Hydrogen Peroxide Disproportionation Activity Is Sensitive to Pyridine Substitutions on Manganese Catalysts Derived from 12-Membered Tetra-Aza Macrocyclic Ligands

Abstract

The abundance of manganese in nature and versatility to access different oxidation states have made manganese complexes attractive as catalysts for oxidation reactions in both biology and industry. Macrocyclic ligands offer the advantage of substantially controlling the reactivity of the manganese center through electronic tuning and steric constraint. Inspired by the manganese catalase enzyme, a biological catalyst for the disproportionation of H₂O₂ into water and O₂, the work herein employs 12-membered tetra-aza macrocyclic ligands to study how the inclusion of and substitution to the pyridine ring on the macrocyclic ligand scaffold impacts the reactivity of the manganese complex as a H₂O₂ disproportionation catalyst. Synthesis and isolation of the manganese complexes was validated by characterization using UV–vis spectroscopy, SC-XRD, and cyclic voltammetry. Potentiometric titrations were used to study the ligand basicity as well as the thermodynamic equilibrium with Mn(II). Manganese complexes were also produced in situ and characterized using electrochemistry for comparison to the isolated species. Results from these studies and H₂O₂ reactivity showed a remarkable difference among the ligands studied, revealing instead a distinction in the reactivity regarding the number of pyridine rings within the scaffold. Moreover, electron-donating groups on the 4-position of the pyridine ring enhanced the reactivity of the manganese center for H₂O₂ disproportionation, demonstrating a handle for control of oxidation reactions using the pyridinophane macrocycle.

Graphical Abstract

INTRODUCTION

Oxidation catalysis with mild oxidants, such as H₂O₂, has been desired due to the possibility of performing important oxidation steps in industry at a lower cost and less impact to the environment. The oxidation reactivity of H₂O₂ can be hindered by a mutually exclusive reaction known as disproportionation, where H₂O₂ is both oxidized and reduced to produce molecular oxygen and water. Such reactivity is common in nature, which has developed the superior catalytic tools to appropriately mitigate the reactivity of H₂O₂ and yield unrivaled turnover frequencies and reaction rates. For comparison, peroxidases use H₂O₂ to catalyze oxidation of organic matter, while catalases disproportionate H₂O₂. Consistent with other metalloenzymes, both peroxidase and catalase contain a metal center in their structure, and their reactivity is modulated by a dynamic environment composed of amino acid residues capable of fine-tuned regulation of the coordination sphere using electronics. As a result, the enzymes can achieve a range of chemical potentials that allow the performance of their target reactions with high catalytic efficiency and selectivity, making the isolation and study of the active site chemistry an attractive challenge.¹

The manganese-based catalase (MnCAT) is one of the most active enzymes in nature, performing disproportionation of H2O2 with a rate of 1.8–8.4 × 10⁶ M⁻¹ min⁻¹ and turnover frequencies in the range of 10⁶ s⁻¹. ^2,3 These remarkable values exemplify the importance of MnCAT for efficient disproportionation of H₂O₂ in nature, which has been a challenge to transfer to the bench. This is, in part, due to the unique redox activity of the binuclear active site (Figure 1) that can cycle between Mn(II)/Mn(II) and Mn(III)/Mn(III) redox states during catalytic turnovers. Moreover, the oxidized and reduced forms of the enzymes differ from one another by the bridging ligands connecting the manganese centers. The reduced form of the enzyme features an active site containing two Mn(II) ions connected by the carboxylate group of a glutamate residue in a syn–syn fashion.⁴ The motif is completed by two bridging μ-OH and μ-OH₂ moieties. In contrast, the oxidized form features two μ-O bridged ligands along with the carboxylate bridging unit.⁴

Figure 1.

Schematic representation of the active site of MnCAT from Lactobacillus plantarum, highlighting the (left) reduced form and (right) oxidized form, which differed by the bridging ligands based on ref 4.

Considerable efforts have been made to develop synthetic mimics of MnCAT for applications in oxidation catalysis and as pharmaceutical agents to regulate excess H₂O₂ and other reactive oxygen species (ROS).⁵ Nevertheless, mimetics are far from reaching analogous activities compared to native enzymes, largely because of the inability to perform the same reactivity using low Mn(II)/Mn(III) oxidation states. Instead, mimetics must reach high oxidation states such as Mn(IV) and Mn(V), which hinder the kinetic efficiency due to the energy required to achieve these highly reactive oxidation states.⁶ New approaches and ligand scaffolds are continuously being developed to improve on the oxidation-type reactions through the high-valent pathways. Catalase-like reactivity is a relatively simple reaction that can be used as a probe to learn the behavior of manganese-based small molecules in oxidation catalysis using H₂O₂ as the main oxidant to classify manganese complexes of diverse structure based on their performance.

The ligand scaffold is important to consider in attempts to stabilize high oxidation states of manganese. Macrocycles stand out due to their high thermodynamic stability of complex formation with catalytically active transition metals, compared to open chain ligands. Moreover, the diversity within the library of macrocyclic ligand frameworks offers the opportunity to adjust electronic effects, cavity size, and basicity to subsequently tune the properties of the metal complex to yield a preferred structure for a target reactivity. For example, inclusion of pyridine rings in pyclen-based pyridinophanes (PyN₃) has shown important results in stabilizing high oxidation states of iron in the literature, as opposed to cyclen (N₄) and cyclam scaffolds. Based on these results, pyridinophanes have been used in oxidation catalysis with manganese.⁷ Previous studies have shown that manganese complexes of PyN₃ were more active than N₄ for the disproportionation of H₂O₂. However, the catalytic efficiencies or any mechanistic insights of the manganese complexes of PyN₃ and N₄ toward H₂O₂ disproportionation have not been described. Moreover, spectroscopic evidence from the preliminary studies suggested the presence of bis-μ-oxo dimers for the ligands N₄ and PyN₃, but identification of such a species with the addition of the second pyridine ring in the Py₂N₂ scaffold of the catalyst remains unclear.⁸

Inspired by the ability of 12-membered pyridinophane macrocycles to modulate the reactivity of manganese and knowledge gathered from DFT calculations on MnCAT mechanisms,^9–11 we hypothesized that manganese complexes of 12-membered tetra-aza macrocyclic ligands would be more efficient for catalytic disproportionation of H₂O₂ if they possess (1) electron-donating groups (EDGs) that facilitate higher oxidation states on the metal center and (2) pyridine moieties that can assist delocalization of the charge to modulate electron density through different oxidation states. With this goal, we prepared a group of ligands that ranged from cyclen (N₄), pyclen (PyN₃) and its pyridine substituted congeners (^OHPyN₃ and ^OMePyN₃) as well as Py₂N₂ (Figure 2), and their manganese complexes to systematically understand what properties were modulated by the differences between each ligand set and how these results related to metrics of H₂O₂ disproportionation.

Figure 2.

Structures of the ligands organized by the variables considered in this study.

RESULTS AND DISCUSSION

Pyridine Modifications on the Ligand Scaffold Tune Their Basicity.

The ligand series studied (Figure 2) was selected to systematically modify 12-membered tetra-aza pyridinophanes by changing the number of pyridine rings on the macrocycle or attaching EDGs to the pyridine. Apart from cyclen (N₄), which was obtained commercially, the ligands were prepared following previously published methods.^12–15 The PyN₃ series were isolated as HCl salts, and the Py₂N₂ ligand as a free base. All ligands are highly water-soluble, which aided greatly in the studies described herein.

Inclusion of pyridine rings within the macrocycle is known to increase the rigidity of the macrocyclic ligand and subsequently a decrease in basicity, which can be quantified by the pH of the protonation events of each ligand.^16,17 Therefore, stepwise protonation constants of the ligand series were obtained by pH-potentiometric titrations (25 °C and 0.15 M NaCl). The pK_a ($\text{log}{K}_i^{\mathrm{H}}$) values for each measurable protonation event within the N₄, PyN₃, and Py₂N₂ series (Table 1) were consistent with the decrease of basicity from secondary amines to pyridine rings.¹⁸ Based on the sum of the pK_a values determined for the N atom donors of each ligand (Σ pK_a,N-donors), basicity decreased from N₄ (21.35) > PyN₃ (21.20) ≫ Py₂N₂ (15.63). This trend shows that substitution of two secondary amines for a pyridine nitrogen has a much bigger impact on the overall basicity of the ligand than just a single substitution, which only decreased the basicity slightly. On the other hand, EDGs on the 4-position of the pyridine ring decreased the total basicity from ^OHPyN₃ (21.97) > PyN₃ (21.20) ≫ ^OMePyN₃ (20.07). Substitutions of EDG on the 4-position of the pyridine ring of the PyN₃ ligand fine-tune the overall basicity of the ligand over a range of roughly two log units (Σ pK_a,N-donors = 20.07–21.97), while substitution of secondary amines with pyridine units results in much more broad range changes (Σ pK_a,N-donors = 15.63–21.35).

