Studies Relevant to the Functional Model of Mo-Cu CODH: In Situ Reactions of Cu(I)-L Complexes with Mo(VI) and Synthesis of Stable Structurally Characterized Heterotetranuclear MoVI2CuI2 Complex

In this study, we report the synthesis, characterization, and reactions of Cu(I) complexes of the general form Cu(L)(LigH2) (LigH2 = xanthene-based heterodinucleating ligand (E)-3-(((5-(bis(pyridin-2-ylmethyl)amino)-2,7-di-tert-butyl-9,9-dimethyl-9H-xanthen-4-yl)imino)methyl)benzene-1,2-diol); L = PMe3, PPh3, CN(2,6-Me2C6H3)). New complexes [Cu(PMe3)(LigH2)] and [CuCN(2,6-Me2C6H3)(LigH2)] were synthesized by treating [Cu(LigH2)](PF6) with trimethylphosphine and 2,6-dimethylphenyl isocyanide, respectively. These complexes were characterized by multinuclear NMR spectroscopy, IR spectroscopy, high-resolution mass spectrometry (HRMS), and X-ray crystallography. In contrast, attempted reactions of [Cu(LigH2)](PF6) with cyanide or styrene failed to produce isolable crystalline products. Next, the reactivity of these and previously synthesized Cu(I) phosphine and isocyanide complexes with molybdate was interrogated. IR (for isocyanide) and 31P NMR (for PPh3/PMe3) spectroscopy demonstrates the lack of oxidation reactivity. We also describe herein the first example of a structurally characterized multinuclear complex combining both Mo(VI) and Cu(I) metal ions within the same system. The heterobimetallic tetranuclear complex [Cu2Mo2O4(μ2-O)(Lig)2]·HOSiPh3 was obtained by the reaction of the silylated Mo(VI) precursor (Et4N)(MoO3(OSiPh3)) with LigH2, followed by the addition of [Cu(NCMe)4](PF6). This complex was characterized by NMR spectroscopy, high-resolution mass spectrometry, and X-ray crystallography.


Introduction
Molybdenum-copper carbon monoxide dehydrogenase (Mo-Cu CODH) is an airstable enzyme that catalyzes the oxidation of CO to CO 2 . Mo-Cu CODH is a key enzyme in the metabolism of aerobic carboxidotrophic bacteria (such as Oligotropha carboxidovorans), allowing it to use CO as a sole source of energy and carbon [1]. The oxidation of CO is very efficient, and the enzyme detoxifies a significant fraction of total CO from the atmosphere [2]. Thus, there is a longstanding interest in the design of structural and/or functional models which could (a) provide insight into the reaction mechanism, and (b) lead to efficient CO oxidation and potentially CO 2 reduction catalysts.
The oxidized active site of Mo-Cu CODH contains square-pyramidal Mo(VI) coordinated to pyranopterindithiolene, [3,4] apical and basal oxo groups, and a basal sulfido bridging to a low-coordinate Cu(I) atom, whose coordination is accomplished by cysteinate and weakly bound water at the resting state (Scheme 1) [5,6]. The reduced active site likely contains Mo(IV)-basal hydroxide in an otherwise similar disposition [7]. Spectroscopic studies suggest that the coordination of CO takes place at the Cu(I) center, replacing weakly bound water [8,9]. Computational studies generally agree with the initial coordination scopic studies suggest that the coordination of CO takes place at the Cu(I) center ing weakly bound water [8,9]. Computational studies generally agree with the in ordination of CO to the coordinatively unsaturated Cu(I) site. However, the sub routes differ between different mechanisms, suggesting bridging thiocarbonate or dioxide intermediates [10][11][12][13][14][15][16][17]. It is noted that the bridging thiocarbonate hypothes to a large extent on a high-resolution structure of the product of the reaction of CODH with alkyl isocyanide (CO analog/inhibitor), which undergoes oxidation sertion into the Cu-S bond to form bridging thiocarbamate [5]. Since the report of t resolution structure of the Mo-Cu CODH enzyme, several structural models of CODH have been reported; in all these models the metals were linked only by b sulfido (or oxo and sulfido) ligands [18][19][20][21][22][23][24]. Although selected closely related st models were able to rationalize structural, electronic, and spectroscopic feature active site, only a few have explored the reactivity with CO or CO-like substrates our previous work, we explored a different approach toward a more robust fu model of Mo-Cu CODH, which brings two metals together using xanthene-based dinucleating ligands [25]. Recently, we have shown that the heterodinucleating LigH2, which combines catecholate chelate for Mo(VI) and dipicolylamine for C lows for the formation of mononuclear Cu(I) and Mo(VI) complexes at the designa [26]. The combination of LigH2 with both metals forms metastable Mo(VI)-Cu(I) c (Et4N)[CuMoO3(Lig)] directly observable in solution by 1 H NMR and high-re mass spectrometry (HRMS). However, our attempts to isolate this complex in a so and/or study its reactivity were not successful, as it underwent facile decompos the present work, we explore a different approach toward a functional model, w volves the initial formation of well-defined Cu(I) complexes containing an oxidiza like (or related) ligand (CNR, PR3, CN -, olefin), followed by treatment with (Et4N) 2 We have also explored a different approach towards stable Mo(VI)-Cu(I) comp involves the use of a different Mo(VI) precursor (Et4N)[MoO3(OSiPh3)]. Herein w the first structurally characterized Mo(VI)-Cu(I) complex in our system.

