Multiporphyrinic rotaxanes: Control of intramolecular electron transfer rate by steering the mutual arrangement of the chromophores

Article abstract of DOI:10.1021/ja002018f

A [2]-rotaxane Zn₂-Au⁺ made from a dumbbell component ended by Zn(II) porphyrin stoppers and a ring component incorporating a Au(III) porphyrin has been assembled in 13% yield using the transition metal templating route. ¹

Full text of DOI:10.1021/ja002018f

11834
J. Am. Chem. Soc. 2000, 122, 11834-11844
Multiporphyrinic Rotaxanes: Control of Intramolecular Electron
Transfer Rate by Steering the Mutual Arrangement of the
Chromophores
Myriam Linke,† Jean-Claude Chambron,*,† Vale´rie Heitz,† Jean-Pierre Sauvage,*,†
Susana Encinas,‡ Francesco Barigelletti,‡ and Lucia Flamigni*,‡
Contribution from the Laboratoire de Chimie Organo-Mine´rale, Institut Le Bel, UniVersite´ Louis-Pasteur,
4, rue Blaise Pascal, 67070 Strasbourg Cedex, France and the Istituto FRAE-CNR, Via P. Gobetti, 101,
40129 Bologna, Italy
ReceiVed June 6, 2000
Abstract: A [2]-rotaxane Zn2-Au+ made from a dumbbell component ended by Zn(II) porphyrin stoppers
and a ring component incorporating a Au(III) porphyrin has been assembled in 13% yield using the transition
metal templating route. 1H NMR studies show that its conformation in solution is very different from those of
its complexes with Cu+, Ag+, and Li+. In particular, removal of the templating metal resulted in a changeover
of the molecule, the threaded macrocycle undergoing a pirouetting motion placing the Au(III) porphyrin in the
cleft formed by the two Zn(II) porphyrin stoppers. At room temperature, the changeover could be either complete
or partial, depending on the solvent used. Photoinduced electron transfer from one of the Zn(II) porphyrins to
the Au(III) porphyrin of the macrocycle was evidenced in the case of the free rotaxane and its Cu(I) complex,
Zn2Cu+Au+. In the former case, the photoinduced electron transfer process could be clearly resolved for an
extended conformation that is characterized by the Zn(II) porphyrins pointing far from the Au(III) porphyrin
electron acceptor, and accounting for 30% of the total in acetonitrile at room temperature. In both Zn2Cu+-
Au+ and Zn2-Au+ rotaxanes, the charge-separated state, in which the Zn(II) porphyrin is a cation radical and
the Au(III) porphyrin a neutral radical, was generated at a rate of 5 × 109 s-1 and disappeared at a rate of 2
× 108 s-1. In the case of Zn2Cu+Au+, the primary step is very likely energy transfer from the Zn(II) porphyrin
singlet excited state to the MLCT state of the central Cu(I) complex, followed by an electron transfer from the
excited Cu(I) unit to the Au(III) porphyrin and a successive charge shift from the Zn(II) porphyrin to the
oxidized Cu(II) complex. [2]-rotaxane Zn2-Au+, in which no bond pathway can be identified between the
donor and the acceptor, is a typical case of electron transfer involving molecular fragments connected by
mechanical bonds.
Introduction.
Our groups have been interested for several years in mimick-
ing the function of the special pair (SP)/bacteriopheophytin
(BPh) dyad of the bacterial photosynthetic reaction center, the
structure of which had been solved by X-ray crystallography.2d,e
SP, as a bacteriochlorophyll dimer, is the primary electron donor
in its excited state, and BPh is the secondary electron acceptor,
being established at least at room temperature, that the primary
electron acceptor is a bacteriochlorophyll (BCh) cofactor located
between SP and BPh.3 We used Zn(II) or free-base porphyrins
as SP mimics and introduced Au(III) porphyrins as functional
models of BPh.4 We developed and studied both linear triadic
arrays assembled around Ru(II) and Ir(III) terpyridine com-
plexes5 and oblique systems connected by a 1,10-phenanthroline
The study of electron transfer (ET) and/or energy transfer
(EN) between porphyrins or analogues is of high biological
relevance.1 Examples of natural ET porphyrins are the cyto-
chromes of the mitochondrial respiratory chain, which convey
electrons to the terminal cytochrome oxidase complex, and
chlorophylls, which, as dimeric structures known as P680 and
under light activation, pump electrons from the oxygen-evolving
complex (OEC) of the photosystem II (PS II) of green plants.2a
As far as natural EN is concerned, the so-called light-harvesting
complexes massively rely on porphyrinic chromophores for
collection and transduction of light energy to the reaction center
(RC).2b,c
† Universite´ Louis-Pasteur.
(3) (a) Scherer, P. O. J.; Fischer, S. F. Chem. Phys. Lett. 1987, 141,
179-185. (b) Warshel, A.; Creighton, S.; Parson, W. W. J. Phys. Chem.
1988, 92, 2696-2701. (c) Dressler, K.; Umlauf, E.; Schmidt, S.; Hamm,
P.; Zinth, W.; Buchanan, S.; Michel, H. Chem. Phys. Lett. 1991, 183, 270-
276. (d) Chan, C.-K.; DiMagno, T. J.; Chen, L. X.-Q.; Norris, J. R.; Fleming,
G. R. Proc. Natl. Acad. Sci. U.S.A. 1991, 88, 11202-11206. (e) Arlt, T.;
Schmidt, S.; Kaiser, W.; Lauterwasser, C.; Meyer, M.; Scheer, H.; Zinth,
W. Proc. Natl. Acad. Sci. U.S.A. 1991, 88, 11757-11761.
(4) (a) Chardon-Noblat, S.; Sauvage, J.-P.; Mathis, P. Angew. Chem.,
Int. Ed. Engl. 1989, 28, 593-595. (b) Heitz, V.; Chardon-Noblat, S.;
Sauvage, J.-P. Tetrahedron Lett. 1991, 32, 197-198. (c) Brun, A. M.;
Harriman, A.; Heitz, V.; Sauvage, J.-P. J. Am. Chem. Soc. 1991, 113, 8647-
8663. (d) Harriman, A.; Heitz, V.; Ebersole, M.; van Willigen, H. J. Phys.
Chem. 1994, 98, 4982-4989.
‡ Istituto FRAE-CNR.
(1) (a) McLendon, G.; Hake, R. Chem. ReV. 1992, 92, 481-490. (b)
Wasielewski, M. R. Chem. ReV. 1992, 92, 435-461. (c) Gust, D.; Moore,
A. T.; Moore, A. L. Acc. Chem. Res. 1993, 26, 198-205. (d) Kurreck, H.;
Huber, M. Angew. Chem., Int. Ed. Engl. 1995, 34, 849-866. (e) Harriman,
A.; Sauvage, J.-P. Chem. Soc. ReV. 1996, 41-48.
(2) (a) Rawn, J. D. Biochemistry; Neil Patterson Publishers: Burlington,
North Carolina, 1989. (b) Ku¨hlbrandt, W.; Wang, D. N.; Fujiyoshi, Y.
Nature 1994, 367, 614-621. (c) McDermott, G.; Prince, S. M.; Freer, A.
A.; Hawthornthwaite-Lawless, A. M.; Papiz, M. Z.; Cogdell, R. J.; Isaacs,
N. W. Nature 1995, 374, 517-521. (d) Deisenhofer, J.; Michel, H. Angew.
Chem., Int. Ed. Engl. 1989, 28, 829-847. (e) Huber, R. Angew. Chem.,
Int. Ed. Engl. 1989, 28, 848-869.
10.1021/ja002018f CCC: $19.00 © 2000 American Chemical Society
Published on Web 12/06/2000

Multiporphyrinic Rotaxanes
J. Am. Chem. Soc., Vol. 122, No. 48, 2000 11835
Figure 2. Principle of transition metal-templated synthesis of a
[2]-rotaxane containing a Au(III) porphyrin-incorporating macrocycle.
Symbols are as in Figure 1. See text for details.