Table 1.

Stepwise Protonation Constants (pK_a) of 12-Membered Pyridinophane Macrocycles (T = 298 K, I = 0.15 M NaCl)

	N4a	Py2N2	PyN3b	OHPyN3b	OMePyN3c
pKa 1	10.81(1)	8.27(4)	11.37(1)	11.38(1)	10.32(2)
pKa 2	9.66(2)	7.36(4)	8.22(5)	8.85(3)	8.00(3)
pKa 3	0.88(2)		1.61(5)	5.59(4)	1.75(4)
pKa 4				1.74(4)
Σ pKa	21.35	15.63	21.20	27.56	20.07
Σ pKa,N-donors	21.35	15.63	21.20	21.97	20.07

^aFrom ref 8.

^bFrom ref 18.

^cFrom ref 19.

Ligand Basicity Modulate Manganese Complexation and Hydrolysis.

Speciation studies of each ligand with Mn(II) were essential, given that complex [(Mn(L)] formation often occurs at pH values that overlap with formation of intractable Mn(OH)_x species, which complicates studies of these complexes in solution.²⁰ Therefore, thermodynamic stability of each ligand with Mn(II) ions was obtained by pH-potentiometric titrations (25 °C and 0.15 M NaCl) to determine the optimal pH for Mn(II) complex formation and to identify the species present in solution at different pH values. This approach significantly aids in the goal of preparing each manganese complex in situ. pH-potentiometric titrations were performed with Mn(II) ions under a N₂ atmosphere since Mn(III) species are known to both easily hydroxylate and disproportionate into Mn(II) and Mn(IV) species, concomitant with formation of MnO_x species.²¹

Thermodynamic stability (log K_ML) of the Mn(II) complexes of the ligand series (Table 2) shows that the PyN₃ scaffold forms the most stable complex regardless of the pyridine substitution on the 4-position, following the basicity trend observed for the protonation steps of the ligand alone: ^OHPyN₃ (10.96) > PyN₃ (10.11) > ^OMePyN₃ (9.70). The effect of the number of pyridine rings within the ligand scaffold on the log K_ML did not follow the trend observed with ligand basicity. Rather, it decreased from PyN₃ (10.11) > N₄ (8.34) ≈ Py₂N₂ (8.26). Since the ligands studied here have the same number of atoms in the cavity size, the difference in stability observed with changing the number of pyridine rings suggests that this observation is a consequence of variability in steric energy required for conformational reorganization of the macrocyclic cavity to complex Mn(II).^22,23

Table 2.

Stability Constants, Protonation Constants, and pMn Values for the Mn(II) Complexes (I = 0.15 M NaCl, T = 25°C)

ligand	logKML	logKMHL	logKM(OH)L	pMn
N4	8.34(2)a		10.22(6)a	5.01
Py2N2	8.26(9)		9.7(2)	5.96
PyN3	10.11(4)a		11.0(5)a	5.25
OHPyN3	10.96(4)	6.40(7)	10.18(6)	5.39
OMePyN3	9.70(9)		9.5(2)	5.63

^aFrom ref 8.

To determine the effect of the ligand basicity on the speciation of Mn(II) ions in solution at different pH values, the pMn (a conditional stability constant) was calculated for each Mn(II) complex. pMn was calculated by pMn = −log[Mn(II)]_free from speciation curves at a concentration of [L] = [M] = 0.01 mM at pH = 7.4.^16,17,24 Higher pMn values reflect the ability of a given ligand to complex with Mn(II) ions at more acidic pH values. The calculated pMn values for complexes with different number of pyridine rings (Table 2) correlated with the basicity of the ligands: the more basic the ligand, the higher the competition of the metal versus protons for the donor sites, pMn for Py₂N₂ (5.96) > PyN₃ (5.25) > N₄ (5.01). Substitution of the 4-position of the pyridine ring in the PyN₃ systems resulted in a slight increase of the pMn values from PyN₃ (5.25) > ^OHPyN₃ (5.39) > ^OMePyN₃ (5.63). This is consistent with the trend observed for log K_ML, as well. The lowest basicity within the PyN₃ ligands was obtained for ^OMePyN₃, which indicates that there is less competition of Mn(II) ions for donor sites, resulting in the highest pMn among the PyN₃ series. Overall, the effect of the ligand basicity on Mn(II) complexation can be visualized in Figure 3, where speciation of free Mn(II) begins to decrease at around pH = 5.5 for Py₂N₂. In contrast, this transition occurs around pH = 6.5 for ^OMePyN₃ and ^OHPyN₃. PyN₃ and N₄ exhibit complexation of Mn(II) at higher pH values (PyN₃ pH = 7, N₄ pH = 8), revealing the lowest pMn in this study.

Figure 3.

Speciation of free Mn(II) ions with 12-membered tetra-aza macrocycles (I = 0.15 M NaCl, T = 25 °C, [L] = [M] = 0.01 mM).

Manganese Complexes Were Isolated as Mononuclear and Dinuclear Species with Distinctive Spectroscopic and Crystallographic Features.

Complexation of manganese was pursued with methods driven by understanding the optimal aqueous formation conditions. The pH for complex formation was a key choice to consider when working with manganese. Speciation diagrams derived from potentiometric titrations can help to circumvent trial and error. As shown in Figures S1–S5, maximum [M(L)] complex formation for N4, PyN₃, and ^OHPyN₃ is achieved at pH = 8, while the population of the corresponding hydroxyl complex [M(OH)L] at pH = 8 was higher for ^OMePyN₃ and Py₂N₂, compared to N₄, PyN₃, and ^OHPyN₃. As predicted from this data, monomeric species were obtained by mixing an aqueous solution of ^OMePyN₃ or Py₂N₂ at pH = 7 with equimolar amounts of the precursor Mn(II) salt without further adjustment of pH, while the pH was set to 8 for ^OHPyN₃. The solution showed slow darkening due to oxidation of Mn(II) to Mn(III) under air atmosphere. Water was evaporated under reduced pressure and products were isolated in MeCN as pink powders. The color of these compounds is characteristic of a mononuclear [MnL]³⁺ manganese complex with ^OHPyN₃, ^OMePyN₃, or Py₂N₂ using previously reported procedures with some variations based on the results of potentiometric titrations.⁸ Dark forest green powder characteristic of dinuclear [Mn₂(μ-O)₂L₂]³⁺ complexes was obtained with ^OHPyN₃ using a reported procedure without modifications, but in the case of ^OMePyN₃ and Py₂N₂, only monomers were obtained.⁸

Monomeric complexes [Mn(^OHPyN₃)Cl₂][ClO₄], [Mn(^OMePyN₃)Cl₂][Cl], and [Mn(Py₂N₂)Cl₂][Cl] were dissolved in MeCN, and each solution exhibited a pink color characteristic of cis-Mn(III) mononuclear complexes in the Literature^25–27. Ligand to metal charge transfer (LMCT) bands were observed from 350 to 400 nm along with a d–d transition band centered at 525 nm (Figure 4, top). The dimer complex [Mn₂(μ-O)₂(^OHPyN₃)₂][ClO₄]₃ was dissolved in water, exhibiting its characteristic dark forest green color. Spectroscopic features matched the bands reported for other bis-μ-oxo complexes of manganese, LMCT absorptions in the region from 350 to 450 nm, the characteristic sharp band at 550 nm corresponding to a Mn(III) d–d transition, broad absorbance centered at 650 nm (LMCT oxo to Mn(IV)), and an intervalence band at low energy (~750 nm) (Figure 4, bottom).^28,29

Figure 4.

Electronic absorption spectrum at 25 °C of (top) monomeric Mn(III) complexes of ^OHPyN₃, ^OMePyN₃, and Py₂N₂ in MeCN, and (bottom) dimeric Mn₂(III,IV) complex of ^OHPyN₃ in water.

Single crystal X-ray diffraction (SC-XRD) analysis was carried out on the crystalline materials obtained from metalation of the ligand series. Mononuclear Mn(III) species were modeled from manganese complexation with ligands ^OMePyN₃ and Py₂N₂, [Mn(^OMePyN₃)F₂][BF₄], and [Mn(Py₂N₂)Cl₂][BF₄] (Figure 5). Throughout the series, Mn(III) is bound in a cis conformation, with two nitrogen atoms in the axial plane (N1 and N3) and two nitrogen atoms in the equatorial plane (N2 and N4). The distorted octahedral coordination is completed by two Cl⁻ ions with Py₂N₂ and two F⁻ ions for ^OMePyN₃. The presence of fluoride ligands was not originally anticipated; however, it is consistent with reports of metal complexes with highly nucleophilic metal centers or by the presence of HF that have produced F⁻ after slow hydrolysis of salts of BF₄ over time.^30–32

Figure 5.

