Effects of Ligand, Metal, and Solvation on the Structure and Stability of

Article Cite This: Organometallics XXXX, XXX, XXX−XXX

Effects of Ligand, Metal, and Solvation on the Structure and Stability of Contact Ion Pairs Relevant to Olefin Polymerization Catalyzed by Rare-Earth-Metal Complexes: A DFT Study Xingbao Wang,†,‡ Guangli Zhou,† Bo Liu,*,§ and Yi Luo*,† †

State Key Laboratory of Fine Chemicals, School of Chemical Engineering, Dalian University of Technology, Dalian 116024, People’s Republic of China ‡ Key Laboratory of Coal Science and Technology (Ministry of Education and Shanxi Province), Training Base of State Key Laboratory of Coal Science and Technology (Jointly Constructed by Shanxi Province and Ministry of Science and Technology), Taiyuan University of Technology, Taiyuan 030024, People’s Republic of China § State Key Laboratory of Polymer Physics and Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, People’s Republic of China S Supporting Information *

ABSTRACT: A full account of theoretical analyses at the DFT level has been reported, focusing on the formation and reactivity of a family of cationic [R-(CH2)n-Py-Sc(CH2SiMe3)]+ catalysts and the effects of counterion and solvation. Two sets of model systems have been considered: (a) structures having identical bridging unit (n = 1) but having varying cyclopentadienyl groups (R = Cp′ (1), R = Ind (2), and R = Flu (3)) and (b) systems with the identical cyclopentadienyl moiety (Flu) but with varying bridging groups (n = 1 (3), n = 0 (4), and n = 2 (5)). For complex 3, various metal ions (Sc, Y, La, Pr, Nd, Gd, Tb, Dy, Ho, Er, Tm, and Lu) were considered to investigate the effect of central metals on the contact ion pairs (CIP). The formation and separation of CIP were found to be influenced by the steric hindrance of the ligand, the electron-donating ability of the cyclopentadienyl group, and the rare-earthmetal ion radius. The separation enthalpy of the CIPs decreases with increasing dielectronic constant of the solvent. The solvation hardly affects the energy barrier for styrene insertion into the Sc−C17H19 bond of the CIP [(Flu-CH2-Py)Sc(C17H19)][B(C6F5)4]. A bulkier and more electron donating ancillary ligand, a smaller ion radius of the rare-earth metal, and a greater polarity of the solvent are more beneficial to the separation of CIP and thus to the monomer coordination, which could contribute to the improvement of polymerization activity.

■

INTRODUCTION Cationic rare-earth-metal monoalkyl complexes have shown excellent catalytic activity and selectivity toward olefin polymerizations.1 Such cationic complexes could be generated by the reaction of the corresponding dialkyl precursors with an appropriate borate compound as cocatalyst, such as [Ph3C][B(C6F5)4] or [PhNMe2H][B(C6F5)4], yielding in situ a contact ion pair (CIP).1,2 A number of studies have indicated that the tacticity, molecular weight, and comonomer incorporation of the resulting polymer could be tuned by suitable modulation of the catalyst and cocatalyst architectures.1−6 Recently, researchers have developed a series of rare-earth-metal complexes bearing cyclopentadienyl (Cp′), indenyl (Ind), and fluorenyl (Flu) ligands for olefin polymerizations.7−12 These complexes were activated with appropriate cocatalysts to yield contact cation−anion pairs, which showed different performances toward olefin polymerizations in various solvents. Experimentally, a limited number of cationic alkyl complexes have been isolated. A major problem that has hampered the isolation and structural characterization of cationic rare-earthmetal alkyl complexes is the facile ligand scrambling in these © XXXX American Chemical Society

complexes. In 2007, Hou and co-workers synthesized a structurally well defined THF-free cationic half-sandwich scandium aminobenzyl complex.13 That work demonstrated the first example of an external Lewis base free half-sandwich rare-earth-metal CIP, (C 5 Me 4 SiMe 3 )Sc(CH 2 C 6 H 4 NMe 2 o)(κ2F-C6F5)-B(C6F5)4, in which the anion [B(C6F5)4] coordinated to the metal center in a κ2F fashion through two adjacent (ortho and meta) F atoms. By treatment of the diallyl complex [(C 5 Me 4 SiMe 3 )Sc-(C 3 H 5 ) 2 ] with 1 equiv of [PhNMe2H][B(C6F5)4] in toluene, the structurally characterizable N,N-dimethylaniline-coordinated CIP [(C5Me4SiMe3)Sc(η3-C3H5)(η6-PhNMe2)][B(C6F5)4] was also isolated by the same group.14 These studies raise one’s understanding of the structures of CIPs and their influence on the polymerization reaction. It is generally difficult for the currently available experimental methods to identify the behavior of CIP during the olefin polymerization catalyzed by rare-earth-metal complexes, since such polymerizations are often rapid. Received: November 29, 2017

A

DOI: 10.1021/acs.organomet.7b00857 Organometallics XXXX, XXX, XXX−XXX

Article

Organometallics Chart 1. Rare-Earth-Metal Bisalkyl Complexes 1−5 as Precursors

Figure 1. Optimized molecular structures of precatalysts (distances in Å and angles in deg).

