 52.53, 52.74, 52.80, 56.88, 57.72, 61.76, 121.91, 122.48, 128.42, 136.76, 148.87. MS (ESI) 290.1 (M+H⁺)⁺. Anal calc. for C₁₆H₂₇N₅: C 66.40%, H 9.40%, N 24.20%; Found: C 66.15%, H 9.61%, N 24.20%.

Synthesis of 4-methyl-10-(pyridin-2-ylmethyl)-1,4,7,10-tetraazabicyclo[5.5.2]tetradecane (CB-MePyCyclen, 4)

Sodium borohydride (2.43 g, 0.064 mol) was added slowly to a solution of 2 (2.35 g, 0.0042 mol) in 95% ethanol (180 mL) and the reaction was stirred for 5 days under nitrogen. 6 M hydrochloric acid was added drop wise until a pH of 1–2 was reached. The ethanol was removed via rotary evaporation. 30% w/w potassium hydroxide solution was added until a pH of 14 was reached. The product was extracted into benzene, dried (NaSO₄) and reduced under vacuum. Yield 0.96 g (68.3%) of 4. ¹H-NMR (300 MHz, CDCl₃): δ 2.43 (m, 3H), 2.70 (m, 16H), 2.97 (m, 3H), 3.82 (s, 2H), 7.27 (td, 1H), 7.41 (d, 1H), 7.77 (td, 1H), 8.34 (d, 1H). ¹³C{¹H} NMR (75.6 MHz, CDCl₃): δ 42.82, 47.04, 51.78, 52.39, 54.38, 55.07, 59.72, 123.10, 123.94, 138.02, 148.16, 158.16. MS (ESI) 304.2 (M+H⁺)⁺. C₁₇H₂₉N₅•0.35H₂O: C 65.92%, H 9.66%, N 22.61%; Found: C 66.04%, H 9.61%, N 22.22%.

General method for synthesis of transition metal complexes (with below exceptions)

In a glove-box 3 (1 mmol, 0.289 g) or 4 (1 mmol, 0.303 g) was dissolved in 10 mL of acetonitrile. 1 mmol of zinc chloride, nickel chloride, copper chloride, and cobalt chloride were added in separate vials containing the ligand. These reactions were left to stir for 2 days at room temperature. The solutions were gravity filtered to remove any un-reacted metal salt. The reaction vials were removed from the glove-box and the acetonitrile removed on a rotary evaporator. The oily residues were dissolved in a minimum of methanol. A solution ammonium hexafluorophosphate (5 mmol, 0.815 g) in methanol was added to each reaction. The metal complex solutions were placed in the freezer to help promote precipitation. All solid precipitates were filtered on glass frits, washed with cold methanol, then ether, and dried prior to characterization. NiCl₂ is not completely soluble in acetonitrile so these reactions were carried out with additional DMF (15 mL) and left to react for an additional 2 days.

The complexation of 3 with CuCl₂ was stirred for only 15 minutes. Leaving the reaction longer resulted in a reduced copper metal precipitation. From this reaction, and other unpublished results, [37] side-bridged macrocycle ligand complexation reactions in acetonitrile with CuCl₂ are susceptible to some kind of uncharacterized redox process whereby the complex is destroyed, and this is accompanied by the production of metallic copper. The secondary NH of the side-bridged ligand is the main difference from cross-bridged, all-tertiary ligands, where this process does not occur. The process can be limited or avoided with SB ligands by using only short reactions times. We have not investigated the specific deleterious reaction, so can only speculate on its mechanism. Since we have overcome this problem with short reaction time, we have not chosen to focus on this mechanism.

The zinc complex of 3 gave a low yield upon precipitation with NH₄PF₆, apparently because the complex is much more soluble in MeOH than its analogues with other metal ions. Since a sufficient amount of product was obtained for characterization and was analytically pure, we have not optimized this reaction.”

[Zn3](PF₆)₂

Yield: 0.025 g (4%) of a beige powder. MS (ESI) 388 (M-2PF₆⁻+Cl⁻)⁺. Anal calc. for [Zn(C₁₆H₂₇N₅)](PF₆)₂: C 29.81%, H 4.22%, N 10.86%; Found: C 30.07%, H 4.10%, N 10.80%.

[Ni3](PF₆)₂

Yield: 0.337 g (53%) of a rust orange powder. MS (ESI) 346 (M-2PF₆⁻)⁺. Anal calc. for [Ni(C₁₆H₂₇N₅)](PF₆)₂ • 0.2 (NH₄PF₆): C 28.66%, H 4.18%, N 10.86%; Found: C 28.43%, H 3.79%, N 10.75%.

[Co3Cl](PF₆)

Yield: 0.212 g (40%) of a brick red powder. MS (ESI) 383 (M-2PF₆⁻+Cl⁻)⁺. Anal calc. for [Co(C₁₆H₂₇N₅)Cl]PF₆•0.5H₂O: C 35.74%, H 5.25%, N 13.02%; Found: C 35.56%, H 5.10%, N 12.99%.