ORTEP representations of solid-state structures of (A) [Mn₂(μ-O)₂(^OHPyN₃)₂][ClO₄]₃, (B) [Mn(^OMePyN₃)F₂][BF₄], and (C) [Mn(Py₂N₂)Cl₂][BF₄]. Counterions and solvents have been removed here for clarity. Atoms are drawn at 50% thermal probability ellipsoids.

Mononuclear Complexes.

Bond lengths and angles were compared with the previously reported [Mn(PyN₃)Cl₂][ClO₄] complex (Table 3).⁸ Bond lengths Mn–N1 and Mn–N3 depict elongation of ~0.2 Å compared to the N2 and N4 bonds, typically observed for d⁴ Mn(III) complexes due to Jahn–Teller distortion. The N1–Mn–N3 axial angle was similar for [Mn(PyN₃)Cl₂][ClO₄] and [Mn(^OMePyN₃)F₂][BF₄] (149.22° and 149.08°, respectively), but [Mn(Py₂N₂)Cl₂][BF₄] shows a smaller angle (145.53°), consistent with the increase of the rigidity by the addition of the second pyridine ring. The equatorial angle N2–Mn–N4 shows modest changes among the complexes with similar trends to the axial angle.

Table 3.

Selected Bond Lengths (Å) and Angles (°) from SC-XRD Analysis of Monomeric Mn(lIl) Complexes

bond length (Å)/angle (°)	PyN3a	OMePyN3b	Py2N2c	Py2MeN2d
Mn–N1	2.247(2)	2.298(1)	2.255(2)	2.328(2)
Mn–N2	2.076(2)	2.078(1)	2.041(2)	2.213(2)
Mn–N3	2.241(3)	2.272(1)	2.255(2)	2.327(2)
Mn–N4	2.059(2)	2.032(1)	2.041(2)	2.198(2)
N1–Mn–N3	149.22(9)	149.08(5)	145.53(12)	143.44(6)
N2–Mn–N4	87.33(10)	86.62(5)	78.01(8)	77.61(6)

^a[Mn(PyN₃)Cl₂][ClO₄], Freire et al.⁸

^b[Mn(^OMePyN₃)F₂][BF₄], this work.

^c[Mn(Py₂N₂)Cl₂][BF₄], this work.

^d[Mn(Py₂^MeN₂)(MeCN)₂][PF₆]₂, Lee et al.³⁶

The bond distance between Mn(III) and the N atom of the pyridine ring can be used to assess the strength of their bonding. Functionalization of the pyridine ring has been shown to modulate the bond length between the pyridine and metal centers such as Fe(III), Ni(II), and Cu(II).^33–35 [Mn(PyN₃)Cl₂][ClO₄] has a longer Mn–N4 bond (2.059 Å) than [Mn(^OMePyN₃)F₂][BF₄] (2.032 Å), which is consistent with the trend observed in the literature for other metal centers with EDGs substituted on the pyridine ring.^33–35 Interestingly, [Mn(Py₂N₂)Cl₂][BF₄] has a bond distance Mn–N4 intermediate (2.041 Å) between [Mn(PyN₃)Cl₂][ClO₄] and [Mn(^OMePyN₃)F₂][BF₄], which could be attributed to the change of the ligand scaffold.

Bond distances for Mn–N1 and Mn–N3 can be used as a readout for the effect of the basicity of the N-donors of the amines as well. For example, [Mn(Py₂N₂)Cl₂][BF₄] provides shorter bond distances (Mn–N1 = Mn–N3 = 2.041 Å) than the previously reported complex [Mn(Py₂^MeN₂)(MeCN)₂][PF₆]₂ (Mn–N1 = Mn–N3 = 2.327 Å),³⁶ consistent with the higher basicity of secondary amines (Py₂N₂) in comparison with tertiary amines (Py₂^MeN₂); nevertheless, the difference in oxidation state of manganese, being Mn(II) for Py₂^MeN₂ and Mn(III) for Py₂N₂, does not allow a clear comparison of ligand basicity due to the differences in acidity of Mn(II) vs Mn(III). Substitutions of the PyN₃ scaffold provide a much better comparison to study the impact of changes to the macrocyclic ligand on the solid-state structure. The structure modeled for [Mn(PyN₃)Cl₂][ClO₄] shows shorter bond distances for Mn–N1 and Mn–N3 (Mn–N1 = 2.247 Å, Mn–N3 = 2.241 Å) compared to [Mn(^OMePyN₃)F₂][BF₄] (Mn–N1 = 2.298 Å, Mn–N3 = 2.272 Å), which is in agreement with the higher basicity of the PyN₃ scaffold obtained with pH-potentiometric titrations.

Dinuclear Complexes.

X-ray quality crystals of a dinuclear Mn₂(III,IV) complex with formulation of [Mn₂(μ-O)₂(^OHPyN₃)₂][ClO₄]₃ was observed for materials obtained from metalation of ^OHPyN₃ with Mn(ClO₄)₂ salts. All attempts to obtain monomeric crystalline congeners were unsuccessful and resulted only in isolation of dimeric species. The geometry around each manganese ion is distorted octahedral, with two cis μ-oxo bridges and four N atoms from the ligand. An inversion center on the O atom results in crystallographically equivalent manganese atoms within this system, consistent with the Mn–N bond lengths that are intermediate between a d³ Mn(IV) and d⁴ Mn(III). This effect has been observed in several other Mn₂(III,IV) dimers of this type, including the previously reported structure [Mn₂(μ-O)₂(N₄)₂][Cl]₃, attributing the effect to static disorder in the crystals.³⁷

In contrast, the previously reported complex [Mn₂(μ-O)₂(PyN₃)₂][ClO₄]₃ shows evidence for two crystallographically distinct metal centers through inspection of the Mn–N bond lengths between the two manganese centers (Table 4). Mn1 is assigned as the d³ Mn(IV) ion with axial and equatorial bond distances relatively similar to one another (Mn1–N1 = 2.086 Å, Mn1–N2 = 2.091 Å, and Mn1–N3 = 2.090 Å). Mn2 is consistent with a d⁴ Mn(III) ion with Jahn–Teller distortion observed in the lengthening of the Mn–N axial bonds (Mn2–N1 = 2.246 Å and Mn2–N3 = 2.266 Å), in contrast with the Mn–N equatorial bonds (Mn2–N2 = 2.110 Å and Mn2–N4 = 2.058 Å).

Table 4.

Selected Bond Lengths (Å) from SC-XRD Analysis of Dimeric Mn₂(III,IV) Complexes

	PyN3a
bond length (Å)	Mn1, d3 Mn(IV)	Mn2, d4 Mn(III)	OHPyN3b	N4c
Mn–N1	2.086(3)	2.246(3)	2.193(15)	2.166(2)
Mn–N2	2.091(3)	2.110(3)	2.104(13)	2.085(2)
Mn–N3	2.090(3)	2.266(3)	2.185(15)	2.175(2)
Mn–N4	1.991(3)	2.058(3)	2.00(12)	2.094(2)
Mn–O1	1.809(2)	1.861(2)	1.834(12)	1.823(1)
Mn–O2	1.800(2)	1.849(2)	1.830(11)	1.812(2)
Mn1–Mn2	2.712(7)		2.709(4)	2.694(1)

^a[Mn₂(μ-O)₂(PyN₃)₂][ClO₄]₃, Freire et al.⁸

^b[Mn₂(μ-O)₂(^OHPyN₃)₂][ClO₄]₃, this work.

^c[Mn₂(μ-O)₂(N₄)₂][Cl]₃, Goodson et al.³⁷

Electrochemistry Is an Analytical Probe To Differentiate Monomeric vs Dimeric Species.

Cyclic voltammetry (CV) was used to evaluate the electronic impact of the ligand scaffold on the manganese center as a monomeric and dimeric bis-μ-oxo complex, when possible. CV data obtained in MeCN were referenced to the redox couple of Fc^+/0, while experiments performed in H₂O were referenced to the redox couple of [Fe(CN)₆]^3−/4−. Monomeric species were analyzed in MeCN, showing a quasi-reversible and diffusion controlled redox couple (Figures 6A and S12–S15). Decreasing the number of pyridine rings from [Mn(Py₂N₂)Cl₂][BF₄] to [Mn(PyN₃)Cl₂][ClO₄] makes the complex harder to reduce, evidenced by the shift of E_1/2 to less positive potentials (Table 5). This result is consistent with previous data obtained from pH-potentiometric titrations showing that Py₂N₂ is a weaker manganese binding ligand compared to PyN₃. However, the redox potential can be fine-tuned through substitutions of the pyridine ring as evidenced by ^OHPyN₃ and ^OMePyN₃ showing a negative shift of E_1/2 in comparison with PyN₃, exhibiting the stabilizing effect of EDGs on the PyN₃ scaffold (Table 5). No monomeric complex of manganese was obtained with N4, which prevented the analysis of electrochemical data of this ligand among this series.

Figure 6.

