Bimetallic Zr,Zr-Hydride Complexes in Zirconocene Catalyzed Alkene Dimerization

Being valuable precursors in the production of adhesives, lubricants, and other high-performance synthetic compounds, alkene dimers and oligomers can be obtained using homogeneous zirconocene catalytic systems. Further advances in such systems require precise control of their activity and chemoselectivity, increasing both the purity and yield of the products. This relies on the process mechanism usually built around the consideration of the hydride complexes as active intermediates in the alkene di- and oligomerization; however, the majority of studies lack the direct evidence of their involvement. Parallel studies on a well-known Cp2ZrCl2-AlR3 or HAlBui2 and a novel [Cp2ZrH2]2-ClAlR2 (R = Me, Et, Bui) systems activated by methylaluminoxane (MMAO-12) have shown a deep similarity both in the catalytic performance and intermediate composition. As a result of the NMR studies, among all the intermediates considered, we proved that new Zr,Zr- hydride complexes having the type x[Cp2ZrH2∙Cp2ZrHCl∙ClAlR2]∙yMAO appear to be specifically responsible for the alkene dimerization with high yield.

Moreover, Marks et al. showed that the zirconium hydride complexes modified with B(C6F5)3 are active in ethylene and propylene polymerization [37,38]. Collins et al. demonstrated significantly higher activity of the complexes (RCp)2ZrH3AlH2 (R = Bu n , TMS) activated with MAO in ethylene polymerization than the corresponding zirconocene dichlorides [39]. Binuclear hydride clusters complexed with organoboron compounds [40][41][42] were found to be highly active initiators of isobutene homopolymerization and isobutene-isoprene copolymerization [42]. It is generally accepted that the activating effect of MAO or organoboron compounds on di-, oligo-and polymerization systems is connected with the formation of highly reactive cationic centers of type [L2M-H] + [34,35,43].
Furthermore, quantum chemical modeling of the possible active sites governing the alkene polymerization also supports the proposed formation of the hydride intermediates and further participation in the reactions. For example, a DFT study of the initial stages of the methyl vinyl ether (MVE) polymerization which runs under the catalytic action of [Me2C(Cp)2Zr(Me)] + and [Me2C(Cp)2Zr(H)] + cations showed that it is the zirconium hydride which preferably initiates the polymer chain growth [44]. In Reference [43], devoted to the modeling of the propylene dimerization and oligomerization processes, the crucial role of the chlorine atom as a structural unit of the active Zr,Al-binuclear hydride intermediates was shown, and the efficiency of the molecular hydrogen as the activator that increases the rate and selectivity of the dimerization was revealed.
Our recent studies regarding the structure of the intermediates formed in the reaction of L2ZrCl2 with XAlBu i 2 (X = H, Cl, Bu i ) [32], as well as the complexes generated in zirconocene dihydride-ClAlR2 (R = Et, Bu i )-methylaluminoxane (MMAO-12) systems [45], showed that both systems L2ZrCl2-XAlBui2 and [Cp2ZrH2]2-ClAlR2 exhibit almost the same set of Zr,Al-hydride complexes. However, in the reaction of [Cp2ZrH2]2 with ClAlR2 (R = Et, Bu i ), new bimetallic Zr,Zr-hydride complexes that can chemically bind to MAO were observed [45].
The goal of this research was to find the conditions for selective alkene dimerization in two catalytic systems: Cp2ZrCl2-(AlR3 or HAlBu i 2 (R = Me, Et, Bu i ))-MMAO-12, and [Cp2ZrH2]2-ClAlR2 (R = Me, Et, Bu i )-MMAO-12, and to reveal the structure of the zirconium hydride intermediates that initiate the alkene transformations. Moreover, Marks et al. showed that the zirconium hydride complexes modified with B(C 6 F 5 ) 3 are active in ethylene and propylene polymerization [37,38]. Collins et al. demonstrated significantly higher activity of the complexes (RCp) 2 ZrH 3 AlH 2 (R = Bu n , TMS) activated with MAO in ethylene polymerization than the corresponding zirconocene dichlorides [39]. Binuclear hydride clusters complexed with organoboron compounds [40][41][42] were found to be highly active initiators of isobutene homopolymerization and isobutene-isoprene copolymerization [42]. It is generally accepted that the activating effect of MAO or organoboron compounds on di-, oligo-and polymerization systems is connected with the formation of highly reactive cationic centers of type [L 2 M-H] + [34,35,43].
