Jean-Paul
Collin
*,
Anne-Chantal
Laemmel
and
Jean-Pierre
Sauvage
*
Laboratoire de Chimie Organo-Minérale, UMR 7513 du CNRS, Uniersité Louis Pasteur, Faculté de Chimie, 4, rue Blaise Pascal, 67070, Strasbourg Cedex, France. E-mail: sauvage@chimie.u-strasbg.fr
First published on 3rd November 2000
A ruthenium complex containing two 1,10-phenanthroline ligands as well as a bipyridine fragment incorporated in a macrocycle (36 atoms) has been synthesized; photoexpulsion of the Ru(phen)2 unit from the macrocycle and its thermal reco-ordination take place efficiently and quantitatively.
We would now like to report another type of photochemically induced movement based on dissociatie photo excit ed states. Ru(bipy)32+ (bipy = 2,2′-bipyridine) and its numerous homologues have mostly been used as photochemical electron or energy transfer agents, in relation to light energy conversion into chemical energy.4 In this respect, the photochemical stability of these compounds represents a real advantage, but a few examples of photochemical dissociation of such complexes have been reported.5 In some cases the “photoexpulsion” process of a given ligand can be relatively efficient6 and it opens the gate to a new way of setting molecular systems into motion. The use of a sterically hindered chelate, such as 6,6′-dimethyl-2,2′-bipyridine (dmbp), turned out to be especially promising since selective and efficient dissociation of this particular chelate was observed under light irradiation.7 In the present system a hindered bipyridine fragment has been incorporated into a macrocycle and the ruthenium(II)-containing unit includes two 1,10-phenanthroline (phen) fragments. Ru(phen)22+ is a stable unit which will not be affected by light irradiation. The principle of the photochemically driven motion and the thermal backward reaction is shown in Scheme 1.
Scheme 1 Principle of the photochemically driven motion and the thermal backward reaction. |
The macrocyclic complex 1 was synthesized from the ruthenium(II) precursor 3 incorporating a 2,2′-bipy ligand substituted at its 6 and 6′ positions by two long chains bearing terminal olefins (see Scheme 2). Compound 2 was prepared in 5 s teps from dmbp8 and its synthesis will be reported elsewhere. The precursor complex 3 was obtained in good yield (75%) from 2 and Ru(phen)2(CH3CN)22+ by heating a stoichiometric mixture of them in ethylene glycol at 140°C for 2 hours. 3 was isolated as its PF6− salt. It is obtained pure as a red powder after column chromatography. The 1H NMR spectrum is in accordance with its structure. The next step consists of cyclizing the two terminal functions of the co-ordinated bipy. Ring-closing metathesis of olefins (RCM), recently proposed by Grubbs and his group,9 has been used in numerous cases, most of the time with remarkable success.10 This catalytic reaction is compatible with a large variety of functions and, in particular, with transition metal complexes such as copper(I) or iron(II) polyimine compounds.11,12 In the present case the RCM reaction turned out to be equally efficient since a 67% yield of the desired complex 1 was obtained, without utilizing high dilution conditions (0.01 M dichloromethane solution of 3, room temperature, 5 days, 7% of ruthenium catalyst [Ru]). The reaction can easily be monitored by 1H NMR since the characteristic signals of the terminal olefins constitute a convenient probe. Gradually, these two sets of signals (δ 5.18 and 5.04 in acetone-d6) are replaced by those corresponding to the cyclic olefin of 1 (5.31 and 5.39), obtained as a mixture of cis and trans isomers. The mass spectrum (FAB-MS) of 1 confirms the expected structure (M − PF6, m/z 1195.3).
Scheme 2 The precursors 2 and 3 of the ruthenium complex 1. |
The photochemical behaviour of complex 1 has been examined by both UV-Visible and 1H NMR spectroscopy in acetonitrile. Under light irradiation (λ>300 nm), it undergoes an efficient and quantitative photolabilization of the macrocycle m36 as demonstrated in Fig. 1 by the different visible spectra recorded during the progress of the reaction. After 6 min of irradiation there is no change in the electronic spectrum and curve 10 is characteristic of the Ru(phen)2(CH3CN)22+ species.13 Also, two well resolved isosbestic points at 342 and 406 nm demonstrate the selectivity of the process. By analysing the absorbance variations s. time, an apparent rate constant can be deduced (t1/2 = 75 s). The efficiency of this photochemical reaction is less than that of a previously studied system7 in which a Ru(phen)2 core is associated to dmbp itself (t1/2 = 28 s under the same conditions). Since the steric hindrance is more prononced in 1 than in Ru(phen)2dmbp2+, it seems that some deco-ordination steps of the chelate unit are probably slower due to the presence of the macrocyclic chain. The cyclic nature of the complex is also expected to favour a cis conformation for the bipy unit, making its stepwise decomplexation slightly more difficult than with its unconstrained analogue, dmbp.
