Article pubs.acs.org/Organometallics
Selective Co-Oligomerization of Ethylene and 1‑Hexene by Chromium-PNP Catalysts: A DFT Study Minglan Gong, Zhen Liu,* Yuanhui Li, Yue Ma, Qiaoqiao Sun, Jialong Zhang, and Boping Liu* State Key Laboratory of Chemical Engineering, East China University of Science and Technology, Meilong Road 130, Shanghai 200237, People’s Republic of China S Supporting Information *
ABSTRACT: The mechanism of selective co-oligomerization of ethylene and 1-hexene by the catalyst [CrCl3(PNPOMe)] (a, PNP OMe = N,N-bis(bis(o-methoxyphenyl)phosphine)methylamine) has been explored in detail using the density functional theory (DFT) method. The full catalytic cycles for the formation of 1-hexene and 1-decenes were calculated on the basis of the metallacyclic mechanism, and the distribution of all decene isomers was explained by locating Gibbs free energy surfaces of various pathways, which is in good agreement with the experimental results. A spin surface crossing through a minimum energy crossing point (MECP) from a sextet to a quartet surface takes place before the formation of metallacyclopentane, which opens up a much lower energy pathway and thus facilitates the following co-oligomerization reactions. It is worth noting that β-hydrogen agostic-assisted hydrogen transfer is of crucial importance for the decomposition of the metallacycle intermediates to give 1-hexene or decenes. Moreover, an analysis of the Cr−O bond distance and NBO charges indicates the important role of a hemiliable methoxy moiety, which acts as a pendant group in the co-oligomerization of ethylene and 1-hexene by CrCl3(PNPOMe) catalyst.
■
INTRODUCTION As catalytic oligomerization of ethylene typically produces a mathematical distribution of LAOs (linear α-olefins), transition-metal-catalyzed ethylene selective oligomerization has received considerable attention from the academic and industrial community.1,2 In particular, trimerization of ethylene to 1-hexene is an area of intense research activity.3,4 It is generally accepted that selective trimerization of ethylene to 1hexene follows a metallacyclic route proposed by Briggs,5 which has been supported by experimental and theoretical studies.6−8 This mechanism, mostly based on chromium catalysts,4,9 titanium catalysts,10,11 and tantalum catalysts,12,13 is believed to involve metallacycle formation by oxidative coupling of two ethylene molecules at the metal center and then ring expansion by insertion of a third ethylene molecule to give a metallacycloheptane that eliminates 1-hexene (Scheme 1). Manyik et al.14,15 first discovered that a significant amount of 1-hexene was formed during ethylene polymerization catalyzed by Cr(III) 2-ethylhexanoate/partially hydrolyzed triisobutylaluminum (PIBAO). In addition to the predominant oligomeric product of 1-hexene, some minor oligomers of 1-octene, 1decene, and branched decenes were also detected in the liquid products. It was suggested that the branched decenes were formed by cotrimerization of two ethylene molecules with a previously formed 1-hexene. In 2000, Yang et al.16 studied ethylene trimerization reactions using a similar catalyst system (Cr(2-ethyl hexanoate)3/2,5-dimethylpyrrole/triethylaluminum/chloro compound) and found that decene isomers are the major byproducts, which were believed to be formed by © XXXX American Chemical Society
Scheme 1. Metallacyclic Mechanism for Ethylene Trimerization
incorporation of 1-hexene in the formation of a 5-membered Cr metallacycle rather than the decomposition of an 11-membered metallacycle. Recently, Do et al.17 published a very detailed Received: December 28, 2015
A
DOI: 10.1021/acs.organomet.5b01029 Organometallics XXXX, XXX, XXX−XXX
Article
Organometallics
experiments by Do et al.17 However, the factors that influence the preferences of the distribution of decenes are ambiguous. Here we have undertaken a detailed theoretical investigation of the cotrimerization of ethylene and 1-hexene on the basis of the complex [CrCl3(PNPOMe)] (a) reported by Do et al.17 and attempt to provide a reasonable mechanistic proposal which could explain experimental observations made with this catalyst. We intended to search for a viable alternative route to traditional nonselective oligomerization from ethylene to decenes and to elucidate the distribution of all decene isomers by means of density functional theory (DFT) calculations.