Figure 1. Schematic representation of metal-complexed [2]-rotaxanes
(a, c, and d) and [2]-catenane (b). The thick lines represent chelating
fragments, the black disk is a metal cation, the empty diamonds are
Zn(II) porphyrins, and the hatched diamonds are Au(III) porphyrins.
observed, whereas in the presence of the templating metal, the
mechanism and kinetics of electron transfer are very different
from those observed in the case of Cu(I) [2]-rotaxane of Figure
1a, despite the fact that both rotaxanes are built from similar
photo- and electroactive components.
spacer.4,6,7 One of our most successful model systems in this
context was the Cu(I) complex of a [2]-rotaxane made from
2,9-diphenyl-1,10-phenanthroline (dpp)-incorporating macro-
cycle threaded onto a dpp-bridged, heterobimetallic Zn(II)/Au-
(III) bis-porphyrin dumbbell molecular component,6 schemati-
cally represented in Figure 1a. Photoinduced electron transfer
from the Zn(II) porphyrin singlet excited state to the Au(III)
porphyrin took place at a rate of (1.7 ps)-1 in butyronitrile, very
close to the (3 ps)-1 global rate that was observed for the natural
system. Furthermore, it was demonstrated that the central metal
played a crucial role in effectively controlling the rate of electron
transfer between the porphyrin chromophores.7 In this model
molecule, the donor and acceptor porphyrins are linked co-
valently, and electron transfer may proceed through bond via
the dpp bridge or through space via the dpp fragment of the
macrocycle, which is interspersed between the two chro-
mophores.
A following step was that of taking advantage of mechanical
bonds. To this aim, catenanes8 and rotaxanes9,10 were synthe-
sized to study through-space electron transfer. At first, systems
involving macrocycles with pendent porphyrins, as schematically
represented in Figure 1b and c, were prepared and their
photochemical properties investigated.9
In this paper, we report on the synthesis, characterization,
and photophysical properties of rotaxanes that were made of a
macrocycle incorporating a Au(III) porphyrin in its backbone
and threaded onto a Zn(II) bis-porphyrin dumbbell component
(Figure 1d).10 In the absence of the central Cu(I) cation,
photoinduced electron transfer across mechanical bonds is
Results and Discussion.
1. Design and Synthesis. 1.1. Principles of Rotaxane
Assembly. Rotaxanes made of Au(III) porphyrin-containing
macrocycles threaded onto Zn(II) bis-porphyrin dumbbell mo-
lecular components were assembled using the transition metal
template technique developed for preparing catenanes,11 rotax-
anes,12 and knots.13 As shown in Figure 2, the metal controls
the threading of the chelating macrocycle (A) onto a comple-
mentary chelate (B) ended by functional groups X, to afford a
prerotaxane metal complex (C). Subsequent construction of the
porphyrin stoppers at the precursor functions X provides the
desired [2]-rotaxane structure, as its complex with the metal
template (D). Removal of the latter by competitive complexation
releases the “free” [2]-rotaxane (E) by suppression of the
coordination bonds between the molecular dumbbell and the
macrocycle. These components nevertheless remain linked by
the so-called mechanical bond, which is nothing more than the
result of steric interactions between the macrocycle and the
stoppers.
1.2. Synthesis of the Precursors and Preparation of the
[2]-Rotaxane. The different precursors to the macrocycle and/
or the rotaxane of this study are represented in Figure 3. At
first, macrocycle Au+, incorporating a Au(III) porphyrin in its
backbone, was prepared as follows. 2,9-Bis(p-hydroxyphenyl)-
1,10-phenanthroline 111b (1 equiv) was reacted with 2-bromo-
ethanol (3 equivs) in the presence of K2CO3 (DMF, 150 °C) to
produce crude 2,9-bis[p-(2-hydroxyethoxy)phenyl]-1,10-phenan-
throline 2, which was used without purification in the reaction
(5) (a) Harriman, A.; Odobel, F.; Sauvage, J.-P. J. Am. Chem. Soc. 1995,
117, 9461-9472. (b) Flamigni, L.; Barigelletti, F.; Armaroli, N.; Collin,
J.-P.; Sauvage, J.-P.; Williams, J. A. G. Chem. Eur. J. 1998, 4, 1744-
1754. (c) Flamigni, L.; Armaroli, N.; Collin, J.-P.; Dixon, I. M.; Sauvage,
J.-P.; Williams, J. A. G. Coord. Chem. ReV. 1999, 190-192, 671-682. (d)
Dixon, I. M.; Collin, J.-P.; Sauvage, J.-P.; Barigelletti, F.; Flamigni, L.
Angew. Chem., Int. Ed. 2000, 39, 1292-1295.
(11) (a) Dietrich-Buchecker, C.; Sauvage, J.-P. Chem. ReV. 1987, 87,
795-810. (b) Dietrich-Buchecker, C.; Sauvage, J.-P. Tetrahedron 1990,
46, 503-512. (c) Sauvage, J.-P. Acc. Chem. Res. 1990, 23, 327-332.
(12) (a) Chambron, J.-C.; Heitz, V.; Sauvage, J.-P. J. Chem. Soc., Chem.
Commun. 1992, 1131-1133. (b) Chambron, J.-C.; Heitz, V.; Sauvage, J.-
P. J. Am. Chem. Soc., 1993, 115, 12378-12384.
(13) (a) Dietrich-Buchecker, C. O.; Sauvage, J.-P. Angew. Chem., Int.
Ed. Engl. 1989, 28, 189-192. (b) Dietrich-Buchecker, C. O.; Sauvage, J.-
P.; Kintzinger, J.-P.; Malte`se, P.; Pascard, C.; Guilhem, J. New J. Chem.
1992, 16, 931-942. (c) Dietrich-Buchecker, C. O.; Nierengarten, J.-F.;
Sauvage, J.-P.; Armaroli, N.; Balzani, V.; De Cola, L. J. Am. Chem. Soc.
1993, 115, 11237-11244. (d) Carina, R.; Dietrich-Buchecker, C.; Sauvage,
J.-P. J. Am. Chem. Soc. 1996, 118, 9110-9116. (e) Dietrich-Buchecker,
C. O.; Sauvage, J.-P.; De Cian, A.; Fischer, J. J. Chem. Soc., Chem.
Commun. 1994, 2231-2232.
(6) Chambron, J.-C.; Harriman, A.; Heitz, V.; Sauvage, J.-P. J. Am. Chem.
Soc. 1993, 115, 6109-6114.
(7) Chambron, J.-C.; Harriman, A.; Heitz, V.; Sauvage, J.-P. J. Am. Chem.
Soc. 1993, 115, 7419-7425.
(8) (a) Amabilino, D. B.; Sauvage, J.-P. Chem. Commun. 1996, 2441-
2442. (b) Amabilino, D. B.; Sauvage, J.-P. New J. Chem. 1998, 22, 395-
409.
(9) (a) Linke, M.; Chambron, J.-C.; Heitz, V.; Sauvage, J.-P. J. Am.
Chem. Soc. 1997, 119, 11329-11330. (b) Andersson, M.; Linke, M.;
Chambron, J.-C.; Davidsson, J.; Heitz, V.; Sauvage, J.-P.; Hammarstro¨m;
L. J. Am. Chem. Soc. 2000, 122, 3526-3527.
(10) Linke, M.; Chambron, J.-C.; Heitz, V.; Sauvage, J.-P.; Semetey,
V. Chem. Commun. 1998, 2469-2470.

11836 J. Am. Chem. Soc., Vol. 122, No. 48, 2000
Linke et al.
Figure 3. Chemical formulas of the precursors of the [2]-rotaxane.
with tosyl chloride (Et3N, CH2Cl2, 0 °C). The overall yield of
ditosylate 3 was 31% after chromatography. Exchange of the
tosylate functional groups of 3 by iodide (NaI, refluxing acetone)
yielded quantitatively the phenanthroline derivative 4. The
porphyrin precursor was prepared following Lindsey’s proce-
dure14 using 3,5-di-tert-butylbenzaldehyde 715 (1 equiv), 4-meth-
oxybenzaldehyde 8 (1 equiv), and pyrrole (2 equivs). 15,20-
bis(3,5-di-tert-butylphenyl)-5,10-bis(4-methoxyphenyl)-
porphyrin 9 was isolated in 6% yield after chromatographic
separation from the other porphyrins produced in this reaction.
