Ethylene dimerization and oligomerization using Bis(phosphino)boryl supported Ni complexes

Depósito de investigación de la Universidad de Sevilla https://idus.us.es/ "This document is the Accepted Manuscript version of a Published Work that appeared in final form in Journal of the American Chemical Society, copyright © American Chemical Society after peer review and technical editing by the publisher. To access the final edited and published work see https://doi.org/10.1021/jacs.2c09471 ." 1 Ethylene Dimerization and Oligomerization Using Bis(phosphino)boryl Supported Ni Complexes Fanji Kong†1, Pablo Ríos†2, Conner Hauck1, Francisco José Fernández-de-Córdova2, Diane A. Dickie1, Laurel G. Habgood3, Amor Rodríguez2* and T. Brent Gunnoe1* †These authors contributed equally to this work. 1Department of Chemistry, University of Virginia; Charlottesville, Virginia 22904, United States. 2Instituto de Investigaciones Quimicas (IIQ), Department of Inorganic Chemistry CSIC and. University of Seville, Center for Innovation in Advanced Chemistry (ORFEO-CINQA), C/Américo Vespucio 49, Seville, 41092, Spain. 3Department of Chemistry, Rollins College, Winter Park, Florida 32789, United States. *Corresponding author. Email: [email protected] *Corresponding author. Email: marodrig[email protected] Abstract We report the dimerization and oligomerization of ethylene using bis(phosphino)boryl supported Ni(II) complexes as catalyst precursors. By using alkylaluminum(III) compounds or other Lewis acid additives, Ni(II) complexes of the type (RPBP)NiBr (R = tBu or Ph) show activity for the production of butenes and higher olefins. Optimized turnover frequencies of 640 molethylene·molNi−1·s−1 for the formation of butenes with 41(1)% selectivity for 1-butene using (PhPBP)NiBr, and 68 molethylene·molNi−1·s−1 for butenes production with 87.2(3)% selectivity for 1- butene using (tBuPBP)NiBr, have been demonstrated. With methylaluminoxane as co-catalyst and (tBuPBP)NiBr as the precatalyst, ethylene oligomerization to form C4 through C20 products was achieved while the use of (PhPBP)NiBr as the pre-catalyst retained selectivity for C4 products. Combined experimental and computational studies indicate that the ethylene dimerization is not 2 initiated by Ni hydride or alkyl intermediates. Rather, our studies point to a mechanism that involves a cooperative B/Ni activation of ethylene to form a key 6-membered borametallacycle intermediate. Thus, a cooperative activation of ethylene by the Ni–B unit of the (RPBP)Ni catalysts is proposed as a key element of the Ni catalysis. Introduction Catalytic oligomerization of ethylene is an important commercial method for the production of linear α-olefins (LAOs), which have extensive uses in fuel, petro-, and fine chemistry.1-4 Each year, > 3.5 million tons of LAOs are produced globally, and the annual growth rate of world consumption of LAOs forecasted to be approximately 4% during the 2019–2024 period.5,6 Since the first discovery of the "nickel effect" by Ziegler and Holzkamp in the 1950s,7 the field of nickel catalyzed olefin oligomerization has become one of intense study. Thus, the development of new catalysts for Ni-catalyzed olefin oligomerization and studies of the reaction mechanisms continue to be of interest to academia and industry.8-10 Another important milestone in this field was the development of the bidentate P–O ligated Ni complex by Keim and coworkers,11 which ultimately led to the commercialization of the Shell higher olefin process (SHOP) that is used for the production of over one million tons of α-olefins every year.2,12-13 This success has encouraged recent studies with the goal of pursuing novel ligand structures to develop new fundamental understanding of how ligand/catalyst structure can impart new types of reactivity.12 The generally accepted mechanism for Ni-mediated ethylene oligomerization (e.g., the Shell Higher Olefin Process) is the Cossee-Arlman mechanism (Scheme 1, left)5,9,14-15 in which the reaction is initiated from a Ni–H/alkyl intermediate followed by ethylene insertion steps (chain growth). The formation of olefin products is generally proposed to occur via β-H elimination from 3 a Ni-alkyl intermediate followed by net olefin dissociation. Chain propagation is normally controlled by the ratio of ethylene insertion and β-H elimination rates,10 and frequently Schulz−Flory distributions of ethylene oligomers are obtained.16-18 Another possible mechanism for Ni-catalyzed ethylene oligomerization involves the formation of a metallacycle similar to the reactions using Cr catalysts,9,19 which is normally proposed in the reaction of a Ni pre-catalyst with a Lewis acid activator (e.g., BF3).20 This type of mechanism has been suggested to be energetically viable based on DFT modeling of phosphine ligated Ni(0) catalyzed ethylene dimerization processes (Scheme 1, right).2,21 Experimental evidence for a Ni metallacycle mechanism (Scheme 1, right) was reported by Grubbs and coworkers using a nickel-based metallacyclopentene complex in the 1970s.22-26 Scheme 1. Proposed mechanisms of Ni-mediated ethylene oligomerization. Pincer ligands have been widely used in transition-metal-mediated catalysis due, in part, to their tunable steric and electronic properties.27-30 As part of the pincer ligand family, examples of bis(phosphino)boryl ligands (RPBP) have been designed and synthesized by the Nozaki and Yamashita groups,31-33 and later, (RPBP) ligands were studied with transition-metals such as Os, 4 