Reversal of enantioselectivity by change of solvent with clay-immobilized bis(oxazoline)–copper catalysts

Reversal of enantioselectivity by change of solvent with clay-immobilized bis(oxazoline)–copper catalysts

Catalysis Communications 2 (2001) 165±170 www.elsevier.com/locate/catcom Reversal of enantioselectivity by change of solvent with clayimmobilized bi...

116KB Sizes 0 Downloads 33 Views

Catalysis Communications 2 (2001) 165±170

www.elsevier.com/locate/catcom

Reversal of enantioselectivity by change of solvent with clayimmobilized bis(oxazoline)±copper catalysts A.I. Fern andez, J.M. Fraile, J.I. Garcõa, C.I. Herrerõas, J.A. Mayoral *, L. Salvatella Departamento de Quõmica Org anica, Instituto de Ciencia de Materiales de Arag on, Universidad de Zaragoza C.S.I.C., Pedro Cerbuna 12, E-50009 Zaragoza, Spain Received 30 April 2001; received in revised form 6 June 2001

Abstract The reduction in the polarity of the solvent induces the approach of the bis(oxazoline)±copper complex to the clay sheet in the immobilized catalyst. This proximity strongly in¯uences the stereochemical course of the cyclopropanation reaction, even leading to a reversal in the enantioselectivity. Ó 2001 Elsevier Science B.V. All rights reserved. Keywords: Asymmetric catalysis; Clays; Immobilization; Solvent e€ects; Supported catalysts

1. Introduction The development of enantioselective heterogeneous catalysts is an area of growing interest due to the inherent advantages of heterogeneous over homogeneous catalysts [1±3]. Many chiral heterogeneous catalysts consist of immobilized chiral metal complexes, prepared in most cases by means of the covalent union of the ligand to the support. The method used to achieve this covalent link has a decisive in¯uence on the performance of the resulting catalyst. In particular, noticeable e€ects on the enantioselectivity have been described [4±7]. These changes are, in general, related to the modi®cation of the conformational preferences of the reactant±catalyst intermediates. In this regard,

* Corresponding author. Tel.: +34-976-762077; fax: +34-976762077. E-mail address: [email protected] (J.A. Mayoral).

immobilization without a covalent bond [8±18], which avoids the chemical modi®cation of the ligand, is believed to reduce the in¯uence of the support. Bis(oxazoline)±copper complexes, which are useful catalysts for cyclopropanation reactions [19±24], have a cationic nature and they can be immobilized by electrostatic interactions with anionic supports, such as clays or na®on±silica nanocomposites [25±27]. One of the main limitations of the cyclopropanation reaction is the ease of dimerization of the carbene intermediate, a process that reduces the yield of cyclopropanes. One method to improve the yield is to use an excess of styrene, even to such an extent that styrene becomes the reaction medium itself. In this respect we have found an unexpected e€ect of the solvent on the magnitude and direction of the enantioselectivity in the case of clay-based catalysts. Although other cases of reversal of enantioselectivity have been described in homogeneous catalysis, for example in hydrogenation reactions [28], our

1566-7367/01/$ - see front matter Ó 2001 Elsevier Science B.V. All rights reserved. PII: S 1 5 6 6 - 7 3 6 7 ( 0 1 ) 0 0 0 2 7 - 9

166

A.I. Fernandez et al. / Catalysis Communications 2 (2001) 165±170

Scheme 1. Cyclopropanation reaction of styrene with diazoacetates.

observations are the ®rst, to our best knowledge, in either homogeneous or heterogeneous cyclopropanation reactions, showing the important role of the solid support. 2. Results and discussion The catalyst was obtained by cation exchange of the sodium form of a synthetic laponite with the complex formed between the bis(oxazoline) 1 and

Cu…OTf†2 [27]. The resulting catalyst was tested in the benchmark cyclopropanation of styrene (3) with ethyl diazoacetate (2a) (Scheme 1). With the homogeneous bis(oxazoline)±Cu catalyst (Table 1, entries 1±6) the only important factor in terms of the enantioselectivity is the coordinating character of the anion. The use of a large excess of styrene only has an e€ect on the yield of the reaction. As far as catalytic activity and selectivity are concerned, laponite displays a behavior interme-

