Vous êtes sur la page 1sur 8

Inorganica Chimica Acta 357 (2004) 30713078 www.elsevier.

com/locate/ica

Heterogenised Rh(I), Ir(I) metal complexes with chiral triaza donor ligands: a cooperative eect between support and complex
Camino Gonzlez-Arellano a
a b

a,c

, Avelino Corma b, Marta Iglesias

a,*

, Flix Snchez e a

c,*

Instituto de Ciencia de Materiales de Madrid, CSIC, Cantoblanco, 28049 Madrid, Spain Instituto de Tecnolog Qu a mica, UPV-CSIC, Avda de los Naranjos s/n 46022, Valencia, Spain c Instituto de Qu mica Orgnica, CSIC, Juan de la Cierva, 3. 28006 Madrid, Spain a Received 15 March 2004; accepted 10 April 2004

Available online

Abstract Rhodium and iridium complexes of the chiral triaza ligands, {N,N0 -bis{[(2S)-(1-benzylpyrrolidinyl)]methyl}amine (2), N,N0 bis{[(2S)-(1-benzylpyrrolidinyl)]methyl}-N-propylamine (3), N,N0 -bis{[(2S)-(1-benzylpyrrolidinyl)]methyl}-N-[3-(triethoxysilyl)pr-opyl]amine (4)}, are described. All ligands form one to one [ML] species with the above metal ions. The structures of these complexes were elucidated by analytical and spectroscopic data (elemental analysis, mass spectroscopy, IR, 1 H and 13 C NMR). The xation of the preformed triethoxysilyl-rhodium and iridium complexes, on mesoporous solids (MCM-41, SBA-15), and their use, under heterogeneous conditions, for the hydrogenation reactions are reported. The catalytic activity and selectivity of heterogenised complexes are higher to that observed under homogeneous conditions, as a consequence of the complex- and/or reagents-to-support interaction. The stable covalent bond between support and complex allows the recovery and recycling of the heterogenised catalysts for a number of cycles, moreover atomic absorption analysis of the reaction solutions shows that there is not any metal leaching into the solutions. 2004 Elsevier B.V. All rights reserved.
Keywords: Rhodium; Iridium; Hydrogenation; Immobilisation; Mesoporous

1. Introduction In 1994 Togni and Venanzi [1] reported very promising results with nitrogen donor ligands in asymmetric catalysis. Six years later, Fache et al. [2] published a precise review on the current state of the investigations in this area and recently, one revision [3] shows the progress accumulated since then in catalytic transformations using tetraaza ligands. Nitrogen-containing ligands have several distinct advantages. First, they are largely available in enantiomerically pure form, both in the chiral pool (quinine, cinchonine, and sparteine) or as cheap industrial chemical intermediates. On the other hand, chirality on the nitrogen atom is dicult to obtain. Contrary to the phosphines, the chiral nitrogen atoms epimerise instantaneously at room temperature;

Corresponding authors. Tel.: +34913349000; fax: +34913720623. E-mail address: marta.iglesias@icmm.csic.es (M. Iglesias).

the formation of a stable chiral centre on a nitrogen atom is, however, possible by using bicyclic structures. The second advantage of the nitrogen-containing ligands lies in the chemistry of the nitrogen functional group itself. The chemistry of these is not always easy, but it has received such abundant attention that there exist, in most cases, numerous synthetic solutions to each possible transformation of these compounds. As a result, these synthetic possibilities allow tailor-made modications for the preparation of ligands with specic physicochemical properties. In particular, the interactions with the transition metals may be widely varied by preparing X-type ligands (amides, sulfonamides), L-type ligands (amines) or p-type ligands (imines). Nitrogen-containing ligands are being used more and more in asymmetric catalysis. They turn out to be suitable for any type of catalysis and especially for heterogeneous catalysis [46], which is one of their main advantages over phosphines. In addition, nitrogencontaining ligands may be used in asymmetric catalysis

0020-1693/$ - see front matter 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.ica.2004.04.024

3072

C. Gonzlez-Arellano et al. / Inorganica Chimica Acta 357 (2004) 30713078 a

3' 2' 7'

4' 5'

N
6' 6

R N
7 2 3

N
5
4

2: R = H, 3: R = (CH2)2CH3, 4: R = (CH2)3Si(OEt)3,

Fig. 1. Chiral triaza ligands.

with transition metals, which are less expensive than noble metals [7]. Starting from the readily available L -proline, our group [812] reported the preparation of the tetraaza C2 -symmetric ligands and their metal-complexes (Rh(I), Ir(I), Cu(I), Mn(II)). These complexes were tested on the asymmetric hydrogenation, cyclopropanation and oxidation reactions, showing that ligands gave only poor enantioselectivities (<20%) but excellent chemical yields (>85%). These catalysts have been heterogenised and the strategy used to preserve as much as possible the coordination sphere of the metal [13]. The heterogenised complexes are still more stable than their homogeneous counterpart and they can be used several times without decreasing activity. Nevertheless, the enantioselectivity achieved with the heterogeneous systems remains close to that obtained with the homogeneous ones. Herein, we present the synthesis and characterisation of a new family of chiral triaza ligands and their Rh(I) and Ir(I) complexes and investigate their co-ordination chemistry and catalytic properties in detail, and how ecient and selective heterogeneous zeolite systems involving this type of ligands could be designed (see Fig. 1).