Ir, Pt, Ru, Rh, Pd and Co.34-45 Nickel complexes with tBuPBP ligands have been isolated and reported to be active for olefin hydrogenation and CO2 reduction.46-49 In addition, experimental evidence has been reported for the formation of a σ-borane (η2-B–H)Ni(0) species from the reaction of (tBuPBP)NiH with 1,5-cyclooctadiene, which suggests that the hydride ligand is capable of migration from the Ni center to the boron center (Scheme 2).50 Recently, the Nozaki group has synthesized a bidentate Ni(0) σ-borane complex {(η2-B–H/P)Ni} and demonstrated its application in catalytic polymerization of ethylene for which the (η2-B–H/P)Ni complex behaves as a masked Ni(II) boryl hydride that selectively produces linear polyethylene but not lower molecular weight ethylene oligomers (Scheme 2).51 These studies suggest that the installation of a non-innocent boryl-moiety in the ligand structure offers unique reactivity, such as serving as a hydride shuttle, and could potentially alter the reaction pathway for catalytic processes. Scheme 2. Examples of previously reported reactions of olefins with PBP-Ni and related PB-Ni complexes and this work. 5 Herein, we report catalytic ethylene oligomerization reactions using a series of (RPBP)NiBr (R = tBu, Ph, or Cy) catalyst precursors. Unique aspects of these catalytic processes include substantial activities for ethylene dimerization with an Al-based or Lewis acid co-catalysts, tunable selectivity for ethylene dimerization versus oligomerization based on ligand structure and cocatalyst identity, and a proposed catalytic reaction pathway that involves a non-innocent boron center on the ligand moiety and does not involve a Ni–H or Ni–alkyl intermediate, which we believe is a unique mechanism for olefin coupling processes. Results and Discussion Initial screening for ethylene dimerization. During the initial screening via in situ 1H NMR spectroscopy, (tBuPBP)NiOAc (1), which was synthesized from (tBuPBP)NiMe and CO2,49 was found to catalyze slow dimerization of ethylene to form 1-butene without a co-catalyst at 90 °C using C6D6 as the solvent (Table 1, entry 1). To test if the acetate group is required for the reaction, (tBuPBP)NiBr (2) was used, and no peaks associated with the formation of butenes were observed in the 1H NMR spectra (Table 1, entry 2). However, when using complex 2 in the presence of AgBF4 as an additive for bromide abstraction, the ethylene dimerization reaction was achieved even at a lower temperature (60 °C) with ~75% selectivity for 1-butene (Table 1, entry 3). In an effort to achieve an in situ Br/OAc metathesis reaction to form complex 1, both TlOAc and AgOAc were examined as the additive for the ethylene dimerization reaction using complex 2 (Table 1, entry 4). The reaction with TlOAc after 2 days provided < 1 TO (turnover) of 1-butene (Table 1, entry 5). Based on these results, it can be concluded that the OAc group is likely not essential for the ethylene dimerization, and we speculated that cationic [(tBuPBP)Ni]BF4, formed through 6 bromide abstraction from 2 with AgBF4, is likely the active species for the ethylene dimerization reaction. Although we were unable to isolate the cationic complex [(tBuPBP)Ni][BF4], a crystal structure of [(tBuPBP)Ni(OH2)][BF4] was obtained, which likely formed during the month-long crystal growing period in an insufficiently dried solvent (Figure S42). Next, different silver and sodium salts were tested as additives (see Supporting Information, Section 2), and only AgBF4, AgSbF6, AgBArF and NaBArF were found to provide active catalysts for ethylene dimerization in the presence of complex 2 (Table 1, entries 3, 6–8). Control experiments using other Ni precursors such as (DME)NiCl2, NiCl2 and Ni(OAc)2 with and without AgBF4 showed no activity for ethylene dimerization (Scheme 3A), which suggests that the tBuPBP ligand is important for the dimerization reaction, leading us to speculate about a possible cooperative role for the Ni and B centers in the catalytic mechanism (see below for more discussion on this point). The Ni(II) complex (PhPBP)NiBr (3) was synthesized and tested for ethylene dimerization, and we found that using 3 as a catalyst precursor yielded butenes at room temperature, but the reaction was less selective for 1-butene (Table 1, entries 9–11). It is possible that changing the tBu group to a less sterically hindered Ph group on the PBP moiety favors the coordination of 1-butene, which might provide access to a more facile isomerization of 1-butene to 2-butenes compared to the Ni catalyst coordinated by the tBuPBP ligand (see below for more discussion and experimental results supporting this suggestion). Propylene was also tested as the substrate using complexes 1–3 in the presence of additives (i.e., AgBF4, MAO); however, no reaction was observed based on the in situ 1H NMR studies (Scheme 3B). We speculate that the dimerization reaction may require the coordination of more than one molecule of ethylene to initiate (see below), and the coordination of two equivalents of propylene seems to be less favorable than ethylene, likely due to sterics, 7 which would inhibit the dimerization of olefins larger than ethylene. However, a mixture of propylene and ethylene resulted in the formation of higher olefins with an odd number of carbon atoms along with products from ethylene oligomerization. These results indicate the likelihood of a Ni-catalyzed coupling reaction between ethylene and propylene, which also indicates that the coupling of ethylene