Synthesis and Characterization of Cu(I)-LigH2 Complexes Featuring an Additional L
As described above, we have targeted complexes of the genera [Cu(L)(LigH2)](PF6), in which L is either directly related to the reactivity of Mo-Cu (L = CNR), or can undergo similar oxidation reactivity (CN − , PR3, olefin). To acce complexes, we have used previously reported Cu(I) complex [Cu(L)(LigH2)](PF6) Scheme 1. Resting site (oxidized) structure of Mo-Cu CODH, and its proposed reactivity with CO and CNR. (R = tert-butyl, [SC(R) = C(R)S] ligand = pyranopterindithiolene). Structures inside frames have been characterized by X-ray crystallography (X = NR).

Results and Discussion
2.1. Synthesis and Characterization of Cu(I)-LigH 2 Complexes Featuring an Additional L Ligand As described above, we have targeted complexes of the general form [Cu(L)(LigH 2 )] (PF 6 ), in which L is either directly related to the reactivity of Mo-Cu CODH (L = CNR), or can undergo similar oxidation reactivity (CN − , PR 3 , olefin). To access these complexes, we have used previously reported Cu(I) complex [Cu(L)(LigH 2 )](PF 6 ) (1(PF 6 )) (Scheme 2) as a precursor [24]. Complex 1(PF 6 ) demonstrates strong coordination of Cu(I) by two pyridine and the imine donors; relatively weak interaction with the central amine was observed. A reaction of complex 1(PF 6 ) with a monodentate ligand L is likely to release Cu(I) from the coordination to the imine, while restoring the H-bonding between the catechol proton and the imine [26,27].
Treatment of complex 1(PF 6 ) with 2,6-dimethylphenyl isocyanide CN(2,6-Me 2 C 6 H 3 ) produced [Cu(CN(2,6-Me 2 C 6 H 3 ))(LigH 2 )](PF 6 ) (2(PF 6 )) in 90% yield as yellow-orange crystals. 2(PF 6 ) was characterized by 1 H and 13 C NMR spectroscopy, FT-IR spectroscopy, HRMS, and X-ray crystallography. The coordination of CN(2,6-Me 2 C 6 H 3 ) to the metal is supported by the presence of the isocyanide methyl groups at 2.32 ppm in 1 H NMR spectrum. Notably, methylene protons (NCH 2 py) appear as a sharp singlet, suggesting weak coordination of the central amine donor. The IR spectroscopy demonstrates the C≡NAr stretch at 2137 cm −1 . This value suggests the coordination of an isocyanide to Cu(I) through mostly σ-donation [25,28,29]. HRMS demonstrates a peak at 848.3577 in the positive ion mode (expected m/z for 2 + is 848.3595), which agrees well with the expected isotopic distribution (see Supplementary Materials). The X-ray structure of 2(PF 6 ) ( Figure 1) confirms the overall composition and is consistent with the spectroscopic features. Coordination of the isocyanide to Cu(I) leads to the release of the metal and the rotation of the Cu(I)-dipicolylamine fragment away from the imine. Copper coordination is replaced by a strong hydrogen bond with one of the catechol-OH groups (OH-N(imine) 1.88 Å). The bond distance between Cu(I) and the central amine is relatively long (Cu1 N1 2.281(1) Å), while the bonds with the pyridine nitrogen are relatively short (2.0289 (14) and 2.0274(14) Å). A short bond (1.844(2) Å) is observed between Cu(I) and isocyanide as well.
