An Asymmetric Dinuclear Bis( ansa -Zirconocene) Complex: Synthesis and Performance in Oleﬁn (co-)Polymerization

: A synthetic strategy to access asymmetric dinuclear bis( ansa -metallocene) pre-catalysts is described. As a key step, the Pd-catalyzed Suzuki cross-coupling of 9,9-bis (trimethylsilyl)-ﬂuoren- 2-yl-boronic acid with a substituted 2-bromo-9 H -ﬂuorene generates an asymmetric 2,2 (cid:48) -biﬂuorene platform, which can be individually functionalized at the two differentiated 9-positions. Herein, as a ﬁrst demonstration of this strategy, we report the asymmetric dinuclear bis( ansa -zirconocene) complex 2,2 (cid:48) -[{Me 2 C(Flu)(Cp)}ZrCl 2 ][{Me 2 C( 7-t Bu Flu)(Cp)}ZrCl 2 ], which has been characterized with NMR spectroscopy and high-resolution mass spectrometry. The performance of this bimetallic pre-catalyst when combined with MAO has been evaluated in ethylene, propylene, and ethylene/1-hexene (co-)polymerization. This pre-catalyst is revealed to be less productive than the mononuclear reference pre-catalyst {Me 2 C( 2,7-t Bu Flu)(Cp)}ZrCl 2 , likely because of higher steric hindrance induced by the linkage at the diﬂuorenyl platform. The resulting (co-)polymers featured only slight differences in terms of molecular weights, tacticity, and comonomer incorporation. No bimodal molecular weight distribution was achieved at any produced polymer; this might have originated from the close similarity of the connected Cp/Flu moieties or a rapid chain-transfer phenomenon between the different active sites which were quite close to each other.