Table 1 Results obtained from the reaction of styrene (3) with ethyl diazoacetate (2a) catalyzed by the (1R)±Cu complexesa Entry

Anion

Solvent

Trans/cisb

ee (%)c Trans

Cis

1 2 3 4 5 6 7 8 9 10 11 12

TfO TfO Cl Cl AcO AcO Laponite Laponite Laponite Laponite Laponite Laponite

CH2 Cl2 Styrene CH2 Cl2 Styrene CH2 Cl2 Styrene CH2 Cl2 Styrene CH2 Cl2 d Toluene n-C6 H14 e n-C7 F16 e

71:29 69:31 62:38 61:39 73:27 67:33 61:39 31:69 60:40 40:60 31:69 32:68

54 (5) 55(5) 17(5) 11(5) 14(5) 10(5) 49(5) 7(5) 45(5) 3(5) 3(5) 1(4)

42(7) 42(7) 13(7) 9(7) 10(7) 3(7) 24(7) 34(6) 22(7) 21(6) 33(6) 30(6)

a

Yield (%)b 33 41 19 59 13 43 28 40 27 15 10 22

Except when styrene was the solvent, the reactions were carried out with equimolecular amounts of reagents and 1% of catalyst. Determined by GC. Total conversion of ethyl diazoacetate. c Determined by GC with a cyclodex-B column. Major product in brackets. d Reaction carried out with the catalyst recovered from entry 8. e Reaction carried out under re¯ux. b

A.I. Fernandez et al. / Catalysis Communications 2 (2001) 165±170

diate between tri¯ate and acetate in methylene chloride (entry 7). The use of styrene as the solvent increases the yield, but it also has unexpected and dramatic e€ects on the selectivities (entry 8). A 70:30 cis-preference is observed, the enantioselectivity in the trans-cyclopropanes is noticeably reduced and, most surprisingly, the absolute con®guration of the major cis-product is reversed. So, whereas in methylene chloride the …1S; 2R† enantiomer (7) is the major cis-product with 24% ee, the …1R; 2S† enantiomer (6) is preferentially obtained with 34% ee in the corresponding reaction carried out in styrene. The recovery of the catalyst used in the styrene reaction and its reuse in methylene chloride (entry 9) clearly shows that the use of a very large excess of styrene does not produce a permanent change in the catalyst. It may be speculated that the use of a large excess of styrene would favor the complexation of this reagent and hence the reaction would take place through a di€erent mechanism. On the other hand, the results may be related to an as yet unreported solvent e€ect. In fact, the results obtained with other solvents of low polarity (entries 10±12) parallel those achieved in styrene, i.e., a cis-preference, a reduction of the enantioselectivity in the trans-cyclopropanes and a reversal in the direction of the asymmetric induction in the cis-products. Moreover, the changes in enantioselectivity are less marked in toluene, a solvent with higher polarity/polarizability. It can therefore be concluded that these results are due to a solvent e€ect. In order to assess the generality of these e€ects we analyzed the reaction between styrene (3) and the bulkier …1R; 2S; 5R†-menthyl diazoacetate (2b) (Table 2). As expected, the yield is lower in comparison to that obtained with the more reactive

167

ethyl diazoacetate (2a) and the increase in size of the ester group improves the trans/cis selectivity. An important match±mismatch e€ect exists between bis(oxazoline) and menthol, a situation in contrast with related catalysts [19,29]. The best asymmetric inductions are obtained with the (S)chiral ligand (entry 1) and, consequently, the claysupported catalyst was prepared with this ligand. With the (S)-bis(oxazoline)±Cu immobilized catalyst, the reaction carried out in methylene chloride (entry 2) led to almost identical results to those obtained in the homogeneous reaction. Indeed, these are the highest asymmetric inductions described for this cyclopropanation reaction when promoted by a heterogeneous catalyst. The reaction carried out in styrene (entry 3) shows a slight cis-preference despite the large size of the ester group, which increases the steric interaction with the phenyl group of the styrene. This result con®rms the strong steric e€ect of the surface. The asymmetric inductions are reduced and the absolute con®guration of the major cis-cyclopropane changes once again. The large e€ect of the clay in this case is remarkable in that there is a change in the de from 85% to 21% of opposite sense. It is not easy to o€er an explanation for this unexpected behavior. Changes in the direction of the enantioselection have been described in Diels± Alder reactions promoted by bis(oxazoline)±Mg catalysts, and these changes have been explained as being due to changes in the geometry of the dienophile±catalyst intermediate [30]. In the cyclopropanation reaction reported in this paper the reactive species is a bis(oxazoline)±Cu(I)±carbene intermediate with a trigonal structure and the attack of styrene on this intermediate determines both the trans/cis and the enantioselectivity [29]. In