(Scheme 2) as a precursor [24]. Complex 1(PF6) demonstrates strong coordination by two pyridine and the imine donors; relatively weak interaction with the centra was observed. A reaction of complex 1(PF6) with a monodentate ligand L is like lease Cu(I) from the coordination to the imine, while restoring the H-bonding betw catechol proton and the imine [26,27].
Treatment of complex 1(PF6) with 2,6-dimethylphenyl isocyanide CN(2,6-M produced [Cu(CN(2,6-Me2C6H3))(LigH2)](PF6) (2(PF6)) in 90% yield as yellow-oran tals. 2(PF6) was characterized by 1 H and 13 C NMR spectroscopy, FT-IR spectr HRMS, and X-ray crystallography. The coordination of CN (2, to the supported by the presence of the isocyanide methyl groups at 2.32 ppm in 1 H NM trum. Notably, methylene protons (NCH2py) appear as a sharp singlet, suggestin coordination of the central amine donor. The IR spectroscopy demonstrates the stretch at 2137 cm −1 . This value suggests the coordination of an isocyanide to Cu(I) mostly σ-donation [25,28,29]. HRMS demonstrates a peak at 848.3577 in the posi mode (expected m/z for 2 + is 848.3595), which agrees well with the expected isoto tribution (see Supplementary Materials). The X-ray structure of 2(PF6) (Figure 1) c the overall composition and is consistent with the spectroscopic features. Coordin the isocyanide to Cu(I) leads to the release of the metal and the rotation of th dipicolylamine fragment away from the imine. Copper coordination is replac strong hydrogen bond with one of the catechol-OH groups (OH-N(imine) 1.88 bond distance between Cu(I) and the central amine is relatively long (Cu1 N1 2.28 while the bonds with the pyridine nitrogen are relatively short (2.0289(14) and 2.0 Å). A short bond (1.844(2) Å) is observed between Cu(I) and isocyanide as well. We have also targeted copper-phosphine complexes (PR3 = PPh3 and PMe3). T thesis and structure of [Cu(PPh3)(LigH2)](PF6) (3(PF6)) habe been reported pre (Scheme 2) [26]. [Cu(PMe3)(LigH2)](PF6) (4(PF6)), featuring more reactive thylphosphine, was obtained by the treatment of 1(PF6) with one equivalent of PM complex was characterized by 1 H, 13 C, and 31 P NMR and HRMS. Our attempts to its solid-state structure were not successful; however, the spectroscopic data for 4 consistent with the spectroscopic data for 3(PF6), suggesting a similar structure. Bo pounds give rise to a catechol proton at a highly downfield chemical shift, 14.07 p 3(PF6) and 13.93 ppm for 4(PF6), suggesting strong (catechol)OH-N(imine) hy bonding. In both cases, methylene arms protons NCH2py appear as broad peaks, s ing weak but not negligible coordination of the central amine. 31   We have also targeted copper-phosphine complexes (PR 3 = PPh 3 and PMe 3 ). The synthesis and structure of [Cu(PPh 3 )(LigH 2 )](PF 6 ) (3(PF 6 )) habe been reported previously (Scheme 2) [26]. [Cu(PMe 3 )(LigH 2 )](PF 6 ) (4(PF 6 )), featuring more reactive trimethylphosphine, was obtained by the treatment of 1(PF 6 ) with one equivalent of PMe 3 . The complex was characterized by 1 H, 13 C, and 31 P NMR and HRMS. Our attempts to obtain its solidstate structure were not successful; however, the spectroscopic data for 4(PF 6 ) is consistent with the spectroscopic data for 3(PF 6 ), suggesting a similar structure. Both compounds give rise to a catechol proton at a highly downfield chemical shift, 14.07 ppm for 3(PF 6 ) and 13.93 ppm for 4(PF 6 ), suggesting strong (catechol)OH-N(imine) hydrogen bonding. In both cases, methylene arms protons NCH 2 py appear as broad peaks, suggesting weak but not negligible coordination of the central amine. 31  In addition to the synthesis of PR3 and isocyanide complexes, we have also attempted the synthesis of the mononuclear cyanide complex [Cu(CN)(LigH2)] and a mononuclear olefin (styrene CH2CHPh) complex. It was previously reported that the reaction of [CuCN]n with LigH2 formed a [Cu2(CN)2(LigH2)] complex featuring two coppers and two cyanides: one terminal and one bridging {Cu(µ2-CN)CuCN} [26]. To prevent the formation of the [CuCN]n chains (likely resulting from the [CuCN]n precursor), complex 1(PF6) was treated with the combination of NaCN and crown ether (18-crown-6). However, the reaction led to a relatively uninformative NMR spectrum containing broad peaks, which suggested product mixture or dynamic processes which could involve species of varying nuclearities. Our attempts to obtain a pure crystalline product from this reaction were not successful. Similarly, the reaction of 1(PF6) with styrene produced a broad NMR spectrum containing excess styrene. Our attempts to purify this compound by crystallization were not successful as well.

