A latent ruthenium based olefin metathesis catalyst with a sterically demanding NHC ligand

Abstract

An olefin metathesis catalyst featuring a SIPr NHC and an ester chelating carbene ligand is introduced. In contrast to its previously published SIMes analogue, only the trans dichloro configurated isomer was obtained. The two counterparts are tested in various olefin metathesis reactions, revealing a striking superiority of the new complex in the cross metathesis of olefins with methyl vinyl ketone allowing for full conversion with only 500 ppm catalyst loading.

---

Olefin metathesis has been virtually revolutionized since the development of potent, well defined ruthenium based initiators. Upon sophisticated ligand tuning, further advancement has been sought ever since, in order to enhance features like activity, selectivity or stability of the complexes.¹ The introduction of sterically demanding NHC ligand SIPr (SIPr = 1,3-bis(2,6-diisopropylphenyl)-4,5-dihydroimidazol-2-ylidene)—or its unsaturated analogue IPr—instead of the conventionally used SIMes (SIMes = 1,3-bis(2,4,6-trimethylphenyl)-4,5-dihydroimidazol-2-ylidene) has been studied regarding various metathesis reactions, especially in the context of indenylidene complexes,² but also for initiators featuring the Hoveyda-concept of a chelating benzylidene carbene.³ Generally, an increased initiation rate due to facilitated phosphine dissociation was stated for complexes with the bulkier NHC ligands, going along with higher conversions in e.g. cross metathesis reactions of α-olefins with acrylates or RCM of easy substrates.^2,4 Also, a study on E/Z selectivity in cross metathesis regarding various NHC ligands has been published, yet revealing no obvious electronically or sterically motivated trends.⁵ However, a decisive influence of the NHC ligand on the pre-catalyst, and also on the active species has been reported, sometimes for the worse, sometimes for the better.

Only recently the synthesis of complex 1, a ruthenium based olefin metathesis catalyst featuring a SIMes NHC ligand and a chelating ester benzylidene ligand, has been presented (cf. Fig. 1).⁶ Now, we disclose its SIPr analogue 2 that exhibits considerable deviations in all respects. The synthesis exclusively furnishes the trans dichloro configurated isomer, while in the case of SIMes three different complexes with the cis dichloro compound 1 as the main product were obtained.⁶ Also, the performance in various metathesis reactions is different, including as a highlight the cross metathesis of 1-hexene with methyl vinyl ketone at ultra-low catalyst loading.

Fig. 1 N-Heterocyclic carbenes and complexes employed.

For the synthesis of 2 the parent complex M32⁷ was dissolved in freshly degassed dichloromethane and the carbene precursor 3 (1.3 eq.) was added (cf. Scheme 1). The mixture was stirred for 48 h at room temperature, whilst the colour turned from red to green. The solvent was reduced to 3 mL and precipitation with n-pentane yielded a light green powder. Several washings with n-pentane using ultrasonic wave completely removed the residual ligand and other impurities. The final complex was isolated in 72% yield. The ¹H-NMR spectrum revealed a “Ru=C–H resonance” at 18.49 ppm, a singlet at 4.12 accounting for the ethylene protons from the NHC core, as well as a single peak accounting for all 8 methyl groups in the NHC ligand superimposing a singlet for the isopropyl ester ligand's two methyl groups (6H) at about 1.25 ppm. In the ¹³C-NMR spectrum, the corresponding carbon resonances appear at 299.3, 54.6, 24.2 and 21.7 ppm respectively. The high symmetry resembles the corresponding spectra of Hov,⁸ but is in strong contrast to complex 1, which in the ¹H-NMR spectrum yields a multiplet for the ethylene group, a single signal for each methyl group and even a diastereotopic splitting of the iso-propyl group.⁶2 is perfectly stable under ambient conditions and no evidence for isomerisation to its cis-dichloro isomer could be retrieved.

Scheme 1 Preparation of 2.

