The slipping approach to self-assembling [n]rotaxanes

Article abstract of DOI:10.1021/ja961817o

A synthetic approach - namely slippage - to self-assembling [n]rotaxanes incorporating π-electron deficient bipyridinium-based dumbbell-shaped components and π-electron-rich hydroquinone- and/or dioxynaphthalene-based macrocyclic polyether components has been developed. The kinetics of rotaxane formation by the slipping procedure were investigated by absorption UV-visible and ¹H-NMR spectroscopies in a range of temperatures and solvents, varying systematically the size of both the stoppers and the macrocyclic components. As expected, the rate constants for these processes are affected by the size complementarity between macrocycles and stoppers. Furthermore, the enthalpic and entropic contributions to the free energies of activation associated with the slippage and the effect of solvent polarity upon the outcome of these processes have been evaluated. In addition, the spectroscopic and electrochemical properties of some of the rotaxanes are presented and discussed with reference to the properties of their chromophoric and electroactive units.

Full text of DOI:10.1021/ja961817o

302
J. Am. Chem. Soc. 1997, 119, 302-310
The Slipping Approach to Self-Assembling [n]Rotaxanes†
Masumi Asakawa,‡ Peter R. Ashton,‡ Roberto Ballardini,§ Vincenzo Balzani,
Martin Beˇlohradsky´,‡ Maria Teresa Gandolfi, Oldrich Kocian,‡ Luca Prodi,
Franc¸isco M. Raymo,‡ J. Fraser Stoddart,*,‡ and Margherita Venturi
Contribution from The School of Chemistry, The UniVersity of Birmingham, Edgbaston,
Birmingham B15 2TT, UK, Istituto FRAE-CNR, Via Gobetti 101, I-40129 Bologna, Italy, and
Dipartimento di Chimica “G. Ciamician” dell’UniVersita`, Via Selmi 2, I-40126 Bologna, Italy
ReceiVed May 30, 1996. ReVised Manuscript ReceiVed October 29, 1996X
Abstract: A synthetic approachsnamely slippagesto self-assembling [n]rotaxanes incorporating π-electron deficient
bipyridinium-based dumbbell-shaped components and π-electron-rich hydroquinone- and/or dioxynaphthalene-based
macrocyclic polyether components has been developed. The kinetics of rotaxane formation by the slipping procedure
1
were investigated by absorption UV-visible and H-NMR spectroscopies in a range of temperatures and solvents,
varying systematically the size of both the stoppers and the macrocyclic components. As expected, the rate constants
for these processes are affected by the size complementarity between macrocycles and stoppers. Furthermore, the
enthalpic and entropic contributions to the free energies of activation associated with the slippage and the effect of
solvent polarity upon the outcome of these processes have been evaluated. In addition, the spectroscopic and
electrochemical properties of some of the rotaxanes are presented and discussed with reference to the properties of
their chromophoric and electroactive units.
Introduction
The slippage approach to self-assembling1 rotaxanes2 is
illustrated schematically in Figure 1. The macrocycle M and
the dumbbell-shaped compound D are formed separately and
then mixed in solution. In order to self-assemble the rotaxane
R, slipping-on of the macrocycle M over the stoppers of D is
required. As a result, the energy barrier ∆Gq has to be
on
overcome. Thus, heating of the solution at an appropriate
temperature T is required. An equilibrium between the newly
formed rotaxane R and the free components M and D is
therefore achieved at T. By contrast, the noncovalent bonding
interactions existing between the dumbbell-shaped and macro-
cyclic components within the rotaxane R make the energy
barrier ∆Gq of the slipping-off process higher than that of
off
the slipping-on pushing the equilibrium toward R [∆Gq
>
off
∆Gqon; the difference between these two values corresponds to
the binding energy ∆G° (∆Gqoff - ∆Gqon ) ∆G°)]. On cooling
the solution down, the energy barrier for both slipping-on and
slipping-off processes cannot be overcome, i.e. ideally, no
interconversion occurs between the rotaxane R and free
components M and D and Vice Versa. As a result, R becomes
kinetically stable and its separation from M and D can be
achieved employing the usual chromatographic techniques.
Figure 1. Schematic illustration of the slipping approach to rotaxanes.