Reactions of Cu-L Complexes (L = CNR, PR3, Styrene) with [MoO4] 2−
The reactions of complexes 2(PF6), 3(PF6), and 4(PF6) with (Et4N)2[MoO4] were investigated by adding the equimolar amounts of the acetonitrile solution of molybdate into the stirring acetonitrile solution of the corresponding complex. These reactions were performed similarly to the previously described synthesis of metastable (Et4N)[CuMoO3(Lig)], [26] suggesting similar incorporation of {Mo VI O3} at the initial step of the reaction. Following the addition, the reaction was analyzed by NMR, IR, and UVvis spectroscopy. Treatment of 2(PF6) with one equivalent of (Et4N)2[MoO4] produced yellow-brown solution. The 1 H NMR spectrum indicated the formation of several new products. The UV-vis spectrum of the crude product does not appear to contain any d-d transitions anticipated for Cu(II), suggesting that no Cu(I) oxidation takes place. Although we were not able to separate the products, IR spectroscopy provides a convenient tool to specifically interrogate the oxidation of isocyanide to isocyanate. An isocyanate is expected to exhibit an intense absorption between 2200 cm −1 and 2300 cm −1 (RN=C=O stretch). However, the IR spectrum of the crude product demonstrates a single sharp resonance around 2129 cm −1 , consistent with the isocyanide stretch in aryl isocyanides (ArN≡C) [30]. As noted above, a stretching frequency of ~2130 cm −1 suggests either a free isocyanide, or an Scheme 2. Synthesis of Cu(I) precursors described in this paper.
In addition to the synthesis of PR 3 and isocyanide complexes, we have also attempted the synthesis of the mononuclear cyanide complex [Cu(CN)(LigH 2 )] and a mononuclear olefin (styrene CH 2 CHPh) complex. It was previously reported that the reaction of [CuCN] n with LigH 2 formed a [Cu 2 (CN) 2 (LigH 2 )] complex featuring two coppers and two cyanides: one terminal and one bridging {Cu(µ 2 -CN)CuCN} [26]. To prevent the formation of the [CuCN] n chains (likely resulting from the [CuCN] n precursor), complex 1(PF 6 ) was treated with the combination of NaCN and crown ether (18-crown-6). However, the reaction led to a relatively uninformative NMR spectrum containing broad peaks, which suggested product mixture or dynamic processes which could involve species of varying nuclearities. Our attempts to obtain a pure crystalline product from this reaction were not successful. Similarly, the reaction of 1(PF 6 ) with styrene produced a broad NMR spectrum containing excess styrene. Our attempts to purify this compound by crystallization were not successful as well.