Furthermore, quantum chemical modeling of the possible active sites governing the alkene polymerization also supports the proposed formation of the hydride intermediates and further participation in the reactions. For example, a DFT study of the initial stages of the methyl vinyl ether (MVE) polymerization which runs under the catalytic action of [Me 2 C(Cp) 2 Zr(Me)] + and [Me 2 C(Cp) 2 Zr(H)] + cations showed that it is the zirconium hydride which preferably initiates the polymer chain growth [44]. In Reference [43], devoted to the modeling of the propylene dimerization and oligomerization processes, the crucial role of the chlorine atom as a structural unit of the active Zr,Al-binuclear hydride intermediates was shown, and the efficiency of the molecular hydrogen as the activator that increases the rate and selectivity of the dimerization was revealed.
The goal of this research was to find the conditions for selective alkene dimerization in two catalytic systems: Cp 2 ZrCl 2 -(AlR 3 or HAlBu i 2 (R = Me, Et, Bu i ))-MMAO-12, and [Cp 2 ZrH 2 ] 2 -ClAlR 2 (R = Me, Et, Bu i )-MMAO-12, and to reveal the structure of the zirconium hydride intermediates that initiate the alkene transformations. On the first step of the catalytic experiments, we studied the effect of MMAO-12 on the activity and chemoselectivity of systems Cp 2 ZrCl 2 -XAlBu i 2 (X = H, Bu i ) that hydroaluminate the terminal alkenes in the absence of the activator [7]. As shown elsewhere, the catalytic system based on HAlBu i 2 and zirconocene dichloride demonstrates low activity in the alkene hydroalumination (hydroalumination product (2)); this experiment was taken as a reference point (Scheme 2, Table 1 (entry 1)) [30][31][32]. Addition of 30-240 eq. of methylaluminoxane to the system does not affect notably the alkene conversion. For example, in the system L 2 ZrCl 2 -HAlBu i 2 -MMAO-12-

Study of 1-Alkene Transformations in Catalytic Systems Cp2ZrY2 (Y = Cl, H)-ОАС-MMAO-12
2.1.1. Activity and Chemoselectivity of System Cp2ZrCl2-(AlR3 or HAlBu i 2)-MMAO-12 with Respect to Alkenes On the first step of the catalytic experiments, we studied the effect of MMAO-12 on the activity and chemoselectivity of systems Cp2ZrCl2-XAlBu i 2 (X = H, Bu i ) that hydroaluminate the terminal alkenes in the absence of the activator [7]. As shown elsewhere, the catalytic system based on HAlBu i 2 and zirconocene dichloride demonstrates low activity in the alkene hydroalumination (hydroalumination product (2)); this experiment was taken as a reference point (Scheme 2, Table 1 (entry 1)) [30][31][32]. Addition of 30-240 eq. of methylaluminoxane to the system does not affect notably the alkene conversion. For example, in the system L2ZrCl2-HAlBu i 2-MMAO-12-1-octene at the ratio  Table  1, (entry 4)).

Scheme 2.
Alkene transformations in catalytic systems I or II.
Unlike HAlBu i 2, triisobutylaluminium is an effective reagent for 1-alkene hydrometalation catalyzed by Cp2ZrCl2 [30,32,46]. As another reference, the hydroalumination products after 3 h of the reaction were obtained at 83% yield (Table 1, entry 2) [32]. The addition of MMAO-12 increases the conversion of 1-alkene up to 98% due to the generation of both hydroalumination (2) and dimerization (4) products (Table 1, entry 3). Meanwhile, minor product 3, the precursor of dimer 4, was observed in this system due to the formation of the metal alkyl, which appeared after incorporation of the second alkene molecule into hydrozirconation product (Scheme 1). Rising the MMAO-12 content from 30 to 240 eq. increases the yield of 1-octene dimers (4b) from 5% to 40% (Table 1, entry 5). Similarly to the case with HAlBu i 2, a decrease in the AlBu i 3 concentration leads to the selective formation of the alkene dimers (97%) within 20 min of the reaction at 20 °C (Table 1, entry 6). A further increase in the temperature accelerates the reaction and helps to reduce the concentration of the catalytic system components without loss of both the chemoselectivity and the main product yield (entries 8,9). These results are consistent with the data obtained for catalytic system L2ZrCl2-AlBu i 3-ClAlR2-MAO (1:20:(1-2):10, 0.05 mol% Zr, 60°C, 1-4 h), in which aluminum alkyl AlBu i 3 and aluminum chloride ClAlR2 (R = Me, Et) activate the system that produces dimers with a high yield up to 94% [5,24,43].
