Macrocyclization of Dienes under Confinement with Cationic Tungsten Imido/Oxo Alkylidene N‐Heterocyclic Carbene Complexes

Macrocyclization reactions are still challenging due to competing oligomerization, which requires the use of small substrate concentrations. Here, the cationic tungsten imido and tungsten oxo alkylidene N‐heterocyclic carbene complexes [[W(N‐2,6‐Cl2‐C6H3)(CHCMe2Ph(OC6F5)(pivalonitrile)(IMes)+ B(ArF)4−] (W1) and [W(O)(CHCMe2Ph(OCMe(CF3)2)(IMes)(CH3CN)+ B(ArF)4−] (W2) (IMes=1,3‐dimesitylimidazol‐2‐ylidene; B(ArF)4−=tetrakis(3,5‐bis(trifluoromethyl)phenyl borate) have been immobilized inside the pores of ordered mesoporous silica (OMS) with pore diameters of 3.3 and 6.8 nm, respectively, using a pore‐selective immobilization protocol. X‐ray absorption spectroscopy of W1@OMS showed that even though the catalyst structure is contracted due to confinement by the mesopores, both the oxidation state and structure of the catalyst stayed intact upon immobilization. Catalytic testing with four differently sized α,ω‐dienes revealed a dramatically increased macrocyclization (MC) and Z‐selectivity of the supported catalysts compared to the homogenous progenitors, allowing high substrate concentrations of 25 mM. With the supported complexes, a maximum increase in MC‐selectivity from 27 to 81 % and in Z‐selectivity from 17 to 34 % was achieved. In general, smaller mesopores exhibited a stronger confinement effect. A comparison of the two supported tungsten‐based catalysts showed that W1@OMS possesses a higher MC‐selectivity, while W2@OMS exhibits a higher Z‐selectivity which can be rationalized by the structures of the catalysts.


Introduction
Catalytic reactions under steric confinement, e. g. in mesoporous systems, mimic enzymes and benefit from the proximity of the catalyst to a pore wall with a defined geometry and polarity.This way, confinement can, e. g., induce a prefolding or preorientation of substrates and, thus, stabilize critical transition states which are difficult to realize otherwise. [1]This can lead to unexpected, sometimes unusually high [2] or inverted selectivities, [3] productivities, and activities. [4]Over the last years we have developed concepts to mimic enzymes by immobilizing well-defined organometallic catalysts inside mesoporous supports, thereby taking advantage of confinement effects.In fact, a catalyst placed inside a small pore adopts a "secondary structure", due to confinement, which allows for tuning a catalyst's reactivity and selectivity by tuning the relevant transition states in the catalytic cycle. [5]5a-d,g,h,6] These catalysts were selectively immobilized inside the mesopores of SBA-15 and other ordered mesoporous silica (OMS) materials and allowed for macrocyclization (MC) reactions at concentrations of up to 100 mM with MC selectivities of up to 98 %.In view of the peculiar reactivity and selectivity of tungsten imido or tungsten oxo alkylidene NHC catalysts [7] we were interested how the analogous tungsten-based catalysts would perform under these conditions and to which extent the concept of confined catalysts could be extended to another class of organometallic catalysts.
In solution, substrates 1-4 were converted with catalysts W1 and W2 into the corresponding macrocycles in 71-74 % yield.MC efficiency was in the range of 27-48 %; the Z-content was in the range of 17-37 %.With the supported catalysts W1@OMS and W2@OMS conversions were in the range of 27-53 % and thus lower than in solution; hindered diffusion can be made accountable for that.Most important, MC selectivity increased substantially upon immobilization of W1 or W2 inside the mesopores of OMS.For substrate 1, whose MC yields the Structures of the parent catalysts W1 and W2, their supported versions W1@OMS and W2@OMS as well as of the substrates 1-4 and the macrocyclic products 5-8.20-membered ring 5, MC selectivity increased from 48 % (W1, W2) to 80 % (W1@OMS 68Å ) using 6.8 nm pores and could be further increased to 86 % (W1@OMS 33Å ) using 3.3 nm pores.