few CIPs of cationic rare-earth-metal complexes and borate anions have been computationally optimized, and their thermodynamic stabilities were compared. However, studies on the effects of ligand, metal, and solvation on the structure and stability of CIPs and on the polymerization activity in rareearth-metal complex systems have remained very limited. The elucidation of these issues is helpful for a better understanding, at the molecular level, of the catalyst activation and the effects of solvent and counteranion on rare-earth-metal-catalyzed olefin polymerization and thus could provide useful information on the development of highly active polymerization systems. In the present work, a full account of theoretical analyses focusing on the formation and reactivity of a family of [R(CH2)n-Py-Sc(CH2SiMe3)]+ catalysts,8,11 including anion and solvation effects, will be reported. Two sets of model systems have been considered (Chart 1): (a) structures having identical an methylene bridging unit but having various cyclopentadienyl groups (R = Cp′ (1), R= Ind (2), and R = Flu (3)) and (b) systems with an identical Flu moiety but with varying bridging units (n = 1 (3), n = 0 (4), and n = 2 (5)). It is noteworthy

Therefore, it is necessary to use computational approaches to investigate the factors affecting CIP and thus catalytic polymerization performance in rare-earth-metal complex systems. Many previous computational studies have provided invaluable theoretical information on the bonding, structure, and energetics of CIPs and on the olefin polymerizations in group 4 catalyst systems.15−21 Among these studies, the structure and reactivity of CIPs have been investigated. In spite of the achievements in the group 4 systems, in-depth studies on the CIPs of cationic rare-earth-metal complexes and borate anions have remained very limited, although such systems produced unique polymers that could not be obtained by group 4 metal complexes, such as styrene−ethylene copolymers having syndiotactic styrene−styrene sequences and almost perfect alternating isoprene−ethylene copolymers.22,23 Moreover, unlike the case for group 4 catalyst systems,24−32 the borate salt [Ph3C][B(C6F5)4] as cocatalyst has more often been used in rare-earth-metal systems in comparison with [B(C6F5)3]. In our previous studies,33−35 a B


Article

Organometallics

Table 1. Computed Bond Lengths (Å) and Angles (deg) of [(Flu-CH2-Py)Ln(CH2SiMe3)2] Precatalysts (3′) as Well as the Effective Ionic Radius of Central Metalsa

metal d(Ln−Cn)b d(Ln−N) d(Ln−Cα) ∠(Cα1−Ln−Cα2) ionic radius (Å)49 a

Sc

Y

La

Pr

Nd

Gd

Tb

Dy

Ho

Er

Tm

Lu

2.25 2.31 2.20 108.4 0.87

2.43 2.48 2.38 110.4 1.02

2.65 2.69 2.69 111.6 1.18

2.60 2.64 2.51 111.1 1.14

2.58 2.62 2.50 110.9 1.12

2.51 2.54 2.44 110.0 1.06

2.50 2.53 2.42 109.8 1.04

2.48 2.51 2.41 109.6 1.03

2.47 2.49 2.40 109.4 1.02

2.45 2.48 2.39 109.3 1.00

2.44 2.47 2.38 109.1 0.99

2.42 2.44 2.35 108.9 0.97

THF molecule was not considered. bThe distance between Cp centroid and central metal.

tetrahedral geometry with the center of the Cp ring as the apex. It is noteworthy that, in the cases of complexes 1−3 and 5, the cyclopentadienyl auxiliaries coordinate to the central metal ions in the η5-cyclopentadienyl/κ1 mode via the Cp carbons and the dangling pyridyl nitrogen atom. However, in complex 4, the cyclopentadienyl auxiliaries coordinate to the central metal ions in the η3-allyl/κ1 mode, which is in agreement with the experimental results.11 In complexes 1−3, the distances between the center of the cyclopentadienyl moieties and the central metal ions are 2.19, 2.22, and 2.25 Å, respectively. As in our previous study,35 this may be related to the donor−acceptor type bond between the metal and ligand. In the case of 3, due to the electron delocalization over the phenyl rings that are electron-withdrawing, the ligand is less electron donating in comparison with 1 and 2. This increases the distances between the center of the cyclopentadienyl moieties and the central metal ions. The Sc···N distances in 1− 3 are the same (2.31 Å). Thus, the electronic properties of the central metal could be mainly tuned by the cyclopentadienyl moieties. The Sc···N contacts in 4 and 5 are shorter and longer in comparison with the other three complexes, respectively, possibly due to the number of bridging methylene units. The distance between the center of the cyclopentadienyl moieties and the central metal is 2.26 Å in 5, which is similar to that in 3. In the present study, the geometries of lanthanide complexes [(Flu-CH2-Py)Ln(CH2SiMe3)2] (3′; Ln = Lu, Tm, Er, Ho, Y, Dy, Tb, Gd, Nd, Pr, La) were also optimized. The calculated results are summarized in Table 1. The results show that, except for the Y analogue, the order of the bond lengths and angles correlates well with the effective ionic radius of the rareearth metals: the larger the ionic radius of the metal, the longer the bond length and the larger the angle in the corresponding complex, where the change in the bond lengths reflects the “lanthanide contraction” effect. Molecular Structures of the Naked Cations. The alkyl abstraction/catalyst activation process to form CIPs, the ionpair separation relevant to the “tightness” of the ion pairing, and the complexation of solvent on the naked cation were investigated in accordance with the equations shown in Scheme 1. The optimized bare cationic species (1+−5+) are shown in Figure 2. In the case of 4+, the ligand binds to the metal atom in an η5-cyclopentadienyl/κ1 mode rather than the η3-allyl/κ1