In previous research,3 we have demonstrated the efficiency
of the slipping approach by self-assembling a number of linear
and branched [2]rotaxanes, [3]rotaxanes, and [4]rotaxanes
incorporating π-electron rich hydroquinone-based macrocyclic
polyethers and π-electron deficient bipyridinium-based dumb-
bell-shaped components. In particular, the dumbbell-shaped
compounds H-D, Me-D, Et-D, and i-Pr-D (Scheme 1), incor-
porating one bipyridinium recognition site and bis(4-tert-
butylphenyl)-4-R-phenylmethane-based stoppers in which the
size of the R groups was varied systematically, were
synthesized.3a,e Heating MeCN solutions of each dumbbell-
†Molecular Meccano, Part 12. For Part 11, see: Asakawa, M.; Ashton,
P. R.; Boyd, S. E.; Brown, C. L.; Gillard, R. E.; Kocian, O.; Raymo, F.
M.; Stoddart, J. F.; Tolley, M. S.; White, A. J. P.; Williams, D. J. J. Org.
Chem. In press.
‡ University of Birmingham.
§ FRAE-CNR.
University of Bologna.
(3) (a) Ashton, P. R.; Beˇlohradsky´, M.; Philp, D.; Stoddart, J. F. J. Chem.
Soc., Chem. Commun. 1993, 1269-1274. (b) Ashton, P. R.; Beˇlohradsky´,
M.; Philp, D.; Spencer, N.; Stoddart, J. F. J. Chem. Soc., Chem. Commun.
1993, 1274-1277. (c) Amabilino, D. B.; Ashton, P. R.; Beˇlohradsky´, M.;
Raymo, F. M.; Stoddart, J. F. J. Chem. Soc., Chem. Commun. 1995, 751-
753. (d) Beˇlohradsky´, M.; Philp, D.; Raymo, F. M.; Stoddart, J. F. In
Organic ReactiVity: Physical and Biological Aspects; Golding, B. T.,
Griffin, R. J., Maskill, H., Eds.; RSC Special Publication No. 148:
Cambridge, 1995; pp 387-398. (e) Ashton, P. R.; Ballardini, R.; Balzani,
V.; Beˇlohradsky´, M.; Gandolfi, M. T.; Philp, D.; Prodi, L.; Raymo, F. M.;
Reddington, M. V.; Spencer, N.; Stoddart, J. F.; Venturi, M.; Williams, D.
J. J. Am. Chem. Soc. 1996, 118, 4931-4951.
X Abstract published in AdVance ACS Abstracts, December 15, 1996.
(1) (a) Whitesides, G. M.; Simanek, E. R.; Mathias, J. P.; Seto, C. T.;
Chin, D. N.; Mammen, M.; Gordon, D. M. Acc. Chem. Res. 1995, 28, 37-
44. (b) Lawrence, D. S.; Jiang, T.; Levett, M. Chem. ReV. 1995, 95, 2229-
2260. (c) Raymo, F. M.; Stoddart, J. F. Curr. Op. Coll. Interf. Sci. 1996, 1,
116-126. (d) Philp, D.; Stoddart, J. F. Angew. Chem., Int. Ed. Engl. 1996,
35, 1154-1196.
(2) (a) Gibson, H. W.; Bheda, M. C.; Engen, P. T. Prog. Polym. Sci.
1994, 19, 843-945. (b) Amabilino, D. B.; Stoddart, J. F. Chem. ReV. 1995,
95, 2725-2828. (c) Beˇlohradsky´, M.; Raymo, F. M.; Stoddart, J. F. Collect.
Czech. Chem. Commun. 1996, 61, 1-43.
S0002-7863(96)01817-3 CCC: $14.00 © 1997 American Chemical Society

The Slipping Approach to Self-Assembling [n]Rotaxanes
J. Am. Chem. Soc., Vol. 119, No. 2, 1997 303
Scheme 1
Scheme 2
shaped compound at 55 °C with 4 molar equiv of the
hydroquinone-based macrocycle HQ-HQ affords3a,e the corre-
sponding [2]rotaxanes in good yields when R is equal to H,
Me, and Et: no rotaxane, however, is isolated when R is equal
should be lowered. Indeed, on heating (Scheme 2) a MeCN
solution of the 1,5-dioxynaphthalene-based macrocycle 1/5DN-
1/5DN and the dumbbell-shaped compound i-Pr-D at 55 °C
during 24 h, the [2]rotaxane 1 self-assembles5 in a yield of 57%.
