A bioinspired model for copper monooxygenase: direct aromatic hydroxylation using O₂

Abstract

A novel copper(I) complex, [Cu^I(L)(CH₃CN)]CF₃SO₃ (1) (L = 1,1,2-tri(pyridin-2-yl)propan-1-ol), has been synthesized, characterized, and investigated as a bioinspired model for copper monooxygenases. Under aerobic conditions in CH₃CN, complex 1 undergoes conversion to a dicopper complex, [(Cu^IIL)(Cu^IIL H)(SO₃CF₃)₂]·CF₃SO₃·H₂O (2), whose molecular structure reveals a Cu–Cu distance of 2.96 Å. A dicopper(II) complex, [(LCu^II)₂(SO₃CF₃)₂] (3), has been synthesized for comparison, which exhibits a similar Cu–Cu distance of 2.97 Å. EPR spectroscopy has ascertained the solution-state geometries of complexes 2 and 3, which displayed g_∥ > g_⊥ values, indicative of distorted square pyramidal geometries consistent with their solid-state structures. Complex 1 selectively hydroxylates benzene in the presence of O₂ and Et₃N, affording 7% phenol based on the substrate, without any side products. However, the use of H₂O₂ as the oxygen source under identical conditions significantly increases the phenol yield to 19%. The catalytically active intermediates generated by the reaction of complex 1 with dioxygen showed an O (π*σ) → Cu ligand-to-metal charge transfer (LMCT) transition at 360 nm and a d–d transition at 650 nm. These spectral features are more pronounced with H₂O₂, showing a new LMCT transition at 360 nm and a very weak d–d transition at 689 nm. This is supported by solution FT-IR spectroscopy, which showed an O–O stretching frequency at 890 cm⁻¹ (DFT spectra at 829 cm⁻¹), corresponding to a Cu–OOH intermediate. The structure of the [(L)Cu^II–OOH]⁺ species was optimized by DFT calculations. Kinetic isotope effect (KIE) studies using C₆H₆/C₆D₆ (1 : 1) (k_H/k_D = 1.03) and isotopic labeling experiments using H₂¹⁸O₂ support our proposed mechanism of benzene hydroxylation. In contrast, dinuclear complexes 2 and 3 exhibited poor benzene hydroxylation activity even with H₂O₂ and yielded only 4% and 6% phenol, respectively, along with by-products such as biphenyl and quinone under identical conditions.

---

Introduction

One of the most challenging reactions so far in the arena of chemistry is the selective C–H functionalization of aromatic hydrocarbons using greener oxidants like O₂,^1–5 H₂O₂,^6–11 and N₂O^12,13 in a single step. In nature, a class of metalloenzymes selectively activates aromatic C–H bonds by utilizing metal-bound oxygen species as crucial catalytic intermediates. These oxygenation reactions are known to play a vital role in the biological metabolism of many living organisms,^14–19 which includes monooxygenase and hydroxylase types of enzymes. Their bioinorganic components have been well documented over the decades by several researchers.^20–24 Of particular interest, non-heme copper monooxygenase enzymes such as particulate methane monooxygenase (pMMO),^15,16,25 dopamine β-monooxygenase (DβM),^17–19 tyrosinase (Ty),^21,26 tyramine β-monooxygenase (TβM), peptidylglycine-α-hydroxylating monooxygenase (PHM), and polysaccharide monooxygenase (PMO) play a crucial role in the functionalization of challenging C–H bonds using dioxygen. They use specially adapted structural motifs to activate O₂ and enable selective C–H functionalization. These enzymes utilize copper centers to elegantly facilitate oxygenation/oxidation reactions of various hydrocarbon substrates under ambient conditions.^15–21 Nevertheless, selective aromatic hydroxylation is challenging and is consistently found to be an enduring area of interest in bioinspired catalysis research. In particular, a direct synthetic methodology for the aromatic hydroxylation of hydrocarbons has not been well explored. Hydroxylated hydrocarbons are synthetic precursors for many drugs and polymer materials. Phenol is one such valuable compound, and it is produced commercially worldwide by a three-step cumene process, which was developed by Hock and Lang in 1944.⁵ This process has a few inherent drawbacks, such as being a multi-step process, requiring high energy consumption, and affording a poor yield of phenol (5%) along with a significant amount of acetone as a by-product. At the same time, over-oxidation of phenol is facile under oxidation conditions because phenol is more sensitive than benzene, as it lacks kinetic control.^27,28

The earlier reports by Karlin and co-workers have focused mainly on intramolecular hydroxylation on auxiliary ligands or electron-rich aromatic systems using bioinspired models of tyrosinase, in which various copper complexes and their reactive intermediates were extensively studied.^4,29,30 In 2011, Perez and co-workers reported [Tp^xCu(NCMe)] [Tp^x = hydrotris(pyrazolyl)borate] as a catalyst for benzene hydroxylation. They achieved a phenol yield of 21%, but with poor selectivity (phenol : benzoquinone = 70 : 30) at 80 °C, and no deeper mechanistic investigation was reported.³¹ Later, in 2017 and 2022, the same group proposed two mechanisms (rebound or electrophilic substitution) through a Wheland intermediate for benzene hydroxylation via experimental and DFT studies.^32,33 In 2017, Kodera and co-workers explored benzene hydroxylation using a dicopper(II) complex [Cu₂(μ-OH)6-hpa]³⁺ [hpa = 2-bis(2-pyridylmethyl)aminomethyl-6-pyridylethane] as the catalyst and achieved 22% conversion of benzene into phenol.³⁴ The mechanism of the oxidation of benzene to phenol with H₂O₂ by copper complexes is still unclear in the literature, such as whether the reaction proceeds directly via a Cu–oxygen species or indirectly via radical species produced by Cu–oxygen species. Our group has developed bioinspired Cu(II) complexes with N₄ ligands and achieved a 37% phenol yield using H₂O₂, with the reaction proceeding via Cu(II)OOH species as an active intermediate.³⁵ Later, Cu(I) complexes reported by us showed benzene hydroxylation with a 29% phenol yield but with reduced selectivity compared to Cu(II) complexes.³⁶ All the above-reported bioinspired aromatic hydroxylation studies extensively used H₂O₂ as an oxygen source. However, none of the reports attempted to use dioxygen as an oxygen source, except for Karlin and co-workers’ study of intramolecular hydroxylation on the ligand moiety.^4,29

This article reports the synthesis of a new derivative of the tris-pyridine-based ligand³⁶ and its copper(I) complexes, serving as a relevant bioinspired model for copper monooxygenase enzymes. The Cu(I) complex catalyst effectively converts benzene to phenol using dioxygen as a greener source, achieving a phenol yield of 7% based on benzene. Furthermore, the yield can be enhanced to 19% when using H₂O₂. Notably, this represents one of the rare examples of intermolecular aromatic hydroxylation by a Cu(I) complex utilizing dioxygen, and only a few reports are available in the literature. Our spectral data and catalysis experiments support the idea that the benzene hydroxylation likely proceeds via active Cu(II)–OOH species, and the experimental results align well with the data obtained from DFT and TDDFT calculations.