Moreover, the conditions that we proved to be effective for the selective 1-octene dimerization were extended to 1-hexene treatment (entries 10,11). These conditions also provide high substrate conversion; however, the product distribution slightly changes due to the appearance of the hydrometalation (2) and alkylation (6) products, and the dimer yield becomes 89%-91%. Unlike HAlBu i 2 , triisobutylaluminium is an effective reagent for 1-alkene hydrometalation catalyzed by Cp 2 ZrCl 2 [30,32,46]. As another reference, the hydroalumination products after 3 h of the reaction were obtained at 83% yield (Table 1, entry 2) [32]. The addition of MMAO-12 increases the conversion of 1-alkene up to 98% due to the generation of both hydroalumination (2) and dimerization (4) products (Table 1, entry 3). Meanwhile, minor product 3, the precursor of dimer 4, was observed in this system due to the formation of the metal alkyl, which appeared after incorporation of the second alkene molecule into hydrozirconation product (Scheme 1). Rising the MMAO-12 content from 30 to 240 eq. increases the yield of 1-octene dimers (4b) from 5% to 40% (Table 1, entry 5). Similarly to the case with HAlBu i 2 , a decrease in the AlBu i 3 concentration leads to the selective formation of the alkene dimers (97%) within 20 min of the reaction at 20 • C (Table 1, entry 6). A further increase in the temperature accelerates the reaction and helps to reduce the concentration of the catalytic system components without loss of both the chemoselectivity and the main product yield (entries 8,9). These results are consistent with the data obtained for catalytic system L 2 ZrCl 2 -AlBu i 3 -ClAlR 2 -MAO (1:20:(1-2):10, 0.05 mol% Zr, 60 • C, 1-4 h), in which aluminum alkyl AlBu i 3 and aluminum chloride ClAlR 2 (R = Me, Et) activate the system that produces dimers with a high yield up to 94% [5,24,43].
Moreover, the conditions that we proved to be effective for the selective 1-octene dimerization were extended to 1-hexene treatment (entries 10,11). These conditions also provide high substrate conversion; however, the product distribution slightly changes due to the appearance of the hydrometalation (2) and alkylation (6) products, and the dimer yield becomes 89%-91%. Further, we studied the influence of different organoaluminum compounds (OAC) on the activity and chemoselectivity of the catalytic system; for this purpose, the isobutylalanes were replaced consequently by AlMe 3 and AlEt 3 . The application of these OACs at 40 • C slightly reduces the alkene conversion to 91%-92% (entries 12,13). In the case of AlEt 3 , a decrease of the dimer yield to 68% was observed due to the increased fraction of the alkylated monomer 6 in the products (entry 13). The formation of 6 is possible through the alkene carbometalation product 5 which is generated by either methyl or ethyl zirconocenes formed during the stage of the alkyl-chloride exchange between Cp 2 ZrCl 2 and either AlMe 3 or AlEt 3 , respectively. It should be emphasized that for the systems like Cp 2 ZrCl 2 -AlMe 3 , these stages, finished by-product 6, are very important since they provide zirconocene hydrides which are the source of dimers 4. Rising of the temperature to 60 • C both increases the substrate conversion and narrows the selectivity towards the dimerization, regardless of the OAC nature (entries [14][15][16][17][18][19][20][21]. Moreover, the amount of the substrate can be increased to 1000 eq. in the case of HAlBu i 2 without loss of the dimer yield (98%, entry 22 On the second step of the catalytic experiments, we studied the performance of system II [Cp 2 ZrH 2 ] 2 -ClAlR 2 -MMAO-12 (R = Me, Et, Bu i ) in the reaction with alkenes having various structures (1a-f). In all the experiments, the main product was dimer 4 (Scheme 2, Table 2).
