Irreversible but Noncovalent Ru(II)-Pyridine Bond: Its Use for the Formation of [2]-Catenanes

Article abstract of DOI:10.1021/ic0302478

The reaction between ligand 1, which consists of two terminal pyridines attached to a central 1,10-phenanthroline (phen), and the complex Ru(phen) ₂(CH₃CN)₂(PF₆)₂ has been studied. A new ruthenium containing metallamacrocycle has been obtained and fully characterized. Despite the relatively poor yield for the cyclization process involving the ruthenium center (20percent), this strategy led to the synthesis of two different kinds of [2]-catenane. The first example reported in this article is a bimetallic Cu(I)/Ru(II) catenane 5³⁺ consisting of a purely organic ring interlocked with the ruthenium(II)-incorporating metallacycle. Complex 5³⁺ was selectively demetalated at the Cu(I) center to lead to the free Ru(II)-containing catenane. A trimetallic Ru(II)/Cu(I)/Ru(II) catenane 8⁵⁺ was also synthesized showing that this approach is reliable and promising for the elaboration of photoactive multicomponent systems.

Full text of DOI:10.1021/ic0302478

Inorg. Chem. 2004, 43, 1895−1901
Irreversible but Noncovalent Ru(II)−Pyridine Bond: Its Use for the
Formation of [2]-Catenanes
B. X. Colasson and J.-P. Sauvage*
Laboratoire de Chimie Organo-Mine´rale, UMR 7513 au CNRS, Institut Le Bel,
UniVersite´ Louis Pasteur, 4, Rue Blaise Pascal, F-67070 Strasbourg, France
Received August 4, 2003
The reaction between ligand 1, which consists of two terminal pyridines attached to a central 1,10-phenanthroline
(phen), and the complex Ru(phen)2(CH CN)2(PF ) has been studied. A new ruthenium containing metallamacrocycle
3
6 2
has been obtained and fully characterized. Despite the relatively poor yield for the cyclization process involving the
ruthenium center (20%), this strategy led to the synthesis of two different kinds of [2]-catenane. The first example
reported in this article is a bimetallic Cu(I)/Ru(II) catenane 53+ consisting of a purely organic ring interlocked with
the ruthenium(II)-incorporating metallacycle. Complex 53+ was selectively demetalated at the Cu(I) center to lead
to the free Ru(II)-containing catenane. A trimetallic Ru(II)/Cu(I)/Ru(II) catenane 85+ was also synthesized showing
that this approach is reliable and promising for the elaboration of photoactive multicomponent systems.
Introduction
Since the pioneering work of Fujita and co-workers on
the preparation of catenanes containing coordinative bonds
such as a Pd(II)-N bond,12 several other systems have been
reported which incorporate transition metals.13 Magnesium
has even been used as a constitutive atom to generate
catenanes from Grignard reagents and crown ethers.14
The field of catenanes has experienced an impressive
development in the course of the past two decades.1-3 One
particularly active branch is that of molecular machines and
motors.4-8 Interlocking rings incorporating transition metals,
either in their coordination sites9 or as components of their
backbones,10,11 represent a special class of such compounds.
A few years ago, our group has reported that catenanes
can be constructed by coordination chemistry alone using a
three-coordination site fragment consisting of a central phen
chelate attached to two peripheral terpy units (phen ) 1,10-
phenanthroline; terpy ) 2,2′:6′,2′′-terpyridine).15
* To whom correspondence should be addressed. E-mail: sauvage@
chimie.u-strasbg.fr.
(1) For early work, see: Schill, G. Catenanes, Rotaxanes and Knots;
Academic Press: New York, 1971.
(2) Molecular Catenanes, Rotaxanes and Knots; Sauvage, J.-P., Dietrich-
Buchecker, C., Eds.; Wiley-VCH: Weinheim, 1999.
(3) For recent review articles, see: Amabilino, D. B.; Stoddart, J. F. Chem.
ReV. 1995, 95, 2725-2828. Chambron, J.-C.; Dietrich-Buchecker, C.;
Sauvage, J.-P. In Templating Self-Assembly, and Self-Organisation;
Sauvage, J.-P., Hosseini, M. W., Eds.; Comprehensive Supramolecular
Chemistry 9; Pergamon: New York, 1996, pp 43-83. Breault, G.
A.; Hunter, C. A.; Mayers, P. C. Tetrahedron 1999, 55, 5265-5293.
Vo¨gtle, F.; Du¨nnwald, T.; Schmidt, T. Acc. Chem. Res. 1996, 29, 451-
460.