Crystals of 2 were obtained by dissolving a small amount of the complex in dichloromethane and overlaying it with n-pentane in a loosely closed glass vial. XRD revealed 1 eq. of DCM in the crystal structure. Complex 2 exhibits a trans dichloride configuration, the ligands being arranged in an almost perfect square-pyramidal alignment (cf. Fig. 2). Apart from the chloride configuration, important bond lengths and angles are very similar to the values obtained for the SIMes analogue 1,⁶ only the imidazolium fragment of the NHC ligand is distinctly more distorted in the SIPr complex (23.2° compared to 6.8° in 1). On the other hand, the buried volumes V_bur of the two different NHC ligands in complexes 1 and 2 that give information about the accessibility to the chelating carbene ligand for an approaching olefin⁹ are almost equal (2: 33.6%, 1: 32.8%). In general, the structural features of the new complex are in good accordance with a SIPr bearing Hoveyda type derivative.^2d

Fig. 2 ORTEP plot of complex 2. Selected bond lengths (Å): Ru(1)–Cl(1) = 2.3401(5), Ru(1)–Cl(2) = 2.3461(5), Ru(1)–C(1) = 1.998(2), Ru(1)–C(28) = 1.820(2), Ru(1)–O(1) = 2.132(1); selected bond angles (°): Cl(1)–Ru(1)–Cl(2) = 159.05(2), Cl(1)–Ru(1)–C(1) = 88.53(5), Cl(1)–Ru(1)–C(28) = 98.47(6), Cl(1)–Ru(1)–O(1) = 84.73(4), Cl(2)–Ru(1)–O(1) = 85.51(4), C(1)–Ru(1)–C(28) = 99.56(8), C(28)–Ru(1)–O(1) = 88.76(7).

DFT calculations were performed to theoretically substantiate the preference of complexes 1 and 2 regarding the formation of either cis or trans dichloro species. This is a challenging issue, because the 1-cis and 2-trans isomers should be slightly favored. Thus, we tested a series of different methods (see Table 1). Inspection of the data reported in Table 1 indicates that all the approaches considered correctly reproduce that 1 is more biased towards the cis isomer than 2, i.e. the trans–cis energy difference for 1 is greater than that for 2, see the almost consistently positive values in the 2–1 column. This indicates that basically all the methods we tested correctly reproduce the experimental fact that the SIPr ligand pushes the chelating alkylidene-ester ligand trans to the NHC ligand. In line with previous results, that indicated that reduced steric bulkiness normally favors the cis dichloride configuration,¹⁰ the tendency of the SIPr ligand to favor the trans isomer can be reasonably ascribed to the steric pressure of the bulkier orthoⁱPr groups of the SIPr ligand on the ester moiety.¹⁰

Table 1 Energy difference, in kcal mol⁻¹, between the trans and cis isomers of 1 and 2. Positive values indicate that the cis isomer is more stable than the trans isomer. E and G indicate potential energy and free energy, respectively

Functional	E trans–cis			G trans–cis
	1	2	2–1	1	2	2–1
a Only the electrostatic component of the solvation free energy is included. b Non-electrostatic contributions (cavitation and dispersion) are added to the solvation free energy.
BP86^a	0.4	−2.6	3.0	−2.0	−4.6	2.6
BP86-D3^a	3.2	1.2	2.0	0.8	−0.8	1.6
M06^a	3.4	1.8	1.6	1.0	−0.2	1.2
BP86^b	−0.8	−2.6	1.8	−3.3	−4.6	1.3
BP86-D3^b	2.0	1.2	0.8	−0.5	0.8	−1.3
M06^b	2.1	0.7	1.4	−0.3	−1.2	0.9

---

Much less consistency, instead, is found when the absolute stability of the cis and trans isomers is concerned. Experiments dictate that E_trans–cis for 1 should be slightly positive, because 1-cis is the major isomer but 1-trans is observed, whereas E_trans–cis for 2 should be somewhat more positive since only 2-trans is observed. Disappointingly, E_trans–cis for 1 > 0, E_trans–cis for 2 < 0, and |E_trans–cis| for 2 > |E_trans–cis| for 1 is found only in very few cases. In terms of internal energy, this condition is only met by the BP86 functional if non-electrostatic solvation terms are not considered (row 1 in Table 1). All other approaches predict that the cis isomer is favored also for 2, with the exception of the BP86 functional when non-electrostatic solvation terms are included in the solvation energy. In this case the cis isomer is favored for both 1 and 2. Moving to free energy, the BP86-D3 is the only approach that gives a qualitative agreement with the experiments, since it correctly predicts that 1-cis and 2-trans are favored. The overall conclusion is that the prediction of such small energy differences remains a challenge.