Cyclic voltammograms of (A) monomeric manganese complexes in MeCN (ν = 100 mV s⁻¹, 0.1 M TBAP), (B) dimeric bis-μ-oxo manganese complexes in H₂O (ν = 100 mV s⁻¹, 0.1 M NaClO₄), (C) dimeric bis-μ-oxo manganese complexes in MeCN (ν = 100 mV s⁻¹, 0.1 M TBAP), and (D) [Mn₂(μ-O)₂(N₄)₂][Cl]₃ in MeCN with addition of H₂O (ν = 100 mV s⁻¹, 0.1 M TBAP). All scans were obtained by scanning in the cathodic direction from the open circuit potential.

Table 5.

Electrochemical Data of Monomeric and Dimeric Isolated Manganese Complexes

complexes	redox wave	solventa	Epa (mV)	Epc (mV)	E1/2 (mV)
[Mn(PyN3)Cl2][ClO4]	Mn(II/III)	MeCN	371	217	294
[Mn(OHPyN3)Cl2][ClO4]			236	61	148
[Mn(OMePyN3)F2][BF4]			329	148	239
[Mn(Py2N2)Cl2][BF4]			484	388	436
[Mn2(μ-O)2(N4)2][Cl]3	Mn(III,IV/IV,IV)	H2O	485	380	433
		MeCN	510	387	448
	Mn(III,IV/III,III)	H2O	−5	−103	−54
		MeCN	−279	−399	−339
[Mn2(μ-O)2(PyN3)2][ClO4]3	Mn(III,IV/IV,IV)	H2O
		MeCN
	Mn(III,IV/III,III)	H2O	193	52	123
		MeCN	67	−16	26
[Mn2(μ-O)2(OHPyN3)2][ClO4]3	Mn(III,IV/IV,IV)	H2O	526	401	464
		MeCN
	Mn(III,IV/III,III)	H2O	155	−41	57
		MeCN

^aManganese complexes in MeCN were referenced vs Fc^+/0, and 0.1 M TBAP was used as electrolyte. Manganese complexes in H₂O were referenced vs [Fe(CN)₆]^3−/4−, and 0.1 M NaClO₄ was used as electrolyte.

First, the dimeric bis-μ-oxo species were analyzed in water. CV results show the traditional two electrochemical waves observed in previous reports of bis-μ-oxo dimers of N₄ (cyclen) and cyclam corresponding to the Mn(III,IV/III,III) reduction couple at more positive potentials, and a Mn(III,IV/IV,IV) oxidation couple at more negative potentials.³⁸ Likewise, both of these two characteristic waves are observed for solutions of [Mn₂(μ-O)₂(^OHPyN₃)₂][ClO₄]₃ and [Mn₂(μ-O)₂(N₄)₂][Cl]₃ (Figure 6B). However, analysis of [Mn₂(μ-O)₂(PyN₃)₂][ClO₄]₃ only results in one reduction wave Mn(III,IV/III,III). All CVs show a quasi-reversible and diffusion controlled behavior (Figures 6B and S16–S20). As the number of pyridine units decreases from PyN₃ to N₄, the E_1/2 shifts to more negative potentials, consistent with data obtained from the monomeric complexes (Table 5). The dimer of ^OHPyN₃ as EDG also shows the same shift observed in monomeric species to more negative potentials in comparison with the unsubstituted congener PyN₃ (Table 5).

To compare electrochemical data from monomeric and dimeric species, electrochemical analysis of dimeric complexes was also performed in MeCN, obtaining quasi-reversible waves that were diffusion controlled for [Mn₂(μ-O)₂(PyN₃)₂][ClO₄]₃ and [Mn₂(μ-O)₂(N₄)₂][Cl]₃ (Figures 6C and S21–S23). [Mn₂(μ-O)₂(^OHPyN₃)₂][ClO₄]₃ was not soluble in MeCN, which prevented the study of this complex. Overall, electrochemical waves were comparatively sharper in MeCN while maintaining the general features observed in water. Since the electrochemical waves were more broad in water, we decided to carry out an experiment for [Mn₂(μ-O)₂(N₄)₂][Cl]₃ where water was added to MeCN in portions to understand the impact of water on the position of the electrochemical waves in MeCN (Figure 6D). As more water is added, the reduction wave of Mn^{III,IV/III,III} shifts to more positive potentials, while the oxidation wave Mn(III,IV/IV,IV) shifts to more negative potentials. A prominent oxidation wave is also observed at positive potentials concomitant with the increase of water aliquots, which is consistent with water oxidation. The results revealed that water impacts both electrochemical events, making both oxidation and reduction waves easier to access.

CV experiments were then performed on manganese complexes formed in situ to assign them as mono- or dinuclear complexes (Figure 7). The setup was selected to keep the conditions as consistent as possible with the H₂O₂ catalytic experiments described in the next section. Complexes of manganese with the library of ligands: N₄, PyN₃, Py₂N₂, ^OHPyN_3, and ^OMePyN₃, were prepared in situ using standardized solutions of MnCl₂ and ligands at pH = 7.5 in a H₂O:MeCN ratio of 1:5 to achieve a total concentration of the complex of 5 mM. Using MeCN in the solvent mixture, the electrochemical behavior of the isolated species could be compared with those prepared in situ. Complexes prepared in situ with MnCl₂ rapidly (~2 min) oxidize, which is evident by the strong color change from transparent to dark green under aerobic conditions. All complexes were allowed to mix aerobically for 10 min, and then the atmospheric air was replaced by a blanket of nitrogen prior to electrochemical measurements.

Figure 7.

Cyclic voltammograms of (top) [Mn(PyN₃)Cl₂][ClO₄] and [Mn₂(μ-O)₂(PyN₃)₂][ClO₄]₃ in MeCN, highlighting the difference between the E_pc waves (ν = 100 mV s⁻¹, 0.1 M TBAP); (bottom) manganese complexes prepared in situ in a solvent mixture of H₂O:MeCN 1:5 (ν = 100 mV s⁻¹, 0.1 M TBAP).

The redox couple of manganese complexes in situ was assigned using electrochemical data of isolated complexes (Table 6) as the reduction wave of bis-μ-oxo complexes, Mn(III,IV/III,III). The cathodic wave of Mn/N₄ is shifted to more negative potentials (E_pc = −343 mV) in comparison with the cathodic wave of [Mn₂(μ-O)₂(N₄)₂][Cl]₃ (E_pc = −399 mV), which is consistent with the shift of the reduction wave to positive potentials observed for [Mn₂(μ-O)₂(N₄)₂][Cl]₃ after addition of water (Figure 6D). Mn/PyN₃ shows electrochemical activity on a region of the chemical potential (E_1/2 = 26 mV) that matches the reduction wave of [Mn₂(μ-O)₂(PyN₃)₂][ClO₄]₃ (E_1/2 = 26 mV).

Table 6.

Electrochemical Data of Manganese Complexes Prepared In Situ^a

complexes	Epa (mV)	Epc (mV)	E1/2 (mV)
Mn/PyN3	90	−39	26
Mn/OMePyN3	72	−98	−13
Mn/OHPyN3	−73	−205	−139
Mn/N4	−171	−343	−257

^aManganese complexes were dissolved in H₂O:MeCN (1:5) and referenced vs Fc^+/0 using 0.1 M TBAP as electrolyte.

Addition of a pyridine ring in the macrocycle going from Mn/N₄ to Mn/PyN₃ shifts the E_1/2 to more positive potentials (Table 5), following the same trend observed for [Mn₂(μ-O)₂(N₄)₂][Cl]₃ and [Mn₂(μ-O)₂(PyN₃)₂][ClO₄]₃ (Table 6). Consistent with isolated complexes, EDGs of manganese complexes prepared in situ (–OMe and –OH substituents) also shifted the E_1/2 to more negative potentials. It was also observed a prominent oxidation wave at ~120 mV that grew as the potential was swapped to more positive values; we believed that this might have prevented the detection of Mn/Py₂N₂ which was expected to show at more positive potentials than Mn/PyN₃, based on the measurements done for [Mn(PyN₃)Cl₂][ClO₄] and [Mn(Py₂N₂)Cl₂][BF₄] (Table 6).

Comparison of redox waves of isolated crystals vs complexes prepared in situ highlights that electrochemical measurements can serve as an analytical probe to differentiate monomeric vs dimeric species of manganese when the chemical potential at which they interact with is sufficiently distinct (201 mV difference in the present study for manganese complexes of PyN₃).

Catalytic H₂O₂ Disproportionation Studies.