The one-sided trimethylsilyl-protection at bifluorene 5 enabled the selective deprotonation of the other (i.e., the tert-butyl-functionalized) fluorenyl moiety, which was then reacted with 6,6-dimethylfulvene to give the mono-(Cp/Flu) compound 6 (Scheme 2, vi) [27,28].Subsequent desilylation of the protected fluorenyl moiety was achieved using potassium hydroxide at elevated temperature (vii) [23].Both mono-(Cp/Flu) bifluorene compounds 6 and 7 were fully characterized using NMR and X-ray diffraction.In particular, the NMR spectra recorded in CDCl 3 (Figures S10-S17) show the presence of two tautomers, as indicated by the presence of two signals for the methylene hydrogens of the cyclopentadienyl fragment (δ 1 H = 3.15-3.30ppm) and for the hydrogen bonded at the sp 3 -carbon of the fluorenyl moiety (δ 1 H = 4.20-4.32ppm).Single crystals suitable for X-ray diffraction analysis of 6 and 7 were obtained from hot saturated heptane solutions at room temperature (Figure 2).The one-sided trimethylsilyl-protection at bifluorene 5 enabled the selective deprotonation of the other (i.e., the tert-butyl-functionalized) fluorenyl moiety, which was then reacted with 6,6-dimethylfulvene to give the mono-(Cp/Flu) compound 6 (Scheme 2, vi) [27,28].Subsequent desilylation of the protected fluorenyl moiety was achieved using potassium hydroxide at elevated temperature (vii) [23].Both mono-(Cp/Flu) bifluorene compounds 6 and 7 were fully characterized using NMR and X-ray diffraction.In particular, the NMR spectra recorded in CDCl3 (Figures S10−S17) show the presence of two tautomers, as indicated by the presence of two signals for the methylene hydrogens of the cyclopentadienyl fragment (δ 1 H = 3.15-3.30ppm) and for the hydrogen bonded at the sp 3 carbon of the fluorenyl moiety (δ 1 H = 4.20-4.32ppm).Single crystals suitable for X-ray diffraction analysis of 6 and 7 were obtained from hot saturated heptane solutions at room temperature (Figure 2).The one-sided trimethylsilyl-protection at bifluorene 5 enabled the selective deprotonation of the other (i.e., the tert-butyl-functionalized) fluorenyl moiety, which was then reacted with 6,6-dimethylfulvene to give the mono-(Cp/Flu) compound 6 (Scheme 2, vi) [27,28].Subsequent desilylation of the protected fluorenyl moiety was achieved using potassium hydroxide at elevated temperature (vii) [23].Both mono-(Cp/Flu) bifluorene compounds 6 and 7 were fully characterized using NMR and X-ray diffraction.In particular, the NMR spectra recorded in CDCl3 (Figures S10−S17) show the presence of two tautomers, as indicated by the presence of two signals for the methylene hydrogens of the cyclopentadienyl fragment (δ 1 H = 3.15-3.30ppm) and for the hydrogen bonded at the sp 3 carbon of the fluorenyl moiety (δ 1 H = 4.20-4.32ppm).Single crystals suitable for X-ray diffraction analysis of 6 and 7 were obtained from hot saturated heptane solutions at room temperature (Figure 2).Further deprotonation of compound 7 with two equivalents of nBuLi and reaction with 6,6-dimethylfulvene gave bis(proligand) 8 in 68% yield (viii) [29].Compound 8 was characterized using NMR spectroscopy, and a complete signal assignment of the recorded 1 H and 13 C NMR spectra was enabled with additional 1 H- 13   Further deprotonation of compound 7 with two equivalents of nBuLi and reaction with 6,6-dimethylfulvene gave bis(proligand) 8 in 68% yield (viii) [29].Compound 8 was characterized using NMR spectroscopy, and a complete signal assignment of the recorded 1 H and 13 C NMR spectra was enabled with additional 1 H- 13   Further deprotonation of compound 7 with two equivalents of nBuLi and reaction with 6,6-dimethylfulvene gave bis(proligand) 8 in 68% yield (viii) [29].Compound 8 was characterized using NMR spectroscopy, and a complete signal assignment of the recorded 1 H and 13 C NMR spectra was enabled with additional 1 H- 13  Regular salt metathesis reaction between the ligand tetra-anion, generated in situ from 8 in diethyl ether, and ZrCl4 returned the dinuclear bis(ansa-zirconocene dichloride) Regular salt metathesis reaction between the ligand tetra-anion, generated in situ from 8 in diethyl ether, and ZrCl 4 returned the dinuclear bis(ansa-zirconocene dichloride) 2,2 -[{Me 2 C(Flu)(Cp)}ZrCl 2 ][{Me 2 C( 7-tBu Flu)(Cp)}ZrCl 2 ] (9) (Scheme 3).After extraction with dichloromethane and removal of lithium chloride, the zirconium complex was isolated in good yield as a characteristically pink, microcrystalline material.Yet, all attempts to grow single crystals suitable for X-ray diffraction analysis remained unsuccessful.Alternatively, high-resolution mass spectrometry unequivocally evidenced the formation of the bimetallic complex (ASAP-MS and MALDI-ToF-MS, Figures S28 and S29).In addition, the complex was analyzed comprehensively using NMR spectroscopy.Notably, once all solvents had evaporated to complete the dryness of complex 9, its solubility decreased dramatically; thus, small amounts of diethyl ether and hexane were detected in both the 1 H and 13 C NMR data in CD 2 Cl 2 , yet without hampering full signal assignment (Figures S22-S27).The signals of the cyclopentadienyl moieties are the most informative, with four sharp apparent quartets in the region of δ 1 H = 5.72-5.84ppm (4H), which correlate with four carbon peaks at δ 13 C = 101.7-102.8ppm, and which with the other signals at δ 1 H = 6.28-6.36ppm (4H) account for the eight individual hydrogens in the 2-5-positions at Cp and Cp .Along with the sum of 44 detected carbon signals, this also means that either only one of the two possible isomers was formed (Scheme 3) or their presence could not be determined with NMR spectroscopy.lated in good yield as a characteristically pink, microcrystalline material.Yet, all attempts to grow single crystals suitable for X-ray diffraction analysis remained unsuccessful.Alternatively, high-resolution mass spectrometry unequivocally evidenced the formation of the bimetallic complex (ASAP-MS and MALDI-ToF-MS, Figures S28 and S29).In addition, the complex was analyzed comprehensively using NMR spectroscopy.Notably, once all solvents had evaporated to complete the dryness of complex 9, its solubility decreased dramatically; thus, small amounts of diethyl ether and hexane were detected in both the 1 H and 13 C NMR data in CD2Cl2, yet without hampering full signal assignment (Figures S22-S27).The signals of the cyclopentadienyl moieties are the most informative, with four sharp apparent quartets in the region of δ 1 H = 5.72-5.84ppm (4H), which correlate with four carbon peaks at δ 13 C = 101.7-102.8ppm, and which with the other signals at δ 1 H = 6.28-6.36ppm (4H) account for the eight individual hydrogens in the 2-5-positions at Cp and Cp´.Along with the sum of 44 detected carbon signals, this also means that either only one of the two possible isomers was formed (Scheme 3) or their presence could not be determined with NMR spectroscopy.Scheme 3. Synthesis of the asymmetric 2,2′-bisfluorene-linked dinuclear bis(ansa-zirconocene) 9 (possible mixture of two diastereomers, "syn"-and "anti"-like).
DFT computations were conducted to assess the possible geometries of the two C1symmetric, "syn"-and "anti"-like diastereomers of 9 (Figure 3) and their relative energies.These calculations returned only 0.9 kcal•mol −1 of energy difference between the two diastereomers, which corresponds to a theoretical "syn"/"anti" ratio of ca.5:1 at room temperature.Note, however, that this minimal energy difference falls within the usually accepted range of accuracy of DFT computations (2-3 kcal•mol −1 ).The two metal centers are located within the same distance (8.726 Å) in both computed diastereomers.Scheme 3. Synthesis of the asymmetric 2,2 -bisfluorene-linked dinuclear bis(ansa-zirconocene) 9 (possible mixture of two diastereomers, "syn"-and "anti"-like).
DFT computations were conducted to assess the possible geometries of the two C 1symmetric, "syn"-and "anti"-like diastereomers of 9 (Figure 3) and their relative energies.These calculations returned only 0.9 kcal•mol −1 of energy difference between the two diastereomers, which corresponds to a theoretical "syn"/"anti" ratio of ca.5:1 at room temperature.Note, however, that this minimal energy difference falls within the usually accepted range of accuracy of DFT computations (2-3 kcal•mol −1 ).The two metal centers are located within the same distance (8.726 Å) in both computed diastereomers.