Au(III) was then inserted by reaction of 9 with KAuCl4 (2
equivs) in the presence of NaOAc (2.5 equivs) in refluxing acetic
acid.4b The Au(III) complex 10 of free base porphyrin 9 was
isolated in nearly quantitative yield as its PF6- salt. The methoxy
groups were cleaved by BBr3 (30 equivs) in CH2Cl2 at -78
°C, affording the Au(III) porphyrin 11 quantitatively. Finally,
the coupling reaction of phenanthroline derivative 4 (1.1 equivs)
and Au(III) porphyrin 11 (1 equiv) in the presence of Cs2CO3
(DMF, 55 °C) afforded macrocycle Au+ as its PF6- salt in 31%
yield.
steps. First, reaction of 2.5 equivs of p-(5,5-dimethyl-1,3-dioxan-
2-yl)lithiobenzene16 with 1,10-phenanthroline, followed by
hydrolysis and rearomatization with excess MnO2 afforded 2,9-
bis[p-(5,5-dimethyl-1,3-dioxan-2-yl)phenyl]-1,10-phenanthro-
line 5 in 47% yield after column chromatography. The desired
dialdehyde 6 was recovered in 95% yield by acidic hydrolysis
of 5.
Template formation of prerotaxane Cu+Au+ of Figure 4 was
carried out by mixing macrocycle Au+ with equimolecular
amounts of Cu(CH3CN)4PF6 and 2,9-bis[p-(formylphenyl)]-1,-
10-phenanthroline 6. Prerotaxane Cu+Au+ was obtained quan-
titatively and used directly in the next step. Porphyrin stoppers
were constructed at the protruding aldehyde functions as follows
(Figure 4): a mixture of prerotaxane Cu+Au+ (1 equiv), 3,5-
di-tert-butylbenzaldehyde 715 (8 equivs), (3,3′-dihexyl-4,4′-
dimethyl-2,2′-dipyrryl)methane 12 (10 equivs),17 and a few
drops of trifluoroacetic acid in CH2Cl2 was stirred at RT
overnight. Tetrapyrrole assembly was fixed by controlled
oxidation of the porphyrinogens with tetrachloro-p-benzo-
quinone in refluxing CH2Cl2. Cu(I)-complexed [2]-rotaxane
(H2)2Cu+Au+ was isolated in 13% yield after chromatographic
purification. Free base porphyrin H2 was also isolated from the
2,9-Bis[p-(formylphenyl)]-1,10-phenanthroline 615 of this
study was prepared from commercial 1,10-phenanthroline in two
(14) (a) Lindsey, J. S.; Hsu, H. C.; Schreiman, I. C. Tetrahedron Lett.
1986, 27, 4969-4970. (b) Lindsey, J. S.; Schreiman, I. C.; Hsu, H. C.;
Kearney, P. C.; Marguerettaz, A. M. J. Org. Chem. 1987, 52, 827-836.
(c) Lindsey, J. S.; Wagner, R. W. J. Org. Chem. 1989, 54, 828-836.
(15) Chardon-Noblat, S.; Sauvage, J.-P. Tetrahedron 1991, 47, 5123-
5132.
(16) Solladie´, N.; Chambron, J.-C.; Sauvage, J.-P. J. Am. Chem. Soc.
1999, 121, 3684-3692.
(17) (a) Arsenault, G. P.; Bullock, E.; McDonald, S. F. J. Am. Chem.
Soc. 1960, 82, 4384-4389. (b) Young, R.; Chang, C. K. J. Am. Chem.
Soc. 1985, 107, 898-909.

Multiporphyrinic Rotaxanes
J. Am. Chem. Soc., Vol. 122, No. 48, 2000 11837
Figure 5. Illustration of the demetalation/remetalation reactions carried
out on Cu(I)-complexed [2]-rotaxane Zn2Cu+Au+ to afford free
[2]-rotaxane Zn2-Au+ and metallo-[2]-rotaxanes Zn2Ag+Au+ and Zn2-
Li+Au+ and showing the pirouetting motion of the Au(III)-incorporating
macrocycle upon removal of the central metal. Indicated with arrows
are the rOe correlations as observed by proton NMR spectroscopy.
Figure 4. Cu(I)-directed template synthesis of [2]-rotaxane Zn2Cu+-
Au+.
reaction mixture. Zn(II) insertion into the porphyrin stoppers
of (H2)2Cu+Au+ was performed by reaction with Zn(OAc)2‚
2H2O in a refluxing mixture of CHCl3 and CH3OH and afforded
nearly quantitatively Cu(I)-complexed [2]-rotaxane Zn2Cu+-
Au+. Reference porphyrin Zn was obtained similarly.
2. Conformations of the [2]-Rotaxanes in Solution: a
Proton NMR Study. 2.1. Dichloromethane Solution. In this
section, we compare the conformations of [2]-rotaxane Zn2-
Au+ and its metal complexes, Zn2Cu+Au+, Zn2Ag+Au+, and
Zn2Li+Au+. We show that removal of the metal template is
followed by a pirouetting motion of the macrocyclic component,
placing the Au(III) porphyrin that it incorporates in the cleft
formed by the two Zn(II) porphyrin stoppers. For the sake of
comparison, the data obtained earlier for the prototypical Cu(I)
[2]-catenate Cu+11b and the related Cu(I) [2]-rotaxane (H2)2-
1.3. Formation of the Free [2]-Rotaxane and Coordination
Chemistry Studies. The metal template (Cu(I)) was selectively
extruded by reacting the [2]-rotaxane complex Zn2Cu+Au+ with
KCN (50 equivs) in a 1:1:4 mixture of CHCl3, H2O, and CH3-
CN at 40 °C for 30 min.18 This decomplexation reaction
liberated the free [2]-rotaxane Zn2-Au+ quantitatively. As
studied in more detail in the next section, and depicted in Figure
5, the template imprint (a bis-dpp, tetrahedral coordination
sphere) is completely vanished by rearrangement of the threaded
macrocycle around its axle. Recomplexation of Zn2-Au+ with
Ag+ or Li+ by reaction with AgBF4 or LiBF4 restored the
template imprint and afforded the Ag+- and Li+-[2]-rotaxane
complexes Zn2Ag+Au+ and Zn2Li+Au+ quantitatively.
19b
Cu+,19a as well as its dumbbell precursor (H2)2 (Figure 6)
will be included in this study. Relevant data are collected in
Table 1. The chart shows the proton labeling of the different
components of the rotaxanes. Protons relevant to the following
discussion are highlighted in Figures 4-6. Let us follow the
sequence of reactions shown in Figures 4 and 5. Diagnosis of
(18) (a) Albrecht-Gary, A.-M.; Saad, Z.; Dietrich-Buchecker, C. O.;
Sauvage, J.-P. J. Am. Chem. Soc. 1985, 107, 3205-3209. (b) Dietrich-
Buchecker, C. O.; Sauvage, J.-P.; Kern, J.-M. J. Am. Chem. Soc. 1984,
106, 3043-3045.
(19) (a) Chambron, J.-C.; Heitz, V.; Sauvage, J.-P. Bull. Soc. Chim. Fr.
1995, 132, 340-347. (b) Hungerford, G.; Van der Auweraer, M.; Chambron,
J.-C.; Heitz, V.; Sauvage, J.-P.; Pierre, J.-L.; Zurita, D. Chem. Eur. J. 1999,
5, 2089-2100.

11838 J. Am. Chem. Soc., Vol. 122, No. 48, 2000
Linke et al.
exerted by the porphyrins on their meso substituents accounts
for the further shift of the m′ protons to 7.50 ppm in Cu(I)
[2]-rotaxane species (H2)2Cu+Au+. Comparison of free base
[2]-rotaxanes (H2)2Cu+ and (H2)2Cu+Au+ (Figures 4 and 6)
shows that all of the chemical shifts of the dumbbell protons
are similar, except for those of the pairs (5′,6′) and (4′,7′), which
are shielded by -0.62 and -0.19 ppm respectively, on going
from an “innocent” to the Au(III) porphyrin-incorporating
macrocycle. Therefore, it is not surprising that the most affected
protons are those located in the vicinity of the Au(III) porphyrin,
that is, those belonging to the “rear” part of the dumbbell
component. We suggest that this shielding is due to the fact
that the proton pairs (5′,6′) and (4′,7′) are placed above the Au-
(III) porphyrin plane, a situation that could be made possible
by a folding of the macrocycle (not represented in the figures).