(4) (a) Collin, J.-P.; Dietrich-Buchecker, C.; Gavin˜a, P.; Jimenez-Molero,
M. C.; Sauvage, J.-P. Acc. Chem. Res., 2001, 34, 477-487 (special
issue on molecular machines). (b) Balzani, V.; Credi, A.; Raymo, F.;
Stoddart, J. F. Angew. Chem., Int. Ed. 2000, 39, 3348-3391.
(5) Sauvage, J.-P. Molecular Machines and Motors, Structure & Bonding;
Springer: Berlin, 2001; Vol. 99.
(9) Dietrich-Buchecker, C. O.; Sauvage, J.-P.; Kintzinger, J.-P. Tetrahe-
dron Lett. 1983, 24, 5091-5094; 1983, 32, 53-61.
(10) Fujita, M. Acc. Chem. Res. 1999, 32, 53-61.
(11) Park, K. M.; Kim, S. Y.; Heo, J.; Whang, D.; Sakamoto, S.;
Yamaguchi, K.; Him, K. J. Am. Chem. Soc. 2002, 124, 2140-2147.
(12) (a) Fujita, M.; Ibukuro, F.; Hagihara, H.; Ogura, K. Nature 1994, 367,
720. (b) Hori, A.; Kumazawa, K.; Kusukawa, T.; Chand, D. K.; Fujita,
M.; Sakamoto, S.; Yamaguchi, K. Chem. Eur. J. 2001, 7, 4142 and
references therein.
(13) (a) Beer, P. D.; Rothin, A. S. Polyhedron 1989, 9, 1251. (b) Piguet,
C.; Bernardinelli, G.; Williams, A. F.; Bosquet, B. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 582. (c) Padilla-Tosta, M. E.; Fox, O. D.;
Drew, M. G. B.; Beer, P. D. Angew. Chem., Int. Ed. 2001, 40, 4235.
(d) Mingos, D. M. P.; Yau, J.; Menzer, S.; Williams, D. J. Angew.
Chem., Int. Ed. Engl. 1995, 34, 1894. (e) Wiseman, M. R.; Marsh, P.
A.; Bishop, P. T.; Brisdon, B. J.; Mahon, M. F. J. Am. Chem. Soc.
2000, 122, 12598. (f) McArdle, C. P.; Irwin, M. J.; Jennings, M. C.;
Vittal, J. J.; Puddephatt, R. J. Chem. Eur. J. 2002, 8, 723.
(6) Feringa, B. L. Molecular Switches; Wiley-VCH: Weinheim, 2001.
(7) Brouwer, A. M.; Frochot, C.; Gatti, F. G.; Leigh, D. A.; Mottier, L.;
Paolucci, F.; Roffia, S.; Wurpel, G. W. H. Science 2001, 291, 2124-
2128.
(8) Shigekawa, H.; Miyake, K.; Sumaoka, J.; Harada, A.; Komiyama, M.
J. Am. Chem. Soc. 2000, 122, 5411-5412.
(14) Gruter, G.-J. M.; de Kanter, F. J. J.; Markes, P. R.; Nomot, T.;
Akkerman, O. S.; Bickelhaupt, F. J. Am. Chem. Soc. 1993, 115, 12179.
10.1021/ic0302478 CCC: $27.50 © 2004 American Chemical Society
Published on Web 02/17/2004
Inorganic Chemistry, Vol. 43, No. 6, 2004 1895

Colasson and Sauvage
(CH3CN)2(PF6)2 (103.3 mg, 0.124 mmol). Purification was per-
formed by column chromatography (silica using CH2Cl2/MeOH as
eluents). The product 53+‚(PF6-)3 was obtained as a red solid (50
mg, 20% yield).
We would now like to describe the use of ruthenium(II)
centers in conjunction with simple monodentate ligands such
as the pyridyl group, to synthesize symmetrical (two identical
rings) or asymmetrical (two different rings) catenanes.