that, in order to compare the effect of ligand structures of the CIPs, complex 5 is designed and computationally modeled here, which has not yet been experimentally synthesized. Scandium metal was considered for complexes 1−5. In order to compare the effect of metal ions, Sc, Y, La, Pr, Nd, Gd, Tb, Dy, Ho, Er, Tm, and Lu metals were considered on the basis of the architecture of complex 3, and the corresponding cationic alkyl species is referred to as 3′.

■

COMPUTATIONAL DETAILS

All calculations were performed with the Gaussian 09 program.36 The B3PW91 hybrid exchange-correlation functional was utilized for geometry optimization and frequency calculations.37−39 The 631G(d) basis set was used for H, C, N, F, and B atoms, and the Si and rare-earth-metal atoms were treated by the Stuttgart/Dresden effective core potential (ECP) and associated basis sets.40,41 One dpolarization function (exponent of 0.284) was augmented for the basis set of the Si atom.42 Such basis sets are donated as BS-I. Geometries of all species were fully optimized without symmetry constraints. The transition states were ascertained by a single imaginary frequency for the correct mode. An intrinsic reaction coordinate (IRC) analysis was conducted for the transition states to verify that they connect two corresponding minima. The minima on the reaction energy profiles were verified to have all real frequencies only. Single-point energy calculations were also performed for the B3PW91/BS-I geometries by using the same functional and the larger basis set BS-II. In BS-II, 6311+G(d,p) was used for H, C, N, F, and B atoms and the basis set for Si and rare-earth-metal atoms is same as that in geometry optimizations. The solvation effect was considered through singlepoint calculations with the PCM solvation model.43 Toluene, chlorobenzene, and dichloromethane (ε = 2.37, 5.70, 8.93) were employed as solvents in the PCM calculations, respectively. The first two solvents are often used for polymerization reactions in experiments, and dichloromethane was chosen here for the purpose of computational investigation of the solvation effects under various polarity environments. For some special cases, the structures were optimized in solution. The reliability of the present computational method was also corroborated by the DLPNO-CCSD(T) method (see Table S1 in the Supporting Information).44−47 The computed structures are illustrated with CYLView.48

■

RESULTS AND DISCUSSION Molecular Structures of the Precatalysts. The schematic representations of several optimized precatalyst structures are shown in Figure 1. All of the metal centers are bonded to one ancillary ligand unit and two alkyl groups, generating a C


Article

Organometallics

single F atom coordinates to the Sc atom were fruitless. The ion pair CIP-3 is chosen as an example for geometric optimization in toluene solution. The computed formation free energy (3+ + B(C6F5)4− → CIP-3) is −20.6 kcal/mol, suggesting that the formation of the CIP is thermodynamically favorable in solution. The geometrical optimizations of CIP-3 at the levels of B3PW91 and B3PW91-D3 indicate that the dispersion correction has no significant effect on the geo-metry of CIP-3 (Figure S2). Due to the coordination of the anion, the agostic interaction between Sc and a methyl group of SiMe 3 disappeared. The optimized CIP structures show that, in comparison with the naked cations, the distances between the center of the cyclopentadienyl moieties and the central metal are obviously elongated. As shown in Figure 3, the distance of the Sc···B contact varies with the steric hindrance of the ancillary ligand: the larger the cyclopentadienyl group and the longer the bridging group, the longer the Sc···B distance. In the case of CIP-3′, except for La analogues, the Sc···B distances correlate with the metal ionic radii: viz., the larger the metal ion, the longer the Sc···B distance (Figure S1). In the case of CIP-3′ (La), the anion coordinates to the metal via two o-F atoms of two C6F6 groups, which could be ascribed to the much larger metal ionic radius. Energetic Aspects for the Formation of Contact Ion Pair. As shown in Table 2, the CIP formation enthalpies roughly vary with the cyclopentadienyl group, bridging group, metal ions, and solvent. Among these, the ancillary ligand has a more significant influence on the formation enthalpies of the CIPs. In the case of 3′, the order of formation enthalpy correlates well with the effective ionic radius of the rare-earth metal: the smaller the metal, the less negative the formation enthalpy of the corresponding complex (Figure 4). The calculated ΔHform values varies with the polarity of the solvent by a few kcal/mol. A solvent with a larger dielectric constant makes the CIP formation less exothermic (less negative ΔHform, Table 2). This is in line with the methide abstraction reaction of [1,2-(CH3)2C5H3]2M(CH3) complexes (M = Zr, Hf) by the organo-Lewis acid cocatalyst B(C6F5)3,16 where lower stabilities (2−8 kcal/mol) of the CIPs were observed in more polar solvents.