Incorporation of Zn(II) into the porphyrin stoppers does not
significantly alter the chemical shifts (compare (H2)2Cu+Au+
and Zn2Cu+Au+), as expected. The same is true for the
exchange of Ag+ or Li+ for Cu+: all of the metallorotaxanes,
Zn2Cu+Au+, Zn2Ag+Au+, and Zn2Li+Au+, have very similar
spectra. Comparison of the data of catenate Cu+ and rotaxanes
(H2)2Cu+ and (H2)2Cu+Au+ (Figures 4 and 6) shows that the
signals of the proton pairs (5,6) and (4,7) belonging to the dpp
portion of the macrocycles are moved upfield upon passing from
Cu+ to the rotaxanes. This result clearly demonstrates that, in
the case of (H2)2Cu+Au+, the phenanthroline nucleus incor-
porated in the macrocycle is, at least partially, embedded in
the shielding field of the porphyrin stoppers, which is in
agreement with the proposed structure (Figure 4).
Drastic modifications of the proton NMR spectrum are
observed upon demetalation of the central coordination site, due
to a complete changeover of the molecule (Figure 5). Protons
o′, m′, and Me(1) of the dumbbell fragment experience a
downfield shift of +1.02, +0.64, and +0.43 ppm, respectively,
whereas the pair (4′,7′) is moved upfield by -0.44 ppm. Upfield
shifts are also observed for pairs (4,7) (-0.20 ppm), and (5,6)
(-0.29 ppm) and protons o′′ (-0.21 ppm), and py1 (-0.50 ppm)
of the macrocycle, whereas o and m of the same component
are moved downfield by ∼+0.70 ppm. These data support the
following picture of free [2]-rotaxane Zn2-Au+: the macrocycle
has executed a pirouetting motion, placing the Au(III) porphyrin
in the cleft formed by the two Zn(II) porphyrins of the dumbbell.
Hence, the result is the upfield shifts of o′′ and py1, which are
likely to be the most efficiently embedded in the shielding field
of the Zn(II) porphyrin stoppers. Upfield shifts of the proton
pairs (4,7) and (5,6) are due to electronic effects: coordination
of Cu(I) to dpp is known to produce downfield shifts of +0.22
and +0.26 ppm, respectively.21 On the other hand, downfield
shifts of the pairs (o,m) and (o′,m′), ranging from +0.64 to
+1.02 ppm, are due to the fact that upon demetalation, the
phenylene spacers become coplanar with the phenanthroline
nuclei.21
All of these interpretations, which rely on shielding and
deshielding effects, are also supported by 2D ROESY experi-
ments. Remarkable rOe cross-peaks can be found in the NMR
maps that correspond to dipolar couplings between protons of
the macrocycle on one hand and protons of the dumbbell
component on the other hand. In the case of the Cu(I) complex
Zn2Cu+Au+, correlations (see Figure 5) are observed between
protons o and o′, as well as between (5′,6′) and py1. In the case
of the free [2]-rotaxane Zn2-Au+, the correlations observed
for the complex give place to correlations between protons o′
Figure 6. Chemical structures of the reference bisporphyrin
(H2)2, [2]-rotaxane (H2)2Cu+, and Cu(I) [2]-catenate Cu+.
successful Cu(I)-directed threading of macrocycle Au+ onto
dialdehyde 6 to afford prerotaxane Cu+Au+ is the observation
of a high-field signal at 6.36 ppm for the protons m of Au+,20
as in [2]-catenate Cu+ (6.02 ppm).11b The same threading
process accounts for the -1.0 and -0.5 ppm upfield shifts of
the m′ and CHO protons, respectively, which are embedded in
the shielding field of the phenanthroline fragment of Au+.
However, the doublet of the m′ protons (7.10 ppm) appears at
lower field than the m protons in Cu+Au+ (6.36 ppm) because
of the deshielding effect of the CHO group. A stronger effect
(20) Comparison to free macrocycle Au+ is not useful because the
spectrum of the latter is concentration-dependent, as observed previously
for a related macrocycle (see ref 8b).
(21) Dietrich-Buchecker, C. O.; Marnot, P. A.; Sauvage, J.-P.; Kintzinger,
J.-P.; Malte`se, P. NouV. J. Chim. 1984, 8, 573-582.

Multiporphyrinic Rotaxanes
J. Am. Chem. Soc., Vol. 122, No. 48, 2000 11839
1
Table 1: Selected H NMR Data for the Rotaxanes and Reference Compoundsa
dumbell protons
macrocycle protons
compd
Cu+11b
5′,6′
4′,7′
3′,8′
o′
m′
Me(1)
meso
5,6
4,7
3,8
o
m
m′′
o′′
py(1) py(2)
8.28 8.66 7.87 7.37 6.02
6
7.88 8.40 8.23 8.63 8.12
7.96 8.52 8.52 9.02 8.30
8.57 9.07 8.41 8.03 7.44
7.80 8.67 8.08 7.60 7.10
10.16b
10.17
10.19
19b
(H2)2
2.41
1.81
(H2)2Cu+19a
Cu+Au+
7.54 8.39 8.11 7.55 6.24
9.66b 7.90 8.45 7.87 7.34 6.36 7.68 8.34 9.27 9.54
(H2)2Cu+Au+ 7.95 8.88 8.50 8.11 7.50
1.72
1.84
1.79
1.74
2.27
10.14
10.13
10.10
10.12
10.07
7.72 8.26 7.92 7.58 6.60 7.74 8.39 9.26 9.57
7.75 8.31 7.97 7.62 6.63 7.78 8.42 9.30 9.60
7.49 8.18 7.93 7.79 6.65 7.75 8.42 9.26 9.61
7.71 8.23 7.84 7.53 6.67 7.82 8.42 9.29 9.60
7.46 8.11 7.99 8.26 7.39 7.91 8.21 8.80 9.56
Zn2Cu+Au+
Zn2Ag+Au+
Zn2Li+Au+
Zn2-Au+
7.98 8.90 8.52 8.13 7.54
7.99 8.82 8.56 8.19 7.58
7.94 8.90 8.48 8.07 7.53
7.90 8.46 8.57 9.15 8.18
a Chemical shifts in ppm downfield from TMS. Solvent: CD2Cl2. b CHO proton.
Chart 1
and m, o′ and py1, and o′ and m′′ which are fully consistent
with the structures depicted in Figure 5.
niques will allow drawing quite a detailed picture of the
conformational behavior of Zn2-Au+ in acetonitrile.
3. Electrochemistry Studies of the [2]-Rotaxanes and their
Models. Table 2 shows selected redox potentials measured in
butyronitrile for the rotaxanes Zn2Cu+Au+ and Zn2-Au+, as
well as for the model compounds Cu+, Zn, Au+ and prerotaxane
Cu+Au+. Only the redox potentials obtained for the first
oxidation processes of the zinc porphyrin and copper complex
subunits and the first reduction process of the gold porphyrin
macrocycle are of interest here to establish the ergonicity of
the various electron-transfer processes and the energy content
of the charge-separated products. Also shown are the excited-
state redox potentials that were calculated using the present data
and the corresponding values of the excited-state energies (see
Table 4) and neglecting Coulombic effects.
2.2. Acetonitrile Solution. In addition to photochemical
investigations that were performed in nitrile solutions, 1H NMR
studies were also carried out on the free [2]-rotaxane Zn2-
Au+ in acetonitrile. At 40 °C, all the peaks were broad. It was
necessary to heat the sample solution to 74 °C to observe
significant peak sharpening and resolution of most of the proton
signals. However, over the 40-74 °C temperature range, there
was no dramatic change in the general appearance of the whole
spectrum. Signals of the probe protons of the CD2Cl2 study were
not dramatically shifted in CD3CN. Indeed, with the exception
of protons m, which undergo a -0.49 ppm upfield shift, the
probe protons were very little affected by the solvent change;
that is, o′, -0.18 ppm; py1, +0.13 ppm; and Me(1), +0.08 ppm.
