Lewis Acid Enhanced Ethene Dimerization and Alkene Isomerization

Nov 9, 2015 - Chile Nanobiotechnology Division at University of Talca, Fraunhofer Chile Research Foundation − Center for Systems Biotechnology, FCR-...
6 downloads 13 Views 2MB Size
Letter pubs.acs.org/acscatalysis

Lewis Acid Enhanced Ethene Dimerization and Alkene IsomerizationESI-MS Identification of the Catalytically Active Pyridyldimethoxybenzimidazole Nickel(II) Hydride Species Manuel A. Escobar,† Oleksandra S. Trofymchuk,† Barbara E. Rodriguez,† Claudia Lopez-Lira,‡ Ricardo Tapia,‡ Constantin Daniliuc,§ Heinz Berke,∥ Fabiane M. Nachtigall,⊥,¶ Leonardo S. Santos,*,⊥,¶ and Rene S. Rojas*,† †

Nucleus Millennium Chemical Processes and Catalysis (CPC), Laboratorio de Química Inorgánica, Facultad de Química, Pontificia Universidad Católica de Chile, Casilla 306, Santiago-22 6094411, Chile ‡ Laboratorio de Química Orgánica, Facultad de Química, Pontificia Universidad Católica de Chile, Casilla 306, Santiago-22 6094411, Chile § Organisch-Chemisches Institut, Universität Münster, Corrensstrasse 40, 48149 Münster, Germany ∥ Chemisches Institut, Universität Zürich, Winterthurerstrasse 190, CH 8057, Zurich, Switzerland ⊥ Laboratory of Asymmetric Synthesis, Chemistry Institute of Natural Resources, University of Talca, P.O. Box 747, Talca 3460000, Chile ¶ Chile Nanobiotechnology Division at University of Talca, Fraunhofer Chile Research Foundation − Center for Systems Biotechnology, FCR-CSB, P.O. Box 747, Talca 3460000, Chile S Supporting Information *

ABSTRACT: A cationic methallyl 2-pyridine-4,7-dimethoxybenzimidazole (L1) nickel precatalyst is highly selective in ethene dimerizations to 1-butene. The same catalyst isomerizes 1-butene and 1-octene to internal olefins. Co-catalytic additives of B(C6F5)3 or BF3·OEt2 coordinate to the catalyst and increase the reaction rates of ethene dimerization. ESI-MS was applied identifying a [L1NiH]+ cation as the catalytically active species.

KEYWORDS: ethene dimerization, nickel catalysis, olefin isomerization, microreactor, mass spectrometry

I

catalysis. When the allyl or methallyl ligands are combined with NN, NO, PP, PN coordinated ligands, the resulting nickel precatalysts are capable to oligomerize ethene (a, c, e, and h, Scheme 1) or to polymerize ethene (b, d, f, g).2 The a−c systems are cationic, whereas nickel compounds d− h are neutral. The latter form zwitterions via Lewis acid attachment, for instance, with B(C6F5)3, coordinating to a ligand donor atom of d−h (Scheme 1). Methallyl nickel complexes were seen to work with (a, d−h) or without (b, c) coactivator but were found to have slow initiation rates.3 Sometimes, when adding larger amounts of the coactivator, the kinetics of the ethene oligomerizations (and of polymerizations) get accelerated.3,4 Frequently reported

n the Shell Higher Olefin Process (SHOP), olefins are oligomerized via neutral Ni(II) catalysts bearing bidentate monoanionic ligands.1 In this context, it is a challenging task to develop alternative catalytic systems, which selectively provide ethene dimerization and can “on demand” isomerize the terminal butene primary product to an internal butene.1 Lowmolecular-weight oligomerizations of ethene require the βhydride elimination step to be fast with respect to the olefin insertion step into the nickel alkyl bond.1g To accomplish this goal within nickel catalysis, we pursued a strategy of constructing cationic nickel(II) hydride centers with neutral NN ligands and to boost their electrophilicity further by cocatalytic addition of strong Lewis acids attached to appropriately functionalized ligand sites. Nickel precatalysts applied in C−C- and C−H-forming processes often contain allyl or methallyl moieties,2 which are anticipated to be split off in the initiation process of the © 2015 American Chemical Society