Scheme 1. Formation and Separation of CIP and Solvent Complexation Processes

mode shown in the corresponding neutral complex (4). This is consistent with our previous study of the Y analogues.35 The coordination manners of the ancillary ligand in the other complexes are similar to those in their corresponding neutral complexes. All of the optimized cationic species show a significant interaction between an Sc atom and a methyl carbon atom of the SiMe3 group, as manifested by the short Sc···Cγ contact (2.44−2.47 Å) and elongated Si−Cγ bond length (1.98−2.00 Å) in comparison with the normal Si−C bond length of 1.89 Å. For complexes 3′+, the metal−ligand contacts also follow the lanthanide contraction trend (Table S2 in the Supporting Information). In the case of 3+, due to the electron delocalization over the phenyl rings, the ligand is less electron donating in comparison with those in 1+ and 2+, and thus the Mulliken charge on the central metal (0.46) is larger than those in 1+ and 2+. This increases the Lewis acidity of the central metal of 3+. The Mulliken charges on the metals in 4+ and 5+ are 0.47 and 0.39, respectively, indicating that the longer bridging group could decrease the Lewis acidity of the metal center. Molecular Structures of Contact Ion Pairs. The structures of the CIPs were optimized, and the results show that the anion coordinates to the Sc center via o- and m-F atoms (Figure 3). This coordination manner (κ2F fashion) was also observed in the solid structure of similar ion pairs.13 Like the result reported in our previous study,35 the structures of CIPs where the anion adopts p- and m-F atoms to coordinate to the metal center are less stable in terms of thermodynamics. Attempts to locate coordination complexes in which only a

Figure 2. Optimized cationic species (distances in Å). D


Article

Organometallics

Figure 3. Optimized structures of CIPs (distances in Å).

specific value, the attraction between cation and anion is no longer obvious (vide infra). The calculated energies for CH2SiMe3 abstraction could reflect a balance between the Ln−CH2SiMe3 bond dissociation enthalpy and the ion-pair interaction enthalpy. It has been found that the Sc···B distance as a function parallels the observed ΔHform trends. This is understandable that such distances are indices of the electrostatic interaction (Table 2 and Figure 3). In all of the precatalyst molecules, the Ln− CH2SiMe3 bond was found to be largely covalent in nature. Therefore, the small change in such bond lengths could cause large energy changes and result in commensurately large destabilization effects. On the other hand, the C−C bond formation of the Ph3C−CH2SiMe3 and the electrostatic interaction between the cation and anion reduce the destabilization effects of the reaction system. For all of the CIP formations, the formation enthalpy of Ph3C−CH2SiMe3 is a constant. Thus, the discrepancies in the stabilization effect are correlative with the cation−anion interaction. In the cases of CIP-1, CIP-2, and CIP-3, the Sc···B distances are 5.47, 5.38, and 5.41 Å and the corresponding ΔHform values are −40.7, −36.7, and −37.5 kcal/mol, respectively. The more negative ΔHform value could be ascribed to the weaker Ln−CH2SiMe3 bond. The Ln−CH2SiMe3 bond length in 1 is longer than those in 2 and 3 (Figure 1). As shown in Figures 1 and 3, in the case of CIP-4 and CIP-5, the less negative ΔHform values could be ascribed to the stronger Ln−CH2SiMe3 bonds (shorter Ln− CH2SiMe3 distances) and greater Sc···B distances, respectively. The greater Sc···B distance in CIP-5 could be partially caused by the steric hindrance of the ligand. Separation Energy of the Contact Ion Pair. The ion pair separation reactions in the gas phase (eq 2 in Scheme 1) are computed to be significantly endothermic (Table 3). However, such endothermic character was suppressed by solvation effects partially due to the resulting charged species (see ΔHips values in the gas phase and solution, Table 3). Moreover, the separation enthalpy differences (ΔH ips (gas phase) − ΔHips(solution)) are about 30, 45, and 50 kcal/mol for the toluene (ε = 2.37), chlorobenzene (ε = 5.71), and dichloro-

Table 2. Calculated CIP Formation Enthalpies (ΔHform, kcal/mol) for the Process [R-(CH2)n-Py-Sc(CH2SiMe3)2] + [Ph3C]·[B(C6F5)4] → [R-(CH2)n-Py-Sc(CH2SiMe3)]· [B(C6F5)4] + Ph3C-CH2SiMe3 + ΔHform, with Respect to Various Mediates CIP

gas phase

C7H8 (ε = 2.37)

C6H5Cl (ε = 5.71)

CH2Cl2 (ε = 9.08)

CIP-1 CIP-2 CIP-3 CIP-4 CIP-5

−40.7 −36.7 −37.5 −33.9 −34.4

−37.1 −32.5 −33.0 −29.4 −30.2

−35.6 −30.8 −30.9 −26.9 −28.5

−35.4 −30.4 −30.4 −26.1 −28.1

Figure 4. Enthalpy profile for the formation of [(Flu-CH2-Py)Ln(CH2SiMe3)]·[B(C6F5)4] species that was calculated in the gas phase and in toluene, chlorobenzene, and dichloromethane solutions.