Reactions of Cu-L Complexes (L = CNR, PR 3 , Styrene) with [MoO 4 ] 2−
The reactions of complexes 2(PF 6 ), 3(PF 6 ), and 4(PF 6 ) with (Et 4 N) 2 [MoO 4 ] were investigated by adding the equimolar amounts of the acetonitrile solution of molybdate into the stirring acetonitrile solution of the corresponding complex. These reactions were performed similarly to the previously described synthesis of metastable (Et 4 N)[CuMoO 3 (Lig)], Ref. [26] suggesting similar incorporation of {Mo VI O 3 } at the initial step of the reaction. Following the addition, the reaction was analyzed by NMR, IR, and UV-vis spectroscopy. Treatment of 2(PF 6 ) with one equivalent of (Et 4 N) 2 [MoO 4 ] produced yellow-brown solution. The 1 H NMR spectrum indicated the formation of several new products. The UV-vis spectrum of the crude product does not appear to contain any d-d transitions anticipated for Cu(II), suggesting that no Cu(I) oxidation takes place. Although we were not able to separate the products, IR spectroscopy provides a convenient tool to specifically interrogate the oxidation of isocyanide to isocyanate. An isocyanate is expected to exhibit an intense absorption between 2200 cm −1 and 2300 cm −1 (RN=C=O stretch). However, the IR spectrum of the crude product demonstrates a single sharp resonance around 2129 cm −1 , consistent with the isocyanide stretch in aryl isocyanides (ArN≡C) [30]. As noted above, a stretching frequency of~2130 cm −1 suggests either a free isocyanide, or an isocyanide coordinated mostly via σ-donation [28][29][30]. Such coordination would be expected for both Mo(VI) and Cu(I). No significant peaks above 2200 cm −1 were observed.
The reactions of the phosphine complexes 3(PF 6 ) and 4(PF 6 ) with molybdate were investigated similarly. Treatment of the cold yellow solution of 3(PF 6 ) and 4(PF 6 ) with molybdate resulted in a slight color change to brown. In both cases, 1 H NMR spectra of the products featured broad resonances. However, 31 P NMR was more informative in both cases, allowing us to focus specifically on the phosphine transformation. Thus, the 31 P NMR spectrum of the product of the reaction of 3(PF 6 ) with molybdate contains a single peak around −3 ppm (free PPh 3 is~−6 ppm, PPh 3 in 3(PF 6 )~3 ppm, P(=O)Ph 3 28 ppm) whereas the product of the reaction of 4(PF 6 ) with molybdate contains a peak around −46 ppm (free PMe 3 is~−62 ppm, PMe 3 in 4(PF 6 )~−49 ppm, P(=O)Me 3 36 ppm). The findings above are consistent with coordinated phosphines and suggest that no oxidation of phosphines to the corresponding phosphine oxides takes place.
The experiments above indicate the lack of in situ reactivity between Cu(I)-bound L (CNR, PR 3 ) and [Mo VI O 3 ] 2− , and that is despite previously observed in situ nucleophilic attack of [Mo VI O 3 ] 2− on the Cu(I)-bound imine [25]. We postulate that the lack of reactivity in the present system is due to the structural difference between the two heterodinucleating ligand systems ( Figure 2). The iminopyridine system led to a relatively accessible threecoordinate Cu(I)-iminopyridine which, due to the coordination to the (xanthene-) amine, was positioned in the vicinity of Mo(VI)-oxo ( Figure 2, left) [25]. In contrast, the coordination of the substrate to the dipicolylamine-coordinate Cu(I) leaves the metal four-coordinate, making it less accessible for attack. Furthermore, the coordination of the substrate L causes the displacement of the imine coordination which, combined with the more significant steric profile of dipicolylamine (vs. iminopyridine), leads to the anti-conformation of the Cu(I) fragment relative to the xanthene bridge (Figure 2, right). This factor is also likely to have a negative effect on reactivity. The reactions of the phosphine complexes 3(PF6) and 4(PF6) with molybdate were investigated similarly. Treatment of the cold yellow solution of 3(PF6) and 4(PF6) with molybdate resulted in a slight color change to brown. In both cases, 1 H NMR spectra of the products featured broad resonances. However, 31 P NMR was more informative in both cases, allowing us to focus specifically on the phosphine transformation. Thus, the 31 P NMR spectrum of the product of the reaction of 3(PF6) with molybdate contains a single peak around −3 ppm (free PPh3 is ~ −6 ppm, PPh3 in 3(PF6) ~ 3 ppm, P(=O)Ph3 ~28 ppm) whereas the product of the reaction of 4(PF6) with molybdate contains a peak around −46 ppm (free PMe3 is ~−62 ppm, PMe3 in 4(PF6) ~ −49 ppm, P(=O)Me3 36 ppm). The findings above are consistent with coordinated phosphines and suggest that no oxidation of phosphines to the corresponding phosphine oxides takes place.