Received: September 9, 2015 Revised: November 6, 2015 Published: November 9, 2015 7338

DOI: 10.1021/acscatal.5b02003 ACS Catal. 2015, 5, 7338−7342

Letter

ACS Catalysis Scheme 1. Reported Methallyl and Allyl Nickel Pre-Catalysts

precatalysts for this purpose are allyl or methallyl nickel complexes bearing neutral NN-type α- or β-diimine ligands. However, there are only very few examples of methallyl or allyl nickel complexes combined with other NN-ligands, which are capable of activating ethene (for example, g).2g Electrospray ionization tandem mass spectrometry was previously used to study the mechanism of Brookhart olefin polymerizations5,6 and also of Ziegler−Natta catalyzes by coupling the ESI-MS analytics online with a microreactor7 and this combination showed significant advantages over the more conventional parallel screening approaches. These experiments provided insight into the ionic reactive species taking a prominent role in the catalytic cycle. Via these studies, the catalytically active species of ethene polymerizations could be identified, which appear in the presence of cocatalytic MAO.5−7 Electrospray ionization tandem mass spectrometry seemed in addition very suited to deepen the mechanistic insight into ethene oligomerizations (and polymerizations) with methallyl nickel catalysts but naturally also into the related cases of ethene dimerization. MS-derived mechanistic information was sought to support the tuning efforts for higher catalytic selectivities, particularly when the olefin dimerization is followed by isomerization processes of the primary terminal product olefins.8 For more conclusive results in the isomerization process, we intended to supplement the butene studies by a study of the nickel-catalyzed isomerization of 1-octene. Recently, R. Breuil et al. put forward that “cationic nickel active species are electronically unsaturated and highly electrophilic”.9 They generally comprise nickel(II) active species, which have one bidentate donor ligand incorporated and possess square planar coordination geometry. However, these species were often difficult to prepare, since they showed high reactivity. Despite such difficulties, cationic π-allyl nickel complexes could already be obtained in the quite early days of organometallic chemistry by Wilke et al. (see Scheme 2).10

On the basis of this notion to build highly electrophilic nickel catalysts, we tried to explore synthetic access to methallyl nickel complexes with the neutral 2-pyridine-4,7-dimethoxybenzimidazole11 (L1) ligand, which in addition was designed to bear methoxy functionalities interacting well with Lewis acidic cocatalysts to foster the Lewis acidity of the nickel center and consecutively boost the catalytic activities and the selectivities for low molecular weight oligomers.3,4 The quite stable, yellow methallyl nickel complex, [L1Ni(η3-methallyl)][B(ArF)4] (1) was prepared by direct reaction of L1 with [Ni(η3-methallyl)Cl]2 in the presence of [NaB(ArF)4]3 in 78% yield (Scheme 3). Scheme 3. Synthesis of the Nickel Complex 1

NMR characterization of 1 was consistent with the formation of a single isomer containing the bidentate L1 and the η3methallyl ligand. The 1H NMR of 1 at room temperature showed the syn and anti allylic hydrogen atoms at δ 2.8, 3.33, and 4.22 ppm (was confirmed by low temperature NMR, see SI page 4). The resonances of the anti allylic hydrogen atoms appeared as broadened signals compared to the syn ones, thus suggesting fluxional behavior of the allyl moiety on the NMR time scale. The 11B and 19F NMR spectra of 1 revealed characteristic singlets for the [B(ArF)4]− counterion at −6.5 (ν1/2 ∼ 11 Hz) and at −62 ppm, respectively. An X-ray diffraction study confirmed the pseudosquare planar structure of 1 in the solid state (N1−C21−C23−N3 dihedral angle is −4.9°), which apparently coincides with the structure in solution (see Figure 1). To explore the catalytic potential of 1 in ethene oligomerization or dimerization, we first performed NMRscale experiments using dichloromethane-d2 as solvent at 25 and 50 °C with and without the use of a Lewis acidic coactivators B(C6F5)3 or BF3·OEt2. From these studies, we concluded that 1 presents single component behavior.12 In an NMR experiment in the absence of a cocatalyst, 1 turned out to slowly react with ethene producing 1-butene, which subsequently isomerized in a still slower reaction to 2butene.