[Cu3](PF₆)₂

Yield: 0.625 g (97%) of a violet powder. MS (ESI) 387 (M-2PF₆⁻+Cl⁻)⁺; 351 (M-2PF₆⁻)⁺. Anal calc. for [Cu(C₁₆H₂₇N₅)](Cl_0.1)(PF₆)_1.9: C 30.41%, H 4.31%, N 11.08%; Found: C 30.37%, H 4.07%, N 10.70%.

[Zn4Cl](PF₆)

Yield: 0.540 g (98%) of a beige powder. MS (ESI) 402 (M-2PF₆⁻+Cl⁻)+. Anal calc. for [Zn(C₁₇H₂₉N₅)Cl]PF₆: C 37.18%, H 5.32%, N 12.75%; Found: C 36.98%, H 5.20%, N 12.41%.

Ni4(PF₆)₂

Yield: 0.429 g (79%) of a purple powder. MS (ESI) 396 (M-2PF₆⁻+Cl⁻)⁺. Anal calc. for [Ni(C₁₇H₂₉N₅)Cl]PF₆•0.5H₂O: C 37.02%, H 5.48%, N 12.70%; Found: C 36.85%, H 5.27%, N 12.36%.

[Co4Cl](PF₆)₂

Yield: 0.532 g (77%) of a pink powder. MS (ESI) 396 (M-2PF₆⁻+Cl⁻)+. Anal calc. for [Co(C₁₇H₂₉N₅)Cl](PF₆)₂: C 29.69%, H 4.25%, N 10.18%; Found: C 29.59%, H 4.23%, N 10.04%.

[Cu4](PF₆)₂

Yield: 0.442 g (67%) of a blue powder. MS (ESI) 401 (M-2PF₆⁻+Cl⁻)+. Anal calc. for [Cu(C₁₇H₂₉N₅)](PF₆)₂: C 31.08%, H 4.45%, N 10.66%; Found: C 30.73%, H 4.33%, N 10.52%.

Figure 3.

Cyclic voltammograms of Ni(4)Cl⁺ and Ni(3)²⁺.

Figure 4.

Cyclic voltammograms of Co(4)Cl²⁺ and Co(3)Cl²⁺.

Figure 5.

Cyclic voltammograms of Cu(4)²⁺ and Cu(3)²⁺.

Table 2.

Crystallographic data. L = PyMeEBC

	[Ni(4)Cl]PF6 sja4_12	[Cu(4)](PF6)2 sja26_12	[Zn(4)Cl]PF6 sja3_12
Chemical Formula	C17 H29 Cl F6 N5 Ni P	C17 H29 Cu F12 N5 P2	C17 H29 Cl F6 N5 P Zn
a = …(esd) Å	11.6458(10)	11.4577(9)	11.584(2)
b = …(esd) Å	11.6930(12)	12.2495(14)	11.729(3)
c = …(esd) Å	15.6867(13)	17.1539(14)	15.730(3)
α = …(esd) degrees	90	90	90
β = …(esd) degrees	97.680(7)	95.837(6)	96.95(2)
γ = …(esd) degrees	90	90	90
V = … Å3	2117.0(3)	2395.1(4)	2121.6(9)
Z =	4	4	4
Formula Weight	542.58	656.93	549.24
Space Group	P21/c	P21/c	P21/c
T = … K	150(2)	150(2)	150(2)
λ = … Å	0.71073	0.71073	0.71073
Dcalcd = … g cm−3	1.702	1.822	1.720
μ = … mm−1	1.186	1.157	1.427
R1(Fo2) =	0.1144	0.1125	0.0598
wR2 (Fo2) =	0.1980	0.1389	0.0660

Table 3.

Selected Bond Lengths [Å] and angles [°].