Results and discussion

Synthesis and characterization

The ligand L = [1,1,2-tri(pyridin-2-yl)propan-1-ol] was synthesized from di-2-pyridyl ketone with lithiated 2-ethyl pyridine in THF at −70 °C and its formation was confirmed by NMR and ESI-MS spectral methods (Scheme 1A, Fig. S1–S3†). The copper(I) complex [Cu^I(L)(CH₃CN)]CF₃SO₃1 was synthesized by the reaction of [Cu(CH₃CN)₄]CF₃SO₃ with L (1 : 1) in a degassed toluene/acetone mixture under a strict Ar atmosphere. Its formation was confirmed by HRESI-MS, ¹H-NMR analysis and elemental analysis (Scheme 1B, Fig. S4 and S5†). Our repeated attempts to crystallize 1 were unsuccessful under various dry conditions. It underwent spontaneous aerial oxidation and structural rearrangements to afford a single crystal of [(Cu^IIL)(Cu^IIL H)(SO₃CF₃)₂]·CF₃SO₃·H₂O 2, which has adopted a dicopper(II) complex (Fig. 1, S21,†Table 1 and Table S1†). Interestingly, one of the pyridyl arms was protonated due to an acidic environment in the solution. For comparison, the copper(II) complex [(LCu^II)₂(SO₃CF₃)₂] 3 was synthesized by the reaction of a 1 : 1 ratio of [Cu(CF₃SO₃)₂] : L in acetonitrile. In fact, attempts were made to obtain a mononuclear complex; however, this always led to the formation of the dicopper(II) complex (Fig. 1, S21, Table S1† and Table 2). Complexes 2 and 3 were confirmed by HRESI-MS, EPR spectral, and single-crystal X-ray studies (Fig. S6 and S7†). However, a better understanding of the structural, spectral, and magnetic properties of 2 and 3 is needed, which is under progress.

Scheme 1 Synthesis of the tris-pyridine-based ligand (L) and its copper complexes.

Fig. 1 Molecular structures of 2 (left) and 3 (right). (POV-Ray figures drawn from a 50% probability factor of thermal ellipsoids). For clarity, the hydrogen atoms and other counter ions are omitted.

Table 1 Selected bond distances^a (Å) and bond angles (°) of complex 2

a Standard deviations are given in parenthesis.
2
Cu(1)–Cu(2)	2.959(13)	Cu(2)–N(2)	1.965(5)
Cu(1)–N(1)	2.026(6)	Cu(2)–O(1)	1.965(5)
Cu(1)–O(1)	1.944(5)	Cu(2)–O(2)	1.908(5)
Cu(1)–O(2)	1.927(5)	Cu(2)–O(4)	2.292(6)
Cu(1)–O(3)	2.272(5)	Cu(2)–N(5)	2.017(6)
Cu(1)–N(4)	1.982(6)
O(2)–Cu(1)–O(1)	79.19(8)	O(2)–Cu(2)–O(1)	79.1(2)
O(3)–Cu(1)–O(1)	87.9(2)	O(4)–Cu(2)–O(2)	91.4(2)
O(3)–Cu(1)–O(2)	99.7(2)	O(4)–Cu(2)–O(1)	102.0(2)
N(1)–Cu(1)–O(1)	81.5(2)	N(2)–Cu(2)–O(1)	91.8(2)
N(1)–Cu(1)–O(2)	150.3(2)	N(2)–Cu(2)–O(2)	169.5(2)
N(1)–Cu(1)–O(3)	101.9(3)	N(2)–Cu(2)–O(4)	95.6(2)
N(4)–Cu(1)–O(1)	170.8(2)	N(5)–Cu(2)–O(1)	150.2(2)
N(4)–Cu(1)–O(2)	91.6(2)	N(5)–Cu(2)–O(2)	81.7(2)
N(4)–Cu(1)–O(3)	94.6(2)	N(5)–Cu(2)–O(4)	101.0(2)
N(4)–Cu(1)–N(1)	106.6(2)	N(5)–Cu(2)–N(2)	104.6(3)

---

Table 2 Selected bond distances^a (Å) and bond angles (°) of complex 3

a Standard deviations are given in parenthesis.
3
Cu(I)–Cu(1)^1	2.971(18)	N(1)–Cu(1)–O(2)	92.6(3)
Cu(I)–O(2)	2.386(7)	O(1)^1–Cu(1)–O(2)	93.0(2)
Cu(1)–N(1)	1.981(7)	O(1)–Cu(1)–O(2)	104.5(3)
Cu(1)–O(1)	1.933(5)	O(1)^1–Cu(1)–N(1)	170.1(2)
Cu(1)–N(3)^1	2.008(7)	O(1)–Cu(1)–N(1)	92.8(2)
		N(3)^1–Cu(1)–O(2)	97.5(3)
		N(3)^1–Cu(1)–N(1)	105.6(3)
		N(3)^1–Cu(1)–O(1)	150.6(2)
		N(3)^1–Cu(1)–O(1)^1	81.8(2)
		S(1)–O(2)–Cu(1)	145.8(5)
		C(1)–N(1)–Cu(1)	121.2(5)
		C(5)–N(1)–Cu(1)	120.2(5)
		C(8)–O(1)–Cu(1)	126.6(4)
		C(8)–O(1)–Cu(1)^1	119.7(4)
		C(14)–N(3)–Cu(1)^1	113.7(5)
		C(18)–N(3)–Cu(1)^1	126.6(6)