and other α-olefins, such as 1-butene, is likely possible (see Supporting Information, Section 9). Thus, the Ni-catalysis is selective toward homo-ethylene coupling or ethylene/α-olefin coupling but is less reactive for the coupling of two α-olefins. Table 1. Initial attempts for ethylene dimerization using (RPBP)Ni complexes.a Entry Ni complex Additive Time (min) Temp. (°C) TOF C4b (s−1) 1-butene (TOs) 2-butenes (TOs) trans cis 1 1 none 1440 90 3×10−5 2.3 N.D. N.D. 2 2 none 240 60 N.D.c N.D. N.D. N.D. 3 2 AgBF4 240 60 1×10−4 1.1 0.24 0.12 4 2 AgOAc 1440 60 N.D. N.D. N.D. N.D. 5 2 TlOAc 2880 60 6×10−6 0.8 N.D. N.D. 6 2 NaBArF 240 60 9×10−5 0.5 0.42 0.36 7 2 AgBArF 240 60 2×10−4 1.3 0.61 0.63 8 2 AgSbF6 30 r.t. 2×10−3 3.4 0.24 0.28 9 3 AgSbF6 15 r.t. 1.6×10−2 3.48 6.75 4.42 10 3 AgBF4 20 r.t. 1.6×10−2 4.92 8.48 5.98 11 3 AgBArF 30 r.t. 9×10−4 0.56 0.70 0.33 a Reaction conditions: (RPBP)NiX (4.53 μmol) in 0.5 mL C6D6, additive (1.0 equiv. relative to Ni precatalyst), 40 psig ethylene. b TOF C4 = molbutenes·molNi−1·s−1. c N.D. = not detected. 8 Scheme 3. (A) Control experiments using other Ni pre-catalysts. (B) Attempts for propylene dimerization using (RPBP)Ni complexes 1–3. Ethylene dimerization/oligomerization using aluminum co-catalyst. Alkylaluminum compounds have been commonly used as co-catalysts for homogeneous ethylene oligomerization reactions using ligated Ni halide catalyst precursors, which generally led to enhanced activities.4,52- 56 Therefore, different alkylaluminums were tested as co-catalysts for our (RPBP)Ni catalysis (Table 2), and the resulting rates of ethylene dimerization are improved. Control experiments using (RPBP)H free ligand, alkylaluminums without Ni, (DME)NiBr2 with alkylaluminums, and (RPBP)H free ligand with alkylaluminums, produce no butenes or only trace amounts of butenes.10,57 When using complex 2 as the pre-catalyst in the presence of 10 equivalents of methylaluminoxane (MAO), ethylene oligomerization was achieved at room temperature with a 0.09(1) s−1 turnover frequency (TOF) for butenes (Table 2, entry 1) and 88(2)% selectivity for 1- butene (among the C4 products), which is ~900 times faster than the reaction using complex 2 with AgBF4 (Table 1, entry 3). However, the reaction products only contain ~12 wt. % of butenes (C4), and a range of linear and 2-ethyl branched α-olefins were observed from C6 to C20 (Scheme 4). 15 Comparison of previously reported Ni catalysts. Table 5 compares selected results of our newly reported catalysis with previously reported homogeneous Ni catalysts which demonstrated activity for ethylene dimerization.2,4,55-56,59-65 Although a direct comparison of previously reported catalysts is not possible since the reactions were performed under different conditions (e.g., pressure, temperature, Ni catalyst concentration, Al/Ni ratio, etc.), the comparative data provide some reasonable comparison points. The activities given in Table 5 were all converted into a commonly used unit goligomers·molNi−1·h−1 for each catalytic system. The overall TOFs given in Table 5 were calculated based on ethylene consumption, for which TOF = activity/(molar mass of C2H4) with a unit of molethylene·molNi−1·s−1. In general, most of the reported highly active Ni catalysts are supported by SHOP-type and related phosphine-sulfur-, phosphine-, nitrogen-based ligand structures. In addition to the ethylene polymerization reaction reported by the Nozaki group,51 there are only a limited number of ligand structures with a central boron center.66 As outlined in Table 5, the new RPBP ligated Ni complexes reported in this work exhibit relatively high activities for the ethylene dimerization reaction, which motivated us to better understand the reaction pathway (see below). Table 5. Comparison of previously reported homogeneous Ni catalysts for ethylene dimerization to our newly reported PBP-Ni catalysis.a Ligand type co-catalyst Activitya g/(molNi·h) TOFethyleneb (s−1) α-C4/C4 (%) C4/Cn (%) Temp. (°C) ref (P,P) EtAlCl2 2.4 × 108 2377 35 82 45 (55) (O,N,S) MAO 1.4 × 108 1411 16 90 0 (56) (N,N) Et3Al2Cl3 4.6 × 107 460 92 88 r.t. (59) (N,N,N) Et2AlCl/PPh3 4.0 × 107 391 12 92 20 (60) (N,N) MAO 1.9 × 107 190 56 90 35 (61) (P,O) None 1.9 × 107 183 99 85 40 (67) (N,N,O) MAO 1.2 × 107 119 18 83 45 (62) (P,N) EtAlCl2 9.1 × 106 90 23 98 40 (63) (N,O) Et2AlCl 6.6 × 106 65 100 100 30 (64) (N,N) Et2AlCl 4.7 × 106 46 > 99 77 45 (65) 16 (tBuP,B,P) EtAlCl2 6.9 × 106 68 87 94 r.t. this work (PhP,B,P) EtAlCl2 6.5 × 107 640 41 89 r.t. a Most of the catalysts activities in this table were originally reported in goligomers·molNi−1·h−1, thus we converted all data to the same units for comparison. b Based on ethylene consumption, for which TOF = activity/(28.05 g/mol) with the unit of molethylene·molNi−1·s−1. Isomerization of 1-butene. As noted above, we speculate that the isomerization process for the conversion of 1-butene to 2-butenes could be a separate competing reaction in the catalytic ethylene dimerization/oligomerization using Ni pre-catalysts 2 or 3. Therefore, a set of experiments using 1-butene as the only substrate with EtAlCl2 or MAO as the co-catalyst with and without Ni pre-catalyst 2 or 3 were performed (Figure 1). Similar to the reaction using propylene as a substrate, no dimerization or oligomerization products of butenes were found based on the GC-MS analysis. While both complex 2 and 3 were able to isomerize 1-butene to 2-butenes in the presence of EtAlCl2 or MAO, control experiments using only EtAlCl2 or