experimental study on the co-oligomerization of ethylene and 1-hexene using the chromium complex [CrCl3(PNPOMe)] (a, where PNPOMe = N,N-bis(bis(o-methoxyphenyl)phosphine)methylamine; Chart 1) activated by modified methylaluminoxChart 1. Structure of the Chromium Precatalyst [CrCl3(PNPOMe)] (a)
■
COMPUTATIONAL MODELS AND METHODS
All DFT calculations were carried out with the Gaussian09 package.25 Geometry optimizations were performed without any constraint using the B3LYP26 functional in combination with the double-ζ basis set BS1, which corresponds to the LANL2DZ valence double-ζ basis set with effective core potential (ECP) on the Cr atom and the fullelectron Pople double-ζ basis set 6-31G(d,p) for all other elements. Each transition state was further verified by full IRC calculations, which showed a direct connection between the corresponding reactant and the product.27 Then the total energies for all the structures were calculated using the large triple-ζ basis set BS2, which denotes the triple-ζ basis set LANL2TZ(f) on the Cr atom and 6-311G(d,p) on the other atoms. On the basis of the optimized geometries, the dispersion corrections have been evaluated using the DFT-D3 (zerodamping) code developed by Grimme and collaborators.28−30 Furthermore, the free energy of solvation was calculated using a SMD solvation model, portraying chlorobenzene as the model solvent. Throughout the paper, all energies are the total energies at the B3LYP/BS2 level of theory corrected by inclusion of the Gibbs free energy at 298 K and 1 atm in the gas phase, the solvation free energies, and dispersion corrections. The bond parameters of the optimized geometry of the precatalyst model at the B3LYP/BS1 level are summarized in Table S1 in the Supporting Information, which showed good agreement with the reported single-crystal data6 of [CrCl3(PNPOMe)] (a). For some chromium catalyzed ethylene trimerization reactions, it has been found that spin crossover facilitates the chromium catalyzed reactions.31−33 Therefore, we calculated the reaction pathways with two possible spin states including a quartet and a sextet for all of the species in the Cr(I)/Cr(III) catalytic cycle. The results showed that the quartet is the ground spin state for Cr(III) species and the sextet is the ground spin state for Cr(I) complexes. For the possible spin surface crossing, all minimum energy crossing points (MECPs) between two adjacent potential energy surfaces were located using the methodology developed by Harvey et al.34,35 The free energy corrections at the MECPs were estimated with the freq = projected keyword available in
ane (MMAO), which also showed high selectivity for ethylene trimerization previously reported by their group.18−20 In their report, the distribution of decene isomers was obtained by detailed gas chromatographic and mass spectrometric analyses, and a possible reaction mechanism for the formation of linear and branched decenes has been proposed. In a very recent work, Zilbershtein et al.21 reported that decene formation was a complex process and was affected by 1-hexene concentration, ethylene pressure, temperature, and other factors, which suggested that this formation was an intrinsic part of the trimerization reaction mechanism. Many experimental studies of ethylene trimerization reactions have considered the existence of cotrimer decenes. In other words, experimentally it is found that co-oligomerization of ethylene with the previously formed 1-hexene could constitute a route to C10 alkene (albeit with a small amount) using ethylene as the only starting reagent, which is called “product recycling”.21−23 When the concentration of the primary product 1-hexene becomes high enough to compete with ethylene coordination and insertion, it is normally seen that trimerization systems can promote 1-hexene incorporation.18,24 Highly selective decene isomers were obtained using both ethylene and 1-hexene as starting materials in the
Scheme 2. Proposed Catalytic Cycle for 1-Hexene and Decene Formation
B
DOI: 10.1021/acs.organomet.5b01029 Organometallics XXXX, XXX, XXX−XXX
Article
Organometallics Scheme 3. Proposed Reaction Pathways for the a/MMAO Catalyst System in This Worka
a
The percentages shown in red correspond to the relative amounts of C10 alkenes in the product.17
alkylation and cationization agent MMAO.45,46 Bercaw and collaborators reported that inactive neutral tetraaryl PNP chromium complexes can catalyze ethylene trimerization after cationization of the catalyst system.6,20,39 Rucklidge,47 as well as Wass,48,49 also suggested that neutral Cr(0)-PNP complexes were active only if they converted to Cr(I) cationic complexes equipped with a weakly coordinating anion. In addition, theoretical work on the nature of the MAO cocatalyst reported by van Rensburg et al. also demonstrated that a dissociated ionpair complex was likely.50 For this reason, the model catalyst studied herein is the cationic complex ((o-MeOC6H4)2PN(CH3)P(o-MeOC6H4)2-Cr(I)+), and the influence of the counteranion has not been considered in the present study. The oxidation states of the active catalyst are largely a matter of debate, and there is evidence for both Cr(I)/Cr(III) and Cr(II)/Cr(IV) mechanisms. The studies on Cr(III)-dialkyl cations which undergo reductive elimination to a Cr(I) cation39 and the related PNP-Cr complexes47,49 suggested that a Cr(I)/ Cr(III) redox cycle is responsible for the activity. Theopold and collaborators also supported a Cr(I)/Cr(III) redox cycle for their studied chromium(I) dinitrogen complex.51 Recently, Yang et al. supported that a Cr(I)/Cr(III) redox cycle was most feasible by comparing activation energies of rate-determining steps in the ethylene trimerization reaction.31,52 Together with the apparent thermodynamic preference for reduction of Cr(II) to Cr(I) complexes, McGuinness et al.53 also suggested that trimerization of ethylene to 1-hexene via a Cr(I)/Cr(III) pathway was quite feasible for the Cr-PNP system. Recently, a detailed work from Britovsek et al. also showed that the Cr(II)/ Cr(IV) redox cycle was less preferred both kinetically and thermodynamically according to a benchmarking study.23 Despite this supporting evidence for the Cr(I)/Cr(III) redox cycle, the Cr(II)/Cr(IV) redox cycle cannot be simply excluded. The formation of CrII species was detected by several spectroscopic techniques following activation of the precatalysts.54 For instance, a CrII cationic complex was isolated after treatment of a Cr(III)-PNP precursor with trimethylaluminum, which is not active in the absence of MAO.55 As Bercaw and collaborators40 demonstrated, however, the majority of
Gaussian09. For all reaction pathways that led to the formation of decene isomers, we calculated the turnover frequency (TOF) in order to compare our calculation results with the corresponding experimental report.17 The TOF of each reaction pathway was calculated using the energetic span model with the user-friendly AUTOF program developed by Kozuch and Shaik.36−38
■
RESULTS AND DISCUSSION
In this work, the catalytic cycle (Scheme 2) of ethylene trimerization and co-oligomerization was designated on the basis of the postulated metallacyclic mechanism proposed by Briggs.5 In order to facilitate the research, we designed a complete reaction network by considering the most probable products that determined by experiments.17 For simplicity, all the intermediates are denoted by Arabic numerals and the decene isomers are labeled by the alphabet letter from the original experimental report,17 as shown in Scheme 3. Although progress has been made in understanding the mechanism of this catalytic system,6,9,39,40 some key problems are still subject to debate. The formation of 1-hexene and decene from the metallacycloheptane can either undergo a direct intramolecular β-hydrogen transfer to the α′-carbon (one-step route) or a β-hydrogen transfer to Cr followed by a reductive elimination (two-step route). Some experimental and theoretical studies6,7,19,41−44 supported 1-hexene elimination via one-step agostic-assisted hydrogen transfer for the chromium catalyst system. Moreover, Zilbershtein et al.21 found that the stepwise elimination mechanism for the decomposition of the chromacycloheptane intermediate failed to explain the distribution of decenes, and therefore an alternative hydrogen shift mechanism should be considered. In this study, every attempt to locate the transition states for liberating 1-hexene via a stepwise elimination step failed, which argues against the stepwise elimination mechanism as well. Thus, the formation of 1-hexene would go exclusively through the one-step route for this catalyst system, which also agrees with the previous studies.6,7,19 For chromium-catalyzed ethylene selective oligomerization, it is reasonable to consider that the catalyst is a cationic chromium compound, because of the requirement of the C
DOI: 10.1021/acs.organomet.5b01029 Organometallics XXXX, XXX, XXX−XXX
Article
Organometallics
Figure 1. Gibbs free energy profiles at 298.15 K for the spin crossover at 62, 63, 68, and 69. The MECPs are marked with a solid circle. Energies are given in kcal/mol and are relative to 61 plus the corresponding number of ethylene and 1-hexene molecules.