Olefin (Co-)Polymerization
The dinuclear bis(ansa-zirconocene) complex 9, in combination with MAO, was evaluated in the homogeneous (co-)polymerization of ethylene, propylene, and ethylene/1hexene (toluene, 4 barg of constant pressure, 20 and 60 • C).Each polymerization experiment was repeated independently two times under the same conditions, revealing good reproducibility in terms of activity (gas uptake) and productivity (polymer yield).For comparison purposes, the catalytic performance of the structurally related mononuclear reference metallocene pre-catalysts {Me 2 C( 2,7-tBu Flu)(Cp)}ZrCl 2 (M) and {Me 2 C(Flu)(Cp)}ZrCl 2 (M ) was determined as well under the same conditions.The productivities of the catalyst are estimated from the polymer yield of the reaction for a short period (15 min) in order to avoid the mass transfer effect.
Selected ethylene polymerization results are summarized in Table 1.For complex 9, a decrease from 5000 to 2000 equivalents of MAO per metal center had no effect on the catalytic performance, and the recovered polymers exhibited similar physicochemical properties (Table 1, entries 1 and 2).The productivities and the molecular weight distributions were slightly lower or comparable to those of the benchmark mononuclear metallocene complexes M and M (entries 4 and 6, respectively).Expectedly, both pre-catalysts 9 and M showed lower polymerization productivity at 20 • C (entries 3 and 5), whereby only the polymer formed by the reference metallocene could be analyzed, showing a 3-fold increased molecular weight and slightly broader molecular weight distribution compared to the polymer produced at 60 • C (entries 4 and 5).

Olefin (Co-)Polymerization
The dinuclear bis(ansa-zirconocene) complex 9, in combination with MAO uated in the homogeneous (co-)polymerization of ethylene, propylene, and hexene (toluene, 4 barg of constant pressure, 20 and 60 °C).Each polymerizat ment was repeated independently two times under the same conditions, reve reproducibility in terms of activity (gas uptake) and productivity (polymer comparison purposes, the catalytic performance of the structurally related m reference metallocene pre-catalysts {Me2C( 2,7-tBu Flu)(Cp)}ZrCl2 ( {Me2C(Flu)(Cp)}ZrCl2 (M′) was determined as well under the same cond productivities of the catalyst are estimated from the polymer yield of the rea short period (15 min) in order to avoid the mass transfer effect.
Selected ethylene polymerization results are summarized in Table 1.For a decrease from 5000 to 2000 equivalents of MAO per metal center had no eff catalytic performance, and the recovered polymers exhibited similar physi properties (Table 1, entries 1 and 2).The productivities and the molecular we butions were slightly lower or comparable to those of the benchmark mononuc locene complexes M and M′ (entries 4 and 6, respectively).Expectedly, both pr 9 and M showed lower polymerization productivity at 20 °C (entries 3 and 5 only the polymer formed by the reference metallocene could be analyzed, sh fold increased molecular weight and slightly broader molecular weight distrib pared to the polymer produced at 60 °C (entries 4 and 5)."syn"-like diastereomer; bottom: "anti"-like diastereomer).Computed Zr . . .Zr distances: 8.726 Å for the "syn"-like diastereomer; 8.727 Å for the "anti"-like diastereomer.