Monitoring  (Table 2, entry 6). A reaction in the presence of dimethyl-or diethylaluminum chlorides is accompanied by the formation of hydro-and carboalumination products with a yield of 22%-23% (Table 2, entries 3,4). As a result, systems based on [Cp 2 ZrH 2 ] 2 , MMAO-12, and ClAlBu i 2 , compared to other ClAlR 2 , were found to be more active and selective. Diisobutylaluminum chloride also showed its effectiveness in the dimerization of other linear substrates: 1-octene and 1-decene. The yield of dimers 4b,c obtained within 3 h at a temperature of 20 • C was 72%-98% (entries 7-18).
When the temperature is elevated to 40 • C, the reaction accelerates, and the induction period vanishes. The yield of 1-hexene and 1-octene dimers reaches 73%-91% in 15 min (entries [19][20][21][22][23][24]. It should be noted that, in this case, the dependency of the system activity on the OAC nature disappears. Moreover, under these conditions, the conversion of 4-methyl-1-pentene (1d) is over 97%, and the yield of the dimerization products is 95% (entry 25). The temperature increase to 60 • C selectively provides 1-hexene dimers at 91%-94% yield within 5 min of the reaction (lines 26-28). Increasing 1-alkene initial concentration to 500 equivalents does not affect the degree of its transformation at 60 • C, and the yield of the dimerization products remains sufficiently high (85%-90%) (entry 32). A significant reduction in the substrate conversion and dimer yield (65% and 57%, correspondingly) was observed only at 1000 equivalents of the alkene (entry 33).
Moreover, the relative amount of MMAO-12 in the catalytic system also affects the activity only up to a certain level. The increase in the MMAO-12 content to 60 eq. (at 20 • C) accelerates the reaction and provides the dimers with a yield of 86% (entries [34][35][36]. Further increase of MMAO-12 content up to 120 eq. does not change both the catalytic activity and the dimerization product yield (entry 37).  It was established that aryl-substituted alkenes 1e and 1f can undergo the dimerization at an elevated temperature of 100 • C. Within 60 min, the dimers were obtained with yields of 61% and 58%, respectively. However, the proximity of the Ar group to the double bond leads to a loss of the reaction regioselectivity. The ratio of the regioisomers of the styrene dimer head-to-tail to tail-to-tail reaches 1.23:1 (entry 39).
Finally, the catalytic systems consisting of zirconocene dihydride, dialkylaluminum chlorides, and MMAO-12, work just as well as the systems based on zirconocene dichloride. Therefore, to identify the intermediates responsible for the alkene dimerization in both systems, we studied the structure and activity of the hydride complexes by the means of NMR spectroscopy.

NMR Study of Hydride Intermediate Structures in Systems Cp 2 ZrY 2 (Y = Cl, H)-OAC-MMAO-12
Earlier, we showed that the system Cp 2 ZrCl 2 -XAlBu i 2 can generate certain bimetallic hydride complexes; among them, the most active species in the hydroalumination reaction are the ones which have an open Zr-H bond [32]. The NMR study showed that in the Cp 2 ZrCl 2 reaction with AlBu i 3 (1:5), alkyl chloride complex 7 is produced and then transformed into complexes 8 and 10c, both undergoing intermolecular exchange via intermediate 9 [30,32] (Scheme 3, Figure 1). In this work, the introduction of MMAO-12 into the reaction mixture, in the 1 H NMR spectra, gives rise to both a doublet at −1.22 ppm (J = 17.6 Hz) and a triplet at −6.40 ppm (J = 17.6 Hz) which appear to be correlated in the COSY HH spectrum ( Figure S5). Moreover, those signals are connected with a singlet of the Cp-rings at 5.50 ppm. The ratio of the signal intensities 2 (Zr-H):1 (Zr-H):20 (Cp) indicates the presence of two ZrCp 2 fragments in the molecule. With an increase in the MMAO-12 concentration, additional doublet at −1.23 ppm and triplet at −6.59 ppm are observed; also, binary Cp-ring signals appear in the downfield part of the spectrum at 5.46 and 5.48 ppm. Taking into account these data, as well as the results obtained earlier [45], the structure of these complexes was assigned as 12 and is presented in Scheme 3. As we have previously shown [45], this type of complexes is formed in the reaction of [Cp 2 ZrH 2 ] 2 with ClAlR 2 (R = Et, Bu i ). In intermediates 12a-c, there is no intermolecular exchange with hydride atoms; this indicates higher stability of these structures compared to complexes 8 and 10c.
