Formation of Mono Oxo Molybdenum(IV) PNP Pincer Complexes: Interplay between Water and Molecular Oxygen

The synthesis of cationic mono oxo MoIV PNP pincer complexes of the type [Mo(PNPMe‐iPr)(O)X]+ (X = I, Br) from [Mo(PNPMe‐iPr)(CO)X2] is described. These compounds are coordinatively unsaturated and feature a strong Mo≡O triple bond. The formation of these complexes proceeds via cationic 14e intermediates [Mo(PNPMe‐iPr)(CO)X]+ and requires both molecular oxygen and water. ESI MS measurements with 18O labeled water (H2 18O) and molecular oxygen (18O2) indicates that water plays a crucial role in the formation of the Mo≡O bond. A plausible mechanism based on DFT calculations is provided. The X‐ray structure of [Mo(PNPMe‐iPr)(O)I]SbF6 is presented.


Introduction
Molybdenum complexes featuring a terminal mono oxo unit comprise an important class of compounds. [1,2] On the one hand, besides of being intrinsically interesting, [3] such complexes are well documented to act as catalysts for various oxidation processes involving for instance molecular oxygen. [4] They are also known to generate hydrogen from water [5] and are applied in various catalytic reactions such as hydrosilylation [6] and sulfur transfer to alkenes and allenes. [7] Moreover, nature efficiently utilizes the Mo=O unit to achieve difficult multielectron redox catalysis with oxotransferases, which catalyze oxygen atom transfer to and from substrates. [8][9][10] High valent Mo=O or Mo≡O species are often generated accidently by trace amounts of O 2 or water contaminations due to the high affinity of molybdenum towards oxygen which may involve proton assisted and/ or water assisted dioxygen cleavage reactions. [11] The oxygen Scheme 1. Possible oxygen-containing species detected by fragmentation of [Mo(PNP Me -iPr)(CO)X 2 ] (1a, 1b) in CH 3 CN in the presence of air and water as established by ESI MS experiments. Structural suggestions are based on DFT calculations. [12] We report here on a rational synthesis of cationic coordinatively unsaturated mono oxo Mo IV PNP pincer complexes of the type [Mo(PNP Me -iPr)(O)X] + (A) which are formed from in-situ prepared [Mo(PNP Me -iPr)(CO)X] + (2a, b) in the presence of molecular oxygen and water. [13]

Results and Discussion
When a solution of [Mo(PNP Me -iPr)(CO)(X)(solv)] + (2a, b) in acetone, prepared in situ by reacting [Mo(PNP Me -iPr)(CO)(X) 2 ] (1a, b) (X = I, Br) with AgSbF 6 followed by removal of AgX, is exposed shortly to air and subsequently treated with an excess of Scheme 2. Formation of mono oxo Mo IV complexes.
water, the cationic mono oxo complexes [Mo(PNP Me -iPr)(O)X] + (3a, b) are afforded in 72 and 66 % isolated yields (Scheme 2). In the absence of air or water, no mono oxo complexes are formed. Accordingly, the formation of the molybdenum oxo bond requires an interplay between these two reagents. NMR and IR monitoring of the reaction with 1a and 1b after addition of the halide scavenger revealed the immediate formation of 2a and 2b, respectively. These intermediates give rise to signals at δ = 183.3 and 189.5 ppm in the 31 P{ 1 H} NMR spectrum and exhibit one strong ν CO band at 1832 and 1840 cm -1 , respectively (cf. 1824 cm -1 in 1a and 1816 cm -1 in 1b). Solvent complexes of the type [Mo(PNP Me -iPr)(CO)(X)(solv)] + (X = Cl, Br, solv = THF, CH 3 CN) were prepared and isolated recently. [12] Upon admission of air and addition of water, new resonances at δ = 149.2 and 145.2 ppm, respectively, were observed in the 31 P{ 1 H} NMR spectrum due to the formation of 3a and 3b and the CO stretching frequencies of 2a and 2b disappeared.