Entry
Precat.The performance of pre-catalysts 9, M, and M in propylene polymerization is summarized in Table 2.A decrease in the [MAO]/[Zr] ratio from 5000 to 2000 did not affect the productivity of pre-catalyst 9 but led to a significant increase in the molecular weight of the resulting PP (9.9 vs. 14 kg mol −1 , entries 7 and 8), while its dispersity stayed constant.A possible explanation for the observed increase in molecular weight at the low [MAO]/[Zr] ratio is the lower amount of AlMe 3 present in the used MAO solution, which is known to act as a chain-transfer reagent.At the polymerization temperature of 60 • C, the dinuclear 9 featured inferior productivity compared to the mononuclear M (5690 vs. 25,870 kg(PP) mol(Zr) −1 h −1 , entries 7 and 10, respectively), though being comparable to that of M (4550 kg(PP) mol(Zr) −1 h −1 , entry 12).Yet, both metallocenes 9 and M featured nearly the same productivity at a polymerization temperature of 20 • C (8300 vs. 7800 kg(PP) mol(Zr) −1 h −1 , entries 11 and 9).As expected, the dinuclear complex 9 produces highly syndiotactic polypropylene similar to the well-investigated benchmark pre-catalyst M; however, its stereoregulation is slightly lower at both polymerization temperatures ([r] at 60 • C: 90.4 vs. 92.2%;at 20 • C: 92.9 vs. 96.8%).Table 3 summarizes the performance observed in ethylene/1-hexene copolymerization.The productivity of the catalytic system based on dinuclear metallocene 9 and the molecular weights of the produced copolymers were affected by the [MAO]/[Zr] ratio (entries 13 and 14); this is comparable to the observations made regarding propylene polymerization.Again, this finding might be due to the variable amount of AlMe 3 acting as a chaintransfer reagent.The productivity of the dinuclear system decreased drastically when copolymerization was carried out at 20 • C, similar to the observations made during ethylene homopolymerization.The copolymerization productivity of the mononuclear catalytic system based on M was less affected by temperature, and this tendency conforms more with the observations made during the propylene polymerization.The copolymer produced with 9/MAO at 20 • C had about the same C 6 content as that obtained from the M system (4.3 and 4.7 mol-%, respectively); yet, the former material had a much broader dispersity than any other copolymer (M w /M n = 3.7 vs. 2.6, respectively), and two distinct melting transitions were observed in the DSC trace (see Figure S45).Although the molecular weight distribution of this material remained apparently monomodal, as indicated by the SEC trace (see Figure S35), these observations suggest the presence of two types of macromolecules and, possibly, that the two centers in the catalyst derived from 9 operated in a differentiated manner.

General Considerations
All manipulations (except polymerizations) were performed under a purified argon atmosphere using standard Schlenk techniques or in a glovebox.Solvents were distilled from Na/benzophenone (THF, Et 2 O) and Na/K alloy (toluene, pentane) under nitrogen, degassed thoroughly, and stored under nitrogen prior to use.C 6 D 6 (>99.5% D, Euroisotop) was vacuum-transferred from Na/K alloy into a storage tube.CDCl 3 and CD 2 Cl 2 were kept over CaH 2 and vacuum-transferred before use.MAO (30 wt-% solution in toluene, Albermale; contains ca. 10 wt-% of free AlMe 3 ) was used as received.Other starting materials were purchased from Alfa, Strem, Acros, and Aldrich and used as received.