Hence, although the “i-Pr-substituted” stoppers are bulky enough
to prevent the slipping-on of the hydroquinone-based macrocycle
HQ-HQ, the passage over them of the larger 1,5-dioxynaph-
thalene-based macrocycle 1/5DN-1/5DN can occur under oth-
erwise identical conditions. Furthermore, the slipping-on
(Scheme 1) of HQ-HQ over the stoppers of the dumbbell-
shaped components H-D, Me-D, and Et-D requires ca. 10 days
for completion and the yields of the corresponding [2]rotaxanes
are 52, 45, and 47%, respectively. On the contrary, the self-
assembly of the [2]rotaxane 1, incorporating 1/5DN-1/5DN as
the macrocyclic component, is achieved in a yield of 57% after
only 1 day.
to i-Pr. Thus, the energy barrier ∆Gq for the slipping-on of
on
HQ-HQ becomes insurmountable at 55 °C on going from Et
to i-Pr substitution. In the wake of these results, we decided to
investigate, in some detail, the kinetics of the slipping approach
to rotaxanes by varying systematically the size of both macro-
cyclic (compound) component and the stoppers at the ends of
the dumbbell-shaped (compound) component. We report here
the results of these kinetic studies, as well as the syntheses,
spectroscopic, and electrochemical properties of rotaxanes
incorporating bipyridinium-based dumbbell-shaped components
and hydroquinone-based and/or dioxynaphthalene-based mac-
rocyclic polyether components.
The “i-Pr-substituted” stoppers are too bulky for the slipping-
on of the HQ-HQ macrocyclesi.e. ∆Gq and, as a result,
Results and Discussion
on
∆Gqoff are too high. However, these stoppers can be employed
to self-assemble rotaxanes, incorporating HQ-HQ as the mac-
rocyclic component, by employing the threading methodology.3e
Reaction (Scheme 3) of the bis(hexafluorophosphate) salt
[BPYPYXY][PF6]2 with the stoppers i-Pr-S in the presence of
only 1.1 molar equiv of HQ-HQ affords5 the [2]rotaxane 2 in
a yield of 19%, after counterion exchange. As a result of the
size of the “i-Pr-substituted” stoppers incorporated within the
Synthesis.4 The energy barrier ∆Gq associated with the
on
slipping-on (Scheme 1) of the hydroquinone-based macrocyclic
polyether HQ-HQ over the stoppers of the dumbbell-shaped
compound i-Pr-D is high at 55 °Csi.e. the “i-Pr-substituted”
stoppers are too bulky for the small hydroquinone-based
macrocycle. Thus, passage of HQ-HQ over the bis(4-tert-
butylphenyl)-4-isopropylphenylmethane-based stoppers is not
possible at this temperature and the corresponding [2]rotaxane
is not formed. However, we reasoned that, since replacement
of both of the hydroquinone rings incorporated within the
macrocyclic polyether by 1,5-dioxynaphthalene units enlarges
[2]rotaxane 2, neither the slipping-on (∆Gq too high) of a
on
second macrocyclic HQ-HQ component nor the slipping-off
(∆Gqoff too high) of the HQ-HQ macrocyclic component already
trapped on the dumbbell-shaped component of 2 can be
achieved under the usual conditions employed to induce
slippingsnamely, 55 °C at ambient pressure. Thus, a 1,5-
dioxynaphthalene-based 1/5DN-1/5DN macrocycle can be slipped-
on over the stoppers of 2, leaving, in place on the dumbbell-
shaped component, the HQ-HQ macrocycle already incorporated
within the [2]rotaxane 2 by employing the usual slippage
conditions. By heating (Scheme 3) a MeCN solution of the
[2]rotaxane 2 in the presence of 3 molar equiv of the 1/5DN-
1/5DN macrocycle at 55 °C during 48 h, the [3]rotaxane 3
incorporating two constitutionally different macrocyclic com-
ponents is self-assembled5 in a yield of 49%.