Complexes 3a and 3b were characterized by a combination of elemental analysis, 1 H, 13 C{ 1 H}, and 31 P{ 1 H} NMR, IR and ESI MS. Characteristic are the Mo≡O stretching frequencies at 955 and 940 cm -1 , respectively. In the ESI-MS the most abundant signals are observed at m/z 604.1 and 556.1, respectively, which correspond to the intact complexes 3a and 3b ([M] + ). In addition to the main products (3a,b), small amounts (ca 10 %) of the known seven-coordinate tricarbonyl complex [Mo(PNP Me -iPr)(CO) 3 X] + are formed as side products due to reaction of 2a and 2b with CO, which is released during the oxidation process (Scheme 2). [14] It has to be noted, that there was no evidence for the formation of CO 2 as a result of CO oxidation by O 2 .
In addition to the NMR, IR and ESI-MS spectroscopic characterization, the crystal structure of 3a was determined by singlecrystal X-ray diffraction. A structural diagram is depicted in Figure 1 with selected bond lengths and angles given in the caption. Complex 3a is best described as having a pseudo square pyramidal structure. The Mo1-O1 bond length of 1.663(2) Å is comparatively short but in the typical range for a Mo≡O triple bond. [15][16][17] This has been investigated by DFT calculations. [18] The frontier orbitals of 3a are represented in Figure 1. The pattern obtained is typical of a d 2 metal complex with a square pyramidal geometry. [19] The HOMO is the xy orbital (the z axis being defined by the Mo-O bond) and the LUMO is mostly centered in the ligand pyridine ring. The two following orbitals (LUMO+1 and LUMO+2) are based on metal yz and xz, respectively (see Figure 2). Those are Mo-O π* orbitals and, thus, are  the two empty antibonding counterparts of π-donation from the oxo ligand to the metal, indicating a Mo≡O triple bond. Finally, the two upper orbitals in Figure 1 are based on the metal z 2 and x 2 -y 2 .
To evaluate the role of water and O 2 as an oxygen source, solutions of 1a in CH 3 CN were subjected to ESI-MS analysis in the positive ion mode in the presence of either 18 18 O 2 ) takes place forming a dioxo product ion at m/z 620 (or m/z 624) together with other products. This clearly shows that in the gas phase ion 2a is the precursor of the dioxo species II or III as shown in Scheme 3. The same  species is also formed and observed in the electrosprayed solution as already reported previously. [12] When the ion-molecule reaction of 2a was performed with H 2 O instead of O 2 , neither [Mo(PNPMe-iPr)(O)I] + (3a) nor other products were observed. This suggests again the need for a cooperation between dioxygen and water that can be realized in solution (sprayed solution) but not in the gas phase where reactants and products are in a rarefied environment (pressure of about 10 -6 -10 -5 Torr that reaches up to 10 -3 Torr with Helium). Solutions of complexes [Mo(PNP Me -iPr)(O)X] + (3a,b) in chlorinated solvents such as CHCl 3 and CH 2 Cl 2 are air sensitive being slowly oxidized to yield the mono oxo Mo VI complex [Mo(κ 2 O,O-ONO Me -iPr)(O)Cl 3 ]SbF 6 (5) (Scheme 4). The same reaction takes place rapidly in the presence of H 2 O 2 yielding quantitatively complex 5 within 10 minutes as monitored by 31 P{ 1 H} NMR spectroscopy. During this reaction, three chloride ligands from the solvent replace both iodo and bromo ligands, while the phosphine moieties are oxidized to the respective phosphine oxides. The pyridine ring is no longer coordinated, while the phosphine oxide moieties are coordinated via the oxygen atoms. In this context, it has to be noted that if the solvent is CH 2 Cl 2 instead of acetone, [Mo(PNP Me -iPr)(CO)-(X)(solv)] + (2a,b) reacts with air and an excess of water to afford the cationic mono oxo complexes [Mo(PNP Me -iPr)(O)X] + (3a,b), but also small amounts of the Mo VI species 5 (ca 15 %). Al-though we could not directly detect H 2 O 2 , this observation suggests that during this reaction H 2 O 2 may be released (vide infra) as this oxidation process is very slow in the presence of oxygen, but fast in the presence of H 2 O 2 . Moreover, H 2 O 2 could disproportionate under these reaction conditions to form water and O 2 which again would form 3a,b from complexes 2a,b. [20] Complex 5 is isolated in essentially quantitative yield and was characterized by elemental analysis, 1 H, 13 C{ 1 H}, and 31 P{ 1 H} NMR spectroscopy. In addition, 5 was characterized by X-ray crystallography.  with A′, the pair of reactants of 2b and O 2 (O 2 being a triplet) producing complex B via transition state TS A′B . In this transition state the new Mo-O bond is only incipient with a distance of 2.81 Å which is still far away from the coordination distance of 2.05 Å in B. The energy barrier is 8.8 kcal/mol. After re-orientation of the O 2 ligand to afford C the process is practically thermoneutral with respect to the initial reagents (C is only 0.6 kcal/ mol less stable than the separated reactants).