spectrometer (ESI-MS, APCI-MS) with positive mode. IR spectra were recorded on a Bruker IFS 66v/S spectrophotometer (range 4000200 cm1 ) in KBr pellets; 1 H, 13 C NMR spectra were taken on a Varian XR300 and a Bruker 200 spectrometers. 1 H NMR chemical shifts are given in ppm using tetramethylsilane as an internal standard. High resolution 13 C MAS or CP/ MAS-NMR spectraof powdered samples, in some cases also with a Toss sequence, in order to eliminate the spinning side bands, were recorded at 100.63 MHz, 6 ls 90 pulse width, 2 ms contact time and 510 recycle delay, using a Bruker MSL 400 spectrometer equipped with an FT unit. The spinning frequency at the magic angle (54440 ) was 4 KHz.Optical rotation values were measured at the sodium-D line (589 nm) with a Perkin Elmer 241 MC polarimeter. Gas chromatography analysis was performed using a Hewlett-Packard 5890 II with a ame ionisation detector in a cross-linked methylsilicone column [18]. The inorganic support for anchoring was a purely siliceous MCM-41 [24] and SBA-15 [19]. 2.1.1. Synthesis of ligands Ligand N,N-bis{[(S)-1-benzylpyrrolidin-2-yl]methyl} amine (2) was prepared according to modied published methods using ethyl chloroformate/trimethyl amine as activating reagent [20]. 2.1.2. Synthesis of N,N-bis{[(S)-1-benzylpyrrolidin-2yl]methyl}propan-1-amine (3) To a solution of amine (2) (0.405 g, 1 mmol) in acetonitrile (25 ml), K2 CO3 (1.42 g, 10 mmol) was added and the mixture stirred for 10 min, then propyl iodide (0.195 ml, 2 mmol) was added dropwise. The mixture was reuxed for 8 h and puried by ash chromatography. (ethyl acetate:ethanol 20:1). The product was isolated as yellow oil (65%); a25 142 (c: 1.49, D CHCl3 ); 1 H NMR (CDCl3 ): d(ppm) 7.407.18 (10H, m, Harom ); 4.10 (2H, d, AB, J 16 Hz); 3.30 (2H, d, AB, J 16.0 Hz); 3.022.90 (2H, m, H5a;5a0 ); 2.672.50 (4H, m, H2;20 , H7a;7a0 ); 2.462.24 (4H, m, H7b;7b0 , H8 ); 2.23 2.12 (2H, m, H5b;5b0 ); 2.071.89 (2H, m, H3a;3a0 ); 1.82 1.56 (6H, m, H3b;3b0 , H4;40 ); 1.44 (2H, sex, H9 , J 8.9 Hz); 0.84 (2H, t, H10 , J 8Hz). 13 C NMR (CDCl3 ): d(ppm) 139.00 (Carom ); 129.02 (Carom ); 128.18 (Carom ); 126.85 (Carom ); 62.50 (C2 ); 60.25 (C7 ); 59.65 (CH2 Ph); 57.65 (C8 ); 54.74 (C5 ); 30.38 (C3 ); 22.49 (C4 ); 20.24 (C9 ); 11.88 (C10 ). EM, m=z (%): 406. 2.1.3. Synthesis of N,N-bis{[(S)-1-benzylpyrrolidin-2yl]methyl}-3-(triethoxysilyl)propan-1-amine (4) To a solution of 2 (0.405 g, 1 mmol) in DMF (25 ml), K2 CO3 (1.42 g, 10 mmol) and NBu4 HSO3 (0.320 g, 0.1 mmol) were added. The solution was stirred for 10 min and iodopropyltriethoxysilane (0.332 g, 1 mmol)) was added dropwise. The mixture was reuxed for 8 h and

2. Experimental 2.1. General Preparation of organometallic complexes was carried out under dinitrogen by conventional schlenk-tube techniques. The starting complexes [Rh(cod)Cl]2 [1416] and [Ir(cod)Cl]2 [17] were prepared using reported methods. All solvents were carefully degassed before use. The silylating agent iodopropyltriethoxysilane was obtained by halogen change starting from the commercial chloropropyltriethoxysilane. C, H and N analysis was carried out by the analytical department of the Institute of Materials Science (CSIC). Metal contents were analysed by atomic absorption using a PerkinElmer AAnalyst 300 atomic absorption and plasma ICP PerkinElmer 40 spectrometers. Mass spectra were performed on a Hewlett-Packard 1100 MSD mass