the size of the cavity of the macrocycle, the energy barrier ∆Gq
on
(4) The compounds bis[4-(4-pyridyl)pyridinium]-p-xylene bis(hexafluo-
rophosphate) and 4,4′-dimethylbipyridinium bis(hexafluorophosphate)s
commonly known as paraquatsare abbreviated to [BPYPYXY][PF6]2
(Scheme 3) and [PQT][PF6]2 (Figure 2), respectively. The compounds
methoxybenzene, 1,4-dimethoxybenzene, 1,5-dimethoxynaphthalene, and
2,7-dimethoxynaphthalene are abbreviated to MB, 1/4DMB, 1/5DMN, and
2/7DMN (Figures 4 and 5), respectively. The acronyms employed to
indicate the dumbbell-shaped compounds depicted in Scheme 1 and Figure
3 are composed of the common alphabetical notation corresponding to the
R groups attached to the stoppers followed by the letter D which stands for
dumbbell. The bis(tert-butylphenyl)-4-isopropylphenylmethane-based chlo-
ride shown in Scheme 3 and the tetraarylmethane-based model compound
shown in Figure 5 are abbreviated to i-Pr-S and R-S where S stands for
stopper. The acronyms, corresponding to the macrocyclic polyethers depicted
in Figure 2, indicate the nature of the two aromatic units incorporated into
each macrocycle. Thus, HQ, 1/5DN, and 2/7DN stand for hydroquinone,
1,5-dioxynaphthalene, 2,7-dioxynaphthalene, and so on. Numbers are
employed for the remaining compounds.
1H-NMR Spectroscopy. The chemical shifts (δ) for the
protons in the R- and â-positions with respect to the nitrogen
(5) Amabilino, D. B.; Ashton, P. R.; Beˇlohradsky´, M.; Raymo, F. M.;
Stoddart, J. F. J. Chem. Soc., Chem. Commun. 1995, 747-750.

304 J. Am. Chem. Soc., Vol. 119, No. 2, 1997
Asakawa et al.
Table 1. Chemical Shiftsa for the Protons in the R- and
â-Positions with respect to the Nitrogen Atoms on the Bipyridinium
Units Incorporated within the Dumbbell-Shaped Compound i-Pr-D
and the Rotaxanes 1, 2, and 3
Scheme 3
compd
R-CH (δ)
â-CH (δ)
i-Pr-D
9.46
9.02
9.32, 9.27
8.79
7.38
8.47, 8.44
1
2b
3c
9.16, 9.09, 9.05, 9.00
8.14-8.07, 7.39, 7.33
a The 1H-NMR spectra were recorded at room temperature in
1
CD3COCD3 on a Bruker AC300 (300 MHz) spectrometer. b The H-
NMR spectrum of 2 shows two pairs of doublets for the R-CH and
1
â-CH protons. c The H-NMR spectrum of 3 shows four doublets for
the R-CH protons and one multiplet and two doublets for the â-CH
protons.
the shielding effect exerted by the two naphthalene rings
sandwiching the bipyridinium unit within the [2]rotaxane
structure. The 1H-NMR spectrum of a CD3COCD3 solution of
the [2]rotaxane 2 incorporating one HQ-HQ macrocycle, but
having two bipyridinium units within the dumbbell-shaped
component, shows only two doublets centered on δ 9.32 and
9.27 for the R-protons and two doublets centered on δ 8.47
and 8.44 for the â-protons when recorded at room temperature.
The simplicity of the spectrum is a result of the shuttling6 of
the macrocyclic component back and forth from one bipyrid-
inium recognition site to the other. This process is fast on the
1H-NMR time scale at room temperature. Thus, the signals of
the two bipyridinium stations are averaged and distinction
between occupied and unoccupied stations is not possible at
this temperature. Slipping-on of the 1/5DN-1/5DN macrocycle
over the stoppers of the [2]rotaxane 2 affords the [3]rotaxane 3
in which each bipyridinium unit is encircled by two constitu-
tionally different macrocyclic components. As a result, the 1H-
NMR spectrum of the [3]rotaxane 3 shows four doublets
centered on δ 9.16, 9.09, 9.05, and 9.00 for the R-protons, and
one multiplet at δ 8.14-8.07 and two doublets centered on δ
7.39 and 7.33 for the â-protons, of the bipyridinium units.