MAO showed much slower rates of isomerization of 1-butene to 2-butenes. Using the Ni pre-catalyst 3 approximately 10-fold faster isomerization of 1-butene to 2-butenes was observed compared to pre-catalyst 2 under the conditions using EtAlCl2, which is consistent with the observed ligand effect on 1- vs 2- butene selectivity of the ethylene dimerization reactions (see Table 1, and Table 4, entry 6 vs 18, 9 vs 21, 10 vs 24, 11 vs 25, 12 vs 26). As noted above, this ligand effect (i.e., 2-butenes vs. 1- butene selectivity as a function of identity of PBP ligand) can be rationalized by the presence of a more sterically hindered tBu group in complex 2 inhibiting 1-butene coordination, and thus retarding the rate of 1-butene isomerization. In addition, using EtAlCl2 with and without Ni precatalyst gave a much faster isomerization compared to MAO, which is also consistent with the observed difference in 1- vs 2-butene selectivity when using EtAlCl2 or MAO as additive (see Table 4, entries 2 vs 7, 3 vs 9, 13 vs 18, 14 vs 19, 15 vs 20). Although the experimental evidence 17 cannot completely rule out the possibility of direct reaction pathways from ethylene to 2-butenes, it seems reasonable to conclude that 2-butenes can be produced via a Ni catalyzed isomerization process of 1-butene, which is a competing reaction during the catalytic ethylene dimerization/oligomerization reaction. Figure 1. Isomerization of 1-butene. Reaction conditions: (RPBP)NiBr (0.905 μmol); EtAlCl2 (9.05 μmol); using toluene as the solvent (1 mL total for each reaction); 10 psig 1-butene at room temperature for 120 min. In situ NMR experiments and kinetic studies. To gain better understanding of the reaction mechanism using Al co-catalyst, experiments have been conducted using (RPBP)NiBr complex 2 and 3 with the presence of EtAlCl2. The in situ NMR studies only showed line broadening after addition of EtAlCl2 (see Supporting information, Sections 4.6–4.8), which could be due to a fluxional process between (RPBP)Ni(Br–LA) and (RPBP)Ni(Cl–LA) that is assisted by the Lewis acidic aluminum salt (LA). Using complex 2 with EtAlCl2, a crystal of (tBuPBP)Ni(AlX4) (X = Br or Cl, 5-AlX4) was isolated and studied by single crystal X-ray diffraction (Figure 2 and Figure S45), in which Br is partially occupying some of the Cl sites to make a mixed halide AlX4 anion. This observation further supports our proposed fluxional process. In addition, the Ni–Cl bond 18 distances in the solid state structures are in the range of 2.378(18) to 2.413(9) Å, longer than that in the (tBuPBP)NiCl complex {2.2399(4) Å}.46 In addition, this fluxional process has also been supported by DFT calculations (Figure S53), where the (tBuPBP)Ni(X–AlCl3) complex seems to be lowest-energy species of the process. Figure 2. ORTEP of (tBuPBP)Ni(AlX4) (X = Br or Cl, 5-AlX4). Ellipsoids are drawn at 50% probability level. The tBu groups were set to stick mode for clarity. Some of the hydrogen atoms, solvent molecules and the other repeated molecules in the unit cell have been removed for clarity. Selected bond lengths (Å): Ni3–Cl13 2.4013(10), Al3–Cl13 2.2303(16), Al3–Cl14 2.1376(18), Al3–Cl2 2.126(9), Al3–Cl16 2.1741(19). The TOFs of ethylene dimerization reactions using complex 3 with EtAlCl2 under variant constant ethylene pressures were monitored (Table S4). As it shown in Figure 3, at low ethylene pressures, a second-order dependence of rate on [C2H4] was observed; while at high ethylene pressures, apparent saturation kinetics and a transition toward a first-order dependence was observed. The saturation kinetics can be rationalized by the proposed reversible coordination of 19 ethylene, which could lead to saturation kinetics at higher [C2H4] (Figure 4). These results are consistent with our proposed mechanism based on DFT calculations (see below). Figure 3. Log-log plot of TOF versus ethylene pressure. Figure 4. Simplified potential rate law of the reaction using (RPBP)NiBr with Al co-catalyst. DFT analysis on the reaction mechanism. The aforementioned experimental data point to a mechanism for which the [(RPBP)Ni]+ fragment plays a fundamental role in the dimerization of ethylene to yield 1-butene. Indeed, control experiments revealed the participation of both the PBP ligand and nickel in the process, and stoichiometric experiments ruled out the likely involvement 20 of nickel acetate, hydride, or alkyl species. Therefore, we conducted DFT studies using cationic complex 5 (Scheme 5), produced from the reaction between bromide species 2 and a halide abstractor (e.g., AgBF4, NaBArF, AgBArF, etc.). Calculations were carried out at the PBE0/def2TZVP/def2QZVP level of theory, including Grimme’s D3 dispersion correction (PBE0-D3, see SI for information and references), using 5 and ethylene as the energy reference. Scheme 5. Simplified pathway for halide abstraction from complex 2 to give cationic complex 5. Coordination of one ethylene molecule to the vacant position of [(tBuPBP)Ni]+ (5) to give [(tBuPBP)Ni(η2-C2H4)]+ (5·C2H4) is isoenergetic (−0.02 kcal mol−1) to the reactants. From this point, several mechanistic scenarios were considered, most of which afforded kinetic barriers too energy-demanding to overcome experimentally (energy profiles for the energetically inaccessible pathways can be found in the Supporting Information). Activation of a C–H bond of bound ethylene (ΔG‡ > 50 kcal mol−1) of 5·C2H4 gave Ni–vinyl and B–H fragments in a less thermodynamically stable geometry than the initial cationic Ni–ethylene