Figure 2. Gibbs free energy profiles for ethylene trimerization. Energies are relative to 61 plus the corresponding number of free ethylene molecules, and energy barriers are indicated in italics. The ligand PNPOMe is omitted for clarity.
hexene over the Cr-PNP catalyst and show that the Cr(I)/ Cr(III) redox cycle is the most facile pathway in comparison to its counterpart Cr(II)/Cr(IV) redox cycle. Spin Transition from Sextet to Quartet Surface. As reported for chromium-catalyzed ethylene trimerization reactions,31 spin crossover could promote the reaction and lead to an increased reactivity for the catalytic cycle on the low spin
chromium species detected by spectroscopic methods are not catalytically active, and the small percentage of CrI or even another undetected minor species could actually be responsible for the activity. Given these experimental difficulties, this system is well suited to gain additional insight from the theoretical studies. Herein, we calculated both Cr(I)/Cr(III) and Cr(II)/Cr(IV) catalytic cycles for the formation of 1D
DOI: 10.1021/acs.organomet.5b01029 Organometallics XXXX, XXX, XXX−XXX
Article
Organometallics
cannot be fully excluded as a potential valence state of the active species on the basis of the limited experimental results, the free energy profiles for ethylene trimerization via Cr(II)/ Cr(IV) cycle are also depicted in Figure S2 in the Supporting Information, which confirms that Cr(I)/Cr(III) is the most plausible catalytic cycle for ethylene trimerization by the CrPNP catalyst. Ethylene Co-Oligomerization. According to the experimental report,17 the [CrCl3(PNPOMe)] catalyst system could facilitate selective cotrimerization of ethylene and 1-hexene with a high yield of C10 olefins: ∼94 wt % of the quantifiable products (i.e., those other than C6). The distribution of C10 isomers is the key factor to understand the mechanism of the cotrimerization of ethylene and 1-hexene. Scheme 3 shows three possible routes leading to the formation of metallacyclopentane 4 with two ethylene molecules or of metallacyclopentanes 10a,b with one ethylene and one 1-hexene. As an ethylene or a 1-hexene molecule coordinates on each of these three five-membered metallacycles (4, 10a,b), there are three variant seven-membered metallacycles formed with two possible routes for each of them. Unsubstituted chromacyclopentane 4 could undergo homotrimerization by insertion of ethylene to give 6 or by insertion of 1-hexene to yield chromacycloheptanes 12a,b, respectively. Similarly, chromacyclopentanes 10a,b could undergo subsequent insertion of an ethylene to yield chromacycloheptanes 12a/12c and 12b/12c, respectively. All of the seven-membered metallacycles could undergo an agostic-assisted β-H shift followed by reductive elimination to release the products as shown in Scheme 3. We will first discuss the formation of the five-membered metallacycles and then the formation of the seven-membered metallacycles followed by ring-opening reactions. Chromacyclopentane Formation. Figure 3 shows the three possible free energy diagrams for the formation of fivemembered metallacycles 4 and 10a,b. The reaction starts with the activated catalyst 1, where an ethylene molecule or a 1hexene coordinated to the cationic Cr(I) species leads to the
potential energy surface. Before the formation of the key intermediate of chromacyclopentane, it is crucial to determine the MECPs with the lowest energy to cross over between two adjacent potential energy surfaces. Since the electrons involved in the spin surface crossing are highly localized on Cr, the strength of the spin−orbit coupling Hamiltonian which mixes the two spin hypersurfaces in the region of crossing is great enough to provide high transmission factors to effectively remove the forbidden nature of the reaction.21 Thus, with a fast crossing rate between the two spin states, the activation barrier of metallacyclopentane formation is effectively reduced through spin acceleration.21,22 With respect to the spin state changing from quartet back to sextet, it occurs after a rate-determining step (metallacycloheptane formation). In other words, spin flipping after the transition states of 1-hexene or decene formation has no significant influence on the reaction rate. In order to find the most facile crossover point, we located two MECPs (6−4CP1 and 6−4CP2) for trimerization pathways and two MECPs (6−4CP3 and 6−4CP4) for co-oligomerization pathways, respectively, which are close to the Cr-PNP adducts with one or two π-coordinated ethylene or 1-hexene molecules (62 and 63, 68 and 69). As a matter of fact, spin surface crossing is revealed by our DFT calculations for the Cr-PNP system, as shown in Figure 1. With coordination of the ethylene molecules, spin flipping at 62 to its analogous quartet 42 requires overcoming an MECP of ca. 10.4 kcal/mol in Gibbs free energy, while the spin transition between 63 and 43 requires a slightly lower barrier of ca. 8.4 kcal/mol. Hence, the spin flipping is predicted to be occurring at complex 63 through the MECP 6−4CP2. Similarly, the spin transition on the cooligomerization pathway is favored at 69 with a barrier of ca. 9.3 kcal/mol, which is slightly lower than that of the spin flipping at 6 8. Since it was reported that the B3LYP functional may slightly overestimate the stability of the high-spin state when referring to spin-state splitting in Cr-based system,53 here we compared the