---

Electronic spectral, redox, and structural properties

Complex 1 [Cu^I(L)(CH₃CN)]CF₃SO₃ showed π–π* transition at 383 nm (ε, 534 M⁻¹ cm⁻¹) in the electronic spectra, with no detectable d–d transitions (Fig. S8†). As anticipated for a d¹⁰ copper(I) center, the complex is EPR silent and showed ¹H NMR signals at 7.31 (d, 8H), 6.75 (d, 4H), and 1.20 (m, 4H). However, due to coordination, the chemical shift values are slightly shifted from the ligand values (Fig. S5†). On the other hand, complexes 2 and 3 showed π–π* transitions at 376–379 nm along with broad d–d transitions at 686 nm (2, ε, 87 M⁻¹ cm⁻¹) and 693 nm (3, ε, 159 M⁻¹ cm⁻¹), respectively (Fig. S8†). The EPR spectra of 2 and 3 were recorded in a CH₃CN/DMF mixture at 70 K and showed an anisotropic g-tensor. Their hyperfine splitting due to Cu (I = 3/2) is resolved in the parallel (g_∥) region (Fig. S9†). The EPR spectra of complex 2 showed g_∥ = 2.26 and g_⊥ = 2.06, which are slightly shifted for 3, g_∥ = 2.37 and g_⊥ = 2.05. Both the complexes exhibited g_∥ > g_⊥, suggesting the presence of an odd electron located at the d_x²−y² orbital. Complex 2 exhibits an f-value of 165 cm⁻¹, higher than the square-based geometries (105–135 cm⁻¹), indicating strong distortion in the geometry. Complex 3 showed a lower f-value of 125 cm⁻¹, corresponding to square-based geometry.^35,37 EPR parameters and the energies of d–d transitions were used to estimate the covalency of in-plane σ-bonds (α²), in-plane π-bonds (β²), and out-plane π-bonds (γ²).³⁸ Complexes 2 and 3 exhibited higher α² (0.818–0.833), β² (0.928–1.08), and γ² (1.124–1.256) values (Table S2†). The α² values indicate a slightly higher degree of in-plane σ-covalent bonding character in Cu–N bonds of 3 than that of 2; in general, the α² value is close to unity for ionic bonding, and this value exhibits a reciprocal correlation with covalent character. The orbital reduction factor K for the complexes 2 (K_∥, 0.898; K_⊥, 0.741) and 3 (K_∥, 0.757; K_⊥, 0.651) was calculated, showing that K_∥ > K_⊥. This reveals that the persistence of a significant amount of out-of-plane π-bonding. The pure σ-bonding K_∥ = K_⊥ and K_∥ < K_⊥ for π-bonding at in-plane have been described in the literature.^39,40 The Fermi contact hyperfine interaction term (K) was calculated as 0.323 and 0.363 for 2 and 3, indicating an almost identical degree of covalent bonding character in Cu–N as predicted from α² values.³⁷ Cyclic voltammetry of 3 exhibited two reduction peaks at −0.25 V and −0.53 V vs. Ag/Ag⁺, and their oxidation peaks are ill-defined, which are attributed to the reduction processes associated with each of the dicopper(II) centers (Fig. S10†).

The molecular structures of complexes 2 and 3 were determined by single-crystal X-ray studies. The unit cells of complexes 2 and 3 showed dicopper centres having distorted square pyramidal geometry (Fig. 1), with τ values for 2 (τ₁, 0.21; τ₂, 0.29) and 3 (τ₁ = τ₂ = 0.23).⁴¹ The slightly varied τ-values of 2 ensure a different electronic environment enforced by the protonation of the pyridyl arm. The variation in the Cu–O bond distances reveals the asymmetric oxygen bridging between Cu1 and Cu2 centres (1.927–1.944 Å). These average Cu–O bond distances are relatively shorter than the average Cu–N_Py bond distances (1.965–2.026 Å) due to the strong overlapping of d-orbital Cu centers with oxygen. Also, it showed almost identical Cu–O_otf bond distances; this anion coordination is labile. The Cu–Cu distances of complexes 2 (2.959) and 3 (2.971 Å) are longer than those of other reported systems (2.38–2.93 Å).^42–44 Interestingly, the crystal structure of 2 showed that one of the pyridine arms is protonated, which was displaced from the coordination sphere (Fig. 1). Similarly, 3 possesses dicopper(II) centres with symmetrical coordination environments. The average bond distance of Cu–N_py (1.994 Å) is relatively longer than the symmetrically bridged Cu–O (1.924 Å) bond distance. The Cu–O_Otf bond distance (2.386 Å) is longer than other bond distances and is highly labile. The Cu–Cu distance of 3 (2.971 Å) is found to be slightly longer than that of 2 (Tables 1 and 2).⁴⁵ The Cu–Cu distances in the range of <3.0 Å are speculated to facilitate C–H activation, and this is observed in the hydroxylase or monooxygenase enzymes.^16,46–48

Model oxygenase reaction and benzene hydroxylation

The model oxygenation reaction was carried out using benzene as a substrate, where complex 1 (0.1 mmol) and benzene (5 mmol) were reacted with excess dioxygen (O₂) in the presence of the sacrificial reducing agent Et₃N (1 : 50 : 100; complex : substrate : base) in acetonitrile. The reaction proceeded at 60 °C for 48 hours, yielding phenol as the sole product with a 7% yield, calculated based on the substrate used (Scheme 2 and Scheme S1-a†). Notably, the over-oxidation products, such as benzoquinones, or self-coupling products, such as biphenyls, were not detected, highlighting the selective hydroxylation of benzene to phenol (Fig. S11†). Increasing the temperature to 70 °C or extending the reaction time showed no improvement in phenol yield. Despite the lower phenol yield, it is interesting because intermolecular aromatic hydroxylation is rarely reported using dioxygen. Karlin and co-workers used several dicopper complexes for intramolecular hydroxylation reactions using O₂; however, direct benzene hydroxylation was rarely explored discretely.^4,29,30 Without Et₃N or a catalyst, the reaction did not yield phenol or other hydroxylation products. Therefore, Et₃N functions as a base and sacrificial reducing agent to stabilize the intermediates formed during the catalytic process. Toluene as a substrate with complex 1 under O₂ resulted in no aromatic hydroxylation; a trace of benzaldehyde was formed (>3%) at 60 °C. On the other hand, using 2 and 3 as catalysts showed no phenol formation or other products from benzene under identical conditions.

Scheme 2 Benzene hydroxylation by complex 1.

Moreover, complex 1 was investigated as a catalyst using H₂O₂ as an oxygen source to substantially enhance the yield of hydroxylation products while maintaining high selectivity for aromatic hydroxylation (Scheme 2). The reaction was optimized using 0.05 mmol of 1, 5 mmol of C₆H₆, 5 mmol of sacrificial reducing agent Et₃N, and 25 mmol of H₂O₂ (1 : 100 : 100 : 500; complex : substrate : base : oxidant) at 60 °C in 3 mL of dry acetonitrile. The formation of product phenol was confirmed and quantified by GC-MS/GC analysis (Fig. S11†). It showed a 19% phenol yield with <99% selectivity (TON, 19) over 8 hours (Table 3).

Table 3 Catalytic benzene hydroxylation^a by 1 and H₂O₂

Complex 1 (mmol)	T (°C)	Oxidant (mmol)	Time (h)	S^c (%)	Y^d (%)	TON^e
a Reaction conditions: benzene (5 mmol), dry Et3N (5 mmol) and hydrogen peroxide (30%) (25 mmol) in dry acetonitrile. b T = temperature. c S = selectivity. d Y = yield. e TON = turnover number. f Reaction conditions: benzene (5 mmol), Et3N (5 mmol) and excess O2 in wet acetonitrile.
0.05	RT	25	8	100	12	12
0.05	60	25	8	>99	19	19
0.05	RT	10	8	100	9	9
0.05	60	10	8	>99	13	13
0.05	0–5	25	8	—	Trace	—
0.025	60	25	8	>99	9	18
0.025	60	25	30	97	21	42
0.025	60	100	8	95	19	38
0.01	60	100	12	>99	14	70
0.001	60	1000	48	>99	7	350
0.0001	60	Excess	120	>94	1.7	850
0.1	60	O 2	48	>99	7	3.5