Table 2 Results obtained from the reaction of styrene (3) with …1R; 2S; 5R†-menthyl diazoacetate (2b) catalyzed by the (1S)±Cu complexesa Entry

Anion

Solvent

Trans/cisb

de (%)b; c Trans

Cis

1 2 3

TfO Laponite Laponite

CH2 Cl2 CH2 Cl2 Styrene

79:21 77:23 47:53

77(5) 73(5) 24(5)

90(7) 85(7) 21(8)

a

Yield (%)b 20 16 26

Except when styrene was used as a solvent, the reactions were carried out with equimolecular amounts of reagents and 2% of catalyst. Determined by GC. Total conversion of diazoacetate. c Major product in brackets. b

168

A.I. Fernandez et al. / Catalysis Communications 2 (2001) 165±170

the homogeneous phase the nature of the anion has an in¯uence on the enantioselectivity [20,31], probably due to changes in the geometry of the intermediate. The results obtained in methylene chloride indicate that the oxygen atoms of the laponite sheets [32] behave as intermediate between tri¯ate and acetate. The absence of a solvent e€ect with both anions in the homogeneous phase indicates that the reaction taking place in the reduced space of a clay is responsible for this particular behavior. It seems logical that a decrease in the dielectric constant of the solvent will increase the electrostatic attraction between the cationic intermediate and the anionic clay sheet. When the reaction is carried out in methylene chloride, the intermediate is not too close to the clay surface and the results are not too dissimilar from those obtained under homogeneous conditions. In less polar solvents the interaction between the intermediate and the clay increases, the complex is situated closer to the surface and, as a consequence, the sheet has an

important in¯uence on the stereochemical course of the reaction. When the intermediate is close to the clay sheet, the orientation in which the ester group points away from the sheet will be favored (Fig. 1). In this intermediate, the formation of the cis-product is disfavored by the interaction between the ester group of the carbene and the phenyl group of the incoming styrene, as occurs in solution, whereas the trans-product will be disfavored by the steric interaction between the clay and the phenyl group of the styrene [32]. Taking into account the change in selectivity, the e€ect of the clay is very important and is in the order of 1 kcal/mol. The in¯uence on the asymmetric induction is much more dicult to explain. It is clear that the proximity to the clay sheet makes the C2 symmetry of the catalytic complexes disappear and, as a consequence, the relative energies of the di€erent transition states and the stereochemical course of the reaction are modi®ed. It may be considered that the sheet acts as a large substituent that disfavors the approach of styrene to the face that is

Fig. 1. Steric interactions in the possible approaches of styrene to the copper-carbene intermediate in the clay-based catalysts.