This suggests that, at least in CD3CN at 74 °C, the conformation
of Zn2-Au+ is very similar to the one observed in CD2Cl2 at
room temperature. Peak broadening at lower temperatures could
be indicative of relatively slow exchange between conforma-
tions. If the driving force for bringing the gold porphyrin cation
between the two zinc porphyrins is due to ion-dipole interac-
tions, these are expected to drop in the more polar acetonitrile
solvent, allowing for the existence of an extended conformation
in free [2]-rotaxane Zn2-Au+. Indeed, as shown in section 4,
combined time-resolved and steady-state spectroscopy tech-
Remarkably, the values measured for the porphyrinic model
compounds are found unchanged in the rotaxanes, which shows
that the positive charge of the central Cu(I) complex has no
influence on the neighboring porphyrins, which behave inde-
pendently.
In the case of the different compounds incorporating a gold
porphyrin subunit, the first reduction wave, which is reversible,
is observed at ∼-1.04 V vs Fc+/0. It corresponds to the
reduction of the Au(III) porphyrin cation to the corresponding
neutral radical: the Au(III) porphyrin is the easiest reducible

11840 J. Am. Chem. Soc., Vol. 122, No. 48, 2000
Linke et al.
Table 2: Electrochemical Data for the Rotaxanes and the Reference Compounds
ground-state redox potentialsa
excited-state redox potentialsb
compd
Cu+
Cu2+/Cu+
Zn+•/Zn
Au+/Au•
Cu2+/Cu+*
Zn+•/1Zn*
Zn+•/3Zn*
3Au+*/Au•
0.14 (70)
Zn
0.17 (65)
-1.95
-1.54
Au+
-1.03 (50)
-1.02 (60)
-1.04 (65)
-1.05 (65)
0.70
0.71
0.71
0.70
Cu+Au+
Zn2Cu+Au+
Zn2-Au+
0.37 (100)
0.43 (140)
-1.40
-1.34
0.18 (60)
0.17 (70)
-1.94
-1.95
-1.52
-1.53
a V vs Fc+/0 (∆Ep, mV). 0.1 M TBAB in butyronitrile. b Calculated by subtracting the excited-state energy level from the corresponding ground-
state redox potential.
Table 3. Absorption Maxima in the Visible Regiona
4. Photophysical Properties of the Rotaxanes Zn2-Au+
and Zn2Cu+Au+. The spectroscopic and photophysical proper-
ties of the models Zn, Au+, and Cu+ are collected in Tables 3
(absorption), 4 (emission), and 5 (transient absorption), together
with those of the rotaxanes Zn2-Au+ and Zn2Cu+Au+, and
are in agreement with those of related compounds.27-30 As
shown in Figure 7, the absorption spectrum of Zn2Cu+Au+
closely matches the addition of the separate components. The
same is true for [2]-rotaxane Zn2-Au+. Therefore, there are
few, if any, interactions between the different chromophores in
the ground state of the rotaxanes.
4.1 Luminescence Properties. The emission of the zinc
porphyrin-based singlet excited state is strongly quenched in
rotaxanes Zn2Cu+Au+ and Zn2-Au+ at 293 or 77 K. The RT
quenching process occurs to a rather different extent (Figure 8)
for both compounds. Time-resolved experiments (λexc ) 532
nm) indicate similar lifetimes for the quenched luminescence:
180 ps for Zn2Cu+Au+ and 170 ps for Zn2-Au+, respectively.
In addition, although in the case of Zn2Cu+Au+ the calculated
radiative rate constant (kr ) ΦF/τ ) 4.5 × 106 s-1) is in
reasonable agreement with that of the model Zn (kr ) 5.4 ×
106 s-1), in the case of Zn2-Au+, it is about 30% that of the
model (1.7 × 106 s-1).
λ
max (nm) (ꢀ × 10-5 [M-1 cm-1])
Znb
415 (3.10)
415 (1.86)
425 (0.02)
415 (5.51)
415 (7.83)
543 (0.17)
525 (0.14)
576 (0.07)
560 (0.04)
Au+
Cu+
Zn2-Au+
Zn2Cu+Au+
540 (0.35)
541 (0.41)
574 (0.12)
573 (0.13)
a In acetonitrile, except otherwise specified. b In butyronitrile/aceto-
nitrile (1/1).
redox subunit and, therefore, is the best electron acceptor. As
expected, oxidation of the Zn porphyrin subunits takes place at
0.17 V vs Fc+/0. Oxidation of the central Cu(I) complex is found
at +0.37 vs Fc+/0 for prerotaxane Cu+Au+ and +0.43 V vs
Fc+/0 for [2]-rotaxane Zn2Cu+Au+. These values are 230-290
mV more positive than that of the model compound Cu+.
Examination of the literature data shows that the oxidation
potentials of copper(I) bis-phenanthroline complexes range from
0.045 to 1.175 V vs Fc+/0 (0.42 to 1.55 V vs SCE22), depending
on both the electronic and steric effects of the substituents
anchored to the 1,10-phenanthroline ligand.23-25 In particular,
bulky substituents enforcing a tetrahedral geometry, particularly
2,9-substituents24,25, or electron-withdrawing groups, such as
CF326 raise the Cu(II)/Cu(I) potential by the stabilization of the
Cu(I) redox state. The present data can be explained in this
frame.
This apparent discrepancy can be explained on the basis of
at least two conformations for Zn2-Au+ (Scheme 1): an
extended one, bringing the gold and zinc porphyrin nuclei to a
large distance from one another [Zn2-Au+(l), l standing for
“long”] and a compact conformation, for which these compo-
nents are likely to be in close proximity [Zn2-Au+(s), s
standing for “short”].31 The luminescence results can be
interpreted if an “immediate” (that is, static) quenching of the
zinc porphyrin emission occurs for Zn2-Au+(s) faster than our
time resolution. By contrast, in the case of Zn2Cu+Au+, no
static quenching can be evidenced. From the time-resolved data
and by comparing the emission quantum yields of Zn2Cu+-
Au+ and Zn2-Au+ (Table 4), it turns out that the fraction of
“immediately” quenched molecules of Zn2-Au+, those in the
Zn2-Au+(s) conformation, is ca. 70%. The remaining 30%
correspond to Zn2-Au+(l), whose emission properties are
identical to those of the extended complex, Zn2Cu+Au+.
4.2 Transient Absorption Spectroscopy. Transient absorp-
tion spectra were obtained at the end of a 35-ps laser pulse (532
Inspection of the excited-state redox potentials shows that
the Zn porphyrin excited state (singlet or triplet) can transfer
an electron to the Au(III) porphyrin ground state (∆G ) -0.90
eV and -0.48 eV, respectively), whereas the triplet Au(III)
porphyrin excited state can abstract an electron from the Zn
porphyrin (∆G ) -0.53 eV). The energy levels of the
•
corresponding charge-separated states, namely + Zn2Cu+Au•
and +•Zn2-Au• can be estimated to lie at 1.22 eV. Importantly,
the Au(III) porphyrin in its ground state can also be reduced
by the central Cu(I) complex in its excited state (MLCT in
nature). The resulting charge-separated state, Zn2Cu2+Au•, lies
at the energy level of 1.47 eV. As a consequence of the relatively
high Cu(II/I) redox potential, E(Cu2+/Cu+) > E(Zn+•/Zn);
therefore, the state Zn2Cu2+Au• can be converted into the fully
charge-separated state +•Zn2Cu+Au• by a charge shift from the
Zn porphyrin to the oxidized central Cu(II) complex.
(22) Note that, with few exceptions, most of the redox potentials of
copper(I) bisphenanthroline complexes have been reported vs SCE in CH3-
CN, which makes direct comparisons to the present systems difficult.
However, data vs SCE can be converted vs Fc+/0, taking into account the
fact that the redox potential of the Fc+/0 couple in CH3CN is +0.375 V vs
SCE (J.-M. Kern, personal communication).
(27) Flamigni, L.; Armaroli, N.; Barigelletti, F.; Chambron, J.-C.; Solladie´
N.; Sauvage J.-P. New J. Chem. 1999, 23, 1151-1158.
(28) Flamigni, L.; Barigelletti, F.; Armaroli, N.; Ventura, B.; Collin, J.-
P.; Sauvage, J.-P.; Williams, J. A. G. Inorg. Chem. 1999, 38, 661-667.
(29) Armaroli, N.; De Cola, L.; Balzani, V.; Sauvage, J.-P.; Dietrich-
Bucheker, C. O.; Kern, J.-M.; Bailal, A. J. Chem. Soc., Dalton Trans. 1993,
3241-3247.