Kinetic Studies. In order to gain further insight into the
slipping approach to rotaxanes, it was decided to investigate
the kinetics of these processes, changing systematically not only
the size of the stoppers but also the size of the cavity of the
macrocyclic components. The macrocyclic polyethers, depicted
in Figure 2, incorporate π-electron-rich aromatic rings such as
hydroquinone, 1,5-dioxynaphthalene, and/or 2,7-dioxynaphtha-
lene and are able to bind7 in solution the bipyridinium-based
bis(hexafluorophosphate) salt [PQT][PF6]2 with pseudorotax-
ane-like geometries. The association constants (Table 2) in
Me2CO at 25 °C range from 414 up to 1190 M-1 and were
evaluated7 by absorption UV-visible spectroscopy, employing
atoms on the bipyridinium units incorporated within the
dumbbell-shaped compound i-Pr-D and the rotaxanes 1, 2, and
3 are listed in Table 1. The 1H-NMR spectrum of a CD3COCD3
solution of the dumbbell-shaped compound i-Pr-D, recorded
at room temperature, shows two doublets centered on δ 9.46
and 8.79 for the R- and â-protons, respectively. After the
slipping-on of the 1,5-dioxynaphthalene 1/5DN-1/5DN macro-
cycle over the stoppers of the dumbbell-shaped compound i-Pr-
D, the 1H-NMR spectrum of the resulting [2]rotaxane 1,
recorded in CD3COCD3 at room temperature, shows dramatic
upfield shifts for both doublets. The resonance associated with
the R-protons moves to δ 9.02, while the signal for the â-protons
is shifted to δ 7.38. The expected upfield shift is a result of
(6) In previous investigations, a free energy of activation (∆Gq) of ca.
10 kcal mol-1 at 215 K for the degenerate site-exchange processsnamely,
the shuttling of the macrocyclic component from one bipyridinium
recognition site to the othersoccurring within [2]rotaxanes incorporating
either tris(4-tert-butylphenylmethyl groups (R ) t-Bu) or phenylbis(4-tert-
butylphenyl)methyl groups (R ) H) as the stoppers was derived by the
coalescence method. See refs 3b,e and: Ashton, P. R.; Philp, D.; Spencer,
N.; Stoddart, J.F. J. Chem. Soc., Chem. Commun. 1992, 1124-1128.
(7) (a) Allwood, B. L.; Spencer, N.; Shahriari-Zavareh, H.; Stoddart, J.
F.; Williams, D. J. J. Chem. Soc., Chem. Commun. 1987, 1064-1066. (b)
Ashton, P. R.; Chrystal, E. J. T.; Mathias, J. P.; Parry, K. P.; Slawin, A. M.
Z.; Spencer, N.; Stoddart, J. F.; Williams, D. J. Tetrahedron Lett. 1987,
28, 6367-6370. (c) Ortholand, J. Y.; Slawin, A. M. Z.; Spencer, N.;
Stoddart, J. F.; Williams, D. J. Angew. Chem., Int. Ed. Engl. 1989, 28,
1394-1397. (d) Anelli, P. L.; Ashton, P. R.; Ballardini, R.; Balzani, V.;
Delgado, M.; Gandolfi, M. T.; Goodnow, T. T.; Kaifer, A. E.; Philp, D.;
Pietraszkiewicz, M.; Prodi, L.; Reddington, M. V.; Slawin, A. M. Z.;
Spencer, N.; Stoddart, J. F.; Vicent, C.; Williams, D. J. J. Am. Chem. Soc.
1992, 114, 193-218. (e) Asakawa, M.; Ashton, P. R.; Boyd, S. E.; Brown,
C. L.; Gillard, R. E.; Kocian, O.; Raymo, F. M; Stoddart, J. F.; Tolley, M.
S.; Williams, D. J. J. Org. Chem. In press.
Figure 2. Complexation of [PQT][PF6]2 by the π-electron-rich
macrocyclic polyethers HQ-HQ, HQ-1/5DN, HQ-2/7DN, 1/5DN-
1/5DN, 1/5DN-2/7DN, and 2/7DN-2/7DN.