Scheme 2. Cationic Model of a Nickel-Based Active Species

7339

DOI: 10.1021/acscatal.5b02003 ACS Catal. 2015, 5, 7338−7342

Letter

ACS Catalysis

agent (see entries 3−6, Table 1), suggesting in these cases a rate increase of the β-hydride elimination step relative to the olefin insertion reaction (ethene insertion, alkene isomerization, and β-hydride elimination steps are depicted in Scheme 4, and the coordinatively highly unsaturated nickel alkyl species 1c and 1e are drawn with β-agostic stabilization16). The B(C6F5)3 cocatalytic entries 3 and 4 show highest reaction rates in TOFs, which may be explained by a strong association equilibrium of B(C6F5)3, particularly to the 1c species fostering the β-hydride elimination step. In addition, there is a temperature effect: for the B(C6F5)3 addition cases of entries 3 and 4, there is an increasing selectivity for internal butenes on the expense of the amount of 1-butene, which could mean that with increasing temperature B(C6F5)3 dissociation does not occur to a great extent and all the crucial reaction steps, including also the olefin insertion step, that is accelerated because of the temperature increase. This idea is also supported by the similar TON values with changes in the catalytic activity merely based on the temperature effect on reaction rates and not on a changed amount of active species. However, for the BF3·OEt2 addition cases of entries 5 and 6, the interpretation of the catalytic reaction course seems much more complex and can in our eyes not be resolved by a simple rational. In contrast to the B(C6F5)3 cocatalytic reactions, we assume stronger dissociation equilibria with the weaker Lewis acid BF3 occurring for the prominent reactive species of crucial reaction steps. The BF3 added reactive species are thus expected to be present in lower amounts when temperature increases but Table 1 demonstrates the opposite observation with an increase in the overall reaction rate in form of the enhancement of the TOF (Table 1). In addition, the yields of hexanes increases at the expense of the butenes upon a raise in temperature, presumably meaning that olefin insertion is enhanced. This contradicting general situation also motivated us to conduct a more detailed search for the catalytically active species in these catalyzes (vide infra). The reactivity differences observed between the oligomerization reactions carried out in a Parr reactor (toluene) and NMR experiment (CD2Cl2) can be explained by the effect of solvent, and therefore, when an experiment was carried out in deuterated toluene as solvent, the reaction rate was slower (see respective 1H NMR spectra in SI, page 24). Additionally, when the reaction time was increased to 30 min (see entry 7, Table 1), the same butene/hexene and 1-butene/2-butene relations were observed; however, the TOF value decreased probably to catalyst decomposition at higher reaction times (see entry 4 and entry 7 in Table 1).

Figure 1. Structure of 1 with thermal ellipsoids drawn at the 30% probability level (side view). Hydrogen atoms and the [B(ArF)4]− counterion are omitted for clarity.

When B(C6F5)3 or BF3·OEt2 was introduced as a cocatalyst, the formation of 1-butene and its subsequent isomerization product 2-butene was observed, however, in much shorter reaction time (30 times faster with B(C6F5)3 and 70 times faster with BF3·OEt2, 1H NMR in CD2Cl2 at room temperature). Adding 5 equiv of BF3·OEt2 to the reaction solution initiated immediate ethylene consumption. In order to clarify whether (a) 1-butene was formed first and then underwent reinsertion and isomerization producing 2-butene, or (b) whether 2-butene was directly formed through a chain walking mechanism,13 an additional ethene dimerization experiment (C6D6, ethene pressure of 12.5 bar) was performed in a Parr reactor. Mainly 1-butene was detected by NMR indicating that 2-butene formation occurs only in absence of ethene or at very low ethene concentrations,14 an observation which confirmed notion (a) because of a competitive inhibition effect of higher ethene concentrations hindering or preventing coordination of butene. Additionally, to probe the capability of 1 to isomerize terminal olefins, model type experiments with 1-octene were carried out at room temperature. Also, in this case, 1 featured single component behavior and isomerized 1-octene without use of a cocatalyst. The products of 1-octene isomerization forming internal olefins were 2-, 3-, and 4-octenes, which were detected by 13C NMR in an approximate ratio of 4:7:1. In the presence of B(C6F5)3 and BF3·OEt2, the isomerization of 1octene occurred instantaneously.15 The up-scaling of the aforementioned ethene dimerization reactions was carried out in a Parr reactor (12.5 bar of ethene) in toluene leading to dimerization selectivities >89% for 1- and 2-butenes in the absence of an activator and with B(C6F5)3 as activator >80%. Low amounts of n-hexenes were formed with trimerization of ethene (see entries 2−4, Table 1). In the case of BF3·OEt2 as a cocatalyst, the selectivities for ethene dimerization were found to be higher compared to that of B(C6F5)3 as an activating