π complex (23.4 kcal mol−1 difference, Figure S47). Including weakly coordinating anions such as BF4− in the calculations led to even higher Gibbs free energy values (Figure S52). The addition of a free ethylene molecule to the coordinated ethylene of 5·C2H4 was also computed, based on the study by Bernardi, Bottoni et al.21 This resulted in the desired 5·1-butene complex (−18.1 kcal mol−1), yet at the expense of very high energy transition states (Figure S48).68 Exploratory calculations 21 involving [2+2] cycloaddition pathways did not lead to chemically meaningful results. Finally, dissociation of one of the phosphine ligands (Figure S49) or coordination of the ethylene molecule across the Ni–B bond (Figure S51) gave energy barriers above 30 kcal mol−1 along with thermodynamically unstable products. However, coordination of a second ethylene molecule to 5·C2H4 opened the door to a new ethylene dimerization mechanism involving participation of the PBP ligand. Figure 5. Excerpt of the Gibbs energy profile for the reaction of two equivalents of ethylene and [(tBuPBP)Ni]+ (5) to form a product 6 with two ethylene ligands and a dissociated phosphine. Relative Gibbs energies at 298 K and 1 M in kcal mol−1. Hydrogen atoms on the (RPBP)Ni moiety have been omitted for clarity. Enthalpy values highlighted in green. Figure 5 shows a calculated pathway for (RPBP)Ni-mediated positioning of two ethylene molecules for subsequent C–C coupling. Ethylene binding to give complex 5·(C2H4)2 is thermodynamically unfavorable (18.5 kcal mol−1 higher than 5·C2H4), as depicted in Figure 5. 22 This is in line with our previous studies on H2 activation by neutral, square planar Ni(II) complexes,50,69 for which the filled dz2 orbital on nickel gave rise to weak σ-H2 complexes.70 Indeed, 5·(C2H4)2 exhibits very little C=C bond elongation upon binding (1.36 Å vs 1.33 Å in free ethylene). Nonetheless, molecular orbital analysis of ethylene and 5·C2H4 (Figure 6) point to potential orbital overlap involving the PBP ligand {HOMO (5·C2H4) → LUMO (ethylene)} that can lead to ethylene functionalization. In fact, this second ethylene molecule binds across the Ni– B bond in transition state TS1 (23.2 kcal mol−1) to form a 4-membered borametallacycle, simultaneously promoting dissociation of one of the phosphine ligands from the metal center, as observed in the product 6 (Figure 5). Figure 6. Frontier molecular orbitals involved in TS1, which provides C–C formation between two equivalents of ethylene. Complex 5 is 16.2 kcal mol−1 above the reactants, which suggests the formation of 5 is possibly reversible. However, the rest of the kinetic barriers are lower than TS1, and the high thermodynamic stability of the product makes the overall process energetically downhill and favorable towards the formation of the experimentally observed 1-butene (see below). The 23 calculated phosphine dissociation is probably due to steric clash between the ethylene moieties and the tert-butyl substituents on the phosphorus atom. In fact, both ethylene ligands in 6 are oriented towards the space previously occupied by the phosphine ligand (Figure 5, inset). The short C···C distance (2.37 Å) observed for the two closer carbon atoms suggests preorganization of both ethylene fragments for subsequent C–C coupling. The C–C bond formation between two ethylene ligands is calculated to proceed through TS2 (17.1 kcal mol−1), located only 0.9 kcal mol−1 higher than complex 6 (Figure 7). The next intermediate after TS2 (species 7, 13.9 kcal mol−1) contains a six-membered B–Cδ–Cγ–Cβ–Cα–Ni ring, with C–C distances ranging from 1.50 to 1.67 Å. Although the distance arrangement might suggest some butadiene character, these C–C bonds are longer than those observed in 1,3-butadiene (1.34–1.45 Å).71 From complex 7, numerous hydrogen atom migrations have been explored to form either B–H or new C–H bonds, giving too energy demanding (ΔG‡ > 30 kcal mol−1) kinetic barriers. However, the Cβ atom can orient one of its hydrogen atoms closer to the metal center with minimal energy cost (TS3 = 12.8 kcal mol−1), in order to achieve a suitable structure for a β-hydride elimination step. The outcome of this rearrangement is complex 8, which exhibits an agostic interaction72 through the Cβ–H bond (Ni– H = 1.78 Å, Ni–C = 2.11 Å, Ni–H–C = 89.2°), located trans to the bound phosphine ligand. Although this complex is rather low in Gibbs free energy (7.9 kcal mol−1), we found a different, more energy-demanding isomer (8', 19.7 kcal mol−1) where the Cβ–H bond and the bound phosphine are in a cis orientation, and the 3c-2e interaction exhibits a much shorter Ni–H bond (1.55 Å), with similar Ni–C (2.11 Å) and Ni–H–C (96.9°) metrics to those observed for 7. Complex 8' proceeds to a β-H elimination transition state (TS4) with a free energy of activation of only 0.3 kcal/mol. 24 Figure 7. Excerpt of the Gibbs energy profile for Ni-mediated C–C coupling of two ethylene ligands (TS2) to ultimately form complex 8, which possesses a Ni/CH agostic interaction. Relative Gibbs energies at 298 K and 1 M in kcal mol−1. Hydrogen atoms on the (RPBP)Ni system have been omitted for clarity. Enthalpy values highlighted in green. The geometry of 8' is very similar to that observed for TS4 (early transition state), which might explain why the first β-hydride elimination step is only 0.3 kcal mol−1 higher in energy (Figure 8). After TS4, intermediate 9 (12.4 kcal mol−1) contains the newly formed C=C double bond (C=C distance = 1.36 Å) bound to