The reaction of Cp 2 ZrCl 2 with HAlBu i 2 is accompanied by the formation of the known complexes 10c [29,30,32] and 11c [30,31]. With reduced HAlBu i 2 content down to 2-3 equivalents (when the initial Cp 2 ZrCl 2 remains unreacted), complex 12c becomes observable; moreover, the complex is formed in this system even in the absence of MMAO-12 ( Figures S7 and S8). Similar to the system with AlBu i 3 , addition of MMAO-12 increases the relative amount of complex 12c, and a new set of signals similar to 12c appear in the NMR spectrum. This phenomenon is possible due to the exchange reaction with AlMe 3 contained in MMAO-12, as we described earlier [45].
Zirconocene dihydride reacts with ClAlMe 2 in a ratio of 1:3 providing a mixture of complexes: 10a, 12a, 13a, and Cp 2 ZrCl 2 (Scheme 3, Figure 2a, Table 3). The structure of complex 10a was identified by analogy with 10b,c [45]. A pair of signals at −0.60 ppm (Zr-H-Al) and 5.75 ppm (Cp) having an intensity ratio of 2:10 were assigned to complex 13a [33]. The broadened nature of the signals belonging to the hydride atom and cyclopentadienyl rings indicates the participation of 13a in the intermolecular exchange; a possible reaction pathway is shown in Scheme 3. Complex 12a is characterized by both triplet and doublet signals at −6.64 ppm (J = 17.6 Hz) and −1.19 ppm (J = 17.6 Hz); these signals relate to that of Cp-rings at 5.52 ppm as 1:2:20. Probably, this complex is a result of the replacement of one hydride atom in the zirconocene dihydride dimer with chlorine. The dialkylaluminum hydride, produced as a result of the chloride-hydride exchange, reacts with Cp 2 ZrH 2 and ClAlR 2 and provides trihydride complexes 10a-c.     Assessment of the NOESY spectrum for 12b opened the possibility to clarify the structure of complexes 12a-c. The spectrum exhibits cross-peaks between the signals of Cp-ring protons and hydride atoms with a quartet signal of ethyl group protons at 0.19 ppm (Figure 3); the relative intensity of the signals indicate the presence of no more than one ClAlEt2 molecule in the structure of 12b.  [45]; (c) ClAlBu i 2 [45].
Assessment of the NOESY spectrum for 12b opened the possibility to clarify the structure of complexes 12a-c. The spectrum exhibits cross-peaks between the signals of Cp-ring protons and hydride atoms with a quartet signal of ethyl group protons at 0.19 ppm (Figure 3); the relative intensity of the signals indicate the presence of no more than one ClAlEt 2 molecule in the structure of 12b. Assessment of the NOESY spectrum for 12b opened the possibility to clarify the structure of complexes 12a-c. The spectrum exhibits cross-peaks between the signals of Cp-ring protons and hydride atoms with a quartet signal of ethyl group protons at 0.19 ppm (Figure 3); the relative intensity of the signals indicate the presence of no more than one ClAlEt2 molecule in the structure of 12b.   Figure 4a) were observed. In the case of complex 12a, the 1 H NMR spectrum exhibited both triplet and doublet signals at −6.56 and −1.08 ppm, respectively, as well as broadened signals at −6.94 and −1.43 ppm. These signals correspond to a broad peak of cyclopentadienyl rings at 5.11-5.33 ppm. Within a few minutes, the opacity of the homogeneous solution and separation of a heavy phase was detected. This corresponds to the previously observed effect for 12b,c [45]. As follows from the NOESY experiment ( Figure S17), adducts 12a·MAO show a negative NOE effect intrinsic to macromolecular compounds [47,48]. The signal broadening could be attributed to the high molecular weight of the particles. The substantial difference in diffusion coefficients between MMAO-12 derivatives ( Figure S15) may indicate a large increase in the molecular weight and particle volume in the course of a heavy fraction formation. The latter is possible in the case of the intermolecular binding of MMAO-12 oligomers by the complexes. As a result, it was found that among a large set of bimetallic hydride complexes, only complexes 12a-c were able to bind chemically to  Probably, the ability of the complexes to bind with the activator depends on the dynamic stability of the structures, and on the possibility to replace the organoaluminum fragment in the molecules. Moreover, the observation of adducts formed via covalent binding with methylaluminoxane suggests the presence of a considerable number of accessible three-coordinated