The decrease in dipole moment of the CIP could account for the less negative ΔHform in the case of a more polar solvent (Figure 4). It is noted that, when the dielectric constant is larger than 5.71 (chlorobenzene and dichloromethane), the change in ΔHform is no longer distinct. This could be ascribed to the situation when the dielectric constant increases to a E


Article

Organometallics Table 3. Ion Pair Separation Enthalpies (kcal/mol) for the Process [R-(CH2)n-Py-Sc(CH2SiMe3)]·[B(C6F5)4] → [R(CH2)n-Py-Sc(CH2SiMe3)]+ + [B(C6F5)4]− + ΔHips, Calculated in the Gas Phase and in Toluene, Chlorobenzene, and Dichloromethane Solutions CIP

gas phase

C7H8 (ε = 2.37)

C6H5Cl (ε = 5.71)

CH2Cl2 (ε = 9.08)

CIP-1 CIP-2 CIP-3 CIP-4 CIP-5

51.8 53.4 52.2 53.2 47.7

18.9 20.2 19.5 20.2 15.0

5.3 6.5 5.8 6.1 1.4

1.8 3.0 2.3 2.3 −2.0

Figure 5. Enthalpy profile for [(Flu-CH2-Py)Ln(CH2SiMe3)]·[B(C6F5)4] dissociation in the gas phase and in toluene, chlorobenzene, and dichloromethane solutions.

methane (ε = 9.08) cases, respectively. This suggests that a more polar solvent could thermodynamically promote the CIP separation. The calculated separation enthalpies are smaller than that for the corresponding ion-pair separation enthalpy of the group 4 complex H2Si(C5H4)(CH3N)Ti(CH3)·H3CB(C6F5)3 by ca. 5 kcal/mol.16 Therefore, [R-(CH2)n-PySc(CH2SiMe3)]·[B(C6F5)4] is a weaker ion pair in comparison with the H2Si(C5H4)(CH3N)Ti(CH3)·H3CB-(C6F5)3 ion pair. The computed results indicate that, while the cyclopentadienyl groups have smaller effect on ΔHips values (CIP1−CIP-3), the side arm has a greater effect on ΔHips (CIP-3− CIP-5). A noticeable difference in ΔHips values calculated in dichloromethane is observed between CIP-3 (2.3 kcal/mol) with one methylene bridging group and CIP-5 (−2.0 kcal/mol) with two ethylidene bridging groups (Table 3). The negative ΔHips value for the CIP-5 case suggests that the CIP-5 could be spontaneously dissociated in dichloromethane. This could be ascribed to the bulky hindrance of the ligand, which also results in a large Sc···B atom distance in CIP-5 (5.59 Å, Figure 3). Although the Sc···B distance in CIP-4 (5.38 Å) is smaller than that for CIP-3 (5.41 Å), their ion pair separation enthalpies are similar to each other. When the cationic species 4+ was checked, it was found that there is a stronger interaction between the Sc atom and a methyl carbon atom of the SiMe3 group, as manifested by the shorter Sc···Cγ contact (2.44 Å) and larger Si−Cγ bond length (2.00 Å) in comparison with the other cationic species (Figure 2). This could stabilize the cationic 4+ and thus be beneficial to the corresponding ion pair separation. In this sense, the agostic interactions play an important role in ion pair separation energetics because they significantly stabilize the cationic species. It is noteworthy that the bulkier ancillary ligand generally resulted in smaller ion pair separation enthalpy (Table 3), which is understandable with respect to the steric effects. It was previously shown that the contact ion pair interaction is predominantly electrostatic in character and can be described by a rather flat potential surface along the metal···B distance vector. Coulombic attractions depend inversely upon the dielectric constant of the medium, E = q+q−/4πεr,17 and therefore, solvation reduces the attraction between the cation and anion, thus rendering the potential energy surface along the metal···B distance even flatter. As shown in Figure 5, the metal ions have a small effect on the computed ΔHips values. It is interesting that, in the gas phase, the order of ion pair separation enthalpy is correlated with the ionic radius of the rare-earth metal: the smaller the metal, the smaller the ion-pair separation enthalpy. However, for the lanthanide series, this trend gradually reversed in the case of a more polar solvent (chlorobenzene and dichloromethane). One could expect that the stabilizing effect of