(23) Miller, M. T.; Gantzel, P. K.; Karpishin, T. B. Inorg. Chem. 1999,
38, 3414-3422.
(30) Armaroli, N.; Rodgers, M. A. J.; Ceroni, P.; Balzani, V.; Dietrich-
Bucheker, C. O.; Kern, J.-M.; Bailal, A.; Sauvage, J.-P. Chem. Phys. Lett.
1995, 241, 555-558.
(24) Federlin, P.; Kern, J.-M.; Rastegar, A.; Dietrich-Buchecker, C.;
Marnot, P.; Sauvage, J.-P. New J. Chem. 1990, 14, 9-12.
(25) Eggleston, M. K.; McMillin, D. R.; Koenig, K. S.; Pallenberg, A.
J. Inorg. Chem. 1997, 36, 172-176.
(31) The structure of the conformation Zn2-Au+(l) is tentatively
suggested to be similar to that of the array Zn2Cu+Au+, but with a lower
distance between the zinc- and gold-porphyrin groups, which is allowed
by a looser structure than that in Zn2Cu+Au+.
(26) Miller, M. T.; Gantzel, P. K.; Karpishin, T. B. Angew. Chem., Int.
Ed. 1998, 37, 1556-1558.

Multiporphyrinic Rotaxanes
J. Am. Chem. Soc., Vol. 122, No. 48, 2000 11841
Table 4. Emission Properties at 298 Ka and 77 Kb,c
298 K
77 K
λmax (nm)
τ (ns)
ΦF
λmax (nm)
τ (ns)
E(eV)
Znc
1Zn
586
1.84
85
0.01
584
726
718
700
584
710
728
584
710
728
2.5
2.12
1.71
1.73
1.77
2.12
1.75
1.70
2.12
1.75
1.70
3Zn
54 × 106
Au+
3Au+
*Cu+
1Zn
15 000; 105
Cu+
710
586
0.0003
0.0003
Zn2-Au+
0.170(90%)-0.580(10%)e
0.300(70%)-2.2(30%)f
11 000; 105
3Aug
3Zn
20 × 106
Zn2Cu+Au+
1Zn
586
0.180
0.0008
0.630(75%)-2.6(25%)f
15 000; 9 × 104
52 × 106
3Au
3Zn
a In acetonitrile, except otherwise specified. b Butyronitrile. c In the second column, the excited state is specified. d Acetonitrile/butyronitrile
(1/1). e The 580-ps component, which accounts for 3% of the total 1Zn in this system (only 30% of the total Zn is, in fact, detected in this
1
experiment), can be assigned to a minor impurity. f Aggregation phenomena can be responsible for nonexponential decay in polar solvents at 77
K.27 g Weak emission.
Figure 7. Absorption spectra of Zn2Cu+Au+ (-), Zn (‚-‚), Au+
Figure 9. Differential transient absorbance at the end of a 35-ps pulse
(‚‚‚) and Cu+ ()) in acetonitrile, 298 K.
(532 nm, 2.5 mJ) in acetonitrile solutions for (‚‚‚) Au+, A ) 0.11; (- -)
Zn2-Au+, A ) 0.41; (+) Zn2Cu+Au+, A ) 0.43; and (s) Zn, A )
0.21. The sample absorbances were adjusted to have the same number
of photons absorbed by the Zn and Au porphyrin moieties at the
excitation wavelength.32 To achieve dissolution of Zn, butyronitrile was
added, up to 70%.
Figure 8. Fluorescence spectra of (s) Zn, A ) 0.16; (- - -) Zn2-
Au+, A ) 0.18; and (‚‚‚) Zn2Cu+Au+, A ) 0.19 in acetonitrile solutions
at 298 K upon excitation at 545 nm. The sample absorbances were
adjusted to have the same number of photons absorbed by the Zn
porphyrin moiety at the excitation wavelength.
Figure 10. Differential transient absorbance of a Zn2Cu+Au+ solution
(A ) 0.43) in acetonitrile after a 35-ps laser pulse (532 nm, 2.5 mJ).
(O) End of pulse; (-) 500 ps; and (b) 3 ns.
Scheme 1
wavelengths are 466 nm for 1Zn; 616 nm for 3Au+, with some
contribution of Au•,33 and 670 nm for Zn+•.34
In Figure 10 are reported the spectra obtained for Zn2Cu+-
Au+ at the end of the laser pulse, and with delays of 500 ps
and 3 ns. The initial spectrum shows the typical 1Zn and 3Au+
features, the latter being present also in the 500-ps spectrum,
with a prominent band around 630 nm, while the 3-ns spectrum
has maxima at 610 and 670 nm and decays on the nanosecond
time scale.
nm) for both rotaxanes and porphyrin models under comparable
conditions32 and are displayed in Figure 9. Representative
(33) Fajer, J.; Borg, D. C.; Forman, A.; Dolphin, D.; Felton, R. H. J.
Am. Chem. Soc. 1970, 92, 3451-3459.
(34) Abou-Gamra, Z.; Harriman, A.; Neta, P. J. Chem. Soc., Faraday
Trans. 2, 1986, 82, 2337-2350.
(32) At 532 nm, the partition of photons is 60% on zinc porphyrin, 37%
on gold porphyrin and 3% on the Cu+ unit for Zn2Cu+Au+. For Zn2-
Au+, 62% is on zinc porphyrin, and 38% is on gold porphyrin.

11842 J. Am. Chem. Soc., Vol. 122, No. 48, 2000
Linke et al.
Table 5. Transient Absorbance Dataa
state
τ (ns)
8 000
1.5
179e
φ
Znb
3Zn
0.95c
1d
Au+
3Au+
3Cu+
Cu+
Zn2-Au+
1Zn
0.160
3Zn
< 0.05c
n.d.
3Au+
•+Zn-Au•
1Zn
n.d.
5.5
0.170
0.3f
Zn2Cu+Au+
3Zn
1700
0.13c
1c
3Au+
•+ZnCu+Au•
1.5
5
1f
a Oxygen-free acetonitrile for τ > 10 ns. b Oxygen-free acetonitrile/
butyronitrile (1/1). c Triplet-state quantum yields. For the rotaxanes,
these are calculated on the basis of the photons absorbed by Zn or
Au+ porphyrin moiety only. d Assumed to be identical to that of the
gold (III) tetraphenylporphyrin.5b e In deaerated CH2Cl2.30 f Relative
yield (see text).
Figure 11. Transient absorbance decays and fittings for 466, 616, and
670 nm for Zn2Cu+Au+ (a) and Zn2-Au+ (b). The conditions are the
same as for Figures 10 and 13.
Figure 13. Differential transient absorbance of a Zn2-Au+ solution
(A ) 0.41) in acetonitrile after a 35-ps laser pulse (532 nm, 2.5 mJ).
(O) End of pulse spectrum; (b) 190 ps; (-) 530 ps; and (+)3 ns.
Figure 12. Normalized transient absorbance signal (average of 10
shots) detected after a 18-ns laser pulse (532 nm, 2 mJ) in acetonitrile
solutions at λ ) 670 nm for Zn2Cu+Au+ and Zn2-Au+. Curve a is
the instrumental response function.
very similar to the initial spectrum and to the one obtained at
the same delay in the case of Zn2Cu+Au+ (Figure 10).
Therefore, it is assigned to the charge-separated state +•Zn2-
Au•. As shown in Figure 12, no residual absorption due to the
Zn porphyrin-localized triplet can be detected after the CS state
has disappeared. The decays of the transient absorbances at the
three representative wavelengths defined above are reported in
Figure 11b. In all three cases, the overall decay can be fitted
by a double exponential function with lifetimes of 160 ps and
5.5 ns. If one assumes similar absorption coefficients for the
two CS states +•Zn2-Au• and +•Zn2Cu+Au•, comparison of
The kinetic traces at the three representative wavelengths
above are reported in Figure 11a. The decays at 466 and 670
nm can be fitted by the same exponentials, ca. 170 ps and 5 ns.