potential energy surface crossing using both B3LYP and M06L functionals. The free energies of 2, 3, and 7−9 are given in Table S2 in the Supporting Information, which shows good agreement in predicting the ground spin state using both B3LYP and M06L functionals. Ethylene Trimerization. Figure 2 shows the free energy pathway of ethylene trimerization catalyzed by the Cr-PNP catalyst. The bare Cr(I) cationic complex shows a sextet ground spin state. With a consecutive coordination of two ethylene molecules, oxidative coupling of bis(ethylene)chromium adduct 3 to give chromacyclopentane 4 needs to overcome a much higher barrier of 32.5 kcal/mol. However, bis(ethylene)chromium adduct 3 can cross over to the quartet surface through a low-lying MECP and the barrier for the oxidative coupling to form the five-membered ring 4 is lowered to 15.9 kcal/mol (TS3-4), where the oxidation state of chromium changes from Cr(I) to Cr(III). Through coordination and subsequent insertion of a third ethylene, the seven-memberedring species 6 is formed with an energy barrier of 13.5 kcal/mol (TS5-6). Finally, 1-hexene is liberated from 6, undergoing direct reductive liberation via an agostic-assisted hydride shift with an energy barrier of 14.3 kcal/mol (TS6-7). Simultaneously, the final reduction from Cr(III) to Cr(I) experiences a spin state change from quartet back to sextet. The formation of the metallacycloheptane could be identified as the ratedetermining step in this system with a general barrier of 18.4 kcal/mol, including the energy for coordination of the third ethylene molecule. Moreover, as mentioned above, since Cr(II)
Figure 3. Gibbs free energy profiles for the formation of chromacyclopentanes. Energies are given in kcal/mol and are relative to 61 plus the corresponding number of free ethylene molecules. The ligand PNPOMe is omitted for clarity. E
DOI: 10.1021/acs.organomet.5b01029 Organometallics XXXX, XXX, XXX−XXX
Article
Organometallics formation of 2 or 8, respectively. Further coordination of an ethylene on 2 yields the bis(ethylene)Cr complex 3. Complex 9 can be attained by either coordination of one 1-hexene on 2 or coordination of one ethylene on 8. Through the two MECPs 6‑4 CP2 and 6‑4CP4, the following oxidative coupling step would generate the two five-membered metallacycles 10a,b by 2,1- or 1,2-insertion of 1-hexene, respectively. The coordination of the first ethylene or 1-hexene to the unsaturated Cr center is exergonic and occurs easily, which is followed by slightly endergonic coordination of the second ethylene on the sextet surface. It can be seen that co-oxidative coupling of ethylene with 1-hexene to isomer 10a is favored with an energy barrier of 17.1 kcal/mol, which is lower than that to form the isomer 10b on the quartet surface. This result argues against Do’s presumption17 that the formation of 10a is less favorable than that to form 10b on the basis of their experimental observations of a small amount of linear C10 alkenes. However, according to our theoretical results, the less favored formation of linear decenes from 10a might be attributed to the high barrier for insertion of the third ethylene into 10a, which will be explained in detail in the next part. Chromacycloheptane Formation and Ring Opening. The formation of various chromacycloheptanes is the key step for understanding the distribution of the C10 products by cotrimerization of ethylene and 1-hexene. Due to the unsymmetrical nature of the chromacycloheptane intermediates, each chromacycloheptane (clockwise structure) 12a−c can be discriminated from their corresponding counterclockwise structures 12a′−c′. Taking 12a as an example shown in Figure 4, the β-hydrogen agostic interaction is contributed by the hydrogen atom from C2 in 12a, while the βhydrogen agostic interaction is contributed by the hydrogen atom from C5 in 12a′. As a comparison of 12a and 12a′, the β-
H(C2)−Cr bond length is around 2.01 Å for 12a while the corresponding distance in 12a′ is 3.39 Å. Actually, the βH(C5)−Cr bond length in 12a′ is even shorter (1.99 Å). Obviously, it is reasonable to infer that the agostic-assisted βhydrogen transfer to the α-carbon is a facile pathway to generate 1-hexene or various decenes. The hydrogen transfer of chromacycloheptane intermediate 12a or 12a′ results in the decene-coordinated product 13a or 13b. The distance of the carbon−carbon double bond to the chromium center is roughly 2.15 Å. Other structures such as 12b,c are shown in Figure S4 in the Supporting Information. Gibbs free energy profiles for these reductive elimination reactions are shown in Figure 5. It is demonstrated that the energy barrier of 14.5 kcal/mol for agostic-assisted β-hydrogen transfer in 12a to give 13a is higher than that of the β-hydrogen transfer in 12a′ to 13b. With regard to 12b,c, there is no apparent difference in Gibbs free energy barriers for the corresponding reductive elimination. Considering the relative position on the free energy surfaces, the reductive elimination from 12b′,c might be more feasible than that from their counterpart structures 12b,c′. As a result, we have chosen the structures of 12a′,b′,c for following discussion and the prime symbol (′) has been neglected for clarity. The Gibbs free energy profiles for the formation of chromacycloheptanes and ring opening to decenes are depicted in Figure 6. There are three possible pathways leading to the formation of decenes from 4 