---

Also, the reaction at 0–5 °C showed a trace amount of phenol formation. Meanwhile, the reaction that used phenol as a substrate exhibited no hydroxylation (catechol formation) or oxidation products. Under identical conditions, the substrate toluene (5 mmol) transformed into 9% hydroxylated products. The selectivity for aromatic hydroxylation dropped to 72% [60 : 39 : 1; o : p : m (cresols)]. Also, it showed the formation of benzoquinone derivatives (9% selectivity) and benzaldehyde (19% selectivity) as by-products (Scheme S1†). However, this result clearly reveals that aromatic hydroxylation is preferred over aliphatic hydroxylation. Similarly, using anisole as a substrate showed a 14% yield of mixed products of methoxyphenol with the selectivity of ortho, 55%; meta, 8%, and para, 37% (Scheme S2†). The endurance test was performed with 1 (0.1 μmol), Et₃N (10 μmol), H₂O₂ (60 mmol), and benzene (5 mmol), which yielded 1.7% of phenol over 120 hours with a maximum TON of 850 without the formation of by-products. Further increase in the reaction time did not improve the yield.

Mechanism of benzene hydroxylation

The kinetic isotopic effect (KIE) experiment was examined using a 1 : 1 ratio of C₆H₆ : C₆D₆ in H₂O₂ and was calculated at 25 °C, (k_H/k_D = 1.03) and 60 °C, (k_H/k_D = 1.09) from the ratio of C₆H₅OH : C₆D₅OH (Fig. S12c†). It is close to the values (0.9–1.0) reported for intramolecular aromatic hydroxylation. In fact, the KIE value of ∼1.0 clearly supports the involvement of metal-bound oxygen species as the key intermediate for the aromatic hydroxylation via an electrophilic aromatic substitution reaction.^32,49 Thus, this data excludes the involvement of the Fenton-type (˙OH) reaction pathway,⁵⁰ where the KIE is reported as 1.7–1.8 or a copper-insertion type reaction.^49,51 The KIE experiment with excess dioxygen was performed in a 1 : 1 ratio of C₆D₆ : C₆H₆ under identical reaction conditions at 25 °C and 60 °C. The KIE value (1.01–1.09) eliminates the existence of the metal insertion C–H bond activation intermediate and proposes metal-oxo/oxyl/hydroproxo species involvement in this present study.^49,51

Isotopic labelling experiments were performed using H₂¹⁸O₂ (90% atom purity), which showed 95.3% incorporation of ¹⁸O into phenol (Fig. S12a†). However, using H₂¹⁸O and H₂O₂ showed almost no ¹⁸O-atom incorporation into the phenol product (Fig. S12b†). These results suggest that utilizing H₂O₂ is crucial for transforming benzene into phenol, and it is the sole source of oxygen.⁵² The reaction of complex 1 (1 × 10⁻³ M) with H₂O₂ (5 equivalents) and Et₃N (2 equivalents) in CH₃CN under an N₂ atmosphere at −20 °C and 25 °C resulted in the formation of an O (π*_σ) → Cu²⁺ ligand-to-metal charge transfer (LMCT) transition at 360 nm, with the observed rate constant of (3.70 ± 0.01) × 10⁻² s⁻¹ and a lifetime (τ) of 27.02 s. Moreover, a weak d–d transition was identified at 689 nm at −20 °C, while its formation at 25 °C proceeded at a relatively slower rate (Fig. 2, S14, S15†). The low intensity of this signal is consistent with the high reactivity and instability of the Cu(II)–peroxo species, which are prone to rapid decomposition shortly after formation under the specified conditions. The identification of the O₂/H₂O₂ reaction with Cu(I)-complexes in the reaction medium presents a challenge, as observed in our system and similarly reported by other research groups.³³ Notably, the spectral features observed in this study closely resemble those attributed to the [(L)Cu^II–OOH]⁺ or [(L)Cu^II–O–O–Cu^II(L)]²⁺ species reported in the literature.^35,53,54 Upon treating benzene (50 equivalents) with the in situ-generated Cu(II)–peroxo species in acetonitrile at −20 °C, a gradual diminution of the copper–oxygen intermediate was observed. This decay is faster than its corresponding formation. The decay of the species is accompanied by a decrease in the ligand-to-metal charge transfer (LMCT) band at 360 nm, which indicates that direct transfer oxygen atom from [(L)Cu^II–OOH]⁺ or via [(L)Cu^II–O–O–Cu^II(L)]²⁺ to benzene (Fig. 4). The GCMS analysis of the reaction mixture showed that phenol was selectively formed. Similarly, exposure of complex 1 (1 × 10⁻³ M) to dioxygen (O₂) and Et₃N (2 equivalents) at −20 °C resulted in a comparable LMCT transition at 360 nm (743 M⁻¹ cm⁻¹) and a very weak d–d transition at 670 nm (26 M⁻¹ cm⁻¹) (Fig. S13†).^54–56 Thus, it suggests that [(L)Cu^II–OOH]⁺ or its dimeric derivative [(L)Cu^II–O–O–Cu^II(L)]²⁺ is likely to be the key intermediate for aromatic hydroxylation, as reported previously.^34–36

Fig. 2 Electronic spectral changes for complex 1 (1 × 10⁻⁴ M) upon reaction with 2 equivalents of Et₃N and 5 equivalents of H₂O₂ (30%) at −20 °C in dry acetonitrile. Inset: plot of time vs. absorbance (370 nm).

Further characterization of the Cu(II)–peroxo species generated with H₂O₂ was conducted using ¹H NMR spectroscopy, which exhibited paramagnetically shifted broad signals at 11.94 (s, 2H), 10.76 (s, 2H), 8.61 (s, 4H), 7.46 (s, 3H), 6.76 (s, 2H), and 1.26 (m, 4H) (Fig. S16†). These shifts are attributed to the oxidation of Cu(I) to Cu(II). The formation of the [(L)Cu^II–OOH]⁺ species was further confirmed by ESI-MS analysis, which displayed a molecular ion peak at m/z 387.0, in agreement with the calculated value (Fig. S17†). Isotopic labeling studies using H₂¹⁸O₂ (90% atom purity) demonstrated a four-unit mass shift (m/z 391.0) in CH₃CN, corresponding to the formation of [(L)Cu^II(¹⁸O–¹⁸OH)]⁺ (Fig. S17 and S18†). However, solution IR spectroscopy of [(L)Cu^II–OOH]⁺ revealed vibrational bands at 890 cm⁻¹ and 648 cm⁻¹, which are assigned to O–O and Cu–O stretching frequencies, respectively.^35,56 Interestingly, replacing H₂O₂ with O₂ resulted in a shift of the O–O stretching frequency to 862 cm⁻¹ (Fig. 3).^4,30 It is known that Cu(II)–peroxo species typically exhibit O–O stretching frequencies within the range of 800–930 cm⁻¹.^{4,30,35,56,57} The observed spectral data suggest that the species formed by the reaction of H₂O₂ may correspond to [(L)Cu^II–OOH]⁺, whereas the reaction with O₂ generates a peroxo-bridged species, [(L)Cu^II–O–O–Cu^IIL]²⁺, with an O–O stretching frequency of 862 cm⁻¹. These vibrational spectral features are consistent with previously reported Cu(II)–oxygen species, both spectroscopically and structurally.^35,58

Fig. 3 Vibrational spectra of Cu(II)–peroxo species (blue) generated from 1 (1 × 10⁻⁴ M) by adding 10 equivalents of H₂O₂/Et₃N (2 equivalents) in dry CH₃CN and by adding O₂/Et₃N (2 equiv.) in CH₃CN (red). Calculated vibrational spectra of [(L)Cu^II–OOH]⁺ using TD-DFT (black).