However, at 670 nm, the shorter component appears as a
formation, that is, with a negative preexponential factor. The
616 nm decay, which also shows a fast component (150 ps), is
dominated by a lifetime of 1.5 ns. On the basis of its spectral
shape and lifetime, it is due to the gold porphyrin-localized
triplet, which is directly formed by the 532-nm excitation.32 The
fast decay (ca. 150-180 ps), which is detected predominantly
around 470 nm, can be assigned to the 1Zn2Cu+Au+ state and
is in reasonable agreement with the fluorescence decay (180
ps). It is coincident with the rate of formation of Zn•+ (670
nm), which then decays with a lifetime on the order of
nanoseconds. The band with a maximum around 610 nm (Au•)
also decays within the same time window. Therefore, we assign
the 3-ns spectrum reported in Figure 10 to the fully charge-
separated state (CS) +•Zn2Cu+Au•. Its decay (ca. 5 ns) is shown
in Figure 12 on a longer time scale. The residual absorbance
Figures 11a and 11b at sufficiently long delay indicates for +•
-
Zn2-Au• a yield which is ca. 30% of that of +•Zn2Cu+Au•.
This strongly suggests that in the free rotaxane the long-lived
CS state is formed in the extended conformer Zn2-Au+(l) only.
Charge separation in the other conformer [Zn2-Au+(s)] is
much faster. The shape of the initial transient absorption spectra
of Zn2-Au+ can be explained by the fast reactivity of the
excited states 1Zn2-Au+(s) and Zn2-3Au+(s) and by the minor
contribution of the states derived from the Zn2-Au+(l) con-
formation because of its lower population (30%). Consequently,
the typical 3Au+ band around 620-630 nm is not so evident as
in the case of Zn2Cu+Au+. The lifetime (160 ps) of the fast
component suggests that it could be assigned to the decay of
1Zn2-Au+(l). However, because the end-of-pulse absorbance
3
after the CS state decay in Figure 12 is due to the Zn2Cu+-
Au+ state, the Zn porphyrin-localized excited triplet (Table 5).
In Figure 13 are reported the spectra recorded for Zn2-Au+
immediately and with delays of 190 ps, 530 ps, and 3 ns after
the laser pulse. These spectra are characterized by prominent
bands at 610 and 670 nm from the very beginning and, in
contrast to the previous case, there is apparently no remarkable
spectral change in the time evolution; the spectrum at 3 ns is
1
is higher than expected from the Zn2-Au+(l) state only, the
existence of an underlying absorption of comparable lifetime,
ascribable to the charge-separated state +•Zn2-Au•(s) must be
considered. Therefore, at least for the initial stages of the time
evolution, some contribution of the transients derived from

Multiporphyrinic Rotaxanes
J. Am. Chem. Soc., Vol. 122, No. 48, 2000 11843
excitation of the zinc porphyrin subunit, 1Zn2-Au+(l) is formed,
followed by electron transfer to the gold porphyrin (∆G ) -0.9
eV), with a rate of 5.3 × 109 s-1. Energy transfer to form Zn2-
3Au+(l) is forbidden for spin multiplicity. Electronic com-
munication between the mechanically linked chromophores is
likely provided by van der Waals interactions between the axle
and the wheel of the rotaxane, as has been observed in donor-
acceptor systems connected by hydrogen bonds39 or in similarly
3
mechanically linked species.40 The yield of Zn2-Au+, the
lowest-energy excited state of the system, is near zero. This
indicates that energy transfer processes are not operative in
Zn2-Au+(l), which supports the electron transfer interpretation.
In the case of Zn2Cu+Au+, sensitization of the Zn2*Cu+-
Au+ state by energy transfer from the zinc porphyrin-localized
singlet excited state (1Zn2Cu+Au+) is likely to occur at first.
The mixed-singlet and -triplet nature of the emitting MLCT state
of the Cu(I) bis-phenanthroline complexes35,36 makes this energy
transfer process allowed by spin selection rules. In addition,
1
the quenching rate of Zn2Cu+Au+ (5 × 109 s-1) is identical
to that observed in a related [3]-rotaxane, for which energy
transfer between the same chromophores was proven unequivo-
cally.27 This process is followed by electron transfer to the
nearby gold porphyrin (∆G ) -0.3 eV) with formation of the
Zn2Cu2+Au• CS state. Subsequently, fast charge shift from a
zinc porphyrin unit (∆G ) -0.25 eV), produces the lowest
energy CS state, +•Zn2Cu+Au•. It was not possible to detect
these consecutive electron-transfer steps within the 20 ps time
resolution. Therefore, the energy transfer step is rate-determining
1
3
in the CS process. Intersystem crossing of Zn2Cu+Au+ to -
Zn2Cu+Au+ is now competitive, because the zinc porphyrin-
localized triplet is formed with a yield close to 0.1.
The CS states +•Zn2Cu+Au• and +•Zn2-Au• decay at a very
similar rate (2 × 108 s-1) to re-form the ground state upon
charge recombination. This is rather surprising because elec-
tronic communication between the oxidized and the reduced
porphyrins is expected to be very different in the two systems.41
This result could be rationalized from a different viewpoint.
It was assumed and confirmed by the data that the Zn2-Au+-
(l) and Zn2-Au+(s) conformations did not interconvert within
the time-scale of forward electron transfer. The same assumption
could not hold true for longer time scales during which the
conformers would be exchanging. In this case, the lifetime of
5.5 ns, which was detected for the charge recombination reaction
in Zn2-Au+, could in fact measure the rate of conversion of
+•Zn2-Au•(l) to +•Zn2-Au•(s).42 The latter CS state would
immediately return to the ground state because of the closer
Figure 14. Schematic energy levels diagrams for Zn2-Au+ (a) and
Zn2Cu+Au+ (b). For the case of Zn2-Au+, only the rate parameters
of the Zn2-Au+(l) conformation are reported.
Zn2-Au+(s), in particular those due to the CS state +•Zn2-
Au•(s), can be envisaged.
5. Photoinduced Electron Transfer. The energy level
diagrams of Zn2-Au+ and Zn2Cu+Au+, using the data of
Tables 2 and 4, are reported in Figure 14. They differ by the
presence, in the case of Zn2Cu+Au+, of a MLCT35,36 excited
state localized at the copper complex fragment (Zn2*Cu+Au+),
and the presence of a low-lying additional charge-separated state,
Zn2Cu2+Au•. This CS state is 0.2 eV higher in energy than the
fully CS state +•Zn2Cu+Au•. Within the weak interaction limit,
the rates of the observed intramolecular processes can be
calculated according to eq 1.
(37) (a) Dexter, D. L. J. Chem. Phys. 1953, 21, 836-850. (b) Fo¨rster,
T. Discuss. Faraday Soc. 1959, 27, 7-17. (c) Closs, G. L.; Johnson, M.
D.; Miller, J. R.; Piotrowiak, P. J. Am. Chem. Soc. 1989, 111, 3751-3753.
(38) (a) Marcus, R. A.; Sutin, N. Biochim. Biophys. Acta 1985, 811,
322-265 and references therein. (b) Closs, G. L.; Miller, J. R. Science
1988, 240, 440-447 and references therein. (c) Piotrowiak, P. Chem. Soc.
ReV. 1999, 28, 143-150.
k ) 1/τ - 1/τ0
(1)
τ and τ0 are the lifetimes of the chromophore unit in the
rotaxanes and in the model, respectively. Photoinduced processes
will be discussed in the frame of current theories of energy37
and electron transfer.38
(39) (a) De Rege, P. J. F.; Williams, S. A.; Therien, M. J. Science 1995,
269, 1409-1413. (b) Roberts, J. A.; Kirby, J. P.; Nocera, D. G. J. Am.
Chem. Soc. 1995, 117, 8051-8052. (c) Ward, M. D.; White, C. M.;
Barigelletti, F.; Armaroli, N.; Calogero, G.; Flamigni, L. Coord. Chem.
ReV. 1998, 171, 481-488.
For Zn2-Au+, the processes can be clearly resolved only in
the case of the extended conformation Zn2-Au+(l). In the case
of the Zn2-Au+(s) conformer, they can not be resolved because
they are faster than the time resolution available (20 ps). Upon
(40) Hu, Y.-Z.; Bossmann, S. H.; Van Loyen, D.; Schwarz, O.; Durr,
H. Chem. Eur. J. 1999, 5, 1267-1277.
(41) The rate of electron transfer, kel, can be expressed38a as kel
)
νe-∆G#/RT, where the preexponential is the so-called “electronic factor” and
the exponential is generally referred to as the “nuclear factor”. The latter
factor is mainly affected by thermodynamic parameters, which is similar
in the two rotaxanes. The electronic factor depends on the overlap between
electronic wave functions of the partners, and the interposed connector is
crucial in determining the degree of overlap.