The next step involves CO dissociation from C. This step has a barrier of 9.8 kcal/mol (TS CD ) and yields a coordinatively unsaturated species with an O 2 ligand and the halide, beside the PNP ligand (D and D′). The transition state TS CD is a late one with a Mo-C(CO) separation of 3.79 Å. The entire process from C to D′ is essentially thermoneutral (ΔG = 0.7 kcal/mol). Dissociation of the CO ligand directly from 2b is unfavorable with ΔG = 34 kcal/mol and thus coordination of dioxygen is required. From D′ there is coordination of the dangling O-atom with formation of a peroxide κ 2 -O 2 ligand corresponding to an oxidative addition with the metal changing from Mo II in D′ to Mo IV in E. This is a facile process with a barrier of only 0.7 kcal/ mol (TS D′E ). In the transition state, the new Mo-O bond is about to be formed with a distance of 2.58 Å. This is significantly longer than the Mo-O bonds in E (1.97 Å). Formation of the peroxide complex is thermodynamically favorable with E being 16.4 kcal/mol more stable than the initial reagents. The reaction then proceeds from E to F with a change in spin state from triplet (S = 1) to singlet (S = 0). That corresponds to a "spinforbidden" or "non-adiabatic" reaction and, thus, its profile goes through a minimum-energy crossing point (MECP) of the two potential energy surfaces (PES) involved. [21,22] The barrier calculated for the spin change of E is 6.4 kcal/mol (CP EF ) but the spin singlet intermediate F is 4.4 kcal/mol less stable than its high spin counterpart, and, thus corresponds to a rather facile but noticeably endergonic step.
Following intermediate F, the reaction profile proceeds along the spin singlet PES ( Figure 6). There are two alternative paths. In one case, there is O-O bond cleavage via an oxidative addition process that leads to the di-oxo Mo VI complex L. This is a single-step process represented on the left side of the profile in Figure 6, being highly exergonic as the product L is 86.2 kcal/ mol more stable than the initial reagents. The barrier associated In this process one H 2 O 2 molecule will be released. This last step has a barrier of 16.4 kcal/mol and is clearly exergonic with ΔG = -12.0 kcal/mol, resulting in a final product 26.5 kcal/mol more stable than A. In the formation of the mono-oxo complex (from F to K) the least stable transition state is TS JK with a free energy 1.9 kcal/mol above the initial reactants. On the other hand, transition state TS FL associated with the formation of the di-oxo product (from F to L) has an energy of 3.1 kcal/mol relative to A. The difference between the total barriers of the two paths is only 1.2 kcal/mol and, thus, they can be considered competitive. The formation of the mono-oxo complex reveal that the formation of these complexes requires an interplay between water and molecular oxygen. The major source of oxygen of the Mo≡O oxo bond appears to be water. The crystal structure of [Mo(PNP Me -iPr)(O)I]SbF 6 is presented. Detailed theoretical studies based on DFT calculations established a reasonable mechanism for the formation of both mono and dioxo molybdenum complexes proceeding via two competitive pathways.