1H NMR (500 MHz, CD2Cl2): δ 9.68 (d, 2H, H2′′, J ) 5.5 Hz);
8.74 (d, 2H, H4′′, J ) 7.1 Hz); 8.68 (d, 2H, H4-7, J ) 8.3 Hz);
8.42-8.37 (m, 6H, H3′′,7′′,9′′); 8.36 (d, 2H, H4′-7′, J ) 8.3 Hz); 8.32
(s, 2H, H5-6); 8.19 (d, 2H, H5′′, J ) 8.9 Hz); 8.14 (d, 4H, Hm4, J
) 1.2 Hz); 8.08 (d, 2H, H6′′, J ) 8.9 Hz); 7.99 (d, 2H, H3-8, J )
8.3 Hz); 7.78 (d, 4H, Hm3, J ) 8.8 Hz); 7.76 (d, 2H, H3′-8′, J ) 8.3
Hz); 7.75 (s, 2H, H5′-6′); 7.66 (m, 10H, Ho1,o4,8′′); 7.32 (d, 4H, Ho′,
J ) 8.7 Hz); 7.15 (d, 4H, Ho3, J ) 8.8 Hz); 7.03 (d, 4H, Ho2, J )
8.8 Hz); 6.94 (d, 4H, Hm2, J ) 8.8 Hz); 6.70 (d, 4H, Hm1, J ) 8.3
Hz); 5.99 (d, 4H, Hm′, J ) 8.7 Hz); 3.86 (s, 4H, Hꢀ); 3.75-3.50
(brs, 16H, HR,â,γ,δ,). ES-MS: m/z 1029.9 [M - 2PF6]2+, 638.2 [M
- 3PF6]3+ (found); 2349.5 [M] (calcd). UV-vis: λmax ) 418 nm
(ꢀ ) 13300 L‚mol-1‚cm-1), λmax ) 452 nm (ꢀ ) 11300
L‚mol-1‚cm-1).
Experimental Section
1H NMR spectra were recorded on either a Bruker AVANCE
300 (300 MHz) or a Bruker AVANCE 500 (500 MHz) spectrom-
eter, with the deuterated solvent as the lock and residual solvent as
the internal reference. A VG BIOQ triple-quadrupole spectrometer
was used for the electrospray mass spectrometry measurements (ES-
MS), in the positive mode. Absorption spectra were recorded with
an Uvikon XS spectrometer.
Synthesis. Oxygen-sensitive reactions were conducted under a
positive pressure of argon, by Schlenk techniques. Light-sensitive
reactions and purifications were performed in the dark. All solvents
and reagents were of the highest quality available and were used
as received without further purification. Ligands 116 and 2,17 and
the complexes Cu(CH3CN)4(PF6)18 and Ru(phen)2(CH3CN)2(PF6)2,19
were prepared as already reported in the literature.
62+‚(PF6-)2. Compound 53+‚(PF6-)3 (10.7 mg) was dissolved
in CH2Cl2. KCN (50 mg) in H2O was added, and the solution was
stirred for 48 h in the dark. The two layers were separated, and the
solvent was removed to afford an orange solid.
3
2+.(PF6-)2. Ligand 1 (101.9 mg, 0,124 mmol) and Ru(phen)2-
(CH3CN)2(PF6)2 (103.3 mg, 0.124 mmol) were suspended in 2-(2-
ethoxyethoxy)ethanol (30 mL). The solution was stirred and heated
at 140°C under argon for 3 h and then allowed to cool to room
temperature. A saturated aqueous solution of KPF6 was then added.
The precipitate was filtered on a sintered glass funnel, dissolved
in acetone, and purified by a short column (silica, acetone as eluent).
This yielded 40 mg (20% yield) of 32+(PF6-)2 as an orange solid.
1H NMR (500 MHz, CD3CN): δ 9.44 (d, 2H, H2′′, J ) 5.5 Hz);
8.77 (d, 2H, H4′′, J ) 7.1 Hz); 8.58 (d, 4H, Hm1, J ) 8.6 Hz); 8.47
(d, 2H, H4-7, J ) 8.4 Hz); 8.44 (d, 2H, H7′′, J ) 8.2 Hz); 8.33 (m,
6H, H3-8,m4); 8.20 (m, 4H, H3′′,5′′); 8.11 (m, 4H, H6′′,9′′); 7.92 (s,
2H, H5-6); 7.85 (d, 4H, Hm1, J ) 8.8 Hz); 7.75 (d, 8H, Ho2-m3);
7.55 (m, 6H, H8′′,o4); 7.15 (d, 4H, Hm2, J ) 8.8 Hz); 7.07 (d, 4H,
Ho3, J ) 8.8 Hz). ES-MS: m/z 1429.3 [M - PF6]+, 642.3 [M -
2PF6]2+ (found); 1574.4 [M] (calcd). UV-vis: λmax) 419 nm (ꢀ
) 14000 L‚mol-1‚cm-1), λmax ) 453 nm (ꢀ ) 11700 L‚mol-1‚cm-1).
4+.(PF6-). Ligand 1 (213.0 mg, 0.26 mmol), 2 (146.7 mg, 0.26
mmol), and Cu(CH3CN)4PF6 (96.4 mg, 0.26 mmol) were combined
in CH3CN/CHCl3 (30 mL) at room temperature under argon. The
solution was stirred for 1 day. The solvent was then removed to
afford a red solid (418 mg) in quantitative yield.