EXAFS Analysis
Tungsten L 3 XANES spectra of W1 and W1@OMS 68Å in comparison to a tungsten foil are shown in Figure 3 together with their 1 st derivatives in the inset.The absorption edge shift by + 1.5 eV after immobilization indicates a significant change of the electronic structure in W1@OMS 68Å .It has to be noticed that this shift is not necessarily indicating an oxidation state change.In fact, since the white line intensity is identical in both W1 and W1@OMS 68Å , identical oxidation states can be assumed, as it reflects the number of d-electron holes. [11]EXAFS analysis was performed to determine the local structure of W1 and W1@OMS 68Å .As shown in Table 2, the first coordination shell is composed of five atoms according to the structural model of W1.To account for oscillation anharmonicity, the first shell scatters were fitted with the third cumulant. [12]Concerning atom types, EXAFS cannot distinguish between them due to almost identical scattering factors and phase shifts.Thus, the first double-degenerated WÀ C scatter at 1.818(9) Å contains contributions from N and C atoms.Likewise, the third doubledegenerated WÀ C scatter at 2.200(4) Å also contains contributions from C and N atoms.The fitting details on further coordination shells are available in Table S9 and the fits are shown in Figure S34.
Compared to W1, the coordination environment in W1@OMS 68Å changes significantly.Upon immobilization, decoordination of the nitrile and a 4-fold coordination is expected.
To create a suitable model for the fit of the according EXAFS spectrum, the structure of W1 was modified accordingly followed by geometry optimization using DFT calculations (Orca 5.0.2 version) [13] with the PBEh-3c method [14] and fixed positions of the Si atoms.The final XYZ coordinates of the optimized structure are provided in Table S11.Results concerning the 1 st coordination shell for EXAFS analysis based on this model are shown in Table 3, while the further shell data is bond with the 1,3-dimesitylimidazol-2-ylidene group at 2.20 Å in W1, is significantly shorter in W1@OMS 68Å (1.90 Å), which contracts a major part of the complex.Interestingly, at 3.75 Å, a WÀ Si backscatter pair is visible, which is affected by strong anharmonicity of oscillations, a row of magnitude higher than for WÀ C and WÀ O oscillations.It can be rationalized by strong movement constraints on the Si atom due to the rigid support type.