Kinetic studies were carried out to quantify the effect of the ligand structure on the performance of manganese to catalyze disproportionation of H₂O₂ into O₂ and H₂O. In our previous report, the ability to disproportionate H₂O₂ was quantified with manganese complexes of N₄ and PyN₃ in terms of turn over number (TON) and turn over frequency (TOF).⁸ Nevertheless, a comparison of our results with the literature is challenging due to the fact that both TON (mol of product/mol of catalyst) and TOF (TON/time) depend on the concentration of substrate and/or catalyst, temperature, pressure, and time frame selected, which are not parameters homogeneous in the literature.³⁹ Instead, comparative studies of MnCAT biomimetics typically show the second-order kinetic constant (M⁻¹ s⁻¹), denoted as $k$ (initial rates method, r_i) or $k$_cat/K_M (Michaelis–Menten model).^3,5,6,40 In this study, we use the kinetic constant $k$ from initial rates method to describe catalytic activity and employ TON to quantify the longevity and overall robustness of the catalyst.

The ability of manganese in the presence of N₄, PyN3, ^OHPyN3, ^OMePyN₃, and Py₂N₂ to catalyze the disproportionation of H₂O₂ into O₂ and H₂O was tested using a one-pot approach to generate manganese complexes in situ, at a pH value optimal among all the ligands for complex formation based on data from potentiometric titrations (Table 2 and Figures S1–S5). The standardized reagent solutions were added to a starting volume of ultrapure water in the following order: ligand, buffer (Tris, pH = 8.0), and then MnCl₂. This one-pot approach avoided complications related to (1) the rapid formation of inactive manganese oxide/hydroxide species from the isolated complexes, and (2) the need to adjust pH after addition of MnCl₂, which would affect the relative concentration of the manganese complex in solution. In our previous study, we demonstrated that neither MnCl₂ (buffered or unbuffered) or the ligand produce any catalase activity under the conditions of our experiment (Figure S28).⁸ Isolated complexes presented similar reactivity to species in situ, thus were preferred to streamline the catalytic experiments in this latest study. This work also showed that different mixing times of MnCl₂ and ligands (instantaneous, 6 h, or overnight) followed by addition of H₂O₂ did not provide a difference in catalase reactivity based on the TON recorded.

Results regarding the number of pyridine rings showed significant changes for the catalytic activity measured with $k$ among the ligands: N₄ (0.6 M s⁻¹) ≪ PyN₃ (1.8 M s⁻¹) ≫ Py₂N₂ (0.5 M s⁻¹) (Table 7 and Figure 8A). Within this group, PyN₃ possess the right scaffold for efficiently catalyze H₂O₂ disproportionation in comparison with N₄ and Py₂N₂.

Table 7.

Kinetic and Reactivity Results of the Disproportionation of H₂O₂ Using Buffered Aqueous Solutions Containing MnCl₂ and the Respective Macrocyclic Ligand ([MnCl₂]₀ = 1.50 mM, [ligand]₀ = 1.53 mM, [Tris] = 50 mM, pH = 8, and T = 298 K)

	N4	PyN3	OHPyN3	OMePyN3	Py2N2
k (M−1 s−1)	0.60(3)	1.8(1)	2.2(1)	2.3(1)	0.5(1)
TONa	11.26(4)	24.67(3)	36.89(1)	37.31(2)	28.85(3)

^aTON values were calculated at 20 min after the initial injection of H₂O₂, and results are representative of an average of five individual samples.

Figure 8.

(A,B) Second-order kinetic constant, $k$, and (C,D) TON values of the catalytic disproportionation of H₂O₂ using buffered aqueous solutions containing MnCl₂ and the respective macrocyclic ligand ([MnCl₂]₀ = 1.50 mM, [ligand]₀ = 1.53 mM, [Tris] = 50 mM, pH = 8, and T = 298 K).

TON values (Figure 8B) indicate that the Py₂N₂ ligand forms the most robust manganese catalyst regarding the series with different number of pyridine rings. Addition of H₂O₂ causes a steady release of ${\mathrm{H}}^{+}$, evidenced by measurements of pH in unbuffered solutions throughout the reaction time. This change eventually leads to dissociation of the manganese complex as evidenced by mass spectrometry measurements (Figures S36–S40). Py₂N₂ is able to consistently sequester Mn(II) ions in a wider range of pH (pMn = 5.96), which could explain the higher TON observed with this complex.

Results regarding substitution on the 4-position of the pyridine ring revealed that the unsubstituted PyN₃ ligand shows a less reactive catalyst when bound to manganese (k = 1.8 M s⁻¹, TON = 24.7) in comparison to ^OHPyN₃ (k = 2.2 M s⁻¹, TON = 36.9) and ^OMePyN₃ (k = 2.3 M s⁻¹, TON = 37.3) (Table 7 and Figure 8C,D). The EDG substitutions on PyN₃ (–OH and –OMe) stabilize higher oxidation states of manganese, based on CV studies performed on species in situ (E_1/2), increasing the donor strength of the N atom in the pyridine ring closer to the strength of the secondary amines as in N₄ (Figure 7), while retaining the conformational reorganization energy of the PyN₃ scaffold. EDGs on the pyridine ring show that the electronic properties of the macrocycle offer a handle to fine-tune the reactivity ($k$ and TON) without considerable changes to the coordination environment, a property that has also been observed with similar macrocycles bound to Fe(III) and Cu(II).^33,35 Therefore, inclusion of EDGs was also hypothesized to increase the catalytic activity of a given manganese complex, by stabilizing key high-valent state intermediates, which have been postulated to play an important role in the mechanism of reactions of this type.^41,42 While MnCAT achieves catalytic reactivity by modulating the protein environment surrounding the manganese binuclear center,⁴³ small molecules depend on the thermodynamic stability of complex formation and electronic properties to modulate the reactivity of the metal center. The results presented in this section show that addition of pyridine rings and their modification with EDGs must play a role in the stabilization of high valent manganese intermediates that may be present in the catalytic cycle, which have been reported in mechanistic studies with heme and nonheme catalase biomimetics.^3,44

An induction time at the onset of the reaction was observed after injection of H₂O₂, which motivated the construction of plots of O₂ production resulting from addition of H₂O₂ to solutions containing manganese complexes of N₄, PyN₃ and Py₂N₂ (Figure 9). This induction time for both N₄ and PyN₃ was roughly 30 s after injection of H₂O₂, while the time for Py₂N₂ was 2 min. These observations show that the manganese complex initially injected must go through steps to produce the active form of the catalyst to perform H₂O₂ disproportionation.

Figure 9.

Time course of partial pressure of O₂ (P_O2) after injecting H₂O₂ to buffered aqueous solutions containing MnCl₂ and ligands (A) N₄, (B) PyN₃, and (C) Py₂N₂ ([MnCl₂]₀ = 1.50 mM, [ligand]₀ = 1.53 mM, [Tris] = 50 mM, pH = 8, T = 298 K).

The effect of the temperature on the rate constant was investigated between 5 and 25 °C to gather thermodynamic data regarding the activation parameters. The activation energy $\left(E_{\mathrm{a}}\right)$ and pre-exponential factor $\left(A\right)$ were calculated using the Arrhenius model, as shown in the following equation.

$$k_{\mathrm{t}}=A{\mathrm{e}}^{-E_{\mathrm{a}}/RT}$$

The plot of ln $k_{\mathrm{o}\mathrm{b}\mathrm{s}}$ vs $T^{-1}$ for each manganese complex was used to obtain the $E_{\mathrm{a}}$ from the slope of the linear regression and $A$ from the intercept (Table 8 and Figures S34–S38).

Table 8.

Rate Constants ($k_{\mathrm{o}\mathrm{b}\mathrm{s}}$) at Different Temperatures, Activation Energy (E_a), and Pre-Exponential Factor (A)

	temperature (°C)	$k_{\mathrm{o}\mathrm{b}\mathrm{s}}$ (s−1)	Ea (kcal mol−1)	A (s−1)
N4	5	0.00084(2)	1.1(2)	0.005(7)
	10	0.00086(2)
	15	0.00091(8)
	25	0.00094(4)
PyN3	5	0.00163(7)	4.5(5)	5.4(2)
	10	0.00185(8)
	15	0.00202(5)
	25	0.00283(2)
Py2N2	5	0.00019(2)	13.4(4)	6.5 × 106(8)
	10	0.00024(1)
	15	0.00068(3)
	25	0.00086(6)
OHPyN3	5	0.00192(5)	4.2(5)	3.9(3)
	10	0.00213(7)
	15	0.00230(1)
	25	0.00321(1)
OMePyN3	5	0.00187(1)	4.6(3)	7.3(2)
	10	0.00217(1)
	15	0.00263(1)
	25	0.00324(2)

The E_a calculated from kinetic data using the Arrhenius equation (Table 8) showed an increment as the number of pyridine rings increases from N₄ (E_a = 1.1 kcal mol⁻¹) < PyN₃ (E_a = 4.5 kcal mol⁻¹) < Py₂N₂ (E_a = 13.4 kcal mol⁻¹), showing how the reorganization energy of the macrocyclic ligand might play a role in achieving the desired reactivity.