In order to investigate the activity of the complexes 1 and 2, they were employed in various standard benchmark olefin metathesis reactions. Ring opening metathesis polymerization was performed with the norbornene derivative endo,exo-dimethylbicyclo[2.2.1]hept-5-ene-2,3-dicarboxylate (4). A monomer to initiator ratio of 300 and a concentration of 0.1 M with respect to the monomer were maintained in all cases (cf. ESI† for details). Polymerizations in dichloromethane at room temperature initiated by 1 or 2 remained incomplete, even after prolonged reaction time (>48 h). At 40 °C, both 1 and 2 were still less active than reference initiator M2 at room temperature, but completed the polymerization within 5 hours. In contrast, at 80 °C (in toluene), both complexes under investigation were promptly activated and finished polymerizations within 10 min (see details in ESI†). Sub-summarizing these results, both complexes 1 and 2 exhibit latent behaviour and their polymerization activity is similar. This similarity in activity despite the different chloride configuration is somehow surprising, as the SIPr complex does not need to isomerize before initiation, following the idea of trans dichloro species being the actual active species in olefin metathesis.^6,11

However, latent catalysts have proved to be usable in ring closing metathesis of challenging substrates. Increased stability in solution at high temperature due to the strongly chelating carbene ligand (strong compared to e.g. the iso-propoxy group in Hov) providing low initiation rates will slowly release the active species during the catalysis and thus increase conversion to the ring-closed product.¹² Thus complex 2 was employed in the RCM of 4-methyl-N,N-bis(2-methylallyl)benzenesulfonamide (5) using a 0.1 M solution of 5 in toluene and heating in the presence of 1 mol% 2 at 110 °C. Complex 1 was tested under the same reaction conditions for comparison. Reproducibility turned out to be a major issue for both complexes, and a series of experiments resulted in averaged conversions of 74 ± 9% for the SIPr complex 2 and 65 ± 15% for SIMes counterpart 1 (cf. ESI† for details). The somewhat higher conversion obtained with 2 go in line with the observation of the higher stability of 2 under these conditions. Runs with 2 maintained the bright green colour of uninitiated 2 for more than 4 hours associated with ongoing conversion after that time, whereas reaction solutions with 1 had all turned brownish within 1 hour and no further conversion was detected after that time. With the herein obtained results for RCM of 5, complexes 1 and 2 cannot compete with results presented by Plenio et al. for the same substrate using catalysts featuring two NHC ligands (full conversion at 0.5 mol% catalyst loading and 80 °C),¹³ but clearly outperform a comparable complex featuring the unsaturated IMes-NHC ligand and the same chelating ester carbene (44% yield at 5 mol% catalyst loading).¹⁴

Most interestingly, the complexes were employed for the cross metathesis of 1-hexene with methyl vinyl ketone (MVK) to yield oct-3-en-2-one as a target molecule (see Scheme 2). This reaction is of particular interest as α,β-unsaturated ketones such as the herein produced oct-3-en-2-one represent many-sided starting materials for the synthesis of e.g. chiral amines and other relevant substance classes.¹⁵ The employment of MVK in ruthenium catalyzed cross metathesis with α-olefins has been rather ineffective until now due to high catalyst loadings of typically 2–5 mol%.^15,16 Based on recent work of Lipshutz et al.,¹⁶ who succeeded in accelerating the cross metathesis by using Cu(I) as a co-catalyst and diethyl ether as the solvent, reaction conditions were selected as follows: to a solution of 3 equiv. of MVK and 1 equiv. of 1-hexene in diethyl ether 3 mol% of CuI (in respect to 1-hexene) and the catalyst 2 (0.5 mol% in respect to 1-hexene) was added and the reaction mixture was heated at 35 °C. After 4 h 1-hexene was completely consumed and was exclusively transformed to the E-isomer A. No evidence for the Z-isomer B could be retrieved. As a by-product, 9% (in respect to A) of the self-metathesis product of MVK C was formed. Intrigued by the success of 2 in this reaction, the performance of catalysts 1, M51, Hov and M2 was evaluated under the same conditions as benchmark catalyst 2 (cf. Table 2). Complex 1 is insoluble in diethyl ether and accordingly hardly any conversion towards A occurred. M51 and Hov gave conversion of less than 50%, while M2 performed similar to 2. Then, the catalyst loading of M2 and 2 was steadily lowered in order to determine the minimal amount necessary for full conversion. Newly introduced 2 clearly outperforms all other complexes under investigation, as a loading of 0.05 mol% (cf. Table 2, entry 8) is sufficient for a conversion of >98% in 16 h. Surprisingly, the performance of reference complex Hov, which is the preferred catalyst for the cross metathesis of acrylates with terminal olefins,¹⁷ lies far behind that of M2 that gave 96% conversion towards A in 16 h at a loading of 0.25 mol%. In all cases where full conversion of 1-hexene was obtained, traces of the self-metathesis product of MVK C were observed. Only in trials resulting in incomplete conversion the self-metathesis product of 1-hexene D was observed in small amounts.