Fig. 4 Electronic spectral changes of the [(L)Cu^II–OOH]⁺ reaction with 50 equivalents of benzene at −20 °C in dry acetonitrile.

Using the radical trapping reagents TEMPO and DMPO at 25 °C and 60 °C resulted in no perceptible changes in phenol yield, eliminating the possibility of a radical-type chain mechanism. On the other hand, it is known that TEMPO can also be oxidized under these conditions due to the presence of aromatic C–H bonds.^34,35,51 The oxidation of the substrate cis-1,2-dimethylcyclohexane was carried out under optimized reaction conditions, which showed 4.2% of the cis-hydroxylated product with a trace of the trans product (0.7%). This stereo-retention of C–H oxidation reactions indicates that Fenton-type species are unlikely to be involved,⁵⁹ which supports our predictions by KIE studies. The benzene hydroxylation reaction in the CH₃CN–CCl₄ (5 : 1) solvent mixture exhibited the formation of a trace amount of chlorobenzene as a by-product, along with phenol (94% selectivity) as the primary product. This result suggests that CCl₄ does not affect the selectivity of phenol and eliminates the Fenton-type reaction pathway.⁶⁰ Cyclohexane as the substrate and CCl₄ as the radical trapping agent showed 96% selectivity for oxidation products of cyclohexanol and cyclohexanone formation with 4% of cyclohexyl chloride, which again supports the exclusion of the radical-type reaction pathway.³⁵ Using cyclohexanal as the substrate, no deformylation reaction was observed under identical conditions, and the substrate was completely recovered. This result suggests that the aldehyde deformylation reaction via nucleophilic attack is not facilitated by the intermediate Cu(II)–peroxo species. The reaction of thioanisole and triphenylphosphine showed the formation of methyl phenyl sulfoxide (56%) and triphenylphosphine oxide (29%), respectively, at 25 °C over 3 h. Thus, this result suggests the electrophilic nature of the key intermediate [(L)Cu^II–OOH]⁺ or [(L)Cu^II–O–O–Cu^II(L)]²⁺. Similarly, benzene-d₁ was employed as a substrate under identical reaction conditions to understand the mechanism better, and no product formation was observed via the NIH shift.⁶⁰ Therefore, it is evident that the initial formation of the Cu(II)–peroxo intermediate may either undergo direct electrophilic hydroxylation or be converted into highly reactive copper–oxyl (Cu–O) species. This Cu–O species possibly reacts with the benzene leading to the formation of phenol through an electrophilic mechanism (Scheme 3). Our repeated attempts to isolate single crystals of the [(L)Cu^II–OOH]⁺ or [(L)Cu^II–O–O–Cu^II(L)]²⁺ species were unsuccessful under various conditions.

Scheme 3 The proposed mechanism for benzene hydroxylation by complex 1.

However, the geometry of intermediate [(L)Cu^II–OOH]⁺ species was optimized using ORCA 4.0 (Fig. 5).⁶¹ The calculations were performed using B3LYP and def2-TZVP, def2/J basis sets.^62,63 The calculated bond distances of Cu–O (1.830 Å) and O–O (1.383 Å) are close to those of previously reported intermediates (Cu–O, 1.888 Å; O–O, 1.460 Å).^35,55 The calculated electronic spectra of [(L)Cu^IIOOH]⁺ by TDDFT showed O (π*_σ) → Cu²⁺ LMCT at 429 nm (Fig. S19†), which is closer to the experimental value (360 nm). Furthermore, this preliminary DFT optimization indicates that the coordination of the remaining pyridine pendant is likely to destabilize the intermediate Cu^II–OOH species. The calculated vibration spectra showed O–O and Cu–O stretching frequencies at 829 cm⁻¹ and 606 cm⁻¹, which are comparable to our experimental data (Fig. 3).³⁵

Fig. 5 Optimized structure of [(L)Cu^II–OOH]⁺ using the hybrid density functional B3LYP and a double-zeta quality split valence basic set with polarization functions (def2-SVP), and the geometry was optimized using longer triple-zeta basis sets (def2-TZVP) for all atoms.

Conclusions

A new tris-pyridine-based mononuclear copper complex 1 was synthesized and characterized as a relevant model for copper monooxygenase. It afforded a dicopper(II) complex 2via typical aerial oxidation with the protonation of one of the pyridyl arms of the ligand. Complex 3 was generated by a separate reaction, and ligand coordination was observed without protonating the pyridyl arms. Single-crystal XRD analysis of dicopper complexes 2 and 3 confirmed the above observation. Both the dicopper complexes showed distorted square pyramidal geometry, and their Cu–Cu distances in both cases were almost identical (2.9 Å). Benzene hydroxylation was performed using copper complexes; copper(I) complex 1 showed the conversion of benzene to phenol (7%) as the sole product at 60 °C under an O₂ atmosphere over 48 hours. The phenol yield increased to 19% using H₂O₂ as the oxidizing agent within 8 hours. However, dicopper complexes showed no conversion under O₂ and relatively poor phenol conversion with H₂O₂ (4–6%). The electronic, vibrational, and mass spectroscopy results support that the hydroxylation reaction mechanism proceeds via Cu(II)–peroxo species. DFT and TD-DFT calculation results are comparable to the experimental data. The KIE investigations further support the involvement of Cu–oxygen species. The structural, spectral, and magnetic properties of 2 and 3 are in progress for a better understanding of them.

Experimental section

Materials

The chemicals di-2-pyridylketone, 2-ethylpyridine, n-BuLi in cyclohexane [Cu^I(CH₃CN)₄]CF₃SO₄, [Cu^II(CF₃SO₄)₂], tetrabutylammonium hexafluorophosphate, benzene, chloroform-d, benzene-d₆, chlorobenzene, nitrobenzene, and toluene were purchased from Sigma-Aldrich and H₂¹⁸O₂ (90% ¹⁸O atom) was purchased from Icon Isotope. Anhydrous acetonitrile, dichloromethane, and hydrogen peroxide (30%) from Merck, India. Dry methanol and phenol were purchased from SD Fine Chemicals.