Discussion
All reactions were run 25 mM in substrate, which is substantially higher than the usual 5 mM up to which MC reactions are usually run to achieve high MC selectivity.In solution, W1 and W2 performed almost identical for a given substrate in terms of conversion and MC selectivity and showed only subtle differences in Z-selectivity.5b] Under confinement, both supported catalysts W1@OMS and W2@OMS allowed for substantially higher MC selectivity and Z-selectivity compared to the homogeneous analogs W1 or W2, particularly when immobilized inside small 3.3 nm mesopores.Moreover, particularly for the least polar substrate 4, the supported catalyst W1@OMS 33Å allowed for a higher MC selectivity than W2@OMS 33Å , which can be rationalized by the more non-polar environment around the catalyst resulting from the presence of the 2,6-dichlorophenylimido ligand compared to the oxo ligand.By contrast, W2@OMS 33Å allowed for a systematically and significantly higher Z-selectivity than W1@OMS 33Å.This is a direct result of the oxo-moiety in W2, which favors the Z-tungstacyclobutane transition state more than the imido-bearing W1 (Figure 4), at least in a trigonal-bipyramidal configure transition state.However, it should be stated that cationic tungsten alkylidene NHC catalysts can also adopt square pyramidal transition states, in which the above-discussed effects become less effective.
Compared to the analogous supported cationic molybdenum imido alkylidene NHC complex [Mo(N-2,6-Me 2 -C 6 H 3 )(CHCMe 2 Ph)(IMes)(�SiO) + B(Ar F ) 4 À ], [5b] both W1@OMS and W2@OMS give very similar results in terms of MC selectivity.5c] Consequently, MC selectivity must be independent of the catalyst used, provided no additional parameters such as different Reactions were run at room temperature in C 6 D 6 for 16 hours using 1 mol-% catalyst and a substrate concentration of 25 mM.a) 2 mol-% catalyst.distance to the pore wall or polarity become effective.By contrast, compared to the analogous supported cationic molybdenum imido alkylidene NHC complex [Mo(N-2,6-Me 2 -C 6 H 3 )(CHCMe 2 Ph)(IMes)(�SiO) + B(Ar F ) 4 À ], [5b] Z-selectivity is slightly higher with both, W1@OMS and W2@OMS.5b] Whether this higher Zselectivity is a result of a more contracted structure as found by XAS (vide supra), resulting in a more constraint geometry around the metal center, remains a likely but speculative explanation.MC selectivity did not substantially change with conversion for both W1@OMS 33Å and W1@OMS 68Å and showed only a minor drop from 90 % to 86 % and from 83 % to 80 % over time (Tables S4, S5, Supporting Information), as exemplified for substrate 1.5b] Thus, the values for the initial turnover frequency (TOF 30 min ) of W1@OMS 68 Å and W1@OMS 33 Å in C 6 D 6 at room temperature for substrate 1 were 0.45 min À 1 and 0.38 min À 1 (Figure S33), respectively, while those for the corresponding Mo-complexes were 0.49 min À 1 , 0.26 min À 1 , respectively. [15]nclusions A cationic tungsten oxo alkylidene NHC and a cationic tungsten imido alkylidene NHC catalyst have successfully been immobilized inside the pores of two different mesoporous silica materials using a pore-selective immobilization protocol.The structure of the supported cationic tungsten imido alkylidene NHC catalyst was confirmed by XAS.Both catalysts allow for the selective macrocyclization of various α,ω-dienes with high macrocyclization efficiency and appreciable Z-selectivity, substantially exceeding the homogenous analogues.Macrocyclization and Z-selectivity are highest with the smallest pores.The results presented here are in line with those obtained previously both in the macrocyclization and ring-opening crossmetathesis (ROCM) [6] with cationic molybdenum imido alkylidene NHC complexes under confinement and illustrate the generality of the confinement approach outlined here.

Experimental
General: All reactions were performed under the exclusion of air and moisture in a N 2 -filled glove box (MBraun Labmaster) unless noted otherwise.Chemicals were purchased from ABCR, Acros Organics, Alfa Aesar, Sigma Aldrich, Fluka and TCI.Poly(ethylene glycol)-b-poly(propylene glycol)-b-poly(ethylene glycol) (PEG-PPG-PEG, Pluronic® P-123), dodecylethyldimethylammonium bromide (� 98 %), tetramethyl orthosilicate (98 %, TMOS) and 1,2-dichlorobenzene (anhydrous) were purchased from Sigma Aldrich.CH 2 Cl 2 , diethyl ether, n-pentane and toluene were dried using an MBraun SPS-800 solvent purification system and stored over 4 Å molecular sieves.Deuterated solvents were stored over activated alumina and 4 Å molecular sieves for a minimum of 24 h prior to use.NMR spectra were recorded on a Bruker Avance III 400 spectrometer at 400 MHz for proton and at 101 MHz for carbon.NMR spectra were internally calibrated to solvent signals. [16]Abbreviations for multiplicities: s (singlet), bs (broad singlet), d (dublet), t (triplet), q (quartet), hept (heptet), m (multiplet).Elemental analyses were measured on a Perkin Elmer 240 device at the Institute of Inorganic Chemistry, University of Stuttgart, Germany.High performance  liquid chromatography (HPLC) was performed at the Institute of Organic Chemistry, University of Stuttgart, Germany using a Knauer K-501 pump, a Knauer K 2400 RI-detector and a Macherey & Nagel VP250/21 Nucleodur 100-5 column.Argon and nitrogen adsorption analyses were performed at 87 K and 77 K, respectively, on a Quantachrome Autosorb iQ MP automatic volumetric instrument (for Argon) Quantachrome QuadraSorb automatic volumetric instrument (for nitrogen).Silica samples were degassed for 11 h at 150 °C under vacuum prior to the gas adsorption studies.Pore size distributions, pore volumes and surface areas were calculated from the desorption branch using the non-local Density Functional Theory (NLDFT) cylindrical adsorption pores for zeolites/silica implemented in the ASiQwin software version 3.01.Small-angle Xray scattering (SAXS) experiments were performed at 25 °C with an Anton Paar SAXSess mc 2 equipped with a Dectris Mythen 1 K detector.CuÀ Kα radiation was generated with an ID 3003 X-ray generator (Seifert) operated at 40 kV and 40 mA and line collimated.Deconvolution of the measured SAXS curves was performed with the software SAXSquant (Anton Paar).The sample to detector distance was calibrated using a sample of powdered cholesteryl palmitate.ICP-OES data were recorded on a Spectro Acros 160 CCD equipped with a Cetec ASX-260 autosampler.Analysis of the samples was carried out with the Software Smart Analyzer Vision 4.02.0834.Scanning electron microscopy (SEM) analysis was performed at the AMICA core facility of the University of Stuttgart.After sputtering with gold, samples were measured with a Zeiss Evo 15 using a secondary electron detector and the associated software SmartSEM 6.07 (Zeiss).