Fig. 1 Electronic spectra of acetonitrile solution of complex 1 after visible light irradiation: 1(0); 2(6); 3(12); 4(30); 5(40); 6(60); 7(90); 8(130); 9(192); 10(370 s). |
Similar results could be obtained using 1H NMR spectroscopy with direct light irradiation of the NMR tube. The characteristic spectrum of complex 1 is gradually replaced by the sum of the spectra of Ru(phen)2(CD3CN)22+ and free macrocyclic m36 .
The thermal back reaction was also proved to be quantitative. An equimolecular mixture of Ru(phen)2(CH3CN)22+ and m36 leads back to the starting complex 1 in quantitative yield (refluxing ethylene glycol; two hours).
These first results are promising for the construction of photomechanical devices in which the macrocycle m36 and the Ru(phen)2 unit will be respectively the ring and the core of the string of a rotaxane.
1: 1H NMR (acetone-d6 , 200 MHz) δ 8.94 (dd, 2 H, H4, J = 8.36 and 1.24) 8.76 (dd, 2 H2, J = 5.41 and 1.23), 8.68 (dd, 2 H, H7, J = 8.37 and 1.23), 8.62 (d, 2 H, H3,3′ , J = 7.38), 8.47 (d, 2 H, H5, J = 8.86), 8.38 (d, 2 H, H6, J = 8.86), 8.09 (dd, 2 H, H4,4′ , J = 8.12 and 7.14), 8.09 (dd, 2 H, H3), 7.92 (dd, 2 H, H9, J = 5.29 and 1.11), 7.67 (dd, 2 H, H8, J = 5.42 and 8.12), 7.43 (d, 2 H, H5,5′, J = 7.88 Hz), 5.39 (m, 2 H, Hcis, 75%), 5.31 (m, 2 H, Htra ns, 25%), 3.90–1.50 (m, 40 H, Hg, Ha,b,c,d,e,f, Hα,β,γ); FAB-MS: m/z = 1195.3, [M − PF6]+, calc. 1195.35, 8%; 1049.3, [M − 2 PF6 + e−]+, calc. 1050.38, 11%; 462.0 [M − 2 PF6-m36 + e−]+, calc. 462.04, 32%. UV-Vis (CH3CN) λ 450 nm (ε 9200 L mol−1 cm−1).
2: 1H NMR (CDCl3, 200 MHz) δ 8.24 (d, 2 H, H3,3′, J = 7.86), 7.68 (t, 2 H, H4,4′, J = 7.75), 7.14 (dd, 2 H, H5,5′, J = 7.62 and 0.98), 5.91 (ddt, 2 H, H′, J = 17.22, 10.32 and 5.66), 5.26 (dtd, 2 H, Ht, J = 17.10, 3.44 and 1.60), 5.16 (m, 2 Hc, J = 10.32, 1.72 and 1), 4.01 (dt, 4 H, Hg, J = 5.66 and 1.36), 3.67–3.50 (m, 28 H, Ha,b,c,d,e,f, 4 Hγ), 2.92 (t, 4 H, Hα, J = 7.50), 2.11 (tt, 4 Hβ, J = 6.52 and 6.88 Hz); FAB-MS: m/z = 617.5, [M + H]+, calc. 617.38, 100%.
3: 1H NMR (acetone-d6 , 200 MHz) δ 8.94 (dd, 2 H, H4, J = 8.15 and 1.10), 8.74 (dd, 2 H, H2, J = 5.17 and 1.22), 8.67 (dd, 2 H, H7, J = 8.35 and 1.22), 8.63 (d, 2 H, H3,3, J = 8.86), 8.47 (d, 2 H, H5, J = 8.86), 8.38 (d, 2 H, H6, J = 8.86), 8.10 (dd, 2 H, H4,4′, J = 7.86 and 7.88), 8.10 (dd, 2 H, H3, J = 5.23 and 8.24), 7.97 (dd, 2 H, H9, J = 5.40 and 1.23), 7.62 (dd, 2 H, H8, J = 5.28 and 8.24), 7.44 (dd, 2 H, H5,5′ , J = 7.87 and 0.99), 5.82 ( ddt, 2 H, H′, J = 17.34, 10.44 and 5.23), 5.18 (m, 2 H, Ht), 5.04 (m, 2 H, Hc), 3.88 (dt, 4 H, Hg, J = 5.42 and 1.54 Hz), 3.60–3.20 (m, 24 H, Ha,b,c,d,e,f), 3.20–1.00 (12 H, Hα,β,γ).
Photoirradiation was performed in a quartz UV cell (P = 1.0 cm) or in a NMR tube (diameter = 5.0 mm) by the use of a Hanimex slide projector (150 W halogen lamp).
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2001 |