The Slipping Approach to Self-Assembling [n]Rotaxanes
J. Am. Chem. Soc., Vol. 119, No. 2, 1997 305
Table 2. Association Constants (Ka) and Free Energies of
Complexation (-∆G°) for the Binding of [PQT]PF6]2 by
Macrocyclic Polyethers Incorporating π-Electron Rich Aromatic
Units such as Hydroquinone, 1,5-Dioxynaphthalene, and/or
2,7-Dioxynaphthalene
compd
HQ-HQ
HQ-1/5DN
Ka (M-1
)
-∆G° (kcal mol-1
)
λmaxa (nm)
730
919
414
1190
852
970
3.90
4.04
3.57
4.19
4.00
4.07
435
456
416
480
436
418
HQ-2/7DN
1/5DN-1/5DN
1/5DN-2/7DN
2/7DN-2/7DN
a Wavelength of the maximum of the charge-transfer band associated
with the interaction between the π-electron-deficient bipyridinium unit
and the π-electron-rich aromatic rings incorporated within the macro-
cyclic polyethers.
the charge-transfer band arising from the interaction between
the π-electron-deficient bipyridinium unit and the π-electron-
rich aromatic rings incorporated within the macrocycles.
In order to test the ability of these macrocyclic polyethers to
slip-on over bis(4-tert-butylphenyl)-4-R-phenylmethane-based
stoppers, each of the dumbbell-shaped compounds (Figure 3)
Et-D, i-Pr-D, cyclohexyl-D, and t-Bu-D was mixed in CD3CN
solution with equimolar amounts of each macrocycle. The
resulting solutions were heated at 50 °C and the self-assembly
of the corresponding [2]rotaxanes was followed by monitoring
the increasing absorbance A of the charge-transfer band associ-
ated with the newly formed [2]rotaxanes in the visible spectra.
However, a constant value of A is achieved after a period of
time, i.e. equilibrium between the [2]rotaxane and the free
macrocyclic and dumbbell-shaped components is reached.
Nonlinear curve fitting of the plot of A against time t can be
achieved by employing eq 1:8
Figure 3. Slipping processes involving the π-electron rich macrocyclic
polyethers HQ-HQ, HQ-1/5DN, HQ-2/7DN, 1/5DN-1/5DN, 1/5DN-
2/7DN, and 2/7DN-2/7DN and the π-electron-deficient dumbbell-
shaped compounds Et-D, i-Pr-D, cyclohexyl-D, and t-Bu-D.
Table 3. Free Energies (kcal mol-1) of Activation (∆Gq and
on
∆Gqoff) and of Association (∆G°) for the Slipping of the π-Electron
Rich Macrocyclic Polyethers over the Stoppers of the
π-Electron-Deficient Dumbbell-Shaped Compounds in CD3CN at
50 °C
macrocycles
HQ- HQ- 1/5DN- HQ- 1/5DN-
dumbbells parameters HQa 1/5DN 1/5DN a 2/7DN 2/7DN
b
Et-D
∆Gq
24.5 22.4
26.1 25.5
21.3
24.9
3.6
23.9
28.6
4.7
e
20.7
22.9
2.2
24.5
26.6
2.1
e
f
f
f
on
c
∆Gq
off
(C02 - Ce2)
-∆G° d 1.6
3.1
e
2
t
ꢀ C0 Ce 1 - ekon
i-Pr-D
∆Gq
∆Gq
e
e
e
e
e
e
21.8
24.4
2.6
24.2
26.3
2.1
b
on
(
)
[
]
Ce
c
e
e
off
A )
(1)
-∆G° d
(C02 - Ce2)
b
cyclohexyl-D ∆Gq
∆Gq
e
e
e
Ce2 - C0 ekon
2
t
on
c
e
e
e
e
off
Ce
-∆G° d
a The molar extinction coefficient (ꢀ) values measured for one of
the [2]rotaxanes depicted in Scheme 1 and the [2]rotaxane 1 in CD3CN
at 50 °C were employed to calculate the kinetic and thermodynamic
parameters for the slipping of HQ-HQ and 1/5DN-1/5DN, respectively,
over the stoppers of the dumbbell-shaped compounds under these
conditions. b Free energy of activation for the slipping-on process
(percentage error e 0.4%). c Free energy of activation for the slipping-
off process (percentage error e 1%). d Free energy of association
(percentage error e 5%). e No rotaxane formation occurssi.e., no
charge-transfer band is detected. When the dumbbell-shaped compound
t-Bu-D is employed in combination with the dumbbell-shaped com-
pounds Et-D, i-Pr-D, and cyclohexyl-D, no rotaxane formation occurs.
f The equilibrium was reached immediately. When the macrocyclic
polyether 2/7DN-2/7DN is employed in combination with any of the
dumbbell-shaped components, the equilibrium is reached immediately.
where A, ꢀ, and t are absorbance, molar extinction coefficient,
and time, respectively, C0 is the initial concentration of the
species, while Ce is the concentration of the [2]rotaxane at
equilibrium, and kon (Figure 1) is the rate constant of the
slipping-on process. A and t are experimentally measurable,
while C0 is known. Ce can be determined9 by recording the
1H-NMR spectra of the solutions at the same temperature. The
exchange between the free species and the [2]rotaxane is slow
1
on the H-NMR time scale. Thus, distinct resonances for the
free species and the [2]rotaxane are observed in the spectra.