C. Gonzlez-Arellano et al. / Inorganica Chimica Acta 357 (2004) 30713078 a

3073

puried by ash chromatography (ethyl acetate:ethanol 1:1). The product was isolated as yellow oil (69%). a25 1:94 (c: 0.77, CHCl3 ). IR (lm, cm1 ): m 1079 D (SiO). 1 H NMR (CDCl3 ): d (ppm) 7.407.18 (10H, m, Harom ); 4.08 (2H, d, AB, J 10 Hz); 3.80 (q, J 27 Hz, OCH 2 CH3 ); 3.22 (2H, d, AB, J 10 Hz); 2.982.85 (2H, m, H5a;5a0 ); 2.652.47 (4H, m, H2a;2a0 , H7a;7a0 ); 2.46 2.28 (4H, m, H7b;7b0 , H8 ); 2.212.18 (2H, m, H5b;5b0 ); 2.031.90 (2H, m, H3a;3a0 ); 1.761.59 (6H, m, H3b;3b0 , H4;40 ); 1.581.47 (2H, m, H9 ); 0.770.43 (3H, m, H10 ). 13 C NMR (CDCl3 ): (ppm) 140.00 (Carom ); 128.96 (Carom ); 128.15 (Carom ); 126.77 (Carom ); 62.36 (C2 ); 60.30 (C7 ); 59.71 (CH2 Ph); 58.65 (OCH2 ); 58.30 (C8 ); 54.76 (C5 ); 30.42 (C3 ); 22.49 (C4 ); 20.36 (C9 ); 18.30 (OCH2 CH3 ); 7.97 (C10 ). EM, m=z (%): 567 (M ,15). 2.2. Preparation of rhodium and iridium complexes (5 10) 2.2.1. Typical procedure Only the preparation of [Rh(2)COD]PF6 CH2 Cl2 (5) is described in detail. AgPF6 (116 mg, 0.46 mmol) in THF (40 ml) was added to [Rh(cod)Cl]2 (114 mg, 0.23 mmol) in THF (10 ml) and the mixture was stirred vigorously at room temperature for 30 min. Precipitated silver chloride was ltered o and the yellow solution was treated with the ligand (166 mg, 0.46 mmol) in THF. The mixture was stirred for 12 h under reux. The solvent was evaporated under reduced pressure to 2 ml. Careful addition of diethyl ether caused the precipitation of a solid which was collected by ltration,washed with diethyl ether and dried under vacuum (103 mmHg) to give the light brown cationic complex. Yield: 67%. m.p.: 100102 C. a25 0:56 (c: 0.60,CHCl3 ). D Found: C, 48.8; H, 5.0; N, 4.9; Rh, 12.5%. Calc. for C33 H47 N3 Cl2 F6 PRh (804): C, 49.1; H, 5.8; N, 5.2; Rh, 12.7. IR (KBr,cm1 ): m 837 (PF); 474 (RhN); 320 (NRhN). 1 H NMR (CDCl3 , ppm): d 8:057.90 (2H, m, Harom ); 7.567.25 (8H, m, Harom ); 5.705.45 (4H, m, cod); 4.003.40 (8H, m, CH2 Ph, H5a;5a0 , H2;20 ); 3.002.70 (4H, m, H7 ,70 ); 2.692.20 (10H, m, H5b;5b0 ,cod); 2.19 1.23 (8H, m, H4;40 , H3;30 ). 13 C NMR (CDCl3 , ppm): d 131:45 (Carom ); 130.17 (Carom ); 129.10 (Carom ); 128.64 (Carom ); 61.43 (C2 ); 57.10 (CH2 Ph); 54.90 (C5 ); 54.31 (C7 ); 28.43 (C3 ); 26.29 (C4 ). UVVis k, nm): 386. KM (X1 cm2 mol1 ) 125.3. EM, m=z (%): 574 ([Rh(2)][PF6 + CH2 Cl2 ],103 Rh). Complexes 610 were prepared through a procedure similar to that given for 5 starting from [MCl(cod)]2 (M Rh, Ir) and using 0.46 equivalents of the appropriate ligand. 2.2.2. [Rh(3)H2 O]PF6 (6) Brown. Yield: 79%. m.p.: >230 C. a25 0:65 (c: D 0.99, CH3 OH). Found: C, 48.0; H, 6.5; N, 6.0; Rh, 15.8%. Calc. for C27 H39 N3 F6 PRh (653): C, 48.2; H, 6.1;