1H NMR (500 MHz, CD2Cl2): δ 8.72 (d, 2H, H4-7, J ) 8.2
Hz); 8.37 (d, 2H, H4′-7′, J ) 8.3 Hz), 8.33 (s, 2H, H5-6); 8.00 (d,
2H, H3-8, J ) 8.2 Hz); 7.86 (s, 2H, H5′-6′); 7.82 (d, 2H, H3′-8′, J
) 8.3 Hz); 7.72 (d, 4H, Hm3, J ) 8.7 Hz); 7.65 (d, 4H, Ho1, J )
8.4 Hz); 7.35 (d, 4H, Ho′, J ) 8.7 Hz); 7.18 (m, 8H, Ho2-o3); 7.10
(d, 4H, Hm2, J ) 8.7 Hz); 6.81 (d, 4H, Hm1, J ) 8.4 Hz); 6.03 (d,
4H, Hm′, J ) 8.7 Hz); 3.86 (s, 4H, Hꢀ); 3.75-3.50 (brs, 16H,
HR,â,γ,δ). ES-MS: m/z 726.9 [M - PF6 + H]2+ (found); 1598.1
[M] (calcd). UV-vis: λmax ) 438 nm (ꢀ ) 4000 L‚mol-1‚cm-1).
53+‚(PF6-)3. The same procedure as for compound 32+ was used,
starting with 4+‚(PF6-) (197.4 mg, 0.124 mmol) and Ru(phen)2-
1H NMR (300 MHz, CD2Cl2): δ 9.25 (d, 2H, J ) 5.5 Hz); 8.59
(d, 2H, J ) 7.9 Hz); 8.51 (m, 6H); 8.42 (m, 6H); 8.34 (d, 2H, J )
8.0 Hz); 8.23 (d, 2H, J ) 8.6 Hz); 8.17 (d, 2H, J ) 8.0 Hz); 8.06
(m, 8H); 7.90 (m, 6H); 7.77 (m, 10H); 7.60 (m, 4H); 7.19 (m, 4H);
7.07 (d, 4H, J ) 8.7 Hz); 6.97 (d, 4H, J ) 8.8 Hz); 6.27 (d, 4H,
J ) 8.7 Hz); 3.86 (s, 4H, Hꢀ); 3.75-3.50 (brs, 16H, HR,â,γ,δ,). ES-
MS: m/z 925.5 [M - 2PF6]2+ (found); 2141.0 [M] (calcd).
-
73+‚(PF6-)3. The same procedure as for compound 4+‚(PF6
)
was used, starting with 32+‚(PF6-)2 (33.1 mg, 0,021 mmol), 1 (17.3
mg, 0.021 mmol), and Cu(CH3CN)4PF6 (7.8 mg, 0.021 mmol). A
52 mg portion of 73+(PF6-)3 was obtained as a red solid (100%
yield).
1H NMR (500 MHz, CD3CN): δ 9.54 (d, 2H, H2′′, J ) 5.5 Hz);
8.82 (d, 2H, H4′′, J ) 8.4 Hz); 8.63 (brs, 4H, Hm4); 8.51 (d, 2H,
H4′-7′, J ) 8.4 Hz); 8.48 (d, 2H, H7′′, J ) 7.2 Hz); 8.45 (d, 2H,
H4-7, J ) 8.4 Hz); 8.41 (d, 4H, Hm4′, J ) 6.0 Hz); 8.26 (m, 4H,
H3′′,5′′); 8.16 (m, 4H, H6′′,9′′); 7.98 (d, 2H, H3′-8′, J ) 8.40 Hz);
7.96 (s, 2H, H5′-6′); 7.95 (d, 2H, H3-8, J ) 8.4 Hz); 7.84 (d, 4H,
m3′, J ) 8.6 Hz); 7.83 (s, 2H, H5-6); 7.80 (d, 4H, Hm3, J ) 8.1
Hz); 7.67-7.60 (m, 18H, Ho1′,o4,o1,o4′,8′′); 7.20-7.13 (m, 20H,
o3′,o3,o2,o2′,m2′); 7.00 (d, 4H, Hm2, J ) 8.7 Hz); 6.80 (d, 4H, Hm1
H
H
,
J ) 8.4 Hz); 6.76 (d, 4H, Hm2′, J ) 8.4 Hz). ES-MS: m/z 1158.0
[M - 2PF6]2+ (found); 2606.0 [M] (calcd). UV-vis: λmax ) 419
nm (ꢀ ) 11300 L‚mol-1‚cm-1), λmax ) 452 nm (ꢀ ) 8900
L‚mol-1‚cm-1).