Table 1. Selected Ethylene Oligomerization Reactions with Complex 1a product (%)c entryb

co-activator (equiv)

T (°C)

C4

C6

C4:1-C4H8/ 2-trans/2-cis

TON (mol product/mol Ni)

TOF h−1

1 2 3 4 5 6 7d

B(C6F5)3 (5) B(C6F5)3 (5) BF3·OEt2 (20) BF3·OEt2 (20) B(C6F5)3 (5)

25 50 25 50 25 50 50

100 89 80 85 100 87 86

n. d. 11 20 15 n. d. 13 14

80/14/6 69/19/12 65/15/20 36/21/43 90/7/3 62/17/21 41/21/38

5 83 1014 1064 29 141 1342

15 252 3072 3195 88 427 2684

Entries 1−6 were carried out in a Parr autoclave reactor in 30 mL of toluene at an ethene pressure of 12.5 bar, reaction time, 20 min. b[Ni] = 1.3 × 10−4 M. cDetermined by combined gas chromatography (GC). dFor entry 7, reaction time was 30 min. a

7340

DOI: 10.1021/acscatal.5b02003 ACS Catal. 2015, 5, 7338−7342

Letter

ACS Catalysis

Scheme 4. Mechanistic Sketch of the Ethene and Terminal Olefin Activation and Isomerization by the Methallyl Nickel Complex 1

As mentioned previously, we sought to have a closer look into the catalytic reaction course of the ethene dimerization with 1 by electrospray ionization tandem mass spectrometry studies to trace the reactive nickel species responsible for ethene dimerization and the terminal olefin isomerization processes. Studies of ethene dimerization and 1-octene isomerization with 1 were performed by in situ catalysis investigations in the presence of a small excess of B(C6F5)3 as cocatalyst, using a microreactor coupled directly to the ESI source of an ion-trap mass spectrometer (Amazon, Bruker), focusing on the direct detection and mass spectrometric characterization of the cationic catalytically active species that are supposed to accumulate in the catalytic cycle before the rate-determining step. An initial test was performed exploring the reaction of ethene with the methallyl nickel complex 1 in the presence of a small excess of B(C6F5)3 as cocatalyst and trying to intercept the active species in toluene as solvent, which revealed activity for ethene dimerization with predominant formation of 1-butene. At room temperature, a yellow solution was obtained that was then analyzed approximately 0.7 s after mixing the reagents by ESI(+)-MS. In that solution, the nickel ion 1a of m/z 368 was observed (see Scheme 4 and Figure 2a). Besides the cation 1a, the bisligated nickel complex Ni(L1)2 of m/z 567 was observed as depicted in Figure 2a. Then, 1 to 6 s after mixing the reagents, the nickel ion 1b (m/z 314) was detected, a product witnessing the separation of the methallyl ligand from the nickel cation 1a via ethene insertion into the σ-methallyl moiety followed by a β-hydride elimination step (Figure 2b). The isotopic pattern of all ions matched the calculated ones for the suggested species, in particular for those containing the Nispecies (see insert in Figure 2b). Moreover, a species of m/z 520 revealed losses of Na[C6F5] and Na[OC6F5] in the ESIMS/MS experiments that produced [1b + O]+ of m/z 330 and [1b]+ of m/z 314, respectively. This species may be one of the products formed during catalyst decomposition when ethylene oligomerization occurs. Then, 6−10 s after mixing of the reagents, the nickel ion 1a disappeared and the intensity of the nickel cation 1b increased (Figure 2b); thus, it appears very probable that the active species responsible for ethylene dimerization is cation 1b (see Figure 2b, 1b, with the matching theoretical isotopic pattern see below Figure 2b). It is noteworthy to mention that by MS we