CONCLUSIONS

This report details the isolation and characterization of a novel library of manganese complexes comprised of 12-membered tetra-aza macrocyclic ligands that bind in a monomeric or bis-μ-oxo dimers to the metal center. The structure of the ligand was selected to evaluate the impact of the number of pyridine rings as well as EDGs bound to the pyridine moiety. Ligand basicity showed a decrease with the number of pyridine rings, but substitutions did not show a straightforward trend. –OH substitution showed an increase in the basicity of the ligand, but tautomerism keto–enol disclosed an important factor in the complexation of Mn(II) ions, presenting a shift of the mesomeric effect of the –OH group. On the other hand, –OMe substitution did not reflect an electron-donating character on the basicity of the ligand, showing a decrease of basicity in comparison with the parent PyN₃ unsubstituted ligand.

The basicity of the ligand played a fundamental role in equilibrium studies performed with Mn(II) ions in aqueous solutions and allowed us to determine the best strategy to synthesize and isolate the corresponding complexes. Depending on the structure of the ligand, some complexes were isolated as monomeric structures or as bis-μ-oxo dimers as shown by SC-XRD. The features of isolated species in solution were evaluated using UV–vis spectroscopy and cyclic voltammetry. Electrochemistry was fundamental to characterize the species observed in solution when manganese complexes were prepared in situ since the potential for monomer vs dimer species is separated by a potential window of 201 mV.

Species in situ were used for catalytic disproportionation of H₂O₂, showing similar results to isolated species. TON values increased consistently with the number of pyridine rings, but the kinetic constant $k$ favored the PyN₃ system. Py₂N₂ showed a decreased of $k$ and a longer induction period. On the other hand, substitutions on the 4-position of the pyridine ring favored both TON and $k$ to a bigger extent, highlighting the importance of these substitution on the reactivity of manganese.

We hope that this study stimulates the field into venturing in the reactivity of dimers of manganese that expose structural properties required for the desired reactivity. Learning how to tune the reactivity of H₂O₂ with manganese catalysts as an oxidant for catalysis or as a substrate for disproportionation would greatly aid in the field of oxidation catalysis. More importantly, studies with similar catalysts under air atmosphere would facilitate their implementation in industrial applications.

EXPERIMENTAL SECTION

General Methods.

Caution! Perchlorate salts are explosive and should be handled with care; such compounds should never be heated as solids. All chemical reagents were purchased from either Millipore Sigma or Alfa Aesar and used without further purification unless otherwise indicated. Combustion elemental analysis was performed by Atlantic Microlab in Norcross, GA. Electronic absorption spectra were collected between 190 and 1100 nm using a Cary 60 spectrophotometer (Agilent) and a 3 mL quartz cuvette with a path length of 1.0 cm. Molar extinction coefficients were calculated using the Beer–Lambert law (A = εbc). Electrospray ionization mass spectrometry (ESI-MS) was performed on an Advion Expression^L CMS instrument.

Synthesis of Ligands.

^OHPyN₃ was prepared following the synthetic procedure of Lincoln et al., with no further modifications.¹⁴ PyN₃, ^OMePyN₃, and Py₂N₂ were prepared by adapting procedures reported previously.^{13,15,45–52} Scheme S1 shows the overall synthetic procedure. Modified methods are described further herein. Chelidamic acid, 3b, and 10 were obtained from commercial sources.

Synthesis of 1.

Small modifications to the previously reported methods were used.^45–47 Thionyl chloride (27.2 mL, 44.4 g, 373 mmol) was added dropwise to a suspension of chelidamic acid monohydrate (25.0 g, 124 mmol) in absolute EtOH (250 mL) at 0 °C. The mixture was allowed to warm at room temperature under stirring (15 min) and then heated at 80 °C for 4h. The solution was concentrated under reduced pressure as a eutectic mixture with toluene (3×, 100 mL), using solid Na₂CO₃ in the receiving flask. An oil was obtained and mixed with H₂O:Et₂O (1:1, 300 mL) to precipitate a brown solid, which was thoroughly washed with Et₂O. A flash column was performed (SiO₂, CH₂Cl₂:MeOH = 98:2) to obtain 1 (22.2 g, 93.0 mmol, 75%) as a yellow solid. ¹H NMR spectrum was consistent with the literature values.

Synthesis of 5a and 5b.

These compounds were synthesized by following the reported literature methods with slight modifications.^13,49,50 11 (5.71 g, 10.1 mmol) and CsCO₃ (13.2 g, 40.4 mmol) were mixed together in DMF (110 mL). A solution of either 4a or 4b (4.52 g, 10.1 mmol) in DMF (50 mL) was added dropwise. The reaction mixture was stirred at room temperature for 24 h. The solvent was removed under reduced pressure as a eutectic mixture with toluene (3x) and dried in vacuo overnight. The resultant solid was redissolved in H₂O (150 mL) and extracted with CHCl₃ (3×, 150 mL). The combined organic layers were washed with brine (150 mL), dried over Na₂SO₄, and concentrated under reduced pressure. The residue was purified by column chromatography (SiO₂, CH₂Cl₂:EtOAc = 40:1) to afford either 5a (5.79 g, 8.28 mmol, 82%) or 5b (5.74 g, 8.58 mmol, 85%) as a white foamy solid. R_f = 0.4 (SiO₂, CH₂Cl₂:EtOAc = 40:1). ¹H NMR spectrum was consistent with the literature values.

Synthesis of 6a (^OMePyN₃·3HCl) and 6b (PyN₃·3HCl).

These compounds were synthesized by following the reported literature methods with slight modifications.^13,49 A solution of either 5a (5.61 g, 8.03 mmol) or 5b (5.37 g, 8.03 mmol) in concentrated H₂SO₄ (15 mL) was stirred at 100 °C for 4 h. The resultant dark brown solution was cooled down to room temperature, followed by addition of H₂O (75 mL). The acidic mixture was extracted with Et₂O (1×, 75 mL), and the aqueous layer was neutralized with a 60% m/v solution of NaOH until reaching pH = 14. The product was extracted with CH₂Cl₂ (3×, 100 mL). The combined organic layers were washed with brine (100 mL), dried over Na₂SO₄, and concentrated under reduced pressure. The resultant yellow oil was mixed with concentrated HCl (1 mL), followed by CH₃CN (5 mL) and Et₂O (100 mL). The mixture was sonicated (20 min) and left in the freezer for 2 h. The precipitate was filtered, washed with Et₂O, and dried under vacuum to obtain either 6a (2.22 g, 6.42 mmol, 80%) or 6b (2.15 g, 6.82 mmol, 85%) as a white solid. ¹H NMR spectrum was consistent with the literature values. X-ray quality crystals of 6a were also obtained by slow evaporation of MeOH as translucid colorless cubes (CCDC #2245877).

Synthesis of 7 and 8.

These compounds were synthesized by following the reported literature methods with slight modifications.¹⁵ 4b (4.00 g, 8.94 mmol) and TsNHNa (3.44 g, 17.8 mmol) were mixed and suspended in 600 mL of MeCN, followed by reflux under stirring for up to 7 days. The solvent was evaporated under reduced pressure. The resultant solid was washed with cold MeOH and filtrated to obtain a mixture of tosyl-protected dimer and trimer macrocycle. The dimeric macrocycle 8 (1.18 g, 2.01 mmol, 45%) was separated using the procedure reported by Wessel et al., with no further modifications.^{15 H}H NMR spectrum was consistent with the literature values.

Synthesis of 11.

This compound was synthesized by following the reported literature methods with slight modifications.^51,52 Compound 10 (9.21 g, 89.3 mmol) was dissolved in H₂O (55 mL) and NaOH powder (10.7 g, 268 mmol) was added in small portions during a 30 min time frame, keeping the temperature of the solution at 20 °C. Et₂O (55 mL) was added, and the reaction mixture was stirred vigorously. TsCl (51.1 g, 268 mmol) was added while cooling the reaction mixture to 0 °C and then stirred for 2h. A white precipitate was filtered and washed with Et₂O (110 mL). Recrystallization from MeOH gave the desired product 11 (50.0 g, 88.4 mmol, 99%) as a white solid. ¹H NMR spectrum was consistent with the literature values.

Potentiometric Measurements.