[Ni(4)Cl]PF6		[Cu(4)](PF6)2		[Zn(4)Cl]PF6
N(1)-Ni(1)	2.117(4)	N(1)-Cu(1)	1.998(6)	N(1)-Zn(1)	2.255(3)
N(2)-Ni(1)	2.057(4)	N(2)-Cu(1)	2.042(5)	N(2)-Zn(1)	2.185(2)
N(3)-Ni(1)	2.153(5)	N(3)-Cu(1)	2.014(6)	N(3)-Zn(1)	2.202(3)
N(4)-Ni(1)	2.077(5)	N(4)-Cu(1)	2.165(7)	N(4)-Zn(1)	2.142(3)
N(5)-Ni(1)	2.085(5)	N(5)-Cu(1)	2.012(5)	N(5)-Zn(1)	2.118(3)
Cl(1)-Ni(1)	2.4302(16)	N(1)-Cu(1)-N(5)	83.9(2)	Cl(1)-Zn(1)	2.3870(8)
N(2)-Ni(1)-N(4)	84.4(2)	N(1)-Cu(1)-N(3)	170.6(2)	N(5)-Zn(1)-N(4)	156.13(11)
N(2)-Ni(1)-N(5)	162.85(18)	N(5)-Cu(1)-N(3)	105.5(2)	N(5)-Zn(1)-N(2)	94.28(11)
N(4)-Ni(1)-N(5)	94.1(2)	N(1)-Cu(1)-N(2)	84.4(2)	N(4)-Zn(1)-N(2)	81.26(11)
N(2)-Ni(1)-N(1)	81.32(16)	N(5)-Cu(1)-N(2)	164.9(3)	N(5)-Zn(1)-N(3)	118.85(10)
N(4)-Ni(1)-N(1)	84.46(19)	N(3)-Cu(1)-N(2)	86.5(2)	N(4)-Zn(1)-N(3)	83.49(11)
N(5)-Ni(1)-N(1)	81.54(17)	N(1)-Cu(1)-N(4)	88.2(2)	N(2)-Zn(1)-N(3)	78.88(11)
N(2)-Ni(1)-N(3)	84.06(18)	N(5)-Cu(1)-N(4)	106.0(2)	N(5)-Zn(1)-N(1)	78.58(11)
N(4)-Ni(1)-N(3)	81.8(2)	N(3)-Cu(1)-N(4)	88.5(2)	N(4)-Zn(1)-N(1)	77.55(9)
N(5)-Ni(1)-N(3)	112.67(18)	N(2)-Cu(1)-N(4)	83.0(2)	N(2)-Zn(1)-N(1)	80.34(10)
N(1)-Ni(1)-N(3)	160.8(2)			N(3)-Zn(1)-N(1)	153.69(10)
N(2)-Ni(1)-Cl(1)	96.69(15)			N(5)-Zn(1)-Cl(1)	87.75(8)
N(4)-Ni(1)-Cl(1)	176.36(13)			N(4)-Zn(1)-Cl(1)	98.00(8)
N(5)-Ni(1)-Cl(1)	85.95(15)			N(2)-Zn(1)-Cl(1)	176.51(9)
N(1)-Ni(1)-Cl(1)	99.13(14)			N(3)-Zn(1)-Cl(1)	97.66(8)
N(3)-Ni(1)-Cl(1)	94.86(17)			N(1)-Zn(1)-Cl(1)	102.86(8)

Table 5.

Redox potentials (vs SHE) with peak separations for complexes of 3, 4, Me₂BCyclen, SB-PyCyclen, and PyMeEBC.

Complex	Redox Process	Potential (V)	Peak Separation (mV)	Ref
Ni(Me2Bcyclen)Cl2	Ni2+ → Ni+	Ered = −2.036	---	[42]
	Ni2+ / Ni3+	E1/2 = +0.863	68
	????	Eox = +1.450	---
Ni(3)2+	Ni2+ → Ni+	Ered = −1.287	---
	Ni2+ → Ni3+	Eox = +1.375	---
Ni(4)Cl+	Ni2+ / Ni3+	E1/2 = +1.034	76
	????	E1/2 = +1.313	87
Ni(PyMeEBC)Cl+	Ni2+ → Ni+	Ered = −1.026	---	[12]
	Ni2+ → Ni3+	Eox = +1.290	---
Co(Me2BCyclen)Cl2	Co2+ → Co+	Ered = −2.202	---	[47]
	Co2+ / Co3+	E1/2 = −0.157	288
	????	Eox = +0.983	---
Co(3)Cl+	Co2+ → Co+	Ered = −1.884	---
	Co2+ → Co3+	Eox = +1.121	---
	????	Eox = +1.411	---
Co(4)Cl2+	Co3+ → Co2+	E1/2 = −0.127	107
Co(PyMeEBC)Cl+	Co2+ / Co+	E1/2 = −1.386	143	[12]
	Co2+ / Co3+	E1/2 = +0.657	154
Cu(Me2BCyclen)2+	Cu2+ → Cu+	Ered = −0.651	---	[31]
Cu(3)2+	Cu2+ → Cu+	Ered = −0.406	---
	Cu2+ → Cu3+	Eox ~ +1.75	---
Cu(4)2+	Cu2+ / Cu+	E1/2 = −0.341	86
	Cu2+ → Cu3+	Eox ~ +1.86	---
Cu(SB-PyCyclam)2+	Cu2+ / Cu+	E1/2 = −0.586	77	[10]
Cu(PyMeEBC)2+	Cu2+ / Cu+	E1/2 = −0.402	80	[12]

Appendix A: Supplementary data

CCDC 1428920, 1428921, and 1428922 contain the supplementary crystallographic data for [Zn(4)Cl]PF₆, [Ni(4)Cl]PF₆, and [Cu(4)](PF₆)₂, respectively. These data can be obtained free of charge via http://www.ccdc.cam.ac.uk/conts/retrieving.html, or from the Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: (+44) 1223-336-033; or e-mail: deposit@ccdc.cam.ac.uk.