Figure 2. In situ ESI-MS monitoring of (a, b) ethylene dimerization and (c) 1-octene isomerization by 1/B(C6F5)3 as the catalytic system: (a) 0.7 s; (b) 1 to 6 s; (c) 20 s after mixing of the reagents.

have not observed any products of ethene insertion into the cation 1b, presumably due to a very short lifetime of the followup species. 7341

DOI: 10.1021/acscatal.5b02003 ACS Catal. 2015, 5, 7338−7342

Letter

ACS Catalysis

Oligomerization/Polymerization of Ethylene. In Studies in Surface Science and Catalysis; Keii, T., Soga, K., Eds.; Elsevier: Amsterdam, 1986; Vol. 25, pp 201−213. (g) Keim, W. Angew. Chem., Int. Ed. Engl. 1990, 29, 235−244. (2) (a) Heinicke, J.; Kohler, M.; Peulecke, N.; Kindermann, M. K.; Keim, W.; Kockerling, M. Organometallics 2005, 24, 344−352. (b) Chen, M.; Zou, W.; Cai, Z.; Chen, C. Polym. Chem. 2015, 6, 2669−2676. (c) Shim, C. B.; Kim, Y. H.; Lee, B. Y.; Dong, Y.; Yun, H. Organometallics 2003, 22, 4272−4280. (d) Liu, W.; Malinoski, J. M.; Brookhart, M. Organometallics 2002, 21, 2836−2838. (e) Lee, B.; Kim, Y.; Shin, H.; Lee, C. Organometallics 2002, 21, 3481−3484. (f) Bonnet, M. C.; Dahan, F.; Ecke, A.; Keim, W.; Schulz, R. P.; Tkatchenko, I. J. Chem. Soc., Chem. Commun. 1994, 055, 615. (g) Lee, B. Y.; Bu, X.; Bazan, G. C. Organometallics 2001, 20, 5425−5431. (h) Trofymchuk, O. S.; Gutsulyak, D. V.; Quintero, C.; Parvez, M.; Daniliuc, C. G.; Piers, W. E.; Rojas, R. S. Organometallics 2013, 32, 7323−7333. (3) Azoulay, J. D.; Koretz, Z.; Wu, G.; Bazan, G. C. Angew. Chem. 2010, 122, 8062−8066. (4) Azoulay, J. D.; Rojas, R. S.; Serrano, A. V.; Ohtaki, H.; Galland, G. B.; Wu, G.; Bazan, G. C. Angew. Chem., Int. Ed. 2009, 48, 1089−1092. (5) Chen, P. Angew. Chem., Int. Ed. 2003, 42, 2832−2847. (6) Santos, L. S.; Metzger, J. O. Rapid Commun. Mass Spectrom. 2008, 22, 898−904. (7) Santos, L. S.; Metzger, J. O. Angew. Chem., Int. Ed. 2006, 45, 977−981. (8) (a) Zhang, J.; Gao, H.; Ke, Z.; Bao, F.; Zhu, F.; Wu, Q. J. Mol. Catal. A: Chem. 2005, 231, 27−34. (b) Cramer, R.; Lindsey, R. V., Jr. J. Am. Chem. Soc. 1966, 88, 3534−3544. (c) D’Aniello, M. J., Jr.; Barefield, E. K. J. Am. Chem. Soc. 1978, 100, 1474−1481. (9) Breuil, P.-A. R.; Magna, L.; Olivier-Bourbigou, H. Catal. Lett. 2015, 145, 173−192. (10) Wilke, G.; Bogdanovic, B.; Hardt, P.; Heimbach, P.; Keim, W.; Kroner, M.; Oberkirch, W.; Tanaka, K.; Steinrucke, E.; Walter, D.; Zimmermann, H. Angew. Chem., Int. Ed. Engl. 1966, 5, 151−266. (11) Ryu, C.; Song, E.; Shim, J.; You, H.; Choi, K.; Choi, I.; Lee, E.; Chae, M. Bioorg. Med. Chem. Lett. 2003, 13, 17−20. (12) (a) Rojas, R. S.; Galland, G. B.; Wu, G.; Bazan, G. C. Organometallics 2007, 26, 5339−5345. (b) Chen, Z.; Mesgar, M.; White, P. S.; Daugulis, O.; Brookhart, M. ACS Catal. 2015, 5, 631− 636. (c) Younkin, T. R.; Connor, E. F.; Henderson, J. I.; Friedrich, S. K.; Grubbs, R. H.; Bansleben, D. A. Science 2000, 80 (287), 460−462. (13) Zhang, Y.; Cao, Y.; Leng, X.; Chen, C.; Huang, Z. Organometallics 2014, 33, 3738−3745. (14) The reaction was carried out in a Parr autoclave reactor (100 mL), loaded inside a glovebox with an appropriate amount of the catalyst ([Ni] = 1.3 × 10−4 M) and the cocatalysts (20 equiv of BF3· Et2O) and benzene-d6 such that the final volume of the solution was 15 mL. The reaction mixture (ethylene pressure 12.5 bar) was stirred at 50 °C for 20 min, then the reaction mixture was cooled to 0 °C and the ethene was vented. The reaction mixture was analyzed by NMR. (15) 1H NMR spectra and GC data can be found in Supporting Information. (16) Leatherman, M. D.; Svejda, S. A.; Johnson, L. K.; Brookhart, M. J. Am. Chem. Soc. 2003, 125, 3068−308.