The MnCl₂ stock solution was prepared from analytical grade commercial sources, and its concentration was determined by complexometric titration with standardized Na₂H₂EDTA and an EBT indicator in the presence of ascorbic acid and triethanol-amine. The concentration of the ligand stock solution was determined by pH-potentiometric titration. The protonation constants were calculated with the following setup: 0.2 M carbonate-free NaOH titrant with 2 mM ligand solution at an initial volume of 6 mL. Titrations were performed at 25.0 ± 0.1 °C and an ionic strength of I = 0.15 M NaCl. An inert atmosphere was provided by a constant passage of dry N₂ through the sample. The protonation constants of the ligand $\left(\mathrm{l}\mathrm{o}\mathrm{g}{K}_i^H\right)$ are defined as

$$K_i^{\mathrm{H}}=\frac{\left[{\mathrm{H}}_i{\mathrm{L}}^{i+}\right]}{[{\mathrm{H}}_{i-1}{\mathrm{L}}^{(i-1)+}][{\mathrm{H}}^{+}]}$$

where $i=1,2$, and 3 and $[{\mathrm{H}}_{i-1}{\mathrm{L}}^{(i-1)+}]$ and $\left[{\mathrm{H}}^{+}\right]$ are the equilibrium concentrations of the ligand in different protonation states and hydrogen ions, respectively. Potentiometric titrations were carried out with a Metrohm 888 Titrando workstation and a Metrohm 6.0234.100 combined electrode in the pH range of 1.7–11.8. For the calibration of the electrode, KH-phthalate (pH = 4.008)⁵³ and borate (pH = 9.177)⁵⁴ buffer standards were used, and the $\left[{\mathrm{H}}^{+}\right]$ was calculated from the measured pH values by applying the method proposed by Irving et al.⁵⁵ A solution of approximately 0.01 M HCl was titrated with 0.2 M NaOH solution (I = 0.15 M NaCl), and the differences between the measured and calculated pH values (pH < 2.2) were used to calculated the $\left[{\mathrm{H}}^{+}\right]$ from the pH values measured in the ligand titrations. The experimental points above pH 11 for the acid–base titration were used to determine the ionic product of water (13.840) in case of our experimental setup. PSEQUAD software was used to calculate the equilibrium constants.⁵⁶

The protonation constants for the ligands were determined by titrating the ligand solution (acidified with a known volume of a standard HCl solution) with 0.2 M NaOH at 0.15 M NaCl ionic strength in the 1.7–11.8 pH range. Equilibrium points 1 and 2 (EP1 and EP2, respectively) were separated using MnCl₂. The $\mathrm{l}\mathrm{o}\mathrm{g}{K}_i^{\mathrm{H}}$ values were calculated from 150 V (mL)−pH data pairs. The initial proton concentration ${\left[{\mathrm{H}}^{+}\right]}_0$ submitted in PSEQUAD was obtained from the sum of the moles of ${\mathrm{H}}^{+}$ added from standardized solutions of HCl and ligand solution.

To determine the stability constants of the manganese complexes, potentiometric titrations were carried out at a 1:1 metal to ligand molar ratio, with 2% ligand excess to prevent the hydrolysis of Mn(II), allowing 1 min for the equilibration to occur, with 150 V (mL)−pH data pairs. The initial proton concentration ${\left[{\mathrm{H}}^{+}\right]}_0$ submitted in PSEQUAD was calculated using the following equation.

$${\left[{\mathrm{H}}^{+}\right]}_0=\frac{\left[\mathrm{E}\mathrm{P}{1}_{\mathrm{M}\mathrm{L}}+\frac{V_{\mathrm{L}}}{V_{\mathrm{M}\mathrm{L}}}\left(\mathrm{\Delta }{\mathrm{E}\mathrm{P}}_{\mathrm{L}}\right)\right]\left[\mathrm{N}\mathrm{a}\mathrm{O}\mathrm{H}\right]}{V_{\mathrm{o},\mathrm{M}\mathrm{L}}}$$

where the subindexes L and ML represent values derived from the ligand and metal–ligand titration, respectively. $V_{\mathrm{L}}$ and $V_{\mathrm{M}\mathrm{L}}$ is the volume of ligand in each titration. $V_{\mathrm{o}}$ is the initial volume. $\mathrm{E}\mathrm{P}{1}_{\mathrm{M}\mathrm{L}}$ corresponds to the first equilibrium point in the ML titration, and $\mathrm{\Delta }{\mathrm{E}\mathrm{P}}_{\mathrm{L}}$ corresponds to the difference between first and second equilibrium points in the ligand titration. This equation allows a more accurate determination of the ${[{\mathrm{H}}^{+}]}_0$ by taking into account the 2% excess of ligand used, in the case of metal complexes that do not present changes in EP1 between ligand and metal–ligand titration.

Synthesis of Manganese Complexes.

[Mn₂(μ-O)₂(N₄)₂][Cl]₃, [Mn(PyN₃)Cl₂][ClO₄], and [Mn₂(μ-O)₂(PyN₃)₂][ClO₄]₃ were synthesized following previously reported procedures with no further modifications.^8,38

Synthesis of [Mn(^OHPyN₃)Cl₂][ClO₄].

An adapted procedure based on work reported elsewhere was used for the synthesis of this compound.^{8 OH}PyN₃·3HCl (100 mg, 0.302 mmol) was dissolved in 5 mL of H₂O. The pH of the solution was adjusted to 8 with a solution of KOH (1 M). A solution of Mn(ClO₄)₂·xH₂O (75.1 mg, 0.296 mmol) in H₂O (5 mL) was added and stirred at room temperature for 12 h. The solution color gradually changed from pale yellow to dark green. The resultant mixture was filtered, and the solvent was removed under reduced pressure. The resultant solid was redissolved in MeCN (20 mL) and stirred at room temperature for 12 h. A bright pink solution was obtained, which was filtered to remove solid impurities and the solvent was removed under reduced pressure to give the desired product [Mn(^OHPyN₃)Cl₂][ClO₄] (99.4 mg, 0.222 mmol, 75%) as a pink powder. *The strong propensity of this class of compounds to form μ-oxo dimers and MnO₂ under oxidizing conditions, i.e., synthesis or combustion analysis, has continued to result in mixtures of compound despite extensive efforts.

Synthesis of [Mn₂(μ-O)₂(^OHPyN₃)₂][ClO₄]₃.

An adapted procedure based on work reported elsewhere was used for the synthesis of this compound.^{8 OH}PyN₃·3HCl (100 mg, 0.302 mmol) and Mn(ClO₄)₂·xH₂O (75.1 mg, 0.296 mmol) were dissolved in H₂O (10 mL). The pH of the solution was adjusted to 8 with a solution of KOH (1 M) and subsequently stirred at room temperature for 12 h. The solution color gradually changed from pale yellow to dark green. The resultant mixture was filtered, and the solvent was removed under reduced pressure. The resultant solid was redissolved in MeCN (20 mL) and stirred at room temperature for 12 h. A dark green solution was obtained, which was filtered to remove solid impurities and the solvent was removed under reduced pressure to give the desired product [Mn₂(μ-O)₂(^OHPyN₃)₂][ClO₄]₃ (58.2 mg, 0.0630 mmol, 42%) as a green powder. The product was redissolved in H₂O and filtered (PTFE, 0.45 μm) to obtain X-ray quality crystals by slow evaporation (CCDC #2241016).

Synthesis of [Mn(^OMePyN₃)F₂][BF₄].

An adapted procedure based on work reported elsewhere was used for the synthesis of this compound.^{8 OMe}PyN₃·3HCl (100 mg, 0.289 mmol) was dissolved in 5 mL of H₂O. The pH of the solution was adjusted to 7 with a solution of KOH (1 M). A solution of MnCl₂·4H₂O (56.0 mg, 0.283 mmol) in H₂O (5 mL) was added and stirred at room temperature for 12 h. The solution color gradually changed from pale yellow to dark green. The resultant mixture was filtered, and the solvent was removed under reduced pressure. The resultant solid was redissolved in MeCN (20 mL) and stirred at room temperature for 12 h. A bright pink solution was obtained, which was filtered to remove solid impurities and the solvent was removed under reduced. The product was redissolved in MeOH (10 mL), mixed with AgBF₄ (55.1 mg, 0.283 mmol), and stirred for 2 h at room temperature to obtain X-ray quality crystals. The pink solution was filtered, and the solvent was removed under reduced pressure to give the desired product [Mn(^OMePyN₃)F₂][BF₄] (17.5 mg, 0.042 mmol, 15%) as a pink powder. The product was redissolved in MeOH and filtered (Nylon, 0.45 μm) to obtain X-ray quality crystals by slow diffusion of Et₂O. Elemental analysis; [Mn(^OMePyN₃)F₂][BF₄] found (calculated): C, 34.87 (34.64); H, 4.88 (4.85); N, 13.05 (13.47) % (CCDC #2205652).

Synthesis of [Mn(Py₂N₂)Cl₂][BF₄].

An adapted procedure based on work reported elsewhere was used for the synthesis of this compound.⁸ Py₂N₂ (100 mg, 0.416 mmol) was dissolved in 5 mL of H₂O. The pH of the solution was adjusted to 7 with a solution of KOH (1 M). A solution of MnCl₂·4H₂O (80.7 mg, 0.408 mmol) in H₂O (5 mL) was added and stirred at room temperature for 12 h. The solution color gradually changed from pale yellow to dark green. The resultant mixture was filtered, and the solvent was removed under reduced pressure. The resultant solid was redissolved in MeCN (20 mL) and stirred at room temperature for 12 h. A bright pink solution was obtained, which was filtered to remove solid impurities and the solvent was removed under reduced pressure. The product was redissolved in MeOH (10 mL), mixed with AgBF₄ (79.4 mg, 0.408 mmol), and stirred for 2 h at room temperature to obtain X-ray quality crystals. The pink solution was filtered, and the solvent was removed under reduced pressure to give the desired product [Mn(Py₂N₂)Cl₂][BF₄] (145 mg, 0.184 mmol, 45%) as a pink powder. The product was redissolved in MeOH and filtered (Nylon, 0.45 μm) to obtain X-ray quality crystals by slow diffusion of Et₂O. Elemental analysis; Mn[(Py₂N₂)F₂][BF₄]·1.9MnCl₂·3.5MeOH·H₂O found (calculated): C, 26.58 (26.63); H, 3.95 (4.09); N, 7.16 (7.10) % (CCDC #2207371).