-
85+‚(PF6-)5. The same procedure as for compound 32+‚(PF6
)
2
was used, starting with 73+‚(PF6-)3 (20.2 mg, 0.0077 mmol) and
Ru(phen)2(CH3CN)2(PF6)2 (6.4 mg, 0.0077 mmol). The crude
materials were dissolved in acetone. The precipitate was filtered,
and the solvent was evaporated to afford 5 mg (19% yield) of an
orange solid.
1H NMR (500 MHz, CD3CN): δ 9.92 (d, 4H, H2′′, J ) 5.5 Hz);
9.01 (d, 4H, H4′′, J ) 8.4 Hz); 8.78 (m, 14H); 8.56 (d, 4H); 8.42
(m, 8H); 8.32 (d, 4H); 8.11 (m, 8H); 7.92 (m, 12H); 7.76 (m, 18H);
7.25 (m, 16H); 7.05 (d, 8H); 6.87 (m, 8H). ES-MS: m/z 525.0 [M
(15) Ca´rdenas, D. J.; Sauvage, J.-P. Inorg. Chem. 1997, 36, 2777. (b)
Ca´rdenas, D. J.; Gavin˜a, P.; Sauvage, J.-P. J. Am. Chem. Soc. 1997,
119, 2656.
(16) Dietrich-Buchecker, C.; Colasson, B.; Fujita, M.; Hori, A.; Geum, N.;
Sakamoto, S.; Yamaguchi, K.; Sauvage, J.-P. J. Am. Chem. Soc. 2003,
125, 5717.
(17) Dietrich-Buchecker, C.; Sauvage, J.-P. Tetrahedron 1990, 46, 503.
(18) Kubas, G. J. Inorg. Synth. 1990, 28, 68.
(19) Sullivan, B. P.; Salman, D. J.; Meyer, T. J. Inorg. Chem. 1978, 17,
3334.
- 5PF6]5+; 694.4 [M - 4PF6- 4+; 973.9 [M - 3 PF6- 3+ (found);
] ]
3357.4 [M] (calcd). UV-vis: λmax ) 420 nm (ꢀ ) 19200
L‚mol-1‚cm-1), λmax ) 454 nm (ꢀ ) 15600 L‚mol-1‚cm-1).
X-ray Structural Study. Orange single crystals suitable for
X-ray analysis could be obtained for complex 32+‚(PF6-)2 by slow
1896 Inorganic Chemistry, Vol. 43, No. 6, 2004

IrreWersible but NoncoWalent Ru(II)-Pyridine Bond
Table 1. Crystallographic Data for 32+
Results and Discussion
formula
mol wt
cryst syst
space group
a (Å)
b (Å)
c (Å)
R (deg)
â (deg)
γ (deg)
V (Å3)
Z
C88H67F12N11O4P2Ru
Strategy. The strategy leading to a Ru(II) containing
metallamacrocycle and two different catenanes is represented
schematically in Figure 1.
The synthesis of ligand 1 and its coordination properties
with Cu(I) and Pd(II) have been previously reported.16,20 The
sequential use of these two metals allowed the quantitative
synthesis of two different catenanes following a strategy
similar to that shown in Figure 1. We then wanted to
investigate the possibility of making Ru(II) analogues of the
Pd(II)-catenanes.
Ru(II) and Pd(II) display very different properties in the
context of coordination chemistry using 1 and related ligands.
Whereas Pd(II) is clearly a labile metal center, leading to
the thermodynamically most stable complexes, Ru(II) is
notoriously substitutionally inert. In addition, the presence
of metal-to-ligand charge transfer (MLCT) states with Ru(II)
1733.59
triclinic
P1h
12.4176(3)
14.3068(5)
26.672(1)
78.491(5)
87.152(5)
84.293(5)
4617.9(3)
2
color
orange
1.25
0.281
173
0.71073
0.096
D
calcd (g cm-3
)
µ (mm-1
temp (K)
)
wavelength (Å)
Ra
Rw
b
0.129
a R ) ∑||Fo| - |Fc||/|Fo|. b Rw ) [∑w(|Fo| - |Fc|)2/∑w(|Fo| )]1/2
evaporation of a CH3CN solution of 32+‚(PF6-)2. Crystal data and
.
2
details of data collection are provided in Table 1.