N, 6.2; Rh, 15.3. IR (KBr, cm1 ): m 837 (PF); 406 (RhN); 318 (NRhN). 1 H NMR (CDCl3 , ppm): d 7:487.45 (5H, m, Harom ); 7.237.16 (5H, m, Harom ); 4.474.43 (2H, m, CH2 Ph); 4.254.19 (2H, m, CH2 Ph); 3.813.71 (4H, m, H2;20 , H5a;5a0 ); 3.553.41 (6H, m, H8 , H7;70 ); 1.901.75 (2H, m, H5b;5b0 ); 1.701.60 (8H, m, H4a;4a0 , H3a;3a0 ); 1.281.15 (2H, m, H9 ); 0.900.70 (3H, m, H10 ). UVVis: kmax 352 nm. KM (X1 cm2 mol1 ) 118.2. m/z: 508 ([Rh(3)][H2 O + PF6 ], 103 Rh). 2.2.3. [Rh(4)H2 O]PF6 (7) Yield: 56%. m.p.: >220 C. Found: C, 47.2; H, 6.3; N, 4.9; Rh, 12.1%. Calc. for C33 H55 N3 F6 O4 PRhSi (653): C, 47.5; H, 6.6; N, 5.0; Rh, 12.3. IR (KBr, cm1 ): m 836 (PF); 475 (RhN). 1 H NMR (CD3 OD, ppm): d 7:507.35 (10H, m, Harom ); 4.404.31 (2H, m, CH2 Ph); 4.124.02 (2H, m, CH2 Ph); 3.523.43 (4H, m, H2;20 , H5a;5a0 ); 3.123.01 (6H, m, H8 , H7;70 ); 2.632.40 (2H, m, H5b;5b0 ); 2.332.24 (2H, m, H3a;3a0 ); 2.101.85 (4H, m, H4;40 ); 1.871.80 (2H, m, H3b;3b0 ); 1.501.40 (2H, m, H9 ); 0.980.85 (2H, m, H10 ). EM, m=z (%): 586 ([Rh(4)][H2 O + PF6 ], 103 Rh, as hydrolyzed species). 2.2.4. [Ir(2)COD] (8) Dark brown. Yield: 79%. m.p.: 138140 C. a25 50 (c: 0.53, CHCl3 ). Found: C, 57.7; H, 6.9; N, 578 6.3; Ir, 29.0%. Calc. for C32 H46 N3 Ir (663): C, 57.6; H, 6.9; N, 5.9; Ir, 10.9. IR (KBr, cm1 ): m 496 (IrN). 1 H NMR (CDCl3 , ppm): d 7:697.26 (10H, m, Harom ); 4.514.43 (2H, m, CH2 Ph); 4.194.08 (2H, m, CH2 Ph); 3.923.73 (2H, m, H2;20 ); 3.723.43 (6H, m, H5a;5a0 , H7;70 ); 2.201.36 (18H, m, H5b;5b0 , cod, H4;40 , H3;30 ). 13 C NMR (CD3 OD, ppm): d 132:66 (Carom ); 131.86 (Carom ); 131.35 (Carom ); 130.60 (Carom ); 67.36 (C2 ); 60.02 (CH2 Ph); 56.45 (C5 ); 50.25 (C7 ); 28.92 (C3 ); 23.73 (C4 ). UVVis k, nm): 433,358. KM (X1 cm2 mol1 ) 10.2. EM, m=z (%): 663 ([Ir(2)COD] ,193 Ir); 661 ([Ir(2)COD] ,191 Ir). 2.2.5. [Ir(3)H2 O]Cl Et2 O (9) Yellow. Yield: 57%. m.p.: 155157 C. a25 1:2 578 (c: 0.69, CH3 OH). Found: C, 51.0; H, 6.8; N, 5.9; Ir, 26.8%. Calc. for C31 H51 N3 ClIrO2 (723.5): C, 51.3; H, 7.0; N, 5.7; Ir, 26.6. IR (KBr, cm1 ): m 482 (IrN). 1 H NMR (CD3 OD, ppm): d 7:607.57 (2H, m, Harom ); 7.467.40 (2H, m, Harom ); 4.47, 43.4 (2H, AB, J 11.1 Hz, d, CH2 Ph); 3.813.68 (2H, m, H2;20 ); 3.533.39 (2H, m, H5a;5a0 ); 3.283.17 (2H, m, H5b;5b0 ); 2.85 2.78 (2H, m, H7a;7a0 ); 2.662.63 (2H, m, H7b;7b0 ); 2.38 2.33 (4H, m, H3a;3a0 ,H8 ); 2.251.99 (4H, m, H4;40 ); 1.971.77 (2H, m, H3b;3b0 ); 1.471.25 (2H, m, H9 ); 0.98 0.86 (3H, m, H10 ). 13 C NMR (CD3 OD, ppm):d 129:69 (Carom ); 128.59 (Carom ); 126.41 (Carom ); 64.60 (C2 ); 57.80 (CH2 Ph); 55.50 (C8 ); 55.30 (C7 ); 54.30 (C5 ); 28.43 (C3 ); 21.29 (C4 ); 18.00 (C9 ); 11.05 (C10 ). UVVis k,