Experimental conditions and physical measurements

All workup procedures were conducted under ambient conditions. All NMR spectra were recorded on a Bruker 300 MHz spectrometer. Chemical shift values are given in parts per million (ppm). Electrospray ionization (ESI-MS) was measured on a Thermo LC-MS instrument. UV-Vis spectra were recorded on an Agilent 8453 spectrometer with a cooling unit by Unisoku (Osaka, Japan). Cyclic Voltammetry (CV) was performed using a three-electrode cell configuration. A platinum sphere (acetonitrile medium), glassy carbon (non-aqueous), platinum wire, and Ag(s)/Ag⁺ were used as the working, auxiliary, and reference electrodes, respectively. The tetrabutylammonium hexafluorophosphate was used as a supporting electrolyte in acetonitrile. GC-MS and GC analyses were performed on an Agilent 5977E GCMSD using an HP-5 MS ultra-inert (30 m × 250 μm × 0.25 μm) capillary column.

Single-crystal X-ray structure analysis

The experiment was performed on an Agilent Technologies Supernova-E CCD diffractometer. The single crystals of 2 and 3 with suitable sizes were selected from the mother liquor, immersed in paraffin oil, and then mounted on the tip of glass fiber. The structures were solved using the direct methods of the SHELXS-2013 program. Refinement and all further calculations were carried out using SHELXL-2013. The H atoms were included in calculated positions and treated as riding atoms using the SHELXL default parameters. The non-H atoms were refined anisotropically, using weighted full-matrix least-squares on F². CCDC 1962712 and 1962738† contain the supplementary crystallographic data for this paper.

Synthesis of the ligand

The ligand (L) was synthesized using the following procedure: a solution of n-BuLi (7 mL, 14.0 mmol of 2 M in hexane) was added dropwise via a syringe over 15 minutes to 2-ethylpyridine (1.6 mL, 12.0 mmol), which was pre-cooled to −78 °C in dry THF (40 mL) under an argon atmosphere. Then, the mixture was stirred vigorously at a temperature below −70 °C. After 30 minutes, the solution turned deep red, and di(2-pyridyl)ketone (1.1 g, 6 mmol) in THF (10 mL) was added while maintaining the temperature below −65 °C. The solution was further stirred for another 3 hours. Finally, the reaction was quenched with water at 0 °C, and the organic layer was extracted using ethyl acetate. The solvent was removed using a rotary evaporator to form a solid with a yellow oil compound. Off-white crystalline solids were obtained by passing through silica column chromatography (7 : 3, hexane : ethyl acetate) with a yield of 69% (1.2 g). HRESI-MS (m/z): 290.14 [M − 1]⁻. ¹H NMR (300 MHz, CDCl₃), δ in ppm, 8.63 (d, 1H), 8.36 (d, 1H), 8.29 (d, 1H), 8.01 (d, 1H), 7.79 (d, 1H), 7.68 (t, 1H), 7.45 (q, 2H), 7.33 (d, 1H), 7.14 (d, 1H), 6.96–6.88 (dt, 2H), 4.78 (q, 1H) and 1.19 (d, 3H). ¹³C NMR (CDCl₃), 164.41, 157.66, 148.66, 126.57, 125.31, 63.61, 60.24, 58.01, 54.48, 27.48, 13.61, 11.31 ppm. Anal. calcd for C₁₈H₁₇N₃O: C, 74.20; H, 5.88; N, 14.42; found: C, 74.19; H, 5.93; N, 14.38.

Synthesis of copper complexes

To a toluene/acetone (1.5 mL/0.5 mL) solution of [Cu^I(CH₃CN)₄]CF₃SO₃ (0.0375 g, 0.1 mmol), L1 of (0.029 g, 0.1 mmol) in a toluene/acetone (1.8 mL/0.2 mL) was added and stirred for 2 hours under an Ar atm. Then, 15 mL of degassed diethyl ether was added and stirred for an hour. A yellowish-white precipitate settled down, corresponding to copper(I) complex 1 (0.04 g, yield 74%). This complex is highly air-sensitive.

[Cu^I(L)(CH₃CN)]CF₃SO₃ (1). HRESI-MS (m/z): 354.0671 [M − CH₃CN]⁺; ¹H NMR (300 MHz, CDCl₃), 7.31 (d, 8H), 6.75 (d, 4H), 1.20 (m, 4H). Anal. calcd for C₂₁H₂₀CuF₃N₄O₄S: C, 46.28; H, 3.70; N, 10.28; found: C, 46.30; H, 3.74; N, 10.25.

[(Cu^IIL)(Cu^IIL H)(SO₃CF₃)₂]·CF₃SO₃·H₂O (2). Furthermore, complex 1 was dissolved in CH₃CN/diethyl ether and maintained for recrystallization, leading to the formation of a copper(II) complex (2). HRESI-MS (m/z): 758.8621 [M − (CF₃SO₃)₃ + H₂O + CH₃OH]⁺. Anal. calcd for C₃₉H₃₃Cu₂F₉N₆O₁₁S₃: C, 40.52; H, 2.88; N, 7.27; found: C, 40.50; H, 2.90; N, 7.28.

[(LCu^II)₂(SO₃CF₃)₂] (3). To a Cu^II(CF₃SO₃)₂ (0.072 g, 0.2 mmol) in acetonitrile, the ligand (0.058 g, 0.2 mmol) in acetonitrile was added dropwise, and the solution was stirred for 2 hours. The formed violet-colored solution was kept for crystallization. The long-standing solution at room temperature afforded single crystals suitable for X-ray diffraction (XRD) analysis. Yield (0.10 g, 58%). HRESI-MS (m/z): 855.03 [M − CF₃SO₃]⁺. Anal. calcd for C₃₈H₃₂Cu₂F₆N₆O₈S₂: C, 45.37; H, 3.21; N, 8.35; found: C, 45.39; H, 3.22; N, 8.31.

General procedure for the catalytic hydroxylation reaction

A catalytic amount of complex 1 (0.05 mmol) was dissolved in dry acetonitrile (3.0 mL). Triethylamine (5.0 mmol) and benzene (5.0 mmol) were introduced slowly into the solution. Under O₂, the reaction mixture was stirred at 60 °C/25 °C for 48 hours. After cooling to room temperature, the catalyst or trimethylamine N-oxide was removed by passing the mixture through a silica gel column. Then, the solution was analyzed and quantified by GC-MS/GC using a known concentration of nitrobenzene, which was added as an internal standard. The conversion and yield of the reaction were determined using GC-MS/GC. Instead of O₂, 30% hydrogen peroxide (2.5 mL, 25 mmol) was used as an oxygen source for a few catalytic reactions. The control experiment without a catalyst showed no benzene oxidation products under identical experimental conditions. Here, the yields were calculated based on the starting substrates.