solvation toward the charged species possibly accounts for such a trend descrepancy. Again, the data shown in Figure 5 demonstrate that a more polar solvent resulted in a smaller ionpair separation enthalpy. To further analyze the ion pair separation, the potential energy surface of the [(Flu-CH2-Py)Sc(CH2SiMe3)]·[B(C6F5)4] as a representative was also investigated along the reaction coordinate for ion-pair dissociation in the gas phase and in solution. As shown in Figure 6, the coordination manner of the anion to the Sc center changes from a κ2F fashion to a κ1F fashion when the Sc···B distance increases from 5.4 to 7.0 Å. When the Sc···B distance is longer than 8.0 Å, the cationic moiety is very close to the naked cation, as suggested by the Sc−Cn (2.12 Å) and Sc−Cγ (2.46 Å) distances, which are same as those in the naked cation (Figure 6). When the Sc···B distance reaches 10.0 Å, there is nearly no interaction between F and the Sc center. Figure 7 displays the changes in the enthalpy of CIP-3 with the Sc···B distance in the gas phase and solution. It can be found that, in the gas phase, the energy continuously increases with the Sc···B distance. The solvation strongly affects the energetics of ion-pair separation (Figure 7). In a less polar solvent such as toluene, the shape of the potential energy surface is similar to that for the gas phase. In the case of more polar solvents such as chlorobenzene and dichloromethane, the enthalpy is nearly unchanged when the Sc···B distance is greater than 8 Å. It is noteworthy that the energy at a Sc···B distance of ca. 6 Å is lower than that at 5.4 Å (CIP-3) in solution. This could be ascribed to the difference in the Sc···B distances of the optimized structures in the gas phase and solution. Actually, the Sc···B distance of CIP-3 optimized in dichloromethane solvent is longer than that in the gas phase by 0.2 Å (5.6 Å vs 5.4 Å). Thus, in more polar solvents, the CIP tends to dissociate into a naked cation and anion. This is beneficial to the coordination of monomer molecules and therefore increases the polymerization activity. Energetics of Solvated Complexes. The structures of naked cations stabilized by solvent coordination were also investigated (eq 3 in Scheme 1). All optimizations were performed in the corresponding solvents. The results show that the coordination of a single solvent molecule is thermodynamically favored (Figure 8). The coordination enthalpies are −1.6 to − 2.0 kcal/mol, suggesting almost isoenergetic coordinations for toluene, chlorobenzene, and dichloromethane, respectively. These energy values are smaller (unsigned data) than the F


Article

Organometallics

Figure 6. Optimized gas-phase structures of [(Flu-CH2-Py)Sc(CH2SiMe3)]·[B(C6F5)4] (CIP-3) with various fixed Sc···B distances. Distances are given in Å.

In the current system, toluene is favorably bound to the Sc atom through three phenylic carbon atoms in an η3 fashion (Figure 8). In contrast, chlorobenzene and dichloromethane coordinate to the metal via a chlorine atom, being beneficial from the lone-pair electrons. Due to the coordination of solvents, the distances between the center of the cyclopentadienyl moieties and the central metal ions increased by 0.09, 0.07, and 0.07 Å (from 2.10 Å to 2.19, 2.17, and 2.17 Å) for the toluene, chlorobenzene, and dichloromethane cases, respectively (Figures 2 and 8). The C−Cl bonds in the solvated complexes elongated from 1.75 and 1.78 Å to 1.78 and 1.81 Å in the chlorobenzene and dichloromethane cases, respectively. The Mulliken charges on the Sc atom increased from 0.33 to 0.55, 0.63, and 0.66 for toluene, chlorobenzene, and dichloromethane, respectively. This suggests that the metal center becomes more acidic after solvation. Effect of Counteranion on Styrene Insertion Reaction. For a better understanding of the counteranion effect on polymerization reactions, an experimentally demonstrated syndiospecific styrene polymerization catalyzed by a Flu-CH2Py ligated scandium complex6,8 was selected here, for which the experimental activation barrier is available.6 Like that in the previous study,6 the cationic complex [(Flu-CH2-Py)Sc(C17H19)]+ was considered as the active species, and the two phenyls of styrene units are distributed on both sides of the metal center. It is noted that the energy barrier for the insertion of styrene into the initial active species [(Flu-CH2-Py)Sc(CH2SiMe3)][B(C6F5)4] is calculated to be 19.1 kcal/mol, which is lower than that for the [(Flu-CH2-Py)Sc-(C17H19)][B(C6F5)4] case (vide infra). Therefore, the effect of [B(C6F5)4]− anion on the chain growth process was considered. As above, the anion electrostatically interacts with its counterion and such an ion pair could be separated during the polymerization process. The calculated electrostatic

Figure 7. Energetic profile (kcal/mol) for [(Flu-CH2-Py)Sc(CH2SiMe3)]·[B(C6F5)4] dissociation calculated in the gas phase and in toluene, chlorobenzene, and dichloromethane solutions.

Figure 8. Molecular structures of naked cation−solvent molecule complexes, taking the cation part of CIP-3 as a representative.

complexation enthalpies of the Ti complex H2Si(C5H4)(tBuN)TiCH3+ with solvent molecules (−6 to − 16 kcal/mol).16 G


Article

Organometallics

affects the energy barrier of the styrene insertion into [(FluCH2-Py)Sc-(C17 H19)][B(C6F 5) 4] but mainly affects the stability of the CIP and has little effect on the styrene insertion reaction. Therefore, the generally observed higher polymerization activity in a more polar solvent could be ascribed to the easier CIP separation (in more polar solvent) rather than the lower energy barrier of olefin insertion. On the basis of the current calculations, it can be concluded that the stronger electron-withdrawing ability of the cyclopentadienyl group, smaller rare-earth-metal ion radius, and greater solvent polarity are beneficial to the separation of CIPs, which could further improve the polymerization activity.