and 10a,b, respectively. Figure 6A shows that insertion of 1-hexene into the chromacyclopentane 4 follows the 1,2-insertion mode. Despite many attempts to locate a transition state TS11a-12a for generating 12a in the 2,1-insertion mode of 1-hexene, the IRC confirms the located TS11a-12a is indeed TS11a-12b. That is to say, 1-hexene inserts into the five-membered metallacycle 4 exclusively in the 1,2-mode, which is consistent with the experimental phenomena.17,39 However, a preference for 2,1-insertion of styrene is common. For example, in the cotrimerization of ethylene and styrene by [CrCl3(PNPOMe)]/MAO, no 1,2-insertion products were observed.56 This is an interesting phenomenon, and relevant studies are being carried out. The coordination and insertion of the third ethylene on 10a,b generates chromacycloheptane species 12a−c via two different routes passing two close transition states, respectively. The insertion of an ethylene into 10b only needs to overcome a barrier of about 13.5−14.3 kcal/mol, while the third ethylene inserts into 10a requiring the passing of a general barrier of about 19.4−20.9 kcal/mol. This explains that the linear decenes formed through 10a are less favorable in comparison to the branched decenes, which agrees with the experimental observation.17 The transition states TS12a-13b, TS12b-13d, and TS12c-13e for the ring opening of chromacycloheptanes are calculated to be very similar. Since the chromacycloheptane 12a is slightly higher in energy on the free energy diagram, the corresponding β-hydrogen shift is relatively easier (with a barrier of 9.8 kcal/mol in Figure 6B) than that from the more stable complexes 12b,c. Figure 7 illustrates the whole free energy diagrams for ethylene co-oligomerization to decenes catalyzed by CrPNPOMe. The general barrier of the rate-determining step is considered by inclusion of the binding energy of ethylene coordination and the energetic barrier for ethylene/1-hexene insertion. In a comparison of Figures 2 and 7, insertion of ethylene into 4 needs to overcome a general barrier of 18.4 kcal/mol, which is slightly lower than that for insertion of 1-
Figure 4. Graphic representation of metallacycloheptanes 12a,a′ and the product complexes 13a,b. Bond distances are given in angstroms. F
DOI: 10.1021/acs.organomet.5b01029 Organometallics XXXX, XXX, XXX−XXX
Article
Organometallics
Figure 5. Gibbs free energy profiles for the decomposition of chromacycloheptanes. Energies are given in kcal/mol and are relative to 61 plus the corresponding number of free ethylene molecules. 4‑6CP5, 4‑6CP6, 4‑6CP7, 4‑6CP8, 4‑6CP9, and 4‑6CP10 represent MECPs close to 613a−f, respectively.
argument.17 12a could only be generated from 10a via the transition state TS11b-12a, which requires a general barrier of 19.4 kcal/mol. Less competitively, insertion of ethylene into the substituted position of 10a to give 12c requires overcoming a general barrier of 20.9 kcal/mol. It is clearly shown in Figure 7 that both pathways via 10a are less favorable in comparison to other reaction routes due to the presence of two high barriers. 12b could be possibly generated from either unsubstituted chromacyclopentane 4 or chroma(butyl)cyclopentane 10b. As discussed above, insertion of 1-hexene into unsubstituted chromacyclopentane 4 requires overcoming a high barrier of 18.8 kcal/mol. Alternatively, through the most favorable route among the five possible routes to chromacycloheptanes, 12b can be generated from 10b via transition state TS11c-12b with a general barrier of 13.5 kcal/mol. Therefore, decomposition of 12b will lead to the formation of the main decene isomers (M and N in Scheme 3), while only a small amount of linear decene isomers H−L could be possibly generated from 12a. The pathway from 10b to 12c also occurs easily with a general barrier of 14.3 kcal/mol, which leads to the production of decene isomers (O and P) in moderate amounts. This agrees with the distribution of decenes in the experimental report by Do et al.17 It is worth mentioning that Britovsek et al.23 also confirmed the feasibility of co-oligomerization by considering the C10 products in their theoretical study of the Cr-PNP ethylene trimerization system. They investigated two pathways for generation and decomposition of chromacycloheptane (here denoted as 12b,c) by focusing on the 3-butylcyclopentane complex (here denoted as 10b). They found that metallacycle formation involving 1-hexene to give a metalla(3-butylcyclopentane) complex shows a barrier very similar to that involving two ethylene units. However, our calculation reveals that formation of the five-membered metallacycle 4 by two ethylene molecules or of 10a by 2,1-insertion of 1-hexene needs to go through the two close-lying transition states TS43-4 and TS4910a. The transition state TS49-10b for the formation of 10b by 1,2-insertion of 1-hexene lies about 3 kcal/mol above TS43-4. The disagreement might arise from the differences in structures of the catalyst and model investigated in our study and in the study by Britovsek et al.23 In order to make a side by side comparison of theoretical and experimental product distributions, we calculated the turnover
Figure 6. Gibbs free energy profiles for the formation of chromacycloheptanes and the product complexes. Energies are given in kcal/mol and are relative to 61 plus the corresponding number of free ethylene molecules. The ligand PNPOMe is omitted for clarity.