Finally, as 1/B(C6F5)3 showed activity in the reaction with 1octene in condensed phase, we performed an experiment trying to intercept the active nickel species in the gas-phase. Thus, approximately 20 s after mixing of the reagents, a yellow solution was obtained that was analyzed by ESI(+)-MS (see Figure 2c). In Figure 2c, complete disappearance of 1a and 1b is witnessed by forming a new cationic species, which was identified and characterized to be 1d, suggesting that 1b is the active species in the catalytic cycle. Species 1d can be addressed to either olefin−hydride species or isomeric agostic n-octyl species perhaps further stabilized by a β-agostic interaction. In conclusion, the mechanism of the ethene dimerization with the highly electrophilic cationic methallyl nickel complex 1 could be made plausible, revealing high selectivity toward formation of butenes. The isomerization of the terminal olefins to internal olefins is a slower process, also exemplified by 1octene isomerization. The mechanism of the dimerization of ethene and the isomerization of terminal olefins was further supported by electrospray ionization tandem mass spectrometry. The cationic nickel hydride 1b is supposed to be the active species driving the catalytic cycle possessing relatively high stability due to the fact that subsequent ethene coordination and insertion are slow and rate determining. During 1-octene isomerization, the formation of the olefin hydride species of type 1d could be identified as a relatively stable species of the interaction of 1b with octene. Octene coordination may be weaker than ethene coordination, but the given observation can only be explained assuming that octene insertion is still slower than ethene insertion into the Ni−H bond.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acscatal.5b02003. Experimental details of the synthesis and characterization of the catalyst and oligomeric products (PDF) Crystallographic data (CIF)



AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors thank prof. Alejandro Duran (CPC) for graphical abstract design. This work was supported by ICM No. 120082 (Nucleus Millenium CPC) and FONDECYT projects Nos. 1130077, 1150307, and 11130086. L.S.S. and F.M.N. thank Anillo ACT 1107, InnovaChile CORFO (Code FCR-CSB 09CEII-6991), and PIEI-UTalca. The authors acknowledge the valuable comments of the referees.



REFERENCES

(1) (a) Canivet, J.; Aguado, S.; Schuurman, Y.; Farrusseng, D. J. Am. Chem. Soc. 2013, 135, 4195−4198. (b) Chen, Y.; Wu, G.; Bazan, G. C. Angew. Chem., Int. Ed. 2005, 44, 1108−1112. (c) Keim, W. Angew. Chem., Int. Ed. 2013, 52, 12492−12496. (d) Keim, W.; Kowaldt, F. H.; Goddard, R.; Krüger, C. Angew. Chem., Int. Ed. Engl. 1978, 17, 466− 467. (e) Keim, W.; Behr, A.; Kraus, G. J. Organomet. Chem. 1983, 251, 377−391. (f) Keim, W. Chelate Complexes of Nickel: Catalysts for the 7342

DOI: 10.1021/acscatal.5b02003 ACS Catal. 2015, 5, 7338−7342