3074

C. Gonzlez-Arellano et al. / Inorganica Chimica Acta 357 (2004) 30713078 a

nm) 433,358. KM (X1 cm2 mol1 ) 129.5. EM, m=z (%): 406 (3). 2.2.6. [Ir(4)H2 O]Cl (10) Yield: 57%. m.p.: >220 C. Found: C, 48.2; H, 6.5; N, 4.9; Ir, 23.1%. Calc. for C33 H55 N3 ClIrO4 Si (812.5): C, 48.7; H, 6.8; N, 5.2; Ir, 23.6. IR (KBr,cm1 ): m 478 (IrN). 1 H NMR (CD3 OD, ppm): d 7:627.38 (10H, m, Harom ); 4.694.51 (2H, m, CH2 Ph); 4.28 (2H, AB, J 14.2 Hz, d, CH2 Ph); 3.803.59 (2H, m, H2;20 ); 3.49 3.23 (8H, m, H5a;5a0 , H7a;7a0 ); 2.251.91 (10H, m, H4;40 , H3;30 ,H8 ); 1.251.14 (2H, m, H9 ); 0.950.79 (2H, m, H10 ). 13 C NMR (CD3 OD, ppm): d 131:90, 131.79, 131.06, 130.35 (Carom ); 71.63 (C2 ); 69.93 (CH2 Ph); 60.90 (C8 ); 59.49 (C7 ); 55.54 (C5 ); 30.58 (C3 ); 27.28 (C4 ); 24.18 (C9 ); 23.11 (C10 ). KM (X1 cm2 mol1 ) 93.2. 2.3. Heterogenisation of M-complexes (Rh-Support, IrSupport) 2.3.1. Anchoring on SBA-15 and MCM-41-zeolite 2.3.1.1. Anchoring on MCM-41-zeolite: Route A. The supported M-complexes (Rh4-MCM-41, Ir4-MCM-41) were prepared as we have previously described [13]. Thus, a solution of metal complex bearing a triethoxysilyl group (0.5 mmol) in ethanol (2 ml) was added to a well-stirred toluene suspension (40 ml) of the inorganic support (1 g) and the mixture was stirred at 80 C for 24 h. The solid was then ltered and Soxhlet-extracted for 724 h to remove the remaining non-supported complex from heterogenised catalyst. The resulting solid was dried in vacuum and analysed. Rh4-(MCM-41): Elemental analysis indicated 0.53 mass % Rh. IR (KBr, cm1 ): m 3410 (OH); 1237 (vs, support); 1120 (CSi); 1080 (vs, support); 802 (PF); 454 (MN). 13 C NMR (solid): d(ppm) 73.7 (C2 ); 61.1 (CH2 CH3 ); 59.040.3 (CH2 CH2 CH2 Si, C5 ); 39.521.0 (CH2 CH2 NH, C3 , C4 , CH3 ); 14.7 (CH2 CH2 Si); 7.8 (CH2 Si).). UVVis (solid) (nm): 210. Ir4-(MCM-41): Elemental analysis indicated 1.2 mass % Ir. IR (KBr, cm1 ): m 3410 (OH); 1237 (vs, support); 1120 (CSi); 1080 (vs, support); 451 (MN). 13 C NMR (solid): d (ppm) 128.9 (Carom ); 70.0 (C2 ); 60.1 (CH2 CH3 ); 59.040.0 (CH2 CH2 CH2 Si, C5 ); 39.521.0 (C3 , C4 , CH3 ); 15.2 (CH2 CH2 Si); 8.1 (CH2 Si).). UVVis (solid) (nm): 209, 260, 310. 2.3.1.2. Anchoring on SBA: Route B. The supported Rh and Ir-complexes (Rh4-(SBA-15), Ir4-(SBA-15)) were prepared as follows. A solution of chloropropyltriethoxy silane (0.5 mmol) in ethanol (2 ml) was added to a wellstirred toluene suspension (40 ml) of the support (1 g) and the mixture was stirred at 80 C for 24 h. The solid was then ltered and Soxhlet-extracted for 724 h to remove the remaining non-supported product from heterogenised

ligand. Then, halogen exchange using KI in acetone, at room temperature for 24 h, leads a white solid that was dried in vacuum and analysed. Alkylation of the amine 2 with this support halide aords the anchored ligand. To a solution of [M(cod)(thf)2 ] (M Rh, Ir) (1 mmol) in dry thf (20 ml) was added a suspension of the heterogenised ligand (1 mmol); the mixture was stirred for 6 h at room temperature and the pale solid ltered, washed with acetonitrile, and dried in vacuum (103 mmHg/8 h) to give the heterogenised complex. Rh4-(SBA-15): Elemental analysis indicated 1.31 mass % Rh. IR (KBr, cm1 ): m 3400 (OH); 1230 (vs, support); 1118 (CSi); 1080 (vs, support); 853 (PF); 458 (M N). 13 C NMR (solid): d (ppm) 129.5 (Carom ); 70.1 (C2 ); 59.1 (CH2 CH3 ); 59.040.0 (C H2 CH2 CH2 Si, C5 ); 4020.0 (C3 , C4 , CH3 ); 13.9 (CH2 CH2 Si); 8.0 (CH2 Si).). UVVis (solid) (nm): 344, 258, 208. Ir4-(SBA-15): Elemental analysis indicated 2.63 mass % Ir. IR (KBr, cm1 ): m 3550 (OH); 1210 (vs, support); 1120 (CSi); 1068 (vs, support); 458 (MN). 13 C NMR (solid) d (ppm) 129.8 (Ph); 67.3 (C2 ); 64.4 (CH2 CH3 ); 57.3 (CH2 CH2 CH2 Si, C5 ); 31.2 (C3 , C4 , CH3 ); 12.8 (CH2 CH2 Si); 7.6 (CH2 Si). UVVis (solid) (nm): 255, 424. 2.4. Catalytic experiments The catalytic properties, in hydrogenation reactions, of the above Rh and Ir complexes were examined under conventional conditions for batch reactions in a reactor (Autoclave Engineers) of 100 ml capacity at 313 K temperature, 4 atm. dihydrogen pressure and a metal:substrate molar ratio of 1/10000. The results were monitored by gas chromatography using an internal standard reference. The kinetics results are shown in Table 1 and Fig. 2.
Table 1 Asymmetric hydrogenation of diethyl 2-benzylidene succinatea

Catalyst 5 6 Rh4-(MCM-41) Rh4-(SBA-15) 8 9 Ir4-(MCM-41) Ir4-(SBA-15)


a

Conv. %b 89.7 94.5 95.9 81 88.8 68.9 72.5 17.8

TOFc;d 13 578 23 384 24 244 13 148 13 510 8334 15 498 1701

All reactions were performed in EtOH at 40 C, S/C 10 000 and 4 atm H2 . b 60 min. c mmol substrate. mmol cat1 h1 . d Optical yield < 10%.

C. Gonzlez-Arellano et al. / Inorganica Chimica Acta 357 (2004) 30713078 a


Rh2 Rh3
Rh4-MCM-41 Rh4-SBA-15 Ir4-MCM-41

3075

100 90 80 70

Ir3

Conversion (%)

60 50 40 30

30 min 45 min 60 min 75 min

Ir4-SBA-15

20 10 0 ---------

Catalyst

Fig. 2. Kinetic prole for the hydrogenation of diethyl 2-benzylidene succinate.