nickel, along with coordination of one of the phosphines and the new hydride ligand. In addition, the fourth Ni-coordination position is stabilized by a weak σ interaction with the B–Cδ bond (B–Ni = 2.76 Å C–Ni = 2.48 Å). Although one might think that 1-butene is practically formed, exploratory calculations involving hydride transfer to Cδ or B to release the experimentally observed product led to high (> 30 kcal mol−1) energy barriers. Nevertheless, TS5 (11.6 kcal mol−1) was found, which reveals the stretch of the B–C → Ni interaction along its imaginary frequency, giving rise to species 10 (10.8 kcal mol−1). In this geometry, an agostic 31 VCO reactor was opened, and the pressure on the gas burette was recorded. After specific reaction time (10 or 20 min), the valve connected to the VCO reactor was closed and the pressure on the gas burette was recorded, followed by placing the VCO reactor into a dry ice/acetone bath. The pressure change of the gas burette was used to calculate ethylene consumption using the ideal gas law. After the reactor was sufficiently cooled, the top pressure was slowly released, followed by adding 1 mL of toluene (undried) to the reactor. For the conditions using 1000 equivalents of alkylaluminum, 1 drop of water was added. Then, 50 μL of tetrahydrofuran were syringed into the reactor as the standard for GC-MS analysis. Butenes were quantified using a Rt-Q-Bond column, and the remainder of the olefins was quantified using with a Rxi-5ms column. Synthesis and characterization of Ni complexes. (tBuPBP)NiOAc (1),49 and (tBuPBP)NiBr (2),50 and (PhPBP)H33 were synthesized based on published procedures. (PhPBP)NiBr (3). To a solution of (DME)NiBr2 (240 mg, 0.778 mmol) in 15 mL of dry toluene under Ar atmosphere, a solution of (PhPBP)H (400 mg, 0.778 mmol) and Et3N (0.24 mL, 1.556 mmol) in 15 mL of dry toluene was cannulated slowly at −78 °C. The reaction mixture was allowed to warm to room temperature slowly and stirred overnight. The solvent was removed in vacuo, then the solid residue was washed with cold pentane (−20 °C, 5 mL × 2). The resulting solid was extracted using dry toluene, and the solution was dried under vacuum to isolate the product as an orange-yellow solid that is sensitive to moisture and oxygen (390 mg, 78% isolated yield). 1H NMR (800 MHz, benzene-d6) δ 7.89 (dd vt, 1JH,H = 7 Hz, 3JH,H = 2 Hz, 8H), 7.15 (AA'XX' dd, partially overlapped with benzene-d6, 2H), 7.01 – 6.96 (m, 12H), 6.90 (AA'XX' dd, JAX = 7.7 32 Hz, JAX' = 1.2 Hz, JAA' = 0.4 Hz, JXX' = 7.6 Hz, 2H), 4.00 (vt, N = 4.8 Hz, 4H, NCH2P). 31P{1H} NMR (243 MHz, benzene-d6) δ 47.26 (s). 13C{1H} NMR (201 MHz, benzene-d6) δ 139.2 (vt, N = 18 Hz), 133.7 (vt, N = 12 Hz), 132.7 (vt, N = 38 Hz), 130.4, 128.8 (vt, N = 9 Hz), 119.4, 109.7, 49.1 (vt, N = 44 Hz, PCH2). Anal. Calcd for C32H28BBrN2NiP2: C, 58.95; H, 4.33; N, 4.30. Found: C, 58.89; H, 4.50; N, 4.18. (CyPBP)NiBr (4). A solution of (CyPBP)H ligand (500 mg, 0.928 mmol) in toluene (10 mL) and Et3N (0.285 mL, 2.04 mmol) at −20 ⁰C was transferred via cannula to a suspension of (DME)NiBr2 (286.4 mg, 0.928 mmol) in toluene (10 mL) at the same temperature. The resulting suspension was allowed to warm to room temperature, and it was stirred at the same temperature for 18 h, after which the stirring was stopped, and the dark yellow solution decanted to a Schlenk flask by using a cannula with a double filter paper. The remaining solid was extracted with toluene (10 mL × 3), and the combined organic phase was evaporated under vacuum to give Ni–Br as a brown solid (578 mg, 0.854 mmol, 92% yield). X-Ray quality crystals can be obtained by diffusion of pentane into a toluene solution of 4. 1H NMR (400 MHz, benzene-d6): δ 7.18 (AA'XX' dd, partially overlapped with benzene-d6, 2H, aromatic CH), 7.01 (AA'XX' dd, JAX = 7.6 Hz, JAX' = 1.3 Hz, JAA' = 0.5 Hz, JXX' = 7.6 Hz, 2H, aromatic CH), 3.41 (vt, N = 4.2 Hz, 4H, NCH2P), 2.30 (m, 4 H, CH2), 2.19 (quint vt, 3JHH = 2.9 Hz, N = 24.6 Hz, 4H, CH-P), 1.80 (m, 4 H, CH2), 1.70 (m, 4 H, CH2), 1.59 (m, 12 H, CH2), 1.36 (m, 4 H, CH2), 1.22 (m, 4 H, CH2), 1.08 (m, 8 H, CH2). 31P{1H} NMR (161 MHz, benzene-d6) δ 66.68 (s). 11B{1H} NMR (128 MHz, benzene-d6): δ 40.3 (br s, boryl). 13C{1H} NMR (100 MHz, 33 benzene-d6): δ 139.7 (vt, N = 16 Hz, ligand aromatic Cq), 119.0 (ligand aromatic CH), 109.3 (ligand aromatic CH), 40.5 (vt, N = 37 Hz, NCH2P), 33.9 (vt, N = 19. Hz, C1), 29.2 (C2, C6), 28.8, 27.1 – 27.3 (m, C3,C5), 26.5 (C4). Anal. Calcd. for C32H52BBrN2NiP2: C, 56.85; H, 7.75; N, 4.14. Found: C, 56.65; H, 8.03; N, 4.26. Associated Content Supporting Information The Supporting Information is available free of charge. Additional experimental procedures and details, additional condition screening results, control experiments, GC methods and representative chromatograms, NMR and HRMS spectra of the complexes, crystal structure data (PDF). Computational details, Cartesian coordinates for the calculated structures and absolute energies. Access Codes CCDC 2203706–2203708, and 2212008 contains the supplementary crystallographic data for this paper. These data can be obtained free of charge via www.ccdc.cam.ac.uk/structures. Author Information Corresponding Authors T. Brent Gunnoe – Department of Chemistry, University of Virginia, Charlottesville, Virginia 22904, United States; orcid.org/0000-0001-5714-3887; Email: [email protected] 34 Amor Rodríguez – Instituto de Investigaciones Quimicas (IIQ), Department of Inorganic Chemistry CSIC and. University of Seville, Center for Innovation