Lewis-acidic aluminum centers in the activator. Since MMAO-12 acts similarly to trimethylaluminum or ClAlR 2 , which can substitute AOC in the complexes 12a-c, it can be assumed that there is a sufficient number of OAlMe 2 groups in the methylaluminoxane structure [49][50][51][52]. stability of the structures, and on the possibility to replace the organoaluminum fragment in the molecules. Moreover, the observation of adducts formed via covalent binding with methylaluminoxane suggests the presence of a considerable number of accessible three-coordinated Lewis-acidic aluminum centers in the activator. Since MMAO-12 acts similarly to trimethylaluminum or ClAlR2, which can substitute AOC in the complexes 12a-c, it can be assumed that there is a sufficient number of OAlMe2 groups in the methylaluminoxane structure [49][50][51][52]. Using the complexes obtained in the [Cp2ZrH2]2-ClAlBu i 2 system as an example, we showed that the addition of 1-hexene to a solution containing a mixture of complexes 10c-13c in the absence of MMAO-12 is accompanied with the appearance of hydrometalation products 2a: zirconocene alkyl chloride and aluminum alkyl (Scheme 4); their identification was carried out on the base of NMR data [30] and mass spectrometry of the deuterolysis products. For this reaction, complexes 10c and Using the complexes obtained in the [Cp 2 ZrH 2 ] 2 -ClAlBu i 2 system as an example, we showed that the addition of 1-hexene to a solution containing a mixture of complexes 10c-13c in the absence of MMAO-12 is accompanied with the appearance of hydrometalation products 2a: zirconocene alkyl chloride and aluminum alkyl (Scheme 4); their identification was carried out on the base of NMR data [30] and mass spectrometry of the deuterolysis products. For this reaction, complexes 10c and 11c interacted first, whereas 12c was no longer observed in the spectra only after 4 h of the process ( Figure S19). As follows from Figure 5, in the presence of MMAO-12, when the system contains MAO-adducts along with the complexes 10c, 12c, 13c, complex 10 is consumed quickly, providing hydrometalation products 2a. The formation of the dimers is accompanied by the vanishing of the MAO-adducts of complex 12; especially their heavy fraction is consumed first. 11c interacted first, whereas 12c was no longer observed in the spectra only after 4 h of the process ( Figure S19). As follows from Figure 5, in the presence of MMAO-12, when the system contains MAOadducts along with the complexes 10с, 12с, 13с, complex 10 is consumed quickly, providing hydrometalation products 2a. The formation of the dimers is accompanied by the vanishing of the MAO-adducts of complex 12; especially their heavy fraction is consumed first.

Discussion
Our study of the catalytic system Cp2ZrCl2-OAC-MMAO-12 proved the ability of this type of system to dimerize 1-alkenes with high chemoselectivity to give head-to-tail products with a

Discussion
Our study of the catalytic system Cp 2 ZrCl 2 -OAC-MMAO-12 proved the ability of this type of system to dimerize 1-alkenes with high chemoselectivity to give head-to-tail products with a vinylidene moiety > C=CH 2 [5,[9][10][11][12]24,53]. The other Ti subgroup metal compounds, for example, postmetallocene Zr and Hf complexes with [ONNO]-type amine bis(phenolate) ligands, activated by B(C 6 F 5 ) 3 , catalyze the 1-hexene oligomerization with the same regioselectivity [54]. Among these complexes, hafnium catalysts showed the best selectivity towards the dimers (up to 97%); moreover, the molecular weight distribution of the oligomers, obtained in Hf-catalyzed reactions, do not correspond to a typical Schulz−Flory regularity. Similar zirconium complexes with aryl-substituted [OSSO]-type bis(phenolate) ligands on the trans-cyclooctanediyl platform, activated by dried modified methylaluminoxane (dMMAO), catalyze 1,2-regioselective oligomerization of 1-hexene at relatively low catalyst loadings and produce preferably head-to-tail dimer. The use of other transition metal complexes leads, first of all, to changes in the reaction regioselectivity. For example, the systems based on Fe [55] and Co [56,57] complexes and Al-coactivators provide linear head-to-head dimers, whereas W catalysts [58] give predominantly methyl-and dimethyl-branched tail-to-tail products. In these studies, the participation of the transition metal hydride complexes as active centers is assumed as given.