Instruments and Measurements
NMR spectra of the organic compounds and complex 9 were recorded, respectively, in regular and Teflon-valved NMR tubes on Bruker AM-300 and AM-400 spectrometers (Bruker AXS Handheld Inc., Kennewick, WA, USA) at 25 • C. Chemical shifts are reported in ppm.Assignment of the resonances was made from 2D 1 H− 1 H COSY, 1 H− 13 C HSQC, and HMBC NMR experiments.Coupling constants are given in Hertz.Elemental analyses (C, H, N) were performed using a Flash EA1112 CHNS Thermo Electron apparatus (Thermo Finnigan Italia S.p.A., Rodano, Italy).DSC measurements were performed on a SETARAM Instrumentation DSC131 differential scanning calorimeter at heating rate of 10 • C•min −1 ; first and second runs were recorded after cooling to 30 • C; the reported melting temperatures correspond to the second run.SEC analyses of polymer samples were carried out in 1,2,4-trichlorobenzene at 135 • C at the TotalEnergies research center in Feluy (Belgium) using polystyrene standards for universal calibration. 13C NMR analyses of polypropylene and poly(ethylene-co-1-hexene) samples were run on a Bruker AM-500 spectrometer (TotalEnergies, Feluy, Belgium) as follows: solutions of ca.200 mg of polymer in trichlorobenzene/C 6 D 6 mixture at 135 • C in 10 mm tubes, inverse-gated experiment, pulse angle = 90 • , delay = 11 s, acquisition time = 1.25 s, number of scans = 6000.

Computational Studies
The calculations were performed using the Gaussian 09 [30] program employing B3PW91 [31,32] functional and using a standard split-valence basis set def2-SVP [33].The solvent effects, in our case for diethyl ether, were taken into account during all the calculations using the SMD model [34].All stationary points were fully characterized via analytical frequency calculations as true minima (all positive eigenvalues).Zero-point vibrational energy corrections (ZPVEs) were estimated with a frequency calculation at the same level of theory, to be considered for the calculation of the total energy values at T = 298 K.

Conclusions
We have developed an effective synthetic strategy to access asymmetric difluorenyllinked dinuclear bis(ansa-metallocene)s.As a first prototypical example, a di-ansa-{Cp/Flu-Cp/Flu } dizirconocene complex has been synthesized.Preliminary investigations on the performance of this bimetallic pre-catalyst in ethylene, propylene, and ethylene/1-hexene (co-)polymerization revealed, in most cases, lower productivity in comparison with a mononuclear reference pre-catalyst {Me 2 C( 2,7-tBu Flu)(Cp)}ZrCl 2 ; this might have resulted from a higher steric hindrance in the dinuclear bis(ansa-metallocene) due to the direct linkage in the difluorenyl platform [9].On the other hand, only slight differences were noticed between the two pre-catalyst systems in terms of molecular weights, tacticity, and comonomer incorporation in the resulting (co-)polymers.No bimodal molecular weight distribution was achieved at any produced polymer.This might have originated from the too-close similarity of the connected Cp/Flu moieties or a rapid chain-transfer phenomenon between the two different active sites which were at a quite close distance from each other.To gain further insights, the synthesis of proligands and the corresponding dinuclear bis(ansa-metallocene)s with more differentiated catalytically active metal centers are the subject of current investigations.
Author Contributions: L.N.J. designed and performed the experimental work and wrote the preliminary draft.T.R. conducted the crystallographic studies.V.C. and A.W. co-supervised the work and contributed to the evolution of the initial project.E.K. conducted the DFT studies.J.-F.C. edited the initial draft.E.K. and J.-F.C. directed the research and finalized the manuscript.All authors contributed to various sections of the manuscript.All authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by TotalEnergies (postdoctoral fellowship to LNJ).Data Availability Statement: All available data have been made available through the Supplementary Material.

Chart 1 .
Examples of common dinuclear platforms encountered in group 4 metallocene-type precatalysts.

Chart 1 .
Examples of common dinuclear platforms encountered in group 4 metallocene-type pre-catalysts.

Figure 1 .
Figure 1.Molecular structure of the asymmetric 2,2´-bifluorene 5. Atomic displacement parameters are set at the 50% probability level.Hydrogen atoms are omitted for clarity.

Figure 1 .
Figure 1.Molecular structure of the asymmetric 2,2´-bifluorene 5. Atomic displacement parameters are set at the 50% probability level.Hydrogen atoms are omitted for clarity.

Figure 1 .
Figure 1.Molecular structure of the asymmetric 2,2 -bifluorene 5. Atomic displacement parameters are set at the 50% probability level.Hydrogen atoms are omitted for clarity.
C HSQC and HMBC NMR experiments.The observation of three signals for the 9-H Flu hydrogens (δH 4.32-4.20)and four signals for the Cp methylene hydrogens at δH 3.28 and 3.15 ppm in the 1 H NMR spectrum likely accounts for the presence of stereogenic C9 centers (hence two diastereomers) and/or the presence of four possible tautomers of the C = C double bonds in the two different Cp rings.