in Advanced Chemistry (ORFEO-CINQA), C/Américo Vespucio 49, Seville, 41092, Spain; orcid.org/0000-0001-9194- 5613; Email: [email protected] Authors Fanji Kong – Department of Chemistry, University of Virginia, Charlottesville, Virginia 22904, United States; orcid.org/0000-0003-4136-1403 Pablo Ríos – Instituto de Investigaciones Quimicas (IIQ), Department of Inorganic Chemistry CSIC and. University of Seville, Center for Innovation in Advanced Chemistry (ORFEO-CINQA), C/Américo Vespucio 49, Seville, 41092, Spain; orcid.org/0000-0003-4467-4157 Conner Hauck – Department of Chemistry, University of Virginia, Charlottesville, Virginia 22904, United States; orcid.org/0000-0003-4136-1403 Francisco José Fernández-de-Córdova – Instituto de Investigaciones Quimicas (IIQ), Department of Inorganic Chemistry CSIC and. University of Seville, Center for Innovation in Advanced Chemistry (ORFEO-CINQA), C/Américo Vespucio 49, Seville, 41092, Spain; orcid.org/0000- 0002-1784-2840 Diane A. Dickie – Department of Chemistry, University of Virginia, Charlottesville, Virginia 22904, United States; orcid.org/0000-0003-0939-3309 Laurel G. Habgood – Department of Chemistry, Rollins College, Winter Park, Florida 32789, United States; orcid.org/0000-0001-5475-1561 35 Acknowledgment The U.S. National Science Foundation (CHE-2102433) provided support for the experimental efforts. Single crystal X-ray diffraction experiments were performed on a diffractometer at the University of Virginia funded by the NSF-MRI program (CHE-2018870). Financial support (FEDER contribution) for the computational studies was provided from the MINECO (Projects PID2019 109312GB-IOO and RED2018-102387-T), Junta de Andalucía (Project PY20_00513) and the CSIC (AEPP-CTQ2016-76267-P). References 1. Pillai, S. M.; Ravindranathan, M.; Sivaram, S., Dimerization of ethylene and propylene catalyzed by transition-metal complexes. Chem. Rev. 1986, 86, 353-399. doi: 10.1021/cr00072a004 2. Olivier-Bourbigou, H.; Breuil, P. A. R.; Magna, L.; Michel, T.; Espada Pastor, M. F.; Delcroix, D., Nickel Catalyzed Olefin Oligomerization and Dimerization. Chem. Rev. 2020, 120, 7919-7983. doi: 10.1021/acs.chemrev.0c00076 3. Ittel, S. D.; Johnson, L. K.; Brookhart, M., Late-Metal Catalysts for Ethylene Homo- and Copolymerization. Chem. Rev. 2000, 100, 1169-1204. doi: 10.1021/cr9804644 4. Bekmukhamedov, G. E.; Sukhov, A. V.; Kuchkaev, A. M.; Yakhvarov, D. G., Ni-Based Complexes in Selective Ethylene Oligomerization Processes. Catalysts 2020, 10, 498. doi: 10.3390/catal10050498 36 5. Breuil, P.-A. R.; Magna, L.; Olivier-Bourbigou, H., Role of Homogeneous Catalysis in Oligomerization of Olefins : Focus on Selected Examples Based on Group 4 to Group 10 Transition Metal Complexes. Catal. Lett. 2015, 145, 173-192. doi: 10.1007/s10562-014-1451-x 6. S&P Global’s Linear Alpha-Olefins Chemical Economics Handbook (CEH). https://www.spglobal.com/commodityinsights/en/ci/products/linear-alpha-olefins-chemical- economics-handbook.html (accessed 2022-08-31). 7. Ziegler, K.; Gellert, H.; Holzkamp, E.; Wilke, G., Entdeckung des Nickel Effekts. Brennst.- Chem. 1954, 35, 321. 8. Fischer, K.; Jonas, K.; Misbach, P.; Stabba, R.; Wilke, G., The “Nickel Effect”. Angew. Chem. Int. Ed. 1973, 12, 943-953. doi: 10.1002/anie.197309431 9. McGuinness, D. S., Olefin Oligomerization via Metallacycles: Dimerization, Trimerization, Tetramerization, and Beyond. Chem. Rev. 2011, 111, 2321-2341. doi: 10.1021/cr100217q 10. Sydora, O. L., Selective Ethylene Oligomerization. Organometallics 2019, 38, 997-1010. doi: 10.1021/acs.organomet.8b00799 11. Keim, W.; Kowaldt, F. H.; Goddard, R.; Krüger, C., Novel Coordination of (Benzoylmethylene)triphenylphosphorane in a Nickel Oligomerization Catalyst. Angew. Chem. Int. Ed. 1978, 17, 466-467. doi: 10.1002/anie.197804661 12. Kuhn, P.; Sémeril, D.; Matt, D.; Chetcuti, M. J.; Lutz, P., Structure–reactivity relationships in SHOP-type complexes: tunable catalysts for the oligomerisation and polymerisation of ethylene. Dalton Trans. 2007, 515-528. doi: 10.1039/B615259G 37 13. Keim, W., Oligomerization of Ethylene to α-Olefins: Discovery and Development of the Shell Higher Olefin Process (SHOP). Angew. Chem. Int. Ed. 2013, 52, 12492-12496. doi: 10.1002/anie.201305308 14. Speiser, F.; Braunstein, P.; Saussine, L., Catalytic Ethylene Dimerization and Oligomerization: Recent Developments with Nickel Complexes Containing P,N-Chelating Ligands. Acc. Chem. Res. 2005, 38, 784-793. doi: 10.1021/ar050040d 15. Cossee, P., Ziegler-Natta catalysis I. Mechanism of polymerization of α-olefins with Ziegler-Natta catalysts. J. Catal. 1964, 3, 80-88. doi: 10.1016/0021-9517(64)90095-8 16. Flory, P. J., Molecular Size Distribution in Linear Condensation Polymers1. J. Am. Chem. Soc. 1936, 58, 1877-1885. doi: 10.1021/ja01301a016 17. Schulz, P. A.; Sudb¢, A. S.; Krajnovich, D. J.; Kwok, H. S.; Shen, Y. R.; Lee, Y. T., Multiphoton Dissociation of Polyatomic Molecules. Annu. Rev. Phys. Chem. 1979, 30, 379-409. doi: 10.1146/annurev.pc.30.100179.002115 18. Britovsek, G. J. P.; Malinowski, R.; McGuinness, D. S.; Nobbs, J. D.; Tomov, A. K.; Wadsley, A. W.; Young, C. T., Ethylene Oligomerization beyond Schulz–Flory Distributions. ACS Catal. 2015, 5, 6922-6925. doi: 10.1021/acscatal.5b02203 19. Overett, M. J.; Blann, K.; Bollmann, A.; Dixon, J. T.; Haasbroek, D.; Killian, E.; Maumela, H.; McGuinness, D. S.; Morgan, D. H., Mechanistic Investigations of the Ethylene Tetramerisation Reaction. J. Am. Chem. Soc. 2005, 127, 10723-10730. doi: 10.1021/ja052327b 20. Titova, Y. Y.; Belykh, L. B.; Rokhin, A. V.; Soroka, O. G.; Schmidt, F. K., Catalysis of dimerization and oligomerization reactions of lower alkenes by systems based on Ni(PPh3)2(C2H4) and Ni(PPh3)nCl (n = 2 or 3). Kinet. Catal. 