Scheme 2 Products and possible by-products of the cross metathesis of 1-hexene with methyl vinyl ketone.

Table 2 Cross metathesis of 1-hexene with MKV^a

Entry	Catalyst	Loading [mol%]	Conversion^b [%]
a 1 equiv. 1-hexene, 3 equiv. MVK, 3 mol% CuI, Et2O, diethyl ether, 35 °C, 16 h. b Conversion towards A, yield of A in the case of entry 8: 95%.
1	2	0.5	>99
2	1	0.5	<2
2	M51	0.5	45
3	Hov	0.5	48
4	M2	0.5	>99
5	M2	0.25	96
6	2	0.25	>99
7	2	0.1	>99
8	2	0.05	>98
9	2	0.025	73

---

In summary, we have presented a new ruthenium catalyst for olefin metathesis featuring an ester chelating carbene ligand and a SIPr NHC ligand. Employment of the SIPr ligand results in the trans dichloro configuration of the complex, which is, according to DFT calculations, the thermodynamically favoured isomer. This finding is in contrast to the preference of the cis dichloro isomer in the analogous SIMes complex and suggests that generally SIPr favours the trans dichloro stereochemistry. Both the SIPr and the SIMes derivative show a similar latent behaviour in ROMP and similar activity in RCM at elevated temperatures. The strong point of the newly introduced catalyst is the cross metathesis of terminal olefins with methyl vinyl ketone. The usage of complex 2 in combination with CuI renders the synthesis of α,β-unsaturated ketones clearly more efficiently since a loading as low as 0.05 mol% (the previous state of the art is 2–5 mol% of catalyst) gives virtually complete and stereoselective conversion of 1-hexene to E-oct-3-en-2-one.

Financial support of this work from the Higher Education Commission of Pakistan (for M.A.) and the European Community (CP-FP 211468-2 EUMET) is gratefully acknowledged. A.P. thanks the Spanish MINECO for a Ramón y Cajal contract (ref. RYC-2009-05226) and the European Commission for a Career Integration Grant (CIG09-GA-2011-293900). The provision of complex M32 by Prof. S. P. Nolan and team (Univ. St. Andrews) is highly appreciated.