2.4.1. Recovery and recycling of catalysts At the end of the hydrogenation process, the mixture of reaction was ltered; the residue of zeolite-containing catalyst was washed with CH2 Cl2 or acetonitrile to completely remove the remains of products and/or reactants and used again.

bis{[(S)-1-benzylpyrrolidin-2-yl]methyl}amine (2) was obtained by reduction of the respective amide with lithium aluminium hydride at reux. All intermediates and nal products have been obtained with a total yield of 7080%. Amide 1 is a white solid while amine 2 was isolated as colourless oils that are stable at low temperature in an inert atmosphere. Amines 3 and 4 were obtained starting from 2 by alkylation with propyliodide or iodopropyltriethoxysilane. The ligands were characterised unequivocally by mass spectrometry, IR and 1 H, 13 C NMR spectroscopy and gave satisfactory elemental analyses. Mass spectrometry shows the highest ions at m/z 364, 406 and 567 which corresponds to the molecular weights of compounds 2, 3, and 4, respectively. The 1 H and 13 C NMR spectra obtained are in agreement with those expected for these triaza ligands. The ligands have two optically active centres; both have the S absolute conguration. 3.2. Preparation of complexes The rhodium and iridium complexes (510) were obtained by treating the dimeric [M(cod)Cl]2 (M Rh, Ir) with two equivalents of AgPF6 in tetrahydrofuran. An equimolar amount of the corresponding multitopic ligands was added to the solution containing the cationic mononuclear [M(cod)(THF)x ]PF6 species (Scheme 2). The mononuclear complexes were prepared in good yield (>60%) by ligand exchange of L with [M(cod)(THF)2 ]PF6 in THF. The reaction of N,Nbis{[(S)-1-benzylpyrrolidin-2-yl]methyl}amine (2) with [Ir(cod)(THF)2 ]PF6 also provided a neutral M-complex, [Ir(N,N-bis{[(S)-1-benzylpyrrolidin-2-yl]methyl}amine1)(cod)] (8) as the main product of the reaction, which presents the ligand as a deprotonated anionic species; starting from rhodium compound only traces of the neutral species were detected as a black solid. The complexes were isolated by precipitation from diethyl ether as microcrystalline air stable solids. Elemental analysis of C, H, N and metal is consistent with the proposed stoichiometry. Mass spectrometry (ESI-MS and APCI-MS with positive mode) indicated a monomeric formula in accordance with the results of IR and NMR studies. The complexes have their highest ions at m/z values corresponding to the loss of F , and PF in 6 the molecular species. The IR spectra of the free ligands exhibit the characteristic bands of amines. The IR spectra of Rh-amine

3. Results and discussion 3.1. Synthesis of ligands The ligands were prepared and well characterised following the procedure showed in Section 2. All reaction steps were ne-tuned for high yield and selectivity. The preparation of diamide N,N0 -bis[(S)-N-benzylprolyl]methyleneamine (1) was achieved following a modied published method starting from the easily available L -proline protected as the N-benzyl derivative, according to Scheme 1. The corresponding amine N,NO N Bz L-Pro HN
H4LiAl

N Bz HN N Bz
2 I-R [MCl(cod)]2

N Bz N M X N Bz
5, 8

N Bz
1

N Bz R N N Bz
[MCl(cod)]2

N Bz R N M X N Bz

1/2[MCl(cod)]2

+ AgPF6

THF

[M(cod)(THF)x]PF6 + AgCl
THF

[M(cod)(THF)x]PF6
3, R = (CH2)2CH3 4, R = (CH2)3Si(OEt)3 6, 9; R: (CH2)2CH3 7, 10; R: (CH2)3Si(OEt)3

+ Ligand

[M(ligand)(L')]X

M = Rh, Ir, ligand = 2, 3, L' = cod, solvent, X = PF6

Scheme 1. Synthesis of ligands and complexes.

Scheme 2.

3076

C. Gonzlez-Arellano et al. / Inorganica Chimica Acta 357 (2004) 30713078 a

complex 5 show m(NH) band shifted at higher frequencies (3400 cm1 ). The m(PF) frequency appears at 835 cm1 as medium bands characteristic of the noncoordinated PF anion. The 13 C NMR spectra of 6 amine-complexes show only small chemical shift dierences compared with the free ligands. The IR spectrum of the mononuclear Ir-complex 8 does not show the expected m(NH) and m(PF) bands. In the Ir-amine complexes 9, 10, the 13 C NMR signals show only small chemical shift dierences compared with the free ligand and band due to m(PF) was not observed. Electronic absorption spectra of the rhodium and iridium complexes, measured between 800 and 200 nm, show a clear maximum at ca. 300 nm (2 6000 8000) and one shoulder at ca. 400 nm (e 15002000). Taking into account their energy position and intensity, the bands at 400 nm could be assigned [21] to dd transitions localised on the metal ion. At shorter wavelengths, the bands at 300 nm may be assigned to the M ! ligand charge transfer transition and intra-ligand transitions. The conductivity data suggest 1:1 electrolytes for all cationic complexes. 3.3. Heterogenisation of complexes The solids employed in the present study as supports of the active chiral complexes were purely siliceous MCM-41 and SBA-15. The two solids were selected to compare topologies and the complex would be presumably located in the interior of the mesoporous channels. The heterogenisation of homogeneous catalysts is a eld of continuing interest: indeed, although some of these organometallic complexes exhibit remarkable catalytic properties (activities and selectivity) they are dicult to separate, intact, from the reaction medium. Thus, unless the activity of the homogeneous catalysts is exceptionally high, their heterogenisation is still currently economical, but also a toxicological and environmental challenge. We have considered one strategy for heterogenisation, which preserves as much as possible the coordination sphere of the metal. This is achieved by anchoring the homogeneous catalyst to an inorganic support (MCM-41, SBA-15) via covalent bonds between the solid (silanol groups SiOH) and one ligand that have appropriate groups (Si(OEt)3 ) at a position remote to the metal centre (Scheme 3(a)). Preparations of heterogenised materials of complexes bearing a triethoxysilylpropyl group were carried out, by controlled hydrolysis of Si-OEt bonds and reaction with the free silanol (SiOH) on the surface of support. The resulting catalytic material is very stable and the species are covalently bonded to the surface showing only minor frequency shifts from those of the corresponding neat complex, as conrmed by IR and UVVis spectroscopy. The elemental analysis of C, H, N and M