Determination of the kinetic deuterium isotope effect

Benzene (0.25 mmol), benzene-d₆ (0.25 mmol), and triethylamine (0.25 μmol) were added to a solution of 1 (0.025 mmol) in dry acetonitrile (3 mL). Aqueous hydrogen peroxide (30%) (5 mmol, 0.50 mL) was then added slowly under constant stirring. The mixture was stirred for 24 hours at 60 °C. After cooling to room temperature, the catalyst was removed by passing the mixture through a silica gel column, then the solution was analyzed, and the product distribution was determined by GC-MS/GC. The kinetic isotope effect (KIE) value for phenol production was calculated using the formula: (moles of phenol)/(moles of phenol-d₅).

Catalytic benzene hydroxylation reaction with H₂¹⁸O₂

Complex 1 (2.5 μmol) was dissolved in dry acetonitrile and benzene (0.12 mmol). Then, triethylamine (0.12 mmol) and H₂¹⁸O₂ (5 equivalents) were slowly added to the reaction mixture. The mixture was stirred for 10 hours at 60 °C. After cooling to room temperature, the catalyst was removed by passing the mixture through a silica gel column, and then the solution was analyzed and quantified by GC-MS/GC.

Catalytic benzene hydroxylation reaction with H₂¹⁸O

Complex 1 (2.5 μmol) was dissolved in H₂¹⁸O/acetonitrile (0.1 mL/1 mL). Triethylamine (0.12 mmol), benzene (0.12 mmol) and aqueous hydrogen peroxide (0.5 mL, 5 mmol) were slowly introduced into the solution. The reaction mixture was stirred at 60 °C for 8 hours. After cooling to room temperature, the catalyst was removed by passing the mixture through a silica gel column, and then the solution was analyzed and quantified by GC-MS/GC.

Catalytic hydroxylation of substituted benzene

The most active catalyst 1 (0.05 mmol) was dissolved in dry acetonitrile (3.0 mL). Triethylamine (5.0 mmol), the substrate (5.0 mmol), and aqueous hydrogen peroxide (2.5 mL, 25 mmol) were slowly added to the solution. The reaction mixture was stirred at 60 °C for 8 hours. The same procedure was employed for the reaction with O₂, where 1 was dissolved in acetonitrile. After cooling to room temperature, the catalyst was removed by passing the mixture through a silica gel column, and the products were analyzed by GC-MS/GC.

Kinetic measurements

Kinetic experiments for the formation of Cu(II)–peroxo species were carried out in a 1 cm path length UV-visible cell on a Unisoku thermostated cell holder, designed for low-temperature measurements. The deaerated solution of the copper complexes (1 × 10⁻⁴ M) in the cell was maintained at the desired temperature for several minutes, and the oxidant (aqueous 30% H₂O₂ (10 equivalents)) was added via a microsyringe in acetonitrile. The formation of the intermediates was determined by time-dependent UV-visible spectroscopy.

DFT methods

The geometry optimizations were performed using density functional theory (DFT) methods. DFT calculations were carried out using the hybrid density functional B3LYP and a double-zeta quality split valence basic set with polarization functions (def2-SVP),³⁵ and the geometry was optimized using longer triple-zeta basis sets (def2-TZVP)³⁵ for all atoms, and TD-DFT calculations were also performed using B3LYP and basic set def2-TZVP³⁵ for all atoms. All DFT calculations were carried out using the Orca 4.0 program.⁶¹

Data availability

Supplementary data associated with this article can be found in the ESI† on the journal homepage.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

We acknowledge the Science and Engineering Research Board (SERB), New Delhi (CRG/2020/003016), for funding, and RR thanks the CSIR, New Delhi, for the senior research fellowship (09/201(0426)/2019-EMR-I).