The experiments above indicate the lack of in situ reactivity between Cu(I)-bound L (CNR, PR3) and [Mo VI O3] 2-, and that is despite previously observed in situ nucleophilic attack of [Mo VI O3] 2-on the Cu(I)-bound imine [25]. We postulate that the lack of reactivity in the present system is due to the structural difference between the two heterodinucleating ligand systems ( Figure 2). The iminopyridine system led to a relatively accessible three-coordinate Cu(I)-iminopyridine which, due to the coordination to the (xanthene-) amine, was positioned in the vicinity of Mo(VI)-oxo (Figure 2, left) [25]. In contrast, the coordination of the substrate to the dipicolylamine-coordinate Cu(I) leaves the metal fourcoordinate, making it less accessible for attack. Furthermore, the coordination of the substrate L causes the displacement of the imine coordination which, combined with the more significant steric profile of dipicolylamine (vs. iminopyridine), leads to the anti-conformation of the Cu(I) fragment relative to the xanthene bridge (Figure 2, right). This factor is also likely to have a negative effect on reactivity.

Synthesis and Characterization of Mo2Cu2 Heterodinuclear Complex
We have previously reported that the reaction of 1(PF6) with molybdate formed redbrown (Et4N)[CuMoO3(Lig)] ((Et4N)5), which had limited stability in solution and therefore was characterized by 1 H NMR, UV-vis spectroscopy, and HRMS; its DFT-optimized structure was in an agreement with spectroscopic data (Scheme 3). Our multiple attempts to crystallize (Et4N)5 resulted in its decomposition. We hypothesized that the instability of (Et4N)5 could be due to a highly reactive nature of catecholate-bound [Mo VI O3] 2-, and thus decided to replace one of the oxo groups with siloxide. Treatment of (Et4N)[MoO3(OSiPh3)] with LigH2, followed by the addition of the resulting solution to [Cu(NCMe)4](PF6) resulted in an orange-brown solution. Recrystallization of the crude product from acetonitrile formed dark orange crystals of [Cu2Mo2O4(μ2-O)(Lig)2] (6). Xray structure determination revealed 6 to be a tetranuclear complex containing two

Synthesis and Characterization of Mo 2 Cu 2 Heterodinuclear Complex
We have previously reported that the reaction of 1(PF 6 ) with molybdate formed redbrown (Et 4 N)[CuMoO 3 (Lig)] ((Et 4 N)5), which had limited stability in solution and therefore was characterized by 1 H NMR, UV-vis spectroscopy, and HRMS; its DFT-optimized structure was in an agreement with spectroscopic data (Scheme 3). Our multiple attempts to crystallize (Et 4 N)5 resulted in its decomposition. We hypothesized that the instability of (Et 4 N)5 could be due to a highly reactive nature of catecholate-bound [Mo VI O 3 ] 2− , and thus decided to replace one of the oxo groups with siloxide. Treatment of (Et 4  6 to be a tetranuclear complex containing two Mo(VI) and two Cu(I) centers (Figure 3 left). The complex is approximately C 2 -symmetric, albeit the symmetry is not crystallographic due, in part, to the presence of only one molecule of HOSiPh 3 that is H-bonded to one of the equatorial oxos (Figure 3 right, see below for the discussion). The coordination geometry at each of the Cu(I) sites is similar to the geometry in 1(PF 6 ), with three relatively short coppernitrogen (pyridine/imine) bond distances (ranging between 2.000(4) and 2.059(4) Å) and one very long copper-nitrogen (central amine) distance (2.432 (4) Molecules 2023, 28, x FOR PEER REVIEW 6 of 13 product from acetonitrile formed dark orange crystals of [Cu2Mo2O4(µ2-O)(Lig)2] (6). Xray structure determination revealed 6 to be a tetranuclear complex containing two Mo(VI) and two Cu(I) centers (Figure 3 left). The complex is approximately C2-symmetric, albeit the symmetry is not crystallographic due, in part, to the presence of only one molecule of HOSiPh3 that is H-bonded to one of the equatorial oxos (Figure 3 right, see below for the discussion). The coordination geometry at each of the Cu(I) sites is similar to the geometry in 1(PF6), with three relatively short copper-nitrogen (pyridine/imine) bond distances (ranging between 2.000(4) and 2.059(4) Å) and one very long copper-nitrogen (central amine) distance (2.432 (4) were previously evaluated by Holm and coworkers as models for Xanthine Oxidoreductase [31][32][33]   As mentioned above, the complex is co-crystallized with one molecule of Ph3SiOH that is hydrogen-bonded to one of the equatorial oxos (Figure 3, right). The presence of one equivalent of triphenylsilanol per two molybdenum ions suggests the dimerization mechanism presented in Scheme 3. Unlike [MoO4] 2− , [MoO3(OSiPh3)] -contains two different protonation sites. Thus, it is feasible that the acid-base reaction of catechol with [MoO3(OSiPh3)] − leads to the formation of siloxo-terminated and hydroxo-terminated products. Subsequent condensation of these complexes results in Mo-(µ2-O)-Mo complex which produces the final product 6 upon the incorporation of two Cu(I) ions.