hexene into 4 with a general barrier of 18.8 kcal/mol. Therefore, insertion of olefin into the unsubstituted chromacyclopentane 4 exhibits a preference for ethylene over 1-hexene with 1,2-insertion, which is consistent with the experimental G
DOI: 10.1021/acs.organomet.5b01029 Organometallics XXXX, XXX, XXX−XXX
Article
Organometallics
Figure 7. Gibbs free energy profiles for ethylene/1-hexene cotrimerization. Energies are given in kcal/mol and are relative to 61 plus the corresponding number of free ethylene molecules. Blue hue lines and red lines illustrate free energy surfaces in the quartet and sextet states, respectively.
agreement with the ratio of 3:52:45 measured in the experiments.17 On the basis of the current calculations, we believe this agreement is rather good, because calculated TOFs are very sensitive to the barrier height of the rate-determining step, which is very difficult to calculate accurately with the DFT method, especially for transition-metal systems.57 Notably, a pendant methoxy moiety of the PNPOMe ligand to the chromium center plays a crucial role throughout the trimerization pathway and co-oligomerization pathways. Taking the intermediates 4 and 5 as an example (Figure 8), the Cr−O1 distance in 4 and 5 is around 2.2 Å, which efficiently displays a strong coordination of the oxygen atom of a methoxy moiety to the chromium center in the course of the ethylene coordination, while the Cr−O4 distance is elongated from 2.8 Å in 4 to 4.5 Å in 5. Since one of those two oxygen atoms has a stable interaction with the chromium center and thus occupies a coordinating site of the chromium atom, the coordination of two ethylene molecules to the chromacyclopentane 4 is prohibited from forming a double-coordinated bis(ethylene)chromacyclopentane intermediate. Recently, Britovsek et al.23 suggested that a double-coordinated bis(ethylene)chromacyclopentane might be the key intermediate for the formation of 1-octene, while the formation of 1-hexene is through a monocoordinated chromacyclopentane intermediate. As a matter of fact, the Cr/PNPOMe catalyst system is a very good ethylene trimerization catalyst with a small amount (even negligible) of tetramerization product, which can be understood from the aforementioned double-coordination mechanism. NBO charge analysis exhibits a good correlation with the Cr−O1 distance (Figure 9). Elongation of the Cr−O1 distance leads to an increase in the charge of the O1 in Figure 8. Referring to Budzelaar’s44 and Yang’s52 proposal for a pendant chlorine, we presume that the coordinating ability may be mainly but not exclusively a consequence of electronic effects. Steric constraints might also play a role in adjusting the Cr−O
frequency (TOF) for all possible pathways that lead to the formation of ethylene cotrimerization products. The calculated TOFs for all nine pathways are listed in Table 1, and the Table 1. Calculated TOFs and Gibbs Free Energy Barriers for the Rate-Determining Step of Each Reaction Pathway for Ethylene Co-Oligomerization pathwaya
ΔG/kcal mol−1
TOF/h−1
TOF/h−1
P1 P2 P3 P4 P5 P6 P7 P8 P9
19.4 19.4 18.8 13.5 13.5 20.9 20.9 14.3 14.3
0.047 0.026 11.0 13.5 7.6 0.0037 0.0021 6.9 3.9
0.073 32.1
10.8
a
Definitions of pathways: P1, 1−2−9−10a−11b−12a; P2, 1−8−9− 10a−11b−12a; P3, 1−2−3−4−11a−12b; P4, 1−2−9−10b−11c− 12b; P5, 1−8−9−10b−11c−12b; P6, 1−2−9−10a−11b−12c; P7, 1−8−9−10a−11b−12c; P8, 1−2−9−10b−11c−12c; P9, 1−8−9− 10b−11c−12c.
notations for each pathway are detailed in the table footnotes. Clearly, pathways P1 and P2 are two unfavorable routes leading to the formation of linear decenes via 12a. Pathways P3−P5 are expected to result in the formation of 12b with a total TOF of about 32 h−1, and thus the corresponding products M and N can be predicted as the main decene isomers. At the current stage it is difficult to determine the exact ratio between M and N or the ratio between O and P. Nevertheless, the calculated TOF values reveal a ratio of 0.2:74.7:25.1 for the linear decene isomers (H−L), 2-butyl-1-hexene (M) and 5-methyl-1-nonene (N), and 4-ethyl-1-octene (O) and 3-propyl-1-heptene (P), respectively. The calculated ratio of 0.2:74.7:25.1 shows good H
DOI: 10.1021/acs.organomet.5b01029 Organometallics XXXX, XXX, XXX−XXX
Article
Organometallics
Figure 8. Optimized structures of complexes 4 and 5, as well as selected bond distances (Å). The spin multiplicities of 4 and 5 are both quartets.
on both the N and P atoms of the PNPOMe ligand, with the aim to discover a better ligand that promotes the generation of decenes in ethylene co-oligomerization reactions.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.organomet.5b01029. Graphical representation as well as the calculated bond distances of [CrCl 3 (PNP OMe )] (PNP OMe = (oMeOC6H4)2PN(CH3)P(o-MeOC6H4)2), Gibbs free energies of 2, 3, and 7−9 under different spin states by B3LYP and M06L, Gibbs free energies of all intermediates and transition states under all possible spin states, and Gibbs free energy diagrams for ethylene trimerization via the Cr(II)/Cr(IV) redox cycle (PDF) Cartesian coordinates for all intermediates, transition states, and MECPs (XYZ)
Figure 9. Cr−O1 distances and charges on O1 in each intermediate and transition state along the trimerization reaction pathway. The O1 here corresponds to the O1 marked in Figure 8.