Mass spectrometric measurements were performed on an Esquire 3000 plus 3D-quadrupole ion trap mass spectrometer (Bruker Daltonics, Bremen, Germany) in positive-ion mode electrospray ionization (ESI-MS). Mass calibration was done with a commercial mixture of perfluorinated trialkyltriazines (ES Tuning Mix, Agilent Technologies, Santa Clara, CA, USA). All analytes were dissolved in CH 3 CN "Lichrosolv" quality (Merck, Darmstadt, Germany) to a concentration of roughly 1 mg/mL and doped with sodium halides (Merck, Darmstadt, Germany) to avoid or suppress dissociation of halogen substituents from the complexes. Direct infusion experiments were carried out using a Cole Parmer model 74900 syringe pump (Cole Parmer Instruments, Vernon Hills, IL, USA) at a flow rate of 2 μL/ min. Full scan and MS/MS-scans were measured in the range m/z 100-1000 with the target mass set to m/z 800. Further experimental conditions include: drying gas temperature: 150°C; capillary voltage: -4 kV; skimmer voltage: 40 V; octapole and lens voltages: according to the target mass set. Helium was used as buffer gas for full scans and as collision gas for MS/MS-scans in the low energy collision induced dissociation (CID) mode. The activation and fragmentation width for tandem mass spectrometric (MS/MS) experiments was set to 10-12 Da to cover the entire isotope cluster for fragmentation. The corresponding fragmentation amplitude ranged from 0.3 to 0.8 V to keep a low abundant precursor ion intensity in the resulting MS/MS spectrum. All mass calculations are based on the lowest mass isotope for molybdenum ( 92 Mo-isotope). Mass spectra and tandem spectra were averaged during data acquisition time of 1 to 2 min and one analytical scan consisted of five successive micro scans resulting in 50 and 100 analytical scans, respectively, for the final mass spectrum or MS/MS spectrum.
The labelling experiments were performed on a LTQ-XL linear ion trap mass spectrometer (Thermo Fisher Scientific) fitted with an electrospray ionization (ESI) source operating in the positive ion mode.

Instrumental Analysis Conditions:
[Mo(PNP Me -iPr)(CO)I 2 ] (1a) was dissolved in acetonitrile to the millimolar concentration and doped with sodium iodide. Sample solutions were infused at a flow rate of 3-5 μL/min via the instrument′s on-board syringe pump directly connected to the ESI source. Typical experimental conditions were: source voltage 4-5 kV, capillary temperature 200°C. Nitrogen was used as sheath and auxiliary gas at a flow rate of 15 and 5 arbitrary units (a.u. ≈ 0.37 L/min). Full scan mass spectra were measured in the m/z range 100-1000 and were the average of 25-50 scans, each resulting from three micro scans. Two sets of separate and different labelling experiments, using either 18 O 2 or H 2 18 O were performed, as described in the following.

Experiment 1 with 18 O 2 :
A flask containing a mixture of [Mo(PNP Me -iPr)(CO)I 2 ] (1a)/NaI was connected to a vacuum system and carefully evacuated. It was then filled with 18 O 2 (760 Torr) and acetonitrile was subsequently added to the solid mixture using a gas tight syringe to avoid contact with air. The solution was stirred and the flask was left at ambient temperature. Samples taken at different period of times (15 min, 1 h, 3 h, 20 h) were infused into the ESI source and analyzed using instrumental conditions as described in the paragraph above. As the ESI source is an atmospheric pressure ionization (API) source, the contact with (moist) air cannot be avoided. 16 O sources can thus come from: (1) residual 16  O v/v) injected through the rubber septum capping the vial. Samples were promptly taken and infused into the ESI source, using experimental parameters as described above. As in the previous experiments with 18 O 2 , sources of 16 O are present, the major ones coming from: (1) residual H 2 16 O is present as impurity in the labelled sample and is possibly present in acetonitrile; (2) O 2 and H 2 O from air are always present in this type of mass spectrometers. The ion-molecule reactions were performed on the LTQ XL linear ion trap mass exploiting an in-house modification that allows the introduction of neutral gases into the ion trap in order to observe ion-molecule reactions of mass-selected ions with the neutral reagent (O 2 and H 2 O), as described in details elsewhere. [24] Ionic species generated in the electrospray source were isolated with an isolation width of 1 m/z and reacted with the neutral of interest for different periods of time. For each reaction time, mass spectra were recorded using an injection time of 200 ms, a normalized collision energy set to zero, and the activation Q value optimized to ensure stable trapping fields for all ions. Spectra were acquired using the MS n function of the Xcalibur 2.0.6 software to mass-select the precursor ion. All the spectra are the average of 10 scans for each acquisition. Method A: A solution of 3a (50 mg, 0.082 mmol) in CH 2 Cl 2 or CHCl 3 (10 mL) was exposed to air for 4 d at room temperature. After that, the solution was filtered and the solvent was removed. The product was obtained as a red solid which was washed twice with n-pentane and then dried under vacuum.