potential maps reveal that the positively charged areas are largely located on the pyridine moiety (Figure S3). Thus, the [B(C6F5)4]− anion could reside on the site of the pyridine moiety in the optimized structures. The insertion of the styrene molecule into [(Flu-CH2-Py)Sc-(C17H19)][B(C6F5)4] was calculated. The calculated results for the most favorable manner are summarized in Table 4, and more detailed data are shown in Table S3 in the Supporting Information. Table 4. Energy Profiles for the Insertion of Styrene into the Ion Pair [(Flu-CH2-Py)Sc-(C17H19)]+[B(C6F5)4]− a

gas (ε = 1.00) C7H8 (ε = 2.37) C6H5Cl (ε = 5.71) CH2Cl2 (ε = 9.08) exptlb

π complex

transition state

kinetic product

ΔG⧧/ΔH⧧

10.6/0.0 10.3/0.0 10.4/0.3

23.3/10.3 23.2/10.2 23.3/10.7

−0.4/−13.8 −0.1/−13.5 0.1/−12.7

23.3/10.3 23.2/10.2 23.3/10.7

10.5/0.4

23.4/10.8

0.1/−12.7

23.4/10.8

■

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.organomet.7b00857. Self-consistent field (SCF) energies (at the B3PW91(PCM)/BS-II//B3PW91/BS-I level), the geometrical structure parameters of 3′+, the optimized structures of CIP-3′, and the electrostatic potential diagrams of the species for the styrene insertion (PDF) Cartesian coordinates of the calculated structures (XYZ)

21.3/9.7

a

The geometry optimizations were performed under vacuum, and the single-point energies were calculated under vacuum or solution. Energies are given in kcal/mol. The free energy (enthalpy) was obtained from the corresponding single-point calculations, including the free energy (enthalpy) correction from the gas-phase frequency calculations. bExperimental data: polymerization reactions carried out in toluene at 288 K (see ref 6).

■

AUTHOR INFORMATION

Corresponding Authors

As shown in Table 4, when the anion was considered, the calculated coordination energies and energy barriers in the gas phase and toluene, chlorobenzene, and dichloromethane solutions are almost the same. These results suggest that the solvation has a negligible effect on the energetics for styrene coordination and insertion. Moreover, when the anion was not considered, the insertion energy barriers of styrene into cationic [(Flu-CH2-Py)Sc-(C17H19)]+ slightly increase with the solvent polarity, which are 22.3, 23.4, 23.8, and 23.9 kcal/mol in the gas phase and toluene, chlorobenzene, and dichloromethane solutions, respectively (Table S4). Therefore, the reason the activity generally increased when the polymerization reaction was carried out in a more polar solvent is that the polar solvent improved the dissociation of contact ion pairs,11 which could be sterically beneficial to the monomer coordination and insertion and thus to the polymerization activity.

*E-mail for B.L.: [email protected]. *E-mail for Y.L.: [email protected]. ORCID

Yi Luo: 0000-0001-6390-8639 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was partially supported by the NSFC (No.21174023, 21429201, and 21674014) and the Fundamental Research Funds for the Central Universities (DUT2016TB08). The authors also thank the Network and Information Center of the Dalian University of Technology for part of the computational resources.

■

■

CONCLUSION The effects of ancillary ligand, metal ions, and solvation on the structure and stability of contact ion pairs (CIPs) in olefin polymerization catalyzed by rare-earth-metal complexes have been investigated by means of DFT calculations. The optimized CIP structure where the B(C6F5)4− anion adopts o- and m-F atoms to coordinate to the metal center is the most thermodynamically stable, except for La analogues, in which the o-F atoms of the two C6F5 groups participate in the coordination because of its largest metal radius. The formation and separation energies of CIPs are found to be influenced by the steric hindrance of the ligand, the electron-donating ability of the cyclopentadienyl group, and the rare-earth-metal ion radius. The solvation effect has been found to be beneficial to the separation of CIPs. Among the solvents considered, the effects of chlorobenzene and dichloromethane solvations are similar, both of which are distinct from that of the less polar toluene. When the anion was considered, the solvation hardly

REFERENCES

(1) Nishiura, M.; Guo, F.; Hou, Z. Acc. Chem. Res. 2015, 48, 2209− 2220. (2) Nishiura, M.; Hou, Z. Nat. Chem. 2010, 2, 257−268. (3) Li, X.; Nishiura, M.; Hu, L.; Mori, K.; Hou, Z. J. J. Am. Chem. Soc. 2009, 131, 13870−13882. (4) Liu, B.; Li, L.; Sun, G.; Liu, J.; Wang, M.; Li, S.; Cui, D. Macromolecules 2014, 47, 4971−4978. (5) Liu, B.; Sun, G.; Li, S.; Liu, D.; Cui, D. Organometallics 2015, 34, 4063−4068. (6) Lin, F.; Wang, X.; Pan, Y.; Wang, M.; Liu, B.; Luo, Y.; Cui, D. ACS Catal. 2016, 6, 176−185. (7) Nishiura, M.; Hou, Z. Nat. Chem. 2010, 2, 257−268. (8) Pan, Y.; Rong, W.; Jian, Z.; Cui, D. Macromolecules 2012, 45, 1248−1253. (9) Gao, W.; Cui, D. J. Am. Chem. Soc. 2008, 130, 4984−4991. (10) Liu, B.; Cui, D. Macromolecules 2016, 49, 6226−6231. (11) Jian, Z.; Cui, D.; Hou, Z. Chem. - Eur. J. 2012, 18, 2674−2684. (12) Zimmermann, M.; Törnroos, K. W.; Anwander, R. Angew. Chem., Int. Ed. 2008, 47, 775−778. H