Determination of the metal loading via ICP-OES:
Quantitative analysis of the W-loading of each silica sample was determined by ICP-OES.For analysis, the corresponding silica (30-50 mg, Table S2) was mixed with KOH (0.38 g, 6.77 mmol) and KNO 3 (0.65 g, 6.42 mmol).The mixture was heated to 450 °C and the temperature was held for 3 hours.After cooling to room temperature, K 2 S 2 O 8 (50.0 mg, 0.18 mmol) was added.The colorless solid was dissolved in a minimum amount of deionized water and 1 M KOH (2 mL) was added.The suspension was filtered, transferred into a 10 mL volumetric flask and filled to the mark with deionized water.The solution was slowly added to a 25 mL volumetric flask with concentrated HCl (5 mL) and filled to the mark with deionized water.This solution was analyzed by ICP-OES for W. W was measured at λ = 207.911nm; the background was measured at λ = 207.84nm-207.87nm and λ = 207.99nm-208.04nm, respectively.The limit of detection (LOD) was 0.0001 mg.L À 1 .For calibration, aqueous W-standards with W concentration of 0.000, 0.100, 0.500, 1.000, 2.500 and 5.000 mg L À 1 were used.A reference, containing the same amount of KOH, KNO 3 , HCl and deionized water was subjected to the same treatment for comparison.

General Procedure for the RCM of α,ω-Dienes (GP-1):
To a stirred solution of the diene (0.75 mmol) in CH 2 Cl 2 (200 mL) was added the 2 nd -generation Grubbs catalyst RuCl 2 (IMes)(PCy 3 )(CHPh) (31.8 mg, 0.0375 mmol, 5 mol-%) at room temperature.After stirring for 14 hours under reflux and under N 2 , the reaction mixture was cooled to room temperature and ethyl vinyl ether (5 mL, 70 eq) was added.The mixture was stirred for a further 2 hours at room temperature.All volatiles were removed under reduced pressure and the obtained crude product was purified via column chromatography on SiO 2 to obtain the corresponding macrocyclic product, whose E/Z isomers were separated by semi-preparative HPLC.