also conrms the M/ligand (1:1) stoichiometry. It is unlikely that the nature of the complex is substantially altered under the relatively mild conditions of the anchoring reaction [22]. Thermogravimetric analysis shows that the total weight loss is associated with the metal complex content corresponding to composition of the organic ligands. The loading of metal was always ca. $12% (0.1%) measured by atomic absorption of metal of the digested samples. These values have been used for calculating the ratio of catalyst/substrate in the reaction tests. IR and electronic spectra (diuse reectance) of heterogenised complexes are coincident with that recorded for homogeneous complexes. The IR spectra of the zeolite-supported catalyst show bands that correspond to the complexes and support. Peaks due to the support dominate the spectra. These include the OH vibration in the range 37003300 cm1 . Some of the bands characteristic of the complexes could be, however, distinguished. Major framework bands appear around 1140, 1040, 960, 785 and 740 cm1 . Vibrational bands due to PF, CN appear around 840 and 450 cm1 , respectively, broadly similar to those of the neat complex. The UVVis electronic spectra data are given in Section 2. The band maximum is not signicantly altered on zeolite heterogenised complex. The positions and relative intensities of these bands are similar to those of the free complexes. Another alternative route was shown in Scheme 2(b) that consists in anchoring of iodopropyltriethoxysilane to the support or by introducing a chloropropyltriethoxysilane moiety in support preparation and halogen exchange with iodide. Alkylation of amine 2 with the halide heterogenised material gives supported amines which react with the starting metal complex and M-support complex was obtained with spectroscopical characteristics similar to those obtained by route A. Also, we have obtained the supported complexes by anchoring the ligand to the support (Scheme 2C) and reaction of this previously heterogenised ligand with the starting complex to give the respective heterogenised complex. These products show analytical and spectroscopical properties similar to that obtained via route A. 3.4. Catalytic experiments 3.4.1. Hydrogenation of alkenes The use of cationic complexes with chiral nitrogencontaining ligands is an attractive approach for the catalytic hydrogenation of alkenes. In the rst stage, metal complexes were evaluated as catalysts in the hydrogenation of diethyl 2-benzylidene succinate to investigate the potential of ligands 2, 3 and 4 for catalysis (Table 1 and Fig. 2). In the presence of 0.1 mol% of catalysts in ethanol at 313 K under a pressure of 4 bar H2 ,the substrate was fully converted to the alkane in 60 min. The results of these studies reveal that cationic

C. Gonzlez-Arellano et al. / Inorganica Chimica Acta 357 (2004) 30713078 a

3077

A
Support

(EtO)3Si
Support

OH OH OH N M N Bz X N Bz

Toluene 60C / 24 h

O O Si O OH

N N

N M X Bz

Bz

Heterogenised complex O O Si O OH O O Si O OH

B
Support Support

OH OH OH Si(OEt)3(CH2)3Cl

Toluene 60C / 24 h

(CH2)3Cl KI

Support

(CH2)3I

N Bz HN Support: MCM-41, SBA-15 O O Si O OH N Bz

Support

Heterogenised complex

[MCl(cod)]2

N N

Bz

Bz

C
Support

OH OH OH

Support

(EtO)3Si
Toluene

O O Si O OH

Bz

N N Bz

N Bz

60C / 24 h

N Bz

[MCl(cod)]2

Heterogenised complex

Scheme 3. Heterogenisation by covalent bond (tethering).

catalysts are very eective for this transformation affording the product with higher rate and comparable enantioselectivity to that achieved with the analogous tetraaza-derivatives [23]. The results show that rhodium complexes are more ecient than iridium complexes. The change of substituents on primary amine nitrogen plays a signicative role on activity that increases when moving from hydrogen to propyl group. A study about the inuence of structural characteristics of ligand and substrates with these promising catalysts in homogeneous media or supported on zeolites is in progress and will be presented extensively in future. When transition metal complexes have been supported on carriers such as polymers or silica, etc., it is generally accepted that a moderate to strong reduction in the reactivity has been observed; however, in our case the turnover numbers for hydrogenation increase indicating the cooperative eect of the support. Soluble complexes could be used only once,because they deteriorate completely by the end of the rst

catalytic run. Supported-complexes could be recovered for recycling and reused retaining most of their catalytic activity. After each cycle, the liquid phase that was separated from the reaction mixture was examined for catalytic activity in the hydrogenation of alkenes under the same conditions used with the solid catalyst. The results show that the liquid phases are inactive for hydrogenation. Moreover, atomic emission analyses did not detect rhodium in the liquid phases from the rst, second and third cycles. Based on the lower detection limit of the instrument, the amount of rhodium leaching into the liquid phases must be less than 0.2% of the Rh on the catalyst. Nature of the support. Other point that is also clear from the results is that the activity for the heterogenised complexes depends on a small but signicant extent on the support. These results are a reection of the dierences in the structure of the solid (topology, surface and silanol nests). A higher activity is obtained when MCM-41 was used as support, when compared with mesoporous SBA-15.