A.I. Fernandez et al. / Catalysis Communications 2 (2001) 165±170

favored in solution. The e€ect of the clay can be estimated to be between 0.75 and 1.75 kcal/mol depending on the diazoacetate used. It is important to note that this e€ect is not permanent and that the catalyst can be reused. This means that the ``same sample'' of catalyst can be used to preferentially obtain one or other isomer in two successive reactions by simply changing the solvent. Finally, we studied the immobilized complex of the bis(oxazoline) with tert-butyl group (8) instead of phenyl groups. The results show the same tendency, with a decrease in trans/cis and enantioselectivities but without reversal. The e€ect is again non-permanent and the recovered and freshly prepared catalysts lead to the same results. The use of a more bulky substituent probably increases the distance between the clay and the Cu centers, thus reducing the in¯uence of the clay sheet. 3. Experimental 3.1. Preparation of the immobilized catalysts The complex for cationic exchange was prepared by mixing Cu…OTf†2 (1 eq) with a solution of bis(oxazoline) (1 eq) in methylene chloride. The solution was ®ltered through a syringe micro®lter and the solvent was evaporated under reduced pressure. The residue was re-dissolved in methanol and laponite was added to this solution. The rest of the method for the preparation and characterization of the clay catalysts has been described elsewhere [25±27]. 3.2. Cyclopropanation reactions with ethyl diazoacetate Ethyl diazoacetate (320 mg, 2.8 mmol) was slowly added with a syringe pump to a solution of styrene (2.8 mmol, except when styrene was used as a solvent) and n-decane (internal standard, 150 mg) in the corresponding solvent (15 ml) containing the copper catalyst (0.028 mmol, 100 mg of 1-laponite or 280 mg of 8-laponite) at room temperature (unless otherwise is indicated in the tables). The reaction was monitored by gas chromatography with DB-1 and cyclodex-B col-

169

umns [27]. After total consumption of the diazoacetate the solid catalyst was ®ltered o€, washed with methylene chloride and air-dried. 3.3. Cyclopropanation reactions with (1R,2S,5R)menthyl diazoacetate The method was analogous to that used for ethyl diazoacetate except that …1R; 2S; 5R†-menthyl diazoacetate was added in one portion. The reaction was monitored by gas chromatography with a DB-1 column (cross-linked methyl silicone, 25 m  0:2 mm  0:33 lm), oven temperature program 100°C (0 min), 4°C/min±200°C (30 min), retention times (min): styrene 2.18, n-decane 2.83, …1R; 2S; 5R†-menthyl diazoacetate 13.23, …1S; 2R†cyclopropane 30.33, …1R; 2S†-cyclopropane 30.59, …1R; 2R†-cyclopropane 32.30, …1S; 2S†-cyclopropane 32.97. 4. Conclusions In summary, we have shown that a non-covalently bonded support does not play a passive role in the enantioselective reactions promoted by heterogeneous catalysts. The right choice of solvent can force the complex to approach the surface, including it in the steric environment of the active center responsible for the asymmetric induction. The right combination of reagent, catalyst, support and conditions allows the direction of the diastereo- and enantio-selectivities to be changed without permanent modi®cation of the catalyst. Taking into account the reuse of the catalyst, it is possible to obtain one or other product with the same sample of catalyst in successive reactions. The lack of parallelism shown between homogeneous and heterogeneous reactions, caused by the use of a restricted space, can lead to selectivities that are dicult to obtain under classical conditions. These results highlight the value of carrying out mechanistic studies, such as those related to solvent e€ects, in heterogeneous reactions. Such studies could help in the design of ligands to speci®cally to prepare more ecient immobilized catalysts. Although the results described in this paper do not have direct practical

170

A.I. Fernandez et al. / Catalysis Communications 2 (2001) 165±170

use, they open the way to the future application of this promising methodology.