Figure 1. Strategies for the synthesis of (a) a ruthenium-containing macrocycle, (b) a Cu/Ru catenane using a covalent chelate-containing macrocycle and
the correlated demetalated ruthenium-containing catenane, and (c) a Ru-Cu-Ru symmetrical catenane. The Ru(II) center is protected by two inert bidentate
2+
ligands (1,10-phenanthroline) so that only two cis positions are involved in the coordination process. Ru(phen)2 was used as a racemate whereas, for
clarity, only the ∆ enantiomer is represented.
Inorganic Chemistry, Vol. 43, No. 6, 2004 1897

Colasson and Sauvage
Scheme 1. Synthesis of Macrocycle 32+
can potentially lead to novel photochemical properties related
to energy and electron transfer process or light-driven
dynamics.
dilution conditions, which is a classical procedure for the
formation of large macrocycles.22 For unknown reasons, these
conditions did not lead to 32+. We then found that reacting
1 and Ru(phen)2(CH3CN)2(PF6)2 in 2-(2-ethoxyethoxy)-
ethanol (the concentration of both reactants is 4 × 10-3 M)
at 140 °C for 3 h provided the best conditions to obtain 32+
(Scheme 1). Compound 32+ was isolated after column
chromatography in 20% yield as its PF6 salt. Most of the
crude product was recovered as an orange insoluble product
which is likely to contain oligomeric coordination species.
Moreover, no evidence for complexation occurring on the
phen chelate was found. The 20% yield clearly indicates that
the chelate effect of the two pyridines of 1 is very weak. It
appears that Ru(II) first binds the most accessible coordina-
tion site, which is a pyridine, and subsequently another one
(either intramolecularly or intermolecularly), without any
possibility of equilibration in favor of the macrocycle, even
at high temperature (some attempts were carried out at 160
°C).
The ES-MS spectrum shows two major peaks at m/z
1429.3 (monocharged species) and m/z 642.3 (dicharged
species) indicating that the product of the complexation has
a 1:1 (1:Ru(phen)2) stoichiometry.
1H NMR of the product provides some important features
concerning the structure of the desired Ru-macrocycle
(Figure 2). The signal corresponding to H2′ appears at high
field due to the presence of the two pyridyl units coordinated
The phen/bis-pyridine ligand 1 has shown its great ability,
under thermodynamic control, to coordinate, via the two
pyridyl units, two cis positions of a metal to form a
metallamacrocycle. The pincer-like shape of ligand 1 led us
to believe that the formation of a metallamacrocycle could
also be favored in a kinetic process. For this purpose, we
chose Ru(II) which is known to form stable and kinetically
inert complexes when it is coordinated to aromatic imine-
based ligands. The appropriate Ru(II) precursor should be
protected so that only two cis positions can be involved in
the complexation process. cis-Ru(phen)2(CH3CN)2(PF6)2 was
chosen as the two acetonitrile ligands can easily be substi-
tuted by two pyridines.
Figure 1a presents the one-to-one coordination reaction
between 1 and Ru(phen)2(CH3CN)2(PF6)2 which should lead
to the macrocycle. Ru(II) is not expected to complex to the
phen chelate of 1 since the pyridines are much more
accessible, although Ru(II) complexes of 2,9-diaryl-1,10-
phen can certainly be obtained.21 Figure 1b represents the
strategy used to make a catenane with two different rings.
Ligand 1 is first threaded inside the phen-containing mac-
rocycle 2 using the classical template effect of copper(I)
+
leading to a CuI(dpp)2 -type complex (dpp is a 2,9-diphenyl-
1,10-phenanthroline). Considering the great stability of four-
coordinate CuI(dpp)2+ complexes, no side reactions involving
the two pyridines should occur. Next, the cyclization reaction
which consists of a 1:1 complexation reaction between the
two dangling pyridines of 1 and the Ru(II) precursor should
lead to the hetero [2]-catenane. The final approach presented
in Figure 1c is the sum of the previous two strategies:
formation of the Ru-macrocycle, threading process using
copper(I) coordination, and cyclization step with the ruthe-
nium precursor. This strategy was preferred to the double
cyclization approach which should obviously proceed with
a lower yield.
Synthesis of Macrocycle 32+. The reaction between 1 and
Ru(phen)2(CH3CN)2(PF6)2 was examined. We first tried high-
(20) Dietrich-Buchecker, C. O.; Geum, N.; Hori, A.; Fujita, M.; Sakamoto,
S.; Yamaguchi, K.; Sauvage, J.-P. Chem. Commun. 2001, 1182.
(21) Laemmel, A.-C.; Collin, J.-P.; Sauvage, J.-P. Eur. J. Inorg. Chem.
1999, 3, 383 and unpublished results.