3078

C. Gonzlez-Arellano et al. / Inorganica Chimica Acta 357 (2004) 30713078 a [3] M.H. Fonseca, Knig, Adv. Synth. Catal. 345 (2003) 1173. o [4] F. Fache, B. Dunjic, P. Gmez, M. Lemaire, Top. Catal. 4 (1997) a 201. [5] G. Jannes, V. Dubois (Eds.), Chiral Reactions in Heterogenous Catalysis, Plenum Press, New York, London, 1995. [6] D.E. Vos, I.F.K. Vankelecom, P.A. Jacobs (Eds.), Chiral Catalyst Inmobilization and Recycling, WileyVCH, Weinheim, 2000. [7] M.J. Alcn, E. Gutirrez-Puebla, M. Iglesias, M.A. Monge, F. o e Snchez, Inorg. Chim. Acta 306 (2000) 116. a [8] M.J. Alcn, M. Iglesias, F. Snchez, J. Organomet. Chem. 601 o a (2000) 284. [9] M.J. Alcn, M. Iglesias, F. Snchez, J. Organomet. Chem. 634 o a (2001) 25. [10] M.J. Alcn, M. Iglesias, F. Snchez, Inorg. Chim. Acta 333 (2002) o a 83. [11] M.J. Alcn, A. Corma, M. Iglesias, F. Snchez, J. Organomet. o a Chem. 655 (2002) 134. [12] M.J. Alcn, A. Corma, M. Iglesias, F. Snchez, J. Mol. Catal. A o a 178 (2002) 253. [13] M.J. Alcn, A. Corma, M. Iglesias, F. Snchez, J. Mol. Catal. A o a 194 (2003) 137. [14] J. Chatt, L.M. Venanzi, J. Chem. Soc. (1957) 4715. [15] D. Drew, J.R. Doyle, Inorg. Synth. 13 (1972) 48. [16] G. Giordano, R.H. Crabtree, Inorg. Synth. 28 (1990) 88. [17] L. Herde, J.C. Lambert, C.V. Seno, Inorg. Synth. 15 (1974) 18. [18] E. Miranda, F. Snchez, J. Sanz, M.I. Mart a inez-Castro, J. High, Resol. Chromatogr. 21 (1998) 225. [19] (a) C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, J.S. Beck, Nature 359 (1992) 710; (b) J.S. Beck, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D. Schmitt, C.T.-W. Chu, K.H. Olson, E. Sheppard, S.B. McCullen, J.B. Higgins, J.L. Schlenk, J. Am. Chem. Soc. 114 (1992) 10834. [20] D.Y. Zhao, J.L. Feng, Q.S. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D. Stucky, Science 279 (1998) 548. [21] K. Ishihara, Y. Karumi, S. Kondo, H. Yamamoto, J. Org. Chem. 63 (1998) 5692. [22] R.D. Peacock, B. Steward, Coord. Chem. Rev. 46 (1982) 129. [23] L.L. Murrell, in: J.J. Burton, R.L. Garten (Eds.), Advanced Materials in Catalysis, Academic Press, New York, 1977 (Chapter 8). [24] M.J. Alcn, M. Iglesias, F. Snchez, I. Viani, J. Organomet. o a Chem. 601 (2000) 284.

Nature of the recovered materials. One of the main advantages of the heterogeneous catalysts is the possibility for their recovery and reuse. This possibility was explored thoroughly with Rh4-(MCM-41). The results obtained indicate that after the rst reaction, the recovered catalyst does not show a signicant degree of leaching of rhodium. The spectral data for recovered material are essentially the same before and after the hydrogenation reaction. 4. Conclusion A series of new microcrystalline air-stable rhodium and iridium complexes with chiral triaza ligands have been synthesised and fully characterised even in optically pure form. We have shown that these cationic Rh(I), Ir(I) complexes are ecient catalysts for the hydrogenation of olens. Heterogenisation on MCM-41 and SBA-15 increases the activity of the homogeneous catalysts for dierent substrates and can be recovered and reused, retaining most of their catalytic activity. Moreover, the heterogenised complexes are signicantly more stable than the corresponding homogeneous complexes over prolonged reaction times and were handled without special care in standard conditions. Acknowledgements Financial support by the Direccin General de Ino vestigacin Cient o ica y Tcnica of Spain (Project e MAT2003-07945-C02-02) is gratefully acknowledged. References
[1] A. Togni, L. Venanzi, Angew. Chem., Int. Ed. Engl. 33 (1994) 497. [2] F. Fache, E. Schulz, M. Lorraine Tommasino, M. Lemaire, Chem. Rev. 100 (2000) 2159.

Vous aimerez peut-être aussi