Acknowledgements Financial support for this work was provided by the Spanish C.I.C.Y.T. (Project MAT99-1176). References [1] H.-U. Blaser, B. Pugin in: G. Jannes, V. Dubois (Eds.), Chiral Reactions in Heterogeneous Catalysis, Plenum Press, New York, 1995, pp. 33±57. [2] L. Pu, Tetrahedron: Asymmetry 9 (1998) 1457. [3] L. Canali, D.C. Sherrington, Chem. Soc. Rev. 28 (1999) 85. [4] D.W.L. Sung, P. Hodge, P.W. Startford, J. Chem. Soc., Perkin Trans. 1 (1999) 1463. [5] B. Altava, M.I. Burguete, E. Garcõa-Verdugo, S.V. Luis, O. Poza, R.V. Salvador, Eur. J. Org. Chem. (1999) 2263. [6] B. Altava, M.I. Burguete, J.M. Fraile, J.I. Garcõa, S.V. Luis, J.A. Mayoral, M.J. Vicent, Angew. Chem. Int. Ed. Engl. 39 (2000) 1503. [7] M.I. Burguete, J.M. Fraile, J.I. Garcõa, E. Garcõa-Verdugo, S.V. Luis, J.A. Mayoral, Org. Lett. 2 (2000) 3905. [8] M.J. Sabater, A. Corma, A. Domenech, V. Fornes, H. Garcõa, Chem. Commun. (1997) 1285. [9] S.B. Ogunwumi, T. Bein, Chem. Commun. (1997) 901. [10] I. Frunza, H. Kosslick, H. Landmosser, E. H oft, R. Frische, J. Mol. Catal. A 123 (1997) 179. [11] I.F.J. Vankelecom, V. Van de Viver, P.A. Jacobs, Angew. Chem. Int. Ed. Engl. 35 (1996) 1346. [12] K.B.M. Jansen, I. Laquiere, W. Dehaen, R.F. Parton, I.F.J. Vankelecom, P.A. Jacobs, Tetrahedron: Asymmetry 8 (1997) 3481. [13] K.T. Wan, M.E. Davies, Nature 370 (1994) 449.

[14] T. Sento, S. Shimazu, N. Ichikumi, T. Uematsu, J. Mol. Catal. A 137 (1999) 263. [15] C. Langham, P. Piaggio, D. Bethell, D.F. Lee, P. McMorn, P.C. Bulman Page, D.J. Willock, C. Sly, F.E. Hancock, F. King, G.J. Hutchings, Chem. Commun. (1998) 1601. [16] C. Langham, S. Taylor, D. Bethell, P. McMorn, P.C. Bulman Page, D.J. Willock, C. Sly, F.E. Hancock, F. King, G.J. Hutchings, J. Chem. Soc., Perkin Trans. 2 (1999) 1043. [17] J.M. Fraile, J.I. Garcõa, J. Massam, J.A. Mayoral, J. Mol. Catal. A 136 (1998) 47. [18] J.M. Fraile, J.I. Garcõa, C.I. Herrerõas, J.A. Mayoral, M.A. Harmer, J. Mol. Catal. A 165 (2001) 211. [19] R.E. Lowenthal, A. Abiko, S. Masamune, Tetrahedron Lett. 31 (1990) 6005. [20] D.A. Evans, K.A. Woerpel, M.M. Hinman, M.M. Faul, J. Am. Chem. Soc. 113 (1991) 726. [21] A.V. Bedekar, E.B. Koroleva, P.G. Anderson, J. Org. Chem. 62 (1997) 2518. [22] A.V. Bedekar, P.G. Andersson, Tetrahedron Lett. 37 (1996) 4073. [23] N. End, A. Pfaltz, Chem. Commun. (1998) 589. [24] R. Boulch, A. Scheurer, P. Mosset, R.W. Saalfrank, Tetrahedron Lett. 41 (2000) 1023. [25] J.M. Fraile, J.I. Garcõa, J.A. Mayoral, T. Tarnai, Tetrahedron: Asymmetry 8 (1997) 2089. [26] J.M. Fraile, J.I. Garcõa, J.A. Mayoral, T. Tarnai, Tetrahedron: Asymmetry 9 (1998) 3997. [27] J.M. Fraile, J.I. Garcõa, J.A. Mayoral, T. Tarnai, M.A. Harmer, J. Catal. 188 (1999) 214. [28] U. Berens, R. Selke, Tetrahedron: Asymmetry 7 (1996) 2055. [29] H. Fritschi, U. Leutenegger, A. Pfaltz, Helv. Chim. Acta 71 (1988) 1553. [30] S. Crosignani, G. Desimoni, G. Faita, P.P. Righetti, Tetrahedron 54 (1998) 1572, and references cited therein. [31] J.M. Fraile, J.I. Garcõa, J.A. Mayoral, T. Tarnai, J. Mol. Catal. A 144 (1999) 85. [32] P.J. Alonso, J.M. Fraile, J. Garcõa, J.I. Garcõa, J.I. Martõnez, J.A. Mayoral, M.C. Sanchez, Langmuir 16 (2000) 5607.