(E/Z)-tert-Butyl(cycloheptadec-9-en-1-yloxy)dimethylsilane (E/Z-8):
To a solution of E/Z-cycloheptadec-9-en-1-ol (12.6 mg, 0.05 mmol) in DMF (1 mL) were subsequently added imidazole (10.2 mg, 0.15 mmol) and tert-butyldimethylsilyl chloride (11.3 mg, 0.075 mmol).After stirring overnight at room temperature, ethanol (1 mL) was added and the reaction mixture was stirred for a further 15 minutes.The mixture was then diluted with n-pentane (10 mL), washed with H 2 O (15 mL) and brine (15 mL), dried over anhydrous Na 2 SO 4 , filtered and all volatiles were removed in vacuo.The obtained crude product was purified by column chromatography over SiO 2 (n-pentane:diethyl ether-100:0!100: 1) to yield the corresponding macrocycle E/Z-8 as an inseparable mixture on form of a colorless oil.Spectral data were in agreement with previous reports. [5d,10] 1 H NMR (C 6 D 6 ) δ 5.39-5.43(m, 0.21H), 5.31-5.37(m, 1.79H), 3.75-3.81(m, 1H), 2.00-2.10(m, 4H), 1.50-1.63(m, 4H), 1.26-1.435b,20] In brief, for the synthesis of OMS 33Å 38 g of TMOS were added to 27 g of 0.1 N HCl in a polypropylene flask.The forming methanol was removed by rotary evaporation for 15 minutes at 40 °C and 280 mbar.The mixture was then added to 18 g of cetyldiethylammonium bromide and stirred with a KPG-stirrer until a homogenous, clear solution was obtained.For the synthesis of OMS 68Å 30.4 g of TMOS were added to 21.6 g of 0.1 N HCl in a polypropylene flask.The forming methanol was removed as described before.Subsequently, the mixture was added to 17.7 g of P123 and 1 g of hexadecane and stirred until homogenous.The two obtained clear solutions were poured into PTFA dishes and cured for two to three days at 80 °C.The now solid materials were milled for 1 min with a ball mill (Spex 8000 Mixer/ Mill, vial and balls made from stainless steel).Afterwards, the powder materials were calcined by heating them to 550 °C with 1 °C•min À 1 and keeping them at this temperature for 6 h to remove all surfactant molecules.An airflow of 14.5 L•h À 1 was applied during calcination.

EXAFS measurements
W L 3 Extended X-Ray Absorption Fine Structure measurement were conducted at the P65 beamline, Petra III, DESY (Hamburg). [21]Energy selection was done with a Si(111) double crystal monochromator; the energy resolution was around 1.0 eV at 10 keV.The beam spot size was 0.3 x 1 mm 2 and the total flux on the sample was 10 12 ph/ s.Signal detection was conducted in the fluorescence mode using a PIPS detector, all at room temperature.For beam focusing and higher harmonic rejections, Rh-coated mirrors were used.Data acquisition was performed in the continuous scanning mode and spectra were rebinned afterwards.Energy calibration was conducted at the first inflection point of the pure W XANES spectrum (10206.9eV).To avoid radiation damage, each EXAFS spectrum was collected at a different sample position.Initial normalization and background removal were conducted with the Athena software; [22] EXAFS fitting was performed with the Artemis package [22] using the Multiple Scattering approach.The analysis was conducted in the krange of 2.0-15.0Å À 1 and real space of 1.0-4.2Å.

Contributions
MRB elaborated the concepts and wrote the manuscript.FZ carried out all immobilizations, catalyst synthesis and macrocyclization experiments.JRB provided the OMS materials.PP helped in analyzing the data and recorded the reaction kinetics.MN and MB contributed the XAS data, BA and JRB recorded and interpreted the gas sorption experiments.

Figure 1 .
Figure1.Structures of the parent catalysts W1 and W2, their supported versions W1@OMS and W2@OMS as well as of the substrates 1-4 and the macrocyclic products 5-8.

Figure 2 .
Figure 2. Multi-step modification of OMS for the pore-selective immobilization of the catalyst inside the mesopores.

Figure 3 .
Figure3.A) Tungsten L 3 XANES spectra of W1 and W1@OMS 68Å in comparison to a W foil. Insert shows corresponding 1 st derivatives of spectra; B) Fouriertransformed EXAFS along with fitted models for W1 (grey/black) and W1@OMS 68Å (blue/dark blue).

Figure 4 .
Figure 4. Structures of Z-tungstacyclobutane transition states responsible for high Z-selectivity derived from W1 and W2, respectively.