Complex 6 was also characterized in solution, by 1 H NMR spectroscopy and mass spectrometry. Although the spectrum in CD3CN demonstrated only broad peaks, the spectrum in DMF-d7 featured relatively sharp resonances consistent with the overall structure of 6. Additional support for the tetranuclear structure of 6 comes from the observation of the peak corresponding to the molecular ion (6 + ) at m/z = 1706.3350 (expected m/z = 1706.3273), albeit at low intensity.

Summary and Conclusions
Towards the development of the functional model of Mo-Cu CODH, we have previously reported the synthesis of Cu(I)-Mo(VI) complex with a xanthene-bridged heterodinucleating ligand LigH2 featuring dipycolylamine chelate for Cu(I) and catecholate for Mo(VI). However, the complex had only limited stability in solution, and could not be obtained in a solid state; therefore, its subsequent reactivity was not investigated. In the present work, we targeted: (1) the in situ reactivity studies combining Lig, Cu(I), Mo(VI), and a potentially oxidizable ligand L (isocyanide, phosphine, cyanide), and (2) a more stable heterodinuclear Mo VI Cu I complex which could be isolated and structurally characterized. Towards the first goal, we have prepared a series of Lig-supported Cu(I)-L complexes, treated them with [MoO4] 2− , and explored the nature of the product (L vs. L=O) using the corresponding spectroscopy. For both phosphine and isocyanide substrates, no oxidation was observed. We propose that the lack of oxidation reactivity can be attributed (at least in part) to the difficulty in accessing a Cu(I)-bound substrate, which is removed from the Mo site and is obscured by the steric congestion at the Cu site. Toward our second goal, we explored the reactivity of the [MoO3(OSiPh3)] − precursor, which is expected to produce more stable Mo(VI) complexes. The incorporation of both Mo(VI) and Cu(I) metals within the ligand framework produced a tetranuclear Mo VI 2Cu I 2 product [Cu2Mo2O4(µ2-O)(Lig)2], in which two heterodinuclear {Mo VI Cu I (Lig 1 )} fragments were further linked by a µ2-oxo ligand at Mo. The formation of the tetranuclear product was As mentioned above, the complex is co-crystallized with one molecule of Ph 3 SiOH that is hydrogen-bonded to one of the equatorial oxos (Figure 3 Complex 6 was also characterized in solution, by 1 H NMR spectroscopy and mass spectrometry. Although the spectrum in CD 3 CN demonstrated only broad peaks, the spectrum in DMF-d 7 featured relatively sharp resonances consistent with the overall structure of 6. Additional support for the tetranuclear structure of 6 comes from the observation of the peak corresponding to the molecular ion (6 + ) at m/z = 1706.3350 (expected m/z = 1706.3273), albeit at low intensity.