(22) Mu¨rner, H.; Belser, P.; von Zelewsky, A. J. Am. Chem. Soc. 1996,
118, 7989.
Figure 2. 1H NMR (300 MHz, 25 °C, CD3CN) of 32+
.
1898 Inorganic Chemistry, Vol. 43, No. 6, 2004

IrreWersible but NoncoWalent Ru(II)-Pyridine Bond
geometry around the oxygens is close to ideal since the
C(69)-O(1)-C(72) and C(36)-O(2)-C(33) angles of the
diphenyl ether links are about 117°.23 The size of the cavity
can be described with three lengths: the O(1)-O(2) distance
is 13.4 Å, and the N(6)(phen)-Ru and N(7)(phen)-Ru
distances are both 17.3 Å.
Synthesis of the Bimetallic Catenane 53+. Catenane 53+
was obtained after two complexation steps (Scheme 2). The
first reaction consisted of the threading of 1 into 2 by
coordination of the two phen chelates on Cu(I). The copper
salt was added to a solution (CH3CN/ CHCl3) of 2. The
+
orange solution of Cu(2)(CH3CN)2 (not represented in
Figure 3. X-ray structure of 32+. Wire frames representation (hydrogens
are omitted for clarity). The compound represented in Figure 3 is the ∆
enantiomer.
Scheme 2) was stirred during 30 min after which a solution
of 1 was added. The color turned immediately red, and the
solution was allowed to stir another 24 h. After evaporation
of the solvents, compound 4+ (PF6 salt) could be obtained
as a red powder in quantitative yield. Eventhough side
reactions are possible due to the presence of the pyridines,
the stability of the threaded complex 4+ shifts the equilibrium
toward this single species. The cyclization step was then
performed using the same conditions as for 32+ (Ru(II)
precursor, solvent, concentrations, temperature, reaction
to the ruthenium. The 1H ROESY experiment clearly displays
spatial interaction between H2′ and Hm4 but shows no
interaction between the ruthenium part and the phen core,
indicating that the complexation occurred on the pyridines.
Furthermore, Hm2 and Ho3, which have the same chemical
shift in 1 because of the conformational freedom around the
oxygen, are differentiated once the macrocycle is formed.
We were able to grow orange single crystals by slow
-
time), and compound 53+‚(PF6 )3 was isolated after chro-
-
evaporation of a CH3CN solution of 32+‚(PF6 )2. The two
matography column in 20% yield, i.e., with approximately
the same efficiency as for the preparation of the simple ring
32+.
enantiomers of 32+ are present in the unit cell but only the
∆ enantiomer is represented in Figure 3. Structural deter-
mination by single crystal X-ray diffraction confirms the
macrocyclic structure of 32+. Analysis of the structure proves
that the ligand is well adapted to the formation of a cyclic
complex such as 32+, since almost no distortion from ideal
geometry is observed. The N(5)-Ru-N(8) angle of the
pyridine-Ru-pyridine coordination fragment is 90.0°. The
Compound 53+ was characterized by ES-MS spectrometry
and 2D 1H NMR analysis. The mass spectrum exhibits two
peaks corresponding to the product at m/z 1029.9 (for [53+
- 2+
+ PF6 ] ) and m/z 638.2 (for [53+]3+). The two coordination
1
steps yielding 53+ could be followed by H NMR (Figure
4). The threaded precursor complex 4+ has the usual high
Scheme 2. Synthesis of Catenane 53+ and Its Precursor 4+
Inorganic Chemistry, Vol. 43, No. 6, 2004 1899

Colasson and Sauvage
Scheme 3. Demetalation of 53+ Using KCN
field shifted signals corresponding to the protons of the
phenyl rings directly attached to the phen of both 1 and 2.
This is due to the phen ring current effect. Surprisingly, the
protons of the pyridines do not appear in the spectrum of
4+. After cyclization, the characteristic shift of H2′′ appears
at δ ) 9.68 ppm. We can also observe a slight difference
for protons Hm2 and Ho3 again due to the restriction of
conformational freedom in the macrocycle. When coordi-
nated to ruthenium, the protons of the pyridines appear at
approximately the same δ as in 32+ (8.14 for Hm4 and 7.66
for Ho4). 1H ROESY experiments prove the spatial proximity
of H2′′ and Hm4.
Demetalation of 53+. As it has been previously reported,
Cu catenanes can be easily demetalated by reaction with
cyanide.17-24 Our aim was to selectively remove the tem-
plating metal (Cu) from compound 53+ without affecting the
Ru complex and, thus, the topological properties of the
catenane. The reaction was performed in dichloromethane
(53+)/water (excess of KCN) in the dark and monitored by
1H NMR (Scheme 3).