distance, but the influence is not dominant. For instance, 3 and 4 have similar steric hindrance in the coordinating sphere but feature notably different Cr−O distances. Thus, it is believed that the pendant methoxy moiety is conductive to high selectivity for ethylene trimerization.9
■
■
CONCLUSIONS In summary, we investigated ethylene co-oligomerization mechanisms for a cationic chromium-PNPOMe system. The calculations support the metallacycle mechanism for the ethylene trimerization/cotrimerization via various chromacycloheptane intermediates followed by a concerted agosticassisted β-hydrogen reductive elimination. We showed the feasibility of selectively producing decenes through cooligomerization of the ethylene and 1-hexene using both ethylene and 1-hexene as starting materials. The rationale leading to the distribution of all major decene isomers was revealed with a detailed PES analysis and TOF calculation, which appears to be in good accord with experimental observations. Moreover, a plausible spin surface crossing between Cr(I)/sextet and Cr(III)/quartet has also been theoretically clarified. The pendant methoxy moiety of the PNPOMe ligand strongly coordinates with the chromium center, which is proposed to be a key factor for 1-hexene selectivity. Ongoing work is focused on investigating various substituents
AUTHOR INFORMATION
Corresponding Authors
*E-mail for Z.L.:
[email protected]. *E-mail for B.L.:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors thank the National Natural Science Foundation of China (No. 21174037, No. 21304033). This work is also financially supported by the Research Program of the State Key Laboratory of Chemical Engineering, the Program of Introducing Talents of Discipline to Universities (B08021), and the Fundamental Research Funds for the Central Universities.
■
REFERENCES
(1) McGuinness, D. S. Chem. Rev. 2011, 111, 2321−2341.
I
DOI: 10.1021/acs.organomet.5b01029 Organometallics XXXX, XXX, XXX−XXX
Article
Organometallics (2) Forestiere, A.; Olivier-Bourbigou, H.; Saussine, L. Oil Gas Sci. Technol. 2009, 64, 649−667. (3) Dixon, J. T.; Green, M. J.; Hess, F. M.; Morgan, D. H. J. Organomet. Chem. 2004, 689, 3641−3668. (4) Blann, K.; Bollmann, A.; Dixon, J. T.; Hess, F. M.; Killian, E.; Maumela, H.; Morgan, D. H.; Neveling, A.; Otto, S.; Overett, M. J. Chem. Commun. 2005, 620−621. (5) Briggs, J. R. J. Chem. Soc., Chem. Commun. 1989, 674−675. (6) Agapie, T.; Schofer, S. J.; Labinger, J. A.; Bercaw, J. E. J. Am. Chem. Soc. 2004, 126, 1304−1305. (7) Janse van Rensburg, W.; Grové, C.; Steynberg, J. P.; Stark, K. B.; Huyser, J. J.; Steynberg, P. J. Organometallics 2004, 23, 1207−1222. (8) Overett, M. J.; Blann, K.; Bollmann, A.; Dixon, J. T.; Haasbroek, D.; Killian, E.; Maumela, H.; McGuinness, D. S.; Morgan, D. H. J. Am. Chem. Soc. 2005, 127, 10723−10730. (9) Agapie, T. Coord. Chem. Rev. 2011, 255, 861−880. (10) Zhang, Y. L.; Ma, H. Y.; Huang, J. L. J. Mol. Catal. A: Chem. 2013, 373, 85−95. (11) Deckers, P. J. W.; Hessen, B.; Teuben, J. H. Organometallics 2002, 21, 5122−5135. (12) Arteaga-Muller, R.; Tsurugi, H.; Saito, T.; Yanagawa, M.; Oda, S.; Mashima, K. J. Am. Chem. Soc. 2009, 131, 5370−5371. (13) Andes, C.; Harkins, S. B.; Murtuza, S.; Oyler, K.; Sen, A. J. Am. Chem. Soc. 2001, 123, 7423−7424. (14) Manyik, R. M.; Walker, W. E.; Wilson, T. P. (Union Carbide Corp.) U.S. Patent 3300458, 1967; p 8. (15) Manyik, R. M.; Walker, W. E.; Wilson, T. P. J. Catal. 1977, 47, 197−209. (16) Yang, Y.; Kim, H.; Lee, J.; Paik, H.; Jang, H. G. Appl. Catal., A 2000, 193, 29−38. (17) Do, L. H.; Labinger, J. A.; Bercaw, J. E. Organometallics 2012, 31, 5143−5149. (18) Carter, A.; Cohen, S. A.; Cooley, N. A.; Murphy, A.; Scutt, J.; Wass, D. F. Chem. Commun. 2002, 858−859. (19) Agapie, T.; Day, M. W.; Henling, L. M.; Labinger, J. A.; Bercaw, J. E. Organometallics 2006, 25, 2733−2742. (20) Schofer, S. J.; Day, M. W.; Henling, L. M.; Labinger, J. A.; Bercaw, J. E. Organometallics 2006, 25, 2743−2749. (21) Zilbershtein, T. M.; Kardash, V. A.; Suvorova, V. V.; Golovko, A. K. Appl. Catal., A 2014, 475, 371−378. (22) Tomov, A. K.; Chirinos, J. J.; Long, R. J.; Gibson, V. C.; Elsegood, M. R. J. Am. Chem. Soc. 2006, 128, 7704−7705. (23) Britovsek, G. J. P.; McGuinness, D. S.; Wierenga, T. S.; Young, C. T. ACS Catal. 2015, 5, 4152−4166. (24) Deckers, P. J.; Hessen, B.; Teuben, J. H. Angew. Chem., Int. Ed. 2001, 40, 2516−2519. (25) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M. L.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M. E.; 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, J. M.; 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, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, Revision A.01; Gaussian, Inc., Wallingford, CT, 2009. (26) Becke, A. D. J. Chem. Phys. 1993, 98, 1372. (27) Fukui, K. Acc. Chem. Res. 1981, 14, 363−368. (28) Ehrlich, S.; Moellmann, J.; Grimme, S. Acc. Chem. Res. 2013, 46, 916−926. (29) Brandenburg, J. G.; Alessio, M.; Civalleri, B.; Peintinger, M. F.; Bredow, T.; Grimme, S. J. Phys. Chem. A 2013, 117, 9282−9292.