Summary and Conclusions
Towards the development of the functional model of Mo-Cu CODH, we have previously reported the synthesis of Cu(I)-Mo(VI) complex with a xanthene-bridged heterodinucleating ligand LigH 2 featuring dipycolylamine chelate for Cu(I) and catecholate for Mo(VI). However, the complex had only limited stability in solution, and could not be obtained in a solid state; therefore, its subsequent reactivity was not investigated. In the present work, we targeted: (1) the in situ reactivity studies combining Lig, Cu(I), Mo(VI), and a potentially oxidizable ligand L (isocyanide, phosphine, cyanide), and (2) a more stable heterodinuclear Mo VI Cu I complex which could be isolated and structurally characterized. Towards the first goal, we have prepared a series of Lig-supported Cu(I)-L complexes, treated them with [MoO 4 ] 2− , and explored the nature of the product (L vs. L=O) using the corresponding spectroscopy. For both phosphine and isocyanide substrates, no oxidation was observed. We propose that the lack of oxidation reactivity can be attributed (at least in part) to the difficulty in accessing a Cu(I)-bound substrate, which is removed from the Mo site and is obscured by the steric congestion at the Cu site. Toward our second goal, we explored the reactivity of the [MoO 3 (OSiPh 3 )] − precursor, which is expected to produce more stable Mo(VI) complexes. The incorporation of both Mo(VI) and Cu(I) metals within the ligand framework produced a tetranuclear Mo VI 2 Cu I 2 product [Cu 2 Mo 2 O 4 (µ 2 -O)(Lig) 2 ], in which two heterodinuclear {Mo VI Cu I (Lig 1 )} fragments were further linked by a µ 2 -oxo ligand at Mo. The formation of the tetranuclear product was accompanied by the release of silanol directly observed in the crystal. The presence of one equivalent of Ph 3 SiOH per two molybdenum centers suggests a dimerization mechanism, which involves the reaction of Mo-OSiPh3 with Mo-OH intermediate to form Mo-O-Mo and Ph 3 SiOH. As in the previously reported DFT-optimized structure of the heterodinuclear [Mo VI Cu I (Lig 1 )] complex, the metals were relatively far from each other (~7 Å away), which is consistent with the lack of heterobimetallic reactivity. Our future studies in this project will focus on heterodinucleating ligands in which the Cu(I) site is anchored to the Mo site by the nearby imino(phenolate) or amino(phenolate) donor.

General
All reactions involving air-sensitive materials were carried out in a nitrogen-filled glovebox. 2,7-Di-tert-butyl-9,9-dimethyl-9H-xanthene-4,5-diamine, (E)-3-(((5-(bis(pyridin-2-ylmethyl) amino)-2,7-di-tert-butyl-9,9-dimethyl-9H-xanthen-4-yl)imino)methyl)benzene-1,2-diol, tetraethylammonium molybdate, 1(PF 6 ), and (Et 4 N)3 were synthesized using previously published procedures [26,[34][35][36]. 2,6-Dimethylphenyl isocyanide, triphenylphosphine, and trimethylphosphine, were purchased from Sigma and used as received. Tetrakis(acetonitrile)copper(I) hexafluorophosphate was purchased from Strem and used as received. All non-deuterated solvents were purchased from Aldrich and were of HPLC grade. The non-deuterated solvents were purified using an MBraun solvent purification system. Dichloromethane-d 2 and acetonitrile-d 3 were purchased from Cambridge Isotope Laboratories. All solvents were stored over 3 Å molecular sieves. Compounds were generally characterized by 1 H and 13 C NMR spectroscopy (including 2D techniques such as 1 H-1 H COSY and HMBC) and high-resolution mass spectrometry. Selected compounds were characterized by X-ray crystallography. Compounds characterization was carried out at the Lumigen Instrument Center at Wayne State University. NMR spectra of metal complexes were recorded on an Agilent DD2-600 MHz spectrometer, and an Agilent 400 MHz spectrometer in CD 3 CN and DMF at room temperature. Chemical shifts and coupling constants (J) were reported in parts per million (δ) and Hertz, respectively. Detailed assignments of the signals in 1 H NMR are given in the ESI. High-resolution mass spectra of the metal complexes (unless otherwise stated) were collected on a Thermofisher Scientific LTQ Orbitrap XL mass spectrometer. The MS survey scan was set from 200 to 2000 m/z. The resolution was set to 60,000. In all cases, only one microscan was used in the analysis.