(35) The spin multiplicity has been omitted in Zn2*Cu+Au+ (the MLCT
excitedstate localized on the copper complex) to underline the mixed nature
of the emitting state. According to current interpretation,36 the emission
3
originates from a 1MLCT in equilibrium at room temperature with a
MLCT excited state.
-
(36) Kirchhoff, J. R.; Gamache, R. E. Jr.; Blaskie, M. W.; Del Paggio,
A. A.; Lengel, R. K.; McMillin, D. R. Inorg. Chem. 1983, 22, 2380-2384.
(42) Deleuze, M. S.; Leigh, D. A.; Zerbetto, F. J. Am. Chem. Soc. 1999,
121, 2364-2379.

11844 J. Am. Chem. Soc., Vol. 122, No. 48, 2000
Linke et al.
) 73 000 M-1cm-1); the singlet depletion method46 was used to
proximity of the oxidized/reduced chromophores. The rate-
determining step of the charge recombination in Zn2-Au+ could
therefore be the interconversion between conformers.
3
3
determine the ∆ꢀ
of Zn, 90 000 M-1cm-1. The yield of Au+ was
470
assumed to be identical to that of the gold(III) tetraphenylporphyrin.28
Yields of formation of the porphyrin triplets (ΦT) in the arrays were
determined by comparing the transient absorbances at 620-650 nm,
the wavelength region out of ground-state absorption, with the
corresponding values of the model compounds Zn and Au+ and by
making corrections for the fraction of light absorbed at 532 nm by the
moiety of interest in the sample. It was assumed that the molar
absorption coefficients of triplets were unaffected in passing from
models to the arrays. For the determination of the yield of
Zn2Cu+3Au+, the absorbance at 620 nm at the end of the pulse was
corrected for the contribution of the charge-separated state.
Emission lifetimes in the micro- to millisecond range were measured
by using as the excitation source a Nd:YAG laser (20-ns pulse) and a
Hamamatsu R936 photomultiplier tube as the detector. The light emitted
was collected at a right angle with respect to the excitation (1-2 mJ at
532 nm), selected by a series of filters, then directly fed into the
photomultiplier, which was coupled to a digital oscilloscope. A
Hamamatsu C1587 streak camera was used as the detector for
fluorescence lifetimes below 2.5 ns. In this case, the excitation source
was the same Nd:YAG laser as was used for absorption (λ ) 532 nm,
ca. 1 mJ, 35-ps pulse). Time and spectral profiles were derived from
the streak image, which resulted from the average of several hundred
shots. The time profiles were analyzed by standard iterative methods
according to a single or a double exponential fitting.
Conclusion.
A detailed comparison of the properties of the free and the
Cu(I)-complexed rotaxanes showed that the role of the Cu
complexation is important in different perspectives: (i) it gives
a geometric constraint which keeps the reacting partners at a
fixed distance; (ii) it connects from an electronic viewpoint the
primary electron donor, one of the two zinc porphyrins, and
the electron acceptor, the gold porphyrin; and (iii) it offers an
energy transfer pathway by means of its MLCT excited state.
The absence of the copper ion in the Zn2-Au+ rotaxane, on
the contrary, gives origin to a loose structure for which the
distance and the mutual orientation of the components can vary,
giving origin to different conformers. From the electronic
perspective, the potential donor and acceptor units in Zn2-Au+
are not connected by chemical bonds, and no alternative energy
1
transfer pathway is offered to the deactivation of Zn2-Au+,
in contrast to the case of 1Zn2Cu+Au+ which can transfer energy
to the MLCT excited-state localized on the copper complex.
Experimental Section
The energies of the electronic levels of the various compounds were
derived from the maxima of the luminescence bands at 77 K.
Experimental uncertainties are estimated to be within 8% for lifetime
determination, 15% for quantum yields, 20% for molar absorption
coefficients, and 3 nm for emission and absorption peaks. The working
temperature was 298 K, unless otherwise stated.
Spectroscopic and Photophysical Measurements. The solvent used
was acetonitrile (C. Erba, spectroscopic grade). Addition of butyronitrile
(Fluka) up to 50-70% was necessary to dissolve the porphyrin Zn at
the desired concentrations. Absorption spectra were recorded using a
Perkin-Elmer Lambda 9 spectrophotometer. Uncorrected emission
spectra were detected by a Spex Fluorolog II spectrofluorimeter
equipped with a Hamamatsu R-928 photomultiplier tube. The 77 K
phosphorescence spectra of the porphyrin triplets were detected by the
same spectrofluorimeter equipped with a phosphorimeter accessory
(1934D, Spex). Parameters to detect 3Zn phosphorescence were
excitation at 550 nm and gate open for 70 ms, with 1 ms delay with
respect to excitation. Parameters used to detect 3Au+ phosphorescence
were excitation at 525 nm and gate open for 200 µs, with a delay of
20 µs with respect to the excitation. Experiments at 77 K were
performed on samples contained in quartz capillary tubes immersed in
a homemade quartz dewar filled with liquid nitrogen. The fluorescence
quantum yield of Zn, determined by the method of Demas and Crosby43
using [Ru(bpy)3]Cl2 in aerated water as standard (Φ ) 2.8 × 10-2),
was 0.01.
Acknowledgment. We warmly thank Vincent Semetey for
his initial contribution to the synthesis of compounds 2-4, and
9 and MALDI-TOF mass spectrometry measurements. We are
grateful to Jean-Daniel Sauer for the high-field NMR spectra,
and Raymond Hueber for the FAB mass spectra. We thank M.
Minghetti for technical assistance and A. Martelli for electronic
support. M.L. thanks the French Ministry of Education, Re-
search, and Technology for a fellowship. We thank the European
Communities for support through TMR contracts FMRX-CT96-
0031 (M.L.) and FMRX-CT98-0226 (S.E.). This research was
funded by CNR (Italy) and CNRS (France).
Supporting Information Available: Preparation of com-
pounds 2 - 6, 9 - 11, [Au+](PF6-), [Cu+Au+](PF6-)2,
[(H2)2Cu+Au+](PF6-)2, [Zn2Cu+Au+](PF6-)2, [Zn2-Au+]-
(PF6-), [Zn2Ag+Au+](PF6-)2, and [Zn2Li+Au+](PF6-)2; elec-
trochemistry. This material is available free of charge via the
Internet at http://pubs.acs.org.
Nanosecond flash photolysis studies were performed using a Nd:
YAG laser (532 nm, 1-2 mJ per pulse). Samples were deaerated by
bubbling argon in home-modified quartz cuvettes. Optical densities of
solutions of the arrays for this experiments were 0.4-1. Transient
absorption spectra in the picosecond time domain were measured by a
pump and probe system. The second harmonic (532 nm) of a Nd:YAG
laser (Continuum PY62-10) with a 35-ps pulse was used to excite the
samples with an energy of 2-3 mJ. Optical density of the samples at
the excitation wavelength was kept between 0.1 and 0.5. Details on
the nanosecond laser flash photolysis and the picosecond pump-probe
system can be found elsewhere.44
JA002018F
(44) (a) Flamigni, L. J. Phys. Chem. 1992, 96, 3331-3337. (b) Flamigni,
L. J. Chem. Soc., Faraday Trans. 1994, 90, 2331-2336. (c) Flamigni, L.;
Armaroli, N.; Barigelletti, F.; Balzani, V.; Collin, J.-P.; Dalbavie, J.-O.;
Heitz, V.; Sauvage, J.-P. J. Phys. Chem. B 1997, 101, 5936-5943.
(45) Tanielian, C.; Wolff, C. J. Phys. Chem. 1995, 99, 9825-9830.
(46) Bensasson, R. V.; Land, E. J.; Truscott, T. G. Flash Photolysis and
Pulse Radiolysis; Pergamon Press: Oxford, 1983; pp 11-14.
3
The yield of Zn was determined by a comparative method using
zinc tetraphenylporphyrin in benzene as standard45 (ΦT ) 0.81; ∆ꢀ 470
(43) Demas, J. N.; Crosby, G. A. J. Phys. Chem. 1971, 75, 991-1024.