2014, 55, 35-46. doi: 10.1134/S0023158414010169 38 21. Bernardi, F.; Bottoni, A.; Rossi, I., A DFT Investigation of Ethylene Dimerization Catalyzed by Ni(0) Complexes. J. Am. Chem. Soc. 1998, 120, 7770-7775. doi: 10.1021/ja980604r 22. Grubbs, R. H.; Miyashita, A., The metallacyclopentane–olefin interchange reaction. J. Chem. Soc., Chem. Commun. 1977, 864-865. doi: 10.1039/c39770000864 23. Grubbs, R. H.; Miyashita, A.; Liu, M.-I. M.; Burk, P. L., The preparation and reactions of nickelocyclopentanes. J. Am. Chem. Soc. 1977, 99, 3863-3864. doi: 10.1021/ja00453a068 24. Grubbs, R. H.; Miyashita, A., Metallacyclopentanes as catalysts for the linear and cyclodimerization of olefins. J. Am. Chem. Soc. 1978, 100, 7416-7418. doi: 10.1021/ja00491a051 25. Grubbs, R. H.; Miyashita, A., The relationship between metallacyclopentanes and bis(olefin)-metal complexes. J. Am. Chem. Soc. 1978, 100, 1300-1302. doi: 10.1021/ja00472a050 26. Grubbs, R. H.; Miyashita, A.; Liu, M.; Burk, P., Preparation and reactions of phosphine nickelocyclopentanes. J. Am. Chem. Soc. 1978, 100, 2418-2425. doi: 10.1021/ja00476a026 27. Gu, S.; Nielsen, R. J.; Taylor, K. H.; Fortman, G. C.; Chen, J.; Dickie, D. A.; Goddard, W. A.; Gunnoe, T. B., Use of Ligand Steric Properties to Control the Thermodynamics and Kinetics of Oxidative Addition and Reductive Elimination with Pincer-Ligated Rh Complexes. Organometallics 2020, 39, 1917-1933. doi: 10.1021/acs.organomet.0c00122 28. Kong, F.; Gu, S.; Liu, C.; Dickie, D. A.; Zhang, S.; Gunnoe, T. B., Effects of Additives on Catalytic Arene C–H Activation: Study of Rh Catalysts Supported by Bis-phosphine Pincer Ligands. Organometallics 2020, 39, 3918-3935. doi: 10.1021/acs.organomet.0c00623 29. Goldberg, J. M.; Wong, G. W.; Brastow, K. E.; Kaminsky, W.; Goldberg, K. I.; Heinekey, D. M., The Importance of Steric Factors in Iridium Pincer Complexes. Organometallics 2015, 34, 753-762. doi: 10.1021/om501166w 39 30. Arora, V.; Narjinari, H.; Nandi, P. G.; Kumar, A., Recent advances in pincer–nickel catalyzed reactions. Dalton Trans. 2021, 50, 3394-3428. doi: 10.1039/d0dt03593a 31. Segawa, Y.; Yamashita, M.; Nozaki, K., Boryllithium: Isolation, Characterization, and Reactivity as a Boryl Anion. Science 2006, 314, 113. doi: 10.1126/science.1131914 32. Segawa, Y.; Yamashita, M.; Nozaki, K., Syntheses of PBP Pincer Iridium Complexes: A Supporting Boryl Ligand. J. Am. Chem. Soc. 2009, 131, 9201-9203. doi: 10.1021/ja9037092 33. Segawa, Y.; Yamashita, M.; Nozaki, K., Diphenylphosphino- or Dicyclohexylphosphino- Tethered Boryl Pincer Ligands: Syntheses of PBP Iridium(III) Complexes and Their Conversion to Iridium−Ethylene Complexes. Organometallics 2009, 28, 6234-6242. doi: 10.1021/om9006455 34. Hill, A. F.; Lee, S. B.; Park, J.; Shang, R.; Willis, A. C., Analogies between Metallaboratranes, Triboronates, and Boron Pincer Ligand Complexes. Organometallics 2010, 29, 5661-5669. doi: 10.1021/om100557q 35. Ogawa, H.; Yamashita, M., Trial for anti-Markovnikov Hydration of 1-Decene Using Platinum Complexes Bearing a PBP Pincer Ligand, Inducing Alkene Isomerization and Decomposition of PBP Ligand. Chem. Lett. 2014, 43, 664-666. doi: 10.1246/cl.131208 36. McQueen, C. M. A.; Hill, A. F.; Sharma, M.; Singh, S. K.; Ward, J. S.; Willis, A. C.; Young, R. D., Synthesis and reactivity of osmium and ruthenium PBP–LXL boryl pincer complexes. Polyhedron 2016, 120, 185-195. doi: 10.1016/j.poly.2016.05.041 37. Ogawa, H.; Yamashita, M., Platinum complexes bearing a boron-based PBP pincer ligand: synthesis, structure, and application as a catalyst for hydrosilylation of 1-decene. Dalton Trans. 2013, 42, 625-629. doi: 10.1039/c2dt31892j 40 38. Hasegawa, M.; Segawa, Y.; Yamashita, M.; Nozaki, K., Isolation of a PBP-Pincer Rhodium Complex Stabilized by an Intermolecular C---H σ Coordination as the Fourth Ligand. Angew. Chem. Int. Ed. 2012, 51, 6956-6960. doi: doi:10.1002/anie.201201916 39. Fang, F.; Xue, M. M.; Ding, M.; Zhang, J.; Li, S.; Chen, X., The Stability of Diphosphino‐ Boryl PBP Pincer Backbone: PBP to POP Ligand Hydrolysis. Chemistry – An Asian Journal 2021, 16, 2489-2494. doi: 10.1002/asia.202100690 40. Lin, T.-P.; Peters, J. C., Boryl-Mediated Reversible H2 Activation at Cobalt: Catalytic Hydrogenation, Dehydrogenation, and Transfer Hydrogenation. J. Am. Chem. Soc. 2013, 135, 15310-15313. doi: 10.1021/ja408397v 41. Ding, Y.; Ma, Q.-Q.; Kang, J.; Zhang, J.; Li, S.; Chen, X., Palladium(II) complexes supported by PBP and POCOP pincer ligands: a comparison of their structure, properties and catalytic activity. Dalton Trans. 2019, 48, 17633-17643. doi: 10.1039/c9dt03954f 42. Tanoue, K.; Yamashita, M., Synthesis of Pincer Iridium Complexes Bearing a Boron Atom and iPr-Substituted Phosphorus Atoms: Application to Catalytic Transfer Dehydrogenation of Alkanes. Organometallics 2015, 34, 4011-4017. doi: 10.1021/acs.organomet.5b00376 43. Hill, A. F.; McQueen, C. M. A., Arrested B–H Activation en Route to Installation of a PBP Pincer Ligand on Ruthenium and Osmium. Organometallics 2014, 33, 1977-1985. doi: 10.1021/om5001106 44. Miyada, T.; Yamashita, M., Oxygenation of a Ruthenium Complex Bearing a PBP-Pincer Ligand Inducing the Formation of a Boronato Ligand with a Weak Ru–O Bond. Organometallics 2013, 32, 5281-5284. doi: 10.1021/om400915x