(30) Brandenburg, J. G.; Grimme, S. Top. Curr. Chem. 2013, 345, 1− 23. (31) Yang, Y.; Liu, Z.; Zhong, L.; Qiu, P. Y.; Dong, Q.; Cheng, R. H.; Vanderbilt, J.; Liu, B. P. Organometallics 2011, 30, 5297−5302. (32) Liu, Z.; Cheng, R.; He, X.; Wu, X.; Liu, B. J. Phys. Chem. A 2012, 116, 7538−7549. (33) Liu, Z.; Cheng, R. H.; He, X. L.; Liu, B. P. ACS Catal. 2013, 3, 1172−1183. (34) Harvey, J. N.; Aschi, M.; Schwarz, H.; Koch, W. Theor. Chem. Acc. 1998, 99, 95−99. (35) Harvey, J. N.; Aschi, M. Phys. Chem. Chem. Phys. 1999, 1, 5555− 5563. (36) Kozuch, S.; Shaik, S. J. Am. Chem. Soc. 2006, 128, 3355−3365. (37) Kozuch, S.; Shaik, S. J. Phys. Chem. A 2008, 112, 6032−6041. (38) Uhe, A.; Kozuch, S.; Shaik, S. J. Comput. Chem. 2011, 32, 978− 985. (39) Agapie, T.; Labinger, J. A.; Bercaw, J. E. J. Am. Chem. Soc. 2007, 129, 14281−14295. (40) Do, L. H.; Labinger, J. A.; Bercaw, J. E. ACS Catal. 2013, 3, 2582−2585. (41) Blom, B.; Klatt, G.; Fletcher, J. C. Q.; Moss, J. R. Inorg. Chim. Acta 2007, 360, 2890−2896. (42) Bhaduri, S.; Mukhopadhyay, S.; Kulkarni, S. A. J. Organomet. Chem. 2009, 694, 1297−1307. (43) Qi, Y.; Dong, Q.; Zhong, L.; Liu, Z.; Qiu, P. Y.; Cheng, R. H.; He, X. L.; Vanderbilt, J.; Liu, B. P. Organometallics 2010, 29, 1588− 1602. (44) Budzelaar, P. H. M. Can. J. Chem. 2009, 87, 832−837. (45) McGuinness, D. S.; Overett, M.; Tooze, R. P.; Blann, K.; Dixon, J. T.; Slawin, A. M. Z. Organometallics 2007, 26, 1108−1111. (46) McGuinness, D. S.; Rucklidge, A. J.; Tooze, R. P.; Slawin, A. M. Z. Organometallics 2007, 26, 2561−2569. (47) Rucklidge, A. J.; McGuinness, D. S.; Tooze, R. P.; Slawin, A. M. Z.; Pelletier, J. D. A.; Hanton, M. J.; Webb, P. B. Organometallics 2007, 26, 2782−2787. (48) Bowen, L. E.; Haddow, M. F.; Orpen, A. G.; Wass, D. F. Dalton Trans. 2007, 1160−1168. (49) Dulai, A.; de Bod, H.; Hanton, M. J.; Smith, D. M.; Downing, S.; Mansell, S. M.; Wass, D. F. Organometallics 2009, 28, 4613−4616. (50) Janse van Rensburg, W.; van den Berg, J.-A.; Steynberg, P. J. Organometallics 2007, 26, 1000−1013. (51) Monillas, W. H.; Young, J. F.; Yap, G. P.; Theopold, K. H. Dalton Trans. 2013, 42, 9198−9210. (52) Yang, Y.; Liu, Z.; Cheng, R. H.; He, X. L.; Liu, B. P. Organometallics 2014, 33, 2599−2607. (53) McGuinness, D. S.; Chan, B.; Britovsek, G. J. P.; Yates, B. F. Aust. J. Chem. 2014, 67, 1481−1490. (54) Bruckner, A.; Jabor, J. K.; McConnell, A. E. C.; Webb, P. B. Organometallics 2008, 27, 3849−3856. (55) Jabri, A.; Crewdson, P.; Gambarotta, S.; Korobkov, I.; Duchateau, R. Organometallics 2006, 25, 715−718. (56) Bowen, L. E.; Wass, D. F. Organometallics 2006, 25, 555−557. (57) Jiang, W.; DeYonker, N. J.; Determan, J. J.; Wilson, A. K. J. Phys. Chem. A 2012, 116, 870−885.
J
DOI: 10.1021/acs.organomet.5b01029 Organometallics XXXX, XXX, XXX−XXX