References

1. N. Herron and C. A. Tolman, J. Am. Chem. Soc., 1987, 109, 2837.
2. T. Kitano, Y. Kuroda, M. Mori, S. Ito, K. Sasaki and M. Nitta, J. Chem. Soc., Perkin Trans. 2, 1993, 981.
3. Y. Liu, K. Murata and M. Inaba, J. Mol. Catal. A: Chem., 2006, 256, 247.
4. E. E. Chufan, S. C. Puiu and K. D. Karlin, Acc. Chem. Res., 2007, 40, 563.
5. H. Hock and S. Lang, Ber. Dtsch. Chem. Ges. A, B, 1944, 77, 257.
6. D. Bianchi, R. D'Aloisio, R. Bortolo and M. Ricci, Appl. Catal., A, 2007, 327, 295.
7. X. Q. Wang, J. P. Wu, M. W. Zhao, Y. F. Lv, G. Y. Li and C. W. Hu, J. Phys. Chem. C, 2009, 113, 14270.
8. J.-F. Bartoli, V. Mouries-Mansuy, K. Le Barch-Ozette, M. Palacio, P. Battioni and D. Mansuy, Chem. Commun., 2000, 827.
9. P. Roy, K. Dhara, M. Manassero and P. Banerjee, Eur. J. Inorg. Chem., 2008, 28, 4404.
10. P. K. Tandon, R. Baboo, A. K. Singh, G. Purwar and M. Purwar, Appl. Organomet. Chem., 2005, 19, 1079.
11. A. Marsella, S. Agapakis, F. Pinna and G. Strukul, Organometallics, 1992, 11, 3578.
12. L. V. Pirutko, V. S. Chernyavsky, A. K. Uriarte and G. I. Panov, Appl. Catal., A, 2002, 227, 143.
13. A. Costine, T. O'Sullivan and B. K. Hodnett, Catal. Today, 2005, 99, 199.
14. K. Weissermehl and H. J. Arpe, Industrielle Organische Chemie, VCH, Weinheim, 1997.
15. C. Citek, B. L. Lin, T. E. Phelps, E. C. Wasinger and T. D. Stack, J. Am. Chem. Soc., 2014, 136, 14405.
16. R. Balasubramanian, S. M. Smith, S. Rawat, L. A. Yatsunyk, T. L. Stemmler and A. C. Rosenzweig, Nature, 2010, 465, 115.
17. W. Kaim and J. Rall, Angew. Chem., Int. Ed., 1996, 35, 43.
18. J. P. Klinman, Chem. Rev., 1996, 96, 2541.
19. J. P. Klinman, J. Biol. Chem., 2006, 281, 3013.
20. P. Chen and E. I. Solomon, Proc. Natl. Acad. Sci. U. S. A., 2004, 101, 13105.
21. S. Itoh and S. Fukuzumi, Acc. Chem. Res., 2007, 40, 592.
22. E. I. Solomon, D. E. Heppner, E. M. Johnston, J. W. Ginsbach, J. Cirera, M. Qayyum, M. T. Kieber-Emmons, C. H. Kjaergaard, R. G. Hadt and L. Tian, Chem. Rev., 2014, 114, 3659.
23. J. S. Woertink, P. J. Smeets, M. H. Groothaert, M. A. Vance, B. F. Sels, R. A. Schoonheydt and E. I. Solomon, Proc. Natl. Acad. Sci. U. S. A., 2009, 106, 18908.
24. E. I. Solomon, Inorg. Chem., 2016, 55, 6364.
25. J. M. Bollinger Jr., Nature, 2010, 465, 40.
26. S. Schindler, Eur. J. Inorg. Chem., 2000, 11, 2311.
27. Y. R. Luo, Comprehensive Handbook of Chemical Bond Energies, 1999, p. 1.
28. P. Zhai, L. Wang, C. Liu and S. Zhang, Chem. Eng. J., 2005, 111, 1.
29. M. Becker, S. Schindler, K. D. Karlin, T. A. Kaden, S. Kaderli, T. Palanché and A. D. Zuberbühler, Inorg. Chem., 1999, 38, 1989.
30. H. R. Lucas, L. Li, A. A. Sarjeant, M. A. Vance, E. I. Solomon and K. D. Karlin, J. Am. Chem. Soc., 2009, 131, 3230.
31. A. Conde, M. M. Diaz-Requejo and P. J. Perez, Chem. Commun., 2011, 47, 8154.
32. L. Vilella, A. Conde, D. Balcells, M. M. Diaz-Requejo, A. Lledos and P. J. Perez, Chem. Sci., 2017, 8, 8373.
33. E. Borrego, L. Tiessler-Sala, J. J. Lazaro, A. Caballero, P. J. Perez and A. Lledos, Organometallics, 2022, 41, 1892.
34. T. Tsuji, A. A. Zaoputra, Y. Hitomi, K. Mieda, T. Ogura, Y. Shiota, K. Yoshizawa, H. Sato and M. Kodera, Angew. Chem., Int. Ed., 2017, 56, 7779.
35. S. Muthuramalingam, K. Anandababu, M. Velusamy and R. Mayilmurugan, Inorg. Chem., 2020, 59, 5918.
36. (a) S. Kumari, S. Muthuramalingam, A. K. Dhara, U. P. Singh, R. Mayilmurugan and K. Ghosh, Dalton Trans., 2020, 49, 13829; (b) R. Ramasubramanian, K. Anandababu, N. C. Mösch-Zanetti, F. Belaj and R. Mayilmurugan, Dalton Trans., 2019, 48, 14326; (c) K. Anandababu, R. Ramasubramanian, H. Wadepohl, P. Comba, N. Britto, M. Jaccob and R. Mayilmurugan, Chem. – Eur. J., 2019, 25, 9540.
37. C. H. Kjaergaard, S. M. Jones, S. Gounel, N. Mano and E. I. Solomon, J. Am. Chem. Soc., 2015, 137, 8783.
38. S. H. Laurie, T. Lund and J. B. Raynor, J. Chem. Soc., Dalton Trans., 1975, 14, 1389.
39. B. J. Hathaway and A. Tomlinso, Coord. Chem. Rev., 1970, 5, 1.
40. U. Sakaguchi and A. W. Addison, J. Chem. Soc., Dalton Trans., 1979, 4, 600.
41. A. W. Addison, T. N. Rao, J. Reedijk, J. van Rijn, G. C. Verschoor and J. Chem, Soc., Dalton Trans., 1984, 7, 1349.
42. T. Tsukihara, H. Aoyama, E. Yamashita, T. Tomizaki, H. Yamaguchi, K. Shinzawa-Itoh, R. Nakashima, R. Yaono and S. Yoshikawa, Science, 1995, 269, 1069.
43. S. Iwata, C. Ostermeier, B. Ludwig and H. Michel, Nature, 1995, 376, 660.
44. M. Wilmanns, P. Lappalainen, M. Kelly, E. Sauer-Eriksson and M. Saraste, Proc. Natl. Acad. Sci. U. S. A., 1995, 92, 11955.
45. R. P. Houser, V. G. Young and W. B. Tolman, J. Am. Chem. Soc., 1996, 118, 2101.
46. J. M. Bollinger Jr, Nature, 2010, 465, 40.
47. C. W. Koo, F. J. Tucci, Y. He and A. C. Rosenzweig, Science, 2022, 375, 1287.
48. V. C. Wang, S. Maji, P. P. Chen, H. K. Lee, S. S. Yu and S. I. Chan, Chem. Rev., 2017, 117, 8574.
49. (a) P. L. Holland, K. R. Rodgers and W. B. Tolman, Angew. Chem., Int. Ed., 1999, 38, 1139; (b) S. P. de Visser, K. Oh, A.-R. Han and W. Nam, Inorg. Chem., 2007, 46, 4632; (c) P. Heim, R. Gericke, G. Spedalotto, M. Lovisari, E. R. Farquhar and A. R. McDonald, Dalton Trans., 2023, 52, 2663.
50. R. Augusti, A. O. Dias, L. L. Rocha and R. M. Lago, J. Phys. Chem. A, 1998, 102, 10723.
51. (a) Q. Xing, H. Lv, C. Xia and F. Li, Chem. Commun., 2016, 52, 489; (b) M. Yamada, K. D. Karlin and S. Fukuzumi, Chem. Sci., 2016, 7, 2856; (c) Y. Yamada, C.-M. Teoh, Y. Toyoda and K. Tanaka, New J. Chem., 2022, 46, 955–958.
52. K. Ohkubo, T. Kobayashi and S. Fukuzumi, Angew. Chem., Int. Ed., 2011, 50, 8652.
53. B. Kim, D. Jeong and J. Cho, Chem. Commun., 2017, 53, 9328.
54. D. Maiti, H. R. Lucas, A. A. Sarjeant and K. D. Karlin, J. Am. Chem. Soc., 2007, 129, 6998.
55. B. Kim, D. Jeong, T. Ohta and J. Cho, Commun. Chem., 2019, 2, 81.
56. M. Kodera, T. Kita, I. Miura, N. Nakayama, T. Kawata, K. Kano and S. Hirota, J. Am. Chem. Soc., 2001, 123, 7715.
57. C. Wurtele, O. Sander, V. Lutz, T. Waitz, F. Tuczek and S. Schindler, J. Am. Chem. Soc., 2009, 131, 7544.
58. M. S. Reynolds and A. Butler, Inorg. Chem., 1996, 35, 2378.
59. L. Gomez, I. Garcia-Bosch, A. Company, J. Benet-Buchholz, A. Polo, X. Sala, X. Ribas and M. Costas, Angew. Chem., Int. Ed., 2009, 48, 5720.
60. (a) K.-B. Cho, X. Wu, Y.-M. Lee, Y. H. Kwon, S. Shaik and W. Nam, J. Am. Chem. Soc., 2012, 134, 20222; (b) J. T. Groves, J. Chem. Educ., 1985, 62, 928; (c) J. T. Groves and T. E. Nemo, J. Am. Chem. Soc., 1983, 105, 6243.
61. F. Neese, Comput. Mol. Sci., 2018, 8, e1327.
62. A. Schäfer, H. Horn and R. Ahlrichs, J. Chem. Phys., 1992, 97, 2571.
63. F. Weigend and R. Ahlrichs, Phys. Chem. Chem. Phys., 2005, 7, 3297.

---

Footnote

† Electronic supplementary information (ESI) available. CCDC 1962712 and 1962738. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d4dt02079k

---