X-ray Crystallography.

A Leica MZ 75 microscope was used to identify samples suitable for analysis. A Bruker APEX-II CCD diffractometer was employed for crystal screening, unit cell determination, and data collection, which was obtained at 100 K. The Bruker D8 goniometer was controlled using the APEX3 software suite.⁵⁷ The samples were optically centered with the aid of a video camera so that no translations were observed as the crystal was rotated through all positions. The X-ray radiation employed was generated from a MoK_α sealed X-ray tube (λ = 0.71076 Å) with a potential of 50 kV and a current of 30 mA; fitted with a graphite monochromator in the parallel mode (175 mm collimator with 0.5 mm pinholes). The crystal-to-detector distance was set to 50 mm, and the exposure time was 10 s per degree for all data sets at a scan width of 0.5°. The frames were integrated with the Bruker SAINT Software package using a narrow frame algorithm.⁵⁸ Data were corrected for absorption effects using the multiscan method SADABS.⁵⁹ Structures were solved using Olex2⁶⁰ with the ShelXS⁶¹ solution program using Direct Methods and refined with the SHELXL⁶² refinement package using least squares minimization. All hydrogen and non-hydrogen atoms were refined using anisotropic thermal parameters. The thermal ellipsoid molecular plots (50%) were produced using Olex2.⁶⁰

Electrochemistry.

Cyclic voltammetry (CV) was carried out with an EC Epsilon potentiostat (C-3 cell stand) purchased from BASi Analytical Instruments (West Lafayette, IN). A glassy carbon (GC) electrode from BASi (MF-2012), 3 mm in diameter, was polished on a white nylon pad (BASi MF-2058) with different sized diamond polishes (15, 6, and 1 μm) to ensure a mirror-like finish between each measurements. All solutions were bubbled with N₂ for at least 15 min prior to experimentation and kept under a N₂ blanket.

For electrochemical analysis of [Mn(PyN₃)Cl₂][ClO₄], [Mn(^OHPyN₃)Cl₂][ClO₄], [Mn(^OMePyN₃)F₂][BF₄], and Mn[(Py₂N₂)-F₂][BF₄], a three-electrode cell configuration was used: GC as the working electrode, Ag wire as the quasi-reference electrode housed in a glass tube (7.5 cm × 5.7 mm) and isolated from the analyte with a porous CoralPor tip, and a Pt wire (7.5 cm) as the counter electrode (BASi MW-1032). For each electrochemical analysis, samples were dissolved in anhydrous MeCN containing 0.1 M tetrabutylammonium perchlorate (TBAP) as supporting electrolyte. The concentration of each sample was 1 mg/mL in aliquots of 5 mL.

For electrochemical analysis of [Mn₂(μ-O)₂(N₄)₂][Cl]₃, [Mn₂(μ-O)₂(PyN₃)₂][ClO₄]_3, and [Mn₂(μ-O)₂(^OHPyN₃)₂][ClO₄]₃, a three-electrode cell configuration was used: GC as the working electrode, a Ag/AgCl reference electrode, and a Pt wire (7.5 cm) as the counter electrode (BASi MW-1032). For each electrochemical analysis, samples were dissolved in H₂O containing 0.1 M NaClO₄ as supporting electrolyte. The concentration of each sample was 1 mg/mL in aliquots of 5 mL.

For electrochemical analysis of Mn/N4, Mn/PyN3, Mn/^OHPyN₃, and Mn/^OMePyN₃, a three-electrode cell configuration was used: GC as the working electrode, Ag wire as the quasi-reference electrode housed in a glass tube (7.5 cm × 5.7 mm) and isolated from the analyte with a porous CoralPor tip, and a Pt wire (7.5 cm) as the counter electrode (BASi MW-1032). For each electrochemical analysis, samples were prepared in situ by mixing the corresponding amounts of stock solution of ligand and MnCl₂·4H₂O (Merck) to achieve a total concentration of 20 mM in 0.5 mL of H₂O. Subsequently, the pH of each sample was adjusted to 7.5 with solid KOH and stirred aerobically for 20 min where it was observed a color change from transparent to dark green overtime. Finally, 4.5 mL of anhydrous MeCN was added to each solution to achieve a total concentration of 2 mM in 5 mL aliquots of a mixed solvent H₂O:MeCN (1:5). 0.1 M TBAP was used as supporting electrolyte. Attempts to assign these manganese species were performed using UV–vis spectroscopy in our previous study, but there was ambiguity due to the similarity of the ~400 nm region between both species, and the difficulty associated with the stability of the band at ~550 nm, characteristic of the dimer, before additions of H₂O₂.⁸

Reaction of Manganese Complexes with H₂O₂ and Quantification of O₂ Evolution.

O₂ evolution was quantified as partial pressure (P) by a Clark-type commercial microsensor (Unisense, Denmark), calibrated using N₂ and air (P_O2 = 0 and 159 mmHg) under atmospheric pressure. The reactions were performed in a sealed pressure vessel (15 mL) equipped with a stirring bar and a three-way manifold valve. Two of the lines (PTFE tubing) were connected to the cell: one was used for N₂ and the other to inject the aliquot of H₂O₂. The pressure vessel was sealed using a PEEK bushing connected to threaded fittings (IDEX, Health and Science) for custom sealing between the gas tubing, microsensor, and reaction vessel. The pressure of the vessel was kept at 1 atm using a snorkel line with minimal loss of headspace during the catalytic studies.

The metal complexes were prepared in buffered aqueous solutions using stock solutions of the ligand of interest, buffer, and a metal solution of MnCl₂·4H₂O (Merck). Tris(hydroxymethyl) amino methane (Tris) was the buffer selected for the desired pH of the solution (pH = 8). The complexes were prepared using a one-pot approach by adding the corresponding amount of the stock solution of ligand, buffer, and metal salt in a vial to achieve the targeted concentration for the H₂O₂ studies ([ligand] = 2.04 mM; [Mn] = 2 mM; [buffer] = 50 mM). The vessel was loaded with 1.5 mL of the solution detailed above and purge with N₂ gas prior to every measurement. The microsensor signal was read continuously and measured every 0.2 s until a steady signal was obtained for 2 min, and then 0.5 mL of H₂O₂ solution was injected (600 mM, 50 mM buffer). The signal was recorded as ΔP_O2 (mmHg) vs time (min) until a plateau was achieved. Initial concentrations before starting the reaction were [ligand] = 1.53 mM (3.06 μmol), [Mn] = 1.5 mM (3 μmol), and [H₂O₂] = 150 mM (300 μmol).

Once the reaction was complete, the results were plotted to determine the turnover number (TON). The reaction was stopped when a visible plateau in the P_O2 was observed, which happened at roughly 20 min for all experiments. P_O2 was converted into [O₂] using the Ideal Gas Law. TON was calculated as the ratio between the moles of O₂ produced during the time frame selected divided by moles of catalyst used. Kinetic measurements were carried out under the same conditions previously described using the initial rates method (r_i). At constant [Cat]₀ = 1.5 mM (Cat = any given manganese complex prepared in situ), the reaction exhibited first-order kinetics with respect to [H₂O₂]₀, and the pseudo-first-order rate constant ($k_{\mathrm{o}\mathrm{b}\mathrm{s}}$) was obtained from the slope of plots of r_i vs [H₂O₂]₀. Similar results were obtained with isolated complexes, but complexes prepared in situ were preferred since the approach was streamlined and easier to reproduce.

At constants [H₂O₂]₀ = 150 mM, the reaction also exhibited a good linear dependence with respect to [Cat]₀, affording a pseudo-first-order rate constant (k_obs′) (Figures S29–S34). Since the conditions described showed first-order rates with respect to each species (Cat and H₂O₂), the second-order catalytic constant ($k$) was obtained by either dividing k_obs/[Cat]₀ or k_obs′/[H₂O₂]₀ and further validated by the slope of either a plot of r_i/[H₂O₂]₀ vs [Cat]₀ or a plot of r_i/[Cat]₀ vs [H₂O₂]₀, showing $k$ values in the same order of magnitude. Catalytic experiments were performed at 5, 10, 15, and 25 °C to calculate activation energy (E_a) for each catalyst using the Arrhenius equation.⁶³