Crystal Structure Determination
Single crystals of 3a and 5·1/2CH 2 Cl 2 were pre-selected, embedded in perfluorinated polyether and mounted on Kapton micro mounts. X-ray diffraction data were measured in a cold stream of nitrogen at T = 100 K on a Bruker APEX-II diffractometer [25] with Mo-K α radiation. After integration of the data with the program SAINT, [25] an absorption correction based on the semi-empirical "multi-scan" approach was performed with the SADABS program. [25] The crystal structures were solved using the dual space approach implemented in SHELXT [26] and was refined using the SHELXL program package. [26] All H atoms were placed geometrically and refined in the riding model approximation, with C-H = 1.00 Å and U iso (H) = 1.2U eq (C) for the CH groups and with C-H = 0.98 Å and U iso (H) = 1.5U eq (C) for the methyl groups. All non-hydrogen atoms were refined anisotropically. Molecular graphics were generated with the program MERCURY. [27] CCDC 1480834 (for 3a) and 1574491 (for 5·1/2CH 2 Cl 2 ) contain the supplementary crystallographic data for this paper. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre.

Computational Details
The computational results presented have been achieved in part using the Vienna Scientific Cluster (VSC). Calculations were performed using the Gaussian 09 software package, [28] and the B3LYP functional, without symmetry constraints. That functional include a mixture of Hartree-Fock [29] exchange with DFT exchange-correlation, given by Becke's three parameter functional [30] with the Lee, Yang and Parr correlation functional, which includes both local and non-local terms. [31,32] The optimized geometries were obtained with the Stuttgart/Dresden ECP (SDD) basis set [33] to describe the electrons of Mo and I, and a standard 6-31G** basis set [34] for the other atoms. Transition state optimizations were performed with the Synchronous Transit-Guided Quasi-Newton Method (STQN) developed by Schlegel et al., [35] following extensive searches of the Potential Energy Surface. Frequency calculations were performed to confirm the nature of the stationary points, yielding one imaginary frequency for the transition states and none for the minima. Each transition state was further confirmed by following its vibrational mode downhill on both sides, and obtaining the minima presented on the energy profiles. The electronic energies were converted to free energy at 298.15 K and 1 atm by using zero-point energy and thermal energy corrections based on structural and vibration frequency data calculated at the same level.
The Minimum Energy Crossing Point (MECP) between the spin singlet (S = 0) and the spin triplet (S = 1) Potential Energy Surfaces (PES) was determined using a code developed by Harvey et al. [36] This code consists of a set of shell scripts and Fortran programs that uses the Gaussian results of energies and gradients of both spin states to produce an effective gradient pointing towards the MECP. This is not a stationary point and, hence, a standard frequency analysis is not applicable. Therefore, the free energy value of the crossing point (CP EF ) was obtained through frequency calculations projected for vibrations perpendicular to the reaction path. [37] Orbital representations were obtained with Molekel. [38] Supporting Information (see footnote on the first page of this article): Atomic coordinates of all optimized species (xyz files).