Notes and references

1. (a) C. Samojowicz, M. Bieniek and K. Grela, Chem. Rev., 2009, 109, 370; (b) G. C. Vougioukalakis and R. H. Grubbs, Chem. Rev., 2010, 110, 1746.
2. (a) H. Clavier, C. A. Urbina-Blanco and S. P. Nolan, Organometallics, 2009, 28, 2848; (b) H. Clavier, J. Broggi and S. P. Nolan, Eur. J. Org. Chem., 2010, 937; (c) C. A. Urbina Blanco, A. Leitgeb, C. Slugovc, X. Bantreil, H. Clavier, A. M. Z. Slawin and S. P. Nolan, Chem.–Eur. J., 2011, 17, 5045; (d) A. Fuerstner, L. Ackermann, B. Gabor, R. Goddard, C. W. Lehmann, R. Mynott, F. Stelzer and O. R. Thiel, Chem.–Eur. J., 2001, 7, 3236.
3. H. Clavier, F. Caijo, E. Borré, D. Rix, F. Boeda, S. P. Nolan and M. Mauduit, Eur. J. Org. Chem., 2009, 4254.
4. F. Boeda, X. Bantreil, H. Clavier and S. P. Nolan, Adv. Synth. Catal., 2008, 350, 2959.
5. C. Lujan and S. P. Nolan, J. Organomet. Chem., 2011, 696, 3935.
6. M. Zirngast, E. Pump, A. Leitgeb, J. H. Albering and C. Slugovc, Chem. Commun., 2011, 47, 2261.
7. C. A. Urbina-Blanco, S. Manzini, J. Pérez Gomes, A. Doppiu and S. P. Nolan, Chem. Commun., 2011, 47, 5022.
8. J. Wappel, C. A. Urbina-Blanco, M. Abbas, J. H. Albering, R. Saf, S. P. Nolan and C. Slugovc, Beilstein J. Org. Chem., 2010, 6, 1091.
9. (a) A. Poater, B. Cosenza, A. Correa, S. Giudice, F. Ragone, V. Scarano and L. Cavallo, Eur. J. Inorg. Chem., 2009, 1759; (b) A. Poater, F. Ragone, S. Giudice, C. Costabile, R. Dorta, S. P. Nolan and L. Cavallo, Organometallics, 2008, 27, 2679.
10. (a) A. Correa and L. Cavallo, J. Am. Chem. Soc., 2006, 128, 13352; (b) F. Ragone, A. Poater and L. Cavallo, J. Am. Chem. Soc., 2010, 132, 4249.
11. (a) Y. Vidavsky, A. Anaby and N. G. Lemcoff, Dalton Trans., 2012, 41, 32; (b) A. Leitgeb, K. Mereiter and C. Slugovc, Monatsh. Chem., 2012, 143, 901.
12. A. Szadkowska, X. Gstrein, D. Burtscher, K. Jarzembska, K. Woźniak, C. Slugovc and K. Grela, Organometallics, 2010, 29, 117.
13. (a) V. Sashuk, L. H. Peeck and H. Plenio, Chem.–Eur. J., 2010, 16, 3983; (b) H. Plenio and S. Wolf, J. Organomet. Chem., 2010, 695, 2418.
14. A. Fürstner, O. R. Thiel and C. W. Lehmann, Organometallics, 2002, 21, 331.
15. (a) T.-L. Choi, C. W. Lee, A. K. Chatterjee and R. H. Grubbs, J. Am. Chem. Soc., 2001, 123, 10417; (b) P. Poulhes, R. Sylvain, P. Perfetti, G. Gil and S. Gastaldi, Synthesis, 2010, 1334; (c) T. J. Donohoe, J. F. Bower and J. A. Bassuto, Nat. Protoc., 2010, 5, 2005; (d) T. J. Donohoe and J. F. Bower, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 3373.
16. (a) K. Voigtritter, S. Ghorai and H. Lipschutz, J. Org. Chem., 2011, 76, 4697; (b) B. H. Lipshutz, S. Ghorai, W. W. Y. Leong and B. R. Taft, J. Org. Chem., 2011, 76, 5061.
17. (a) M. Abbas and C. Slugovc, Tetrahedron Lett., 2011, 52, 2560; (b) M. Abbas and C. Slugovc, Monatsh. Chem., 2012, 143, 669; (c) X. Miao, R. Malacea, C. Fischmeister, C. Bruneau and P. H. Dixneuf, Green Chem., 2011, 13, 2911; (d) A. Rybak and M. A. R. Meier, Green Chem., 2008, 10, 1099; (e) A. Rybak and M. A. R. Meier, Green Chem., 2007, 9, 1356; (f) S. Gessler, S. Randl and S. Blechert, Tetrahedron Lett., 2000, 41, 9973; (g) M. D. Schulz and K. B. Wagener, ACS Macro Lett., 2012, 1, 449; (h) M. Abbas, J. Wappel and C. Slugovc, Macromol. Symp., 2012, 311, 122.

---

Footnote

† Electronic supplementary information (ESI) available: Preparation and characterisation of 2, procedures and results of catalytic testing, computational details and cartesian coordinates for all DFT-optimized structures. CCDC 866162 (2). For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c2cy20311a

---