However, in the literature, information regarding the direct evidence of the alkene dimerization under the action of metal hydrides is very limited [59,60]. For example, it was shown that the binuclear hydride complex [(Ind') 2 Y(µ-H)] 2 catalyzes regio-and stereoselective homodimerization of various α-olefins at 80 • C, as well as head-to-head codimerization of styrene with other α-olefins [59]. In this work, we show that the activity and chemoselectivity of the systems based on zirconocene dichloride and dihydride are comparable to each other; this outlines the zirconium hydrides as the probable precursors of active intermediates of the dimerization reaction.
As follows from the NMR study, Zr,Al-hydride clusters 10a-c, even in the presence of MMAO-12 activator, give only hydrometalation products, while the dimerization pathway is determined by biszirconium hydride complexes 12a-c. The comparison of the action of complexes 10 and 12 shows that the first stage of the reaction, as presented in Scheme 1, is the hydrometalation stage. Further, the process flow is possible in two directions: either transmetalation with the transfer of the alkyl fragment to aluminum or the introduction of a second alkene molecule into the hydrozirconation product with the formation of the dimer. Therefore, for the bimetallic Zr,Al-hydride complexes 10a-c, the processes of the alkene hydrometalation and subsequent transmetalation proceed in the most rapid manner. Indeed, as shown by quantum chemical modeling [61], the limiting stages of the reaction are either the stage of the alkene incorporation into zirconocene hydrochloride (∆G = 10.4 kcal/mol), or the stages of Zr−Cl and Zr−H bridge bond breaking in Zr,Al-bimetallic complexes (17.8 kcal/mol and 19.8 kcal/mol) before the alkene coordination. The transmetalation stages have insignificant or even no activation barriers. Complex 12 is also, apparently, like 10, capable for the alkene hydrometalation; however, its activation with methylaluminoxane gives active species which selectively provide dimers. Specifically, the stages of the consecutive incorporation of the two alkene molecules and the chain termination through β-H elimination proceed fast, and without the direct involvement of the organoaluminum fragment which, when located close to hydride atoms, would facilitate chain transfer to aluminum.
In broad understanding, this case can be considered as an example of biszirconium catalysis similar to the alkene polymerization in the presence of Group 4 bimetallic complexes [62][63][64]. Further development of this work into establishing the activation mechanism of the biszirconium hydride complexes can be fruitful in uncovering the reasons for the selective formation of alkene diand oligomers.
The yields of compounds 2-5 were determined from the yields of hydrolysis and deuterolysis products, which were calculated relative to the amount of the initial olefin. The products were analyzed using a gas chromatograph mass spectrometer GCMS-QP2010 Ultra (Shimadzu, Tokyo, Japan) equipped with the GC-2010 Plus chromatograph (Shimadzu, Tokyo, Japan), TD-20 thermal desorber (Shimadzu, Tokyo, Japan), and an ultrafast quadrupole mass-selective detector (Shimadzu, Tokyo, Japan).

Conclusions
In summary, our studies on the alkene transformations under the action of MAO-activated systems Cp 2 ZrCl 2 -(AlR 3 or HAlBu i 2 ) and [Cp 2 ZrH 2 ] 2 -ClAlR 2 (R = Me, Et, Bu i ) show their capability to provide dimeric products with high yield and selectivity at the low content of the OAC and methylaluminoxane. Parallel studies on the systems have shown a deep similarity both in the catalytic performance and intermediate composition. As a result of the NMR studies, among all the intermediates considered, we proved for the first time that new Zr,Zr-hydride complexes having type x[Cp 2 ZrH 2 ·Cp 2 ZrHCl·ClAlR 2 ]·yMAO appear to be specifically responsible for the alkene dimerization. Further study of the reaction mechanism will uncover the process of activation of biszirconium complexes by methylaluminoxane, and this will explain the unique selectivity of these intermediates in the dimerization pathway.