Article

Organometallics (13) Li, X.; Nishiura, M.; Mori, K.; Mashiko, T.; Hou, Z. Chem. Commun. 2007, 4137−4139. (14) Yu, N.; Nishiura, M.; Li, X.; Xi, Z.; Hou, Z. Chem. - Asian J. 2008, 3, 1406−1414. (15) Lanza, G.; Fragalà, I. L.; Marks, T. J. J. Am. Chem. Soc. 1998, 120, 8257−8258. (16) Lanza, G.; Fragalà, I. L.; Marks, T. J. J. Am. Chem. Soc. 2000, 122, 12764−12777. (17) Lanza, G.; Fragalà, I. L.; Marks, T. J. Organometallics 2002, 21, 5594−5612. (18) Nifant’ev, I. E.; Ustynyuk, L. Y.; Besedin, D. V. Organometallics 2003, 22, 2619−2629. (19) Nifant’ev, I. E.; Ustynyuk, L. Y.; Laikov, D. N. Organometallics 2001, 20, 5375−5393. (20) Correa, A.; Cavallo, L. J. Am. Chem. Soc. 2006, 128, 10952− 10959. (21) Rowley, C. N.; Woo, T. K. Organometallics 2011, 30, 2071− 2074. (22) Luo, Y.; Baldamus, J.; Hou, Z. J. Am. Chem. Soc. 2004, 126, 13910−13911. (23) Li, X.; Nishiura, M.; Hu, L.; Mori, K.; Hou, Z. J. Am. Chem. Soc. 2009, 131, 13870−13882. (24) Deck, P. A.; Beswick, C. L.; Marks, T. J. J. Am. Chem. Soc. 1998, 120, 1772−1784. (25) Chen, E. Y.-X.; Marks, T. J. Chem. Rev. 2000, 100, 1391−1434. (26) Beswick, C. L.; Marks, T. J. J. Am. Chem. Soc. 2000, 122, 10358− 10370. (27) Hayes, P. G.; Piers, W. E.; Parvez, M. Chem. - Eur. J. 2007, 13, 2632−2640. (28) Ciancaleoni, G.; Fraldi, N.; Budzelaar, P. H. M.; Busicoc, V.; Macchioni, A. Dalton Trans. 2009, 8824−8827. (29) McInnis, J. P.; Delferro, M.; Marks, T. Acc. Chem. Res. 2014, 47, 2545−2557. (30) Xu, Z.; Vanka, K.; Firman, T.; Michalak, A.; Zurek, E.; Zhu, C.; Ziegler, T. Organometallics 2002, 21, 2444−2453. (31) Yang, S.-Y.; Ziegler, T. Organometallics 2006, 25, 887−900. (32) Tomasi, S.; Razavi, A.; Ziegler, T. Organometallics 2007, 26, 2024−2036. (33) Kang, X.; Song, Y.; Luo, Y.; Li, G.; Hou, Z.; Qu, J. Macromolecules 2012, 45, 640−651. (34) Kang, X.; Luo, Y.; Zhou, G.; Wang, X.; Yu, X.; Hou, Z.; Qu, J. Macromolecules 2014, 47, 4596−4606. (35) Wang, X.; Lin, F.; Qu, J.; Hou, Z.; Luo, Y. Organometallics 2016, 35, 3205−3214. (36) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; heeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö .; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, Revision A.02; Gaussian, Inc., Wallingford, CT, 2009. (37) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652. (38) Lee, C. T.; Yang, W. T.; Parr, R. G. Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 37, 785−789. (39) Perdew, J. P.; Burke, K.; Wang, Y. Phys. Rev. B: Condens. Matter Mater. Phys. 1996, 54, 16533−16539. (40) Andrae, D.; Haussermann, U.; Dolg, M.; Stoll, H.; Preuss, H. Theor. Chim. Acta 1990, 77, 123−141. (41) Martin, J. M.; Sundermann, A. J. Chem. Phys. 2001, 114, 3408− 3420.

(42) Höllwarth, A.; Böhme, M.; Dapprich, S.; Ehlers, A. W.; Gobbi, A.; Jonas, V.; Köhler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G. Chem. Phys. Lett. 1993, 208, 237−240. (43) Tomasi, J.; Mennucci, B.; Cammi, R. Chem. Rev. 2005, 105, 2999−3094. (44) Riplinger, C.; Neese, F. J. Chem. Phys. 2013, 138, 034106. (45) Riplinger, C.; Sandhoefer, B.; Hansen, A.; Neese, F. J. Chem. Phys. 2013, 139, 134101. (46) Minenkov, Y.; Chermak, E.; Cavallo, L. J. Chem. Theory Comput. 2015, 11, 4664−4676. (47) Bistoni, G.; Riplinger, C.; Minenkov, Y.; Cavallo, L.; Auer, A. A.; Neese, F. J. Chem. Theory Comput. 2017, 13, 3220−3227. (48) Legault, C. Y. CYLview, 1.0b; Université de Sherbrooke, 2009; http://www.cylview.org. (49) CRC Handbook of Chemistry and Physics, 90th ed.; Lide, D. R., Ed.; CRC Press/Taylor and Francis: Boca Raton, FL, 2009.

I


Effects of Ligand, Metal, and Solvation on the Structure and Stability of

Recommend Documents