Figure 4. 1H NMR monitoring for the preparation of catenane 53+
(aromatic region, 300 MHz, 25 °C): (a) pre-catenane 4+ in CD2Cl2; (b)
catenane 53+ in CD2Cl2.
) + 0.27 ppm) and Hm′ (∆δ ) + 0.28 ppm) (see Scheme
3). These downfield shifts are not as large as those observed
in the case of classical covalent catenanes but are comparable
to the demetalation of a Ru(terpy)2 containing catenane.15
The ES-MS spectrum of the product confirmed the demeta-
lated structure of 62+. The predominant peak at m/z 925.5
corresponds to [62+]. A smaller one at m/z 642.3 corresponds
to the fragment [32+] of 62+. This peak becomes more intense
as the cone voltage increases (until the peak at m/z 925.5
After 48 h, the signal of H2′′ at 9.68 ppm had completely
disappeared and a doublet emerged at 9.25 ppm. This was
correlated with the downfield shift of the protons Hm1 (∆δ
Scheme 4. Synthesis of Catenane 85+ and Its Precursor 73+
1900 Inorganic Chemistry, Vol. 43, No. 6, 2004

IrreWersible but NoncoWalent Ru(II)-Pyridine Bond
-
-
completely disappears). This indicates that we could selec-
tively remove the copper(I) cation without affecting the
ruthenium complex.
Synthesis of the Trimetallic Catenane 85+. We followed
the final strategy (part c in Figure 1) to synthesize the
symmetrical trimetallic catenane 85+. We repeated the
sequence used for catenane 53+, threading reaction using Cu-
(I) and cyclization reaction with the Ru(II) precursor, but
the phen containing macrocycle involved in this sequence
was 32+ instead of 2 (Scheme 4).
sponding to 85+‚(PF6 )5 with the loss of 3, 4, and 5 PF6
anions (respectively, m/z 973.9, 694.4, and 525.0). As only
5 mg of product 85+ (PF6 salt, 19% yield) could be obtained,
no further experiment (demetalation, separation of the two
diastereoisomers) could be performed.
Conclusion
We have investigated the coordination properties of the
phen/bis-pyridine ligand 1 with Ru(II). A metallamacrocycle
constructed around a Ru(II)(phen)2(pyridine)2 fragment could
be obtained. The cyclization step appeared to have a low
yield because of the weak chelate effect of ligand 1.
Nevertheless, this cyclization based on the coordination
reaction between Ru(phen)2(CH3CN)2(PF6)2 and the two
dangling pyridines of 1 was reliable since the yield was
constant (20%) for the three different systems studied
(metallamacrocycle and two catenanes). We were able to
incorporate a Ru(phen)2 fragment in a rigid structure. The
large cavity of macrocycle 32+ with its aromatic walls could
provide interesting behavior in host-guest chemistry. In
addition, the different strategies used should enable us to
construct related systems in which we can take advantage
of both this rigidity and the photophysical properties of the
metals. In this respect, we are now working on the synthesis
of hetero-trimetallic systems with a catenane structure.
The threaded complex 73+ was formed quantitatively after
(1) addition of the copper salt to a CH3CN/CH3Cl solution
of 32+ and (2) addition of ligand 1. Complex 73+ (PF6 salt)
was isolated and fully characterized by 1D and 2D 1H NMR
and ES-MS. Reaction of 73+ with Ru(phen)2(CH3CN)2(PF6)2
in 2-(2-ethoxyethoxy)ethanol gave an orange solid. The solid
was then dissolved in acetone, and insoluble species were
1
filtered off. H NMR of the soluble orange solid shows a
well-resolved doublet at δ ) 9.92 ppm. Two kinds of signals
can be distinguished: sharp and broad ones. We assume that
the sharp signals belong to the protons of the phenanthrolines
coordinated on the Ru and the broad signals are the
superposition of the spectra of the two diastereoisomers
(racemic mixture ∆-∆ and Λ-Λ and meso compound
∆-Λ). ES-MS spectrum of the product indicated that the
desired product has been made: three major peaks corre-
Supporting Information Available: Crystallographic data in
CIF format. This material is available free of charge via the Internet
at http://pubs.acs.org.
(23) Horn, E.; Zhang, S.-Z.; Furukawa, N. New Cryst. Struct. 2000, 215,
557.
(24) Dietrich-Buchecker, C.; Sauvage, J.-P. Chem. ReV. 1987, 87, 795.
IC0302478
Inorganic Chemistry, Vol. 43, No. 6, 2004 1901