Integration of these signals affords the Ce value and, as a result,
ꢀ can be determined (Ae ) ꢀlCe, Ae is the absorbance at
equilibrium and l is the optical path) by measuring the
absorbance of the charge-transfer bands at saturationsi.e. at
equilibrium. Hence, the only unknown is the rate constant kon
which can be calculated by using a nonlinear curve fitting
program. By employing this methodology, the free energies
of activation (∆Gqon and ∆Gqoff) and of association (∆G°) listed
in Table 3 for the slipping processes in CD3CN at 50 °C,
involving the dumbbell-shaped compounds Et-D, i-Pr-D, cy-
clohexyl-D, and t-Bu-D, were determined. As expected, in the
case of Et-D, the free energies of activation ∆Gq for the
on
(8) Equation 1 was determined as reported in the Supporting Information.
For further information, see: Laidler, K. J. Chemical Kinetics; Harper
Collins Publisher: New York, 1987.
slipping-on process decrease by ca. 3.8 kcal mol-1 on going
from the small HQ-HQ to the large 1/5DN-1/5DN macrocycle
(from left to right in Table 3). However, when the even larger
macrocyclic polyether 1/5DN-2/7DN is employed the equilib-
rium is reached immediately. In the case of i-Pr-D, no slipping-
on occurssi.e. no charge-transfer band is detectedswhen the
HQ-HQ and HQ-1/5DN macrocycles are employed, suggesting
that the size of the cavities of these two macrocycles is too
(9) When the macrocyclic polyether HQ-HQ or 1/5DN-1/5DN is
employed, Ce can be calculated from the equation Ae ) ꢀlCe, where l is the
optical path and is equal to 1 cm, by measuring the absorbance at the
equilibrium Ae and the molar extinction coefficient ꢀ of one of the
[2]rotaxanes shown in Scheme 1 or the [2]rotaxane 1, respectively, under
the same conditions of the experiment. The value of ꢀ is not affected by
the nature of the R groups linked to the tetraarylmethane-based stoppers.
See ref 3e.

306 J. Am. Chem. Soc., Vol. 119, No. 2, 1997
Asakawa et al.
small for their passage over the “i-Pr-substituted” stoppers.
However, when the larger macrocycles HQ-2/7DN, 1/5DN-
1/5DN, and 1/5DN-2/7DN are employed, ∆Gqon values of 23.9,
24.5, and 21.8 kcal mol-1, respectively, are calculated. In the
case of cyclohexyl-D, only the macrocycle 1/5DN-2/7DN is
able to slip-on over the “cyclohexyl-substituted” stoppers. No
charge-transfer band was detected in the case of the smaller
macrocycles. When the dumbbell-shaped compound t-Bu-D
is employed, no charge-transfer band is detected for any of the
macrocycles HQ-HQ, HQ-1/5DN, 1/5DN-1/5DN, and 1/5DN-
2/7DN, suggesting that the “t-Bu-substituted” stoppers are too
bulky to permit the passage of these macrocycles over them at
50 °C in CD3CN. However, the macrocyclic polyether 2/7DN-
2/7DN possesses a cavity larger than that of any of the other
macrocycles. Thus, when 2/7DN-2/7DN is heated at 50 °C with
any of the dumbbell-shaped compounds Et-D, i-Pr-D, cyclo-
hexyl-D, and t-Bu-D, a charge-transfer band is detected and
no change of its absorbance is observed with time, suggesting
that the equilibrium is reached immediately in all cases. These
results suggest that the energy barrier ∆Gqon (Figure 1) for the
slipping-on process is related to the size of the cavity of the
macrocyclic polyether. Thus, the size and the substitution
pattern of the two aromatic rings incorporated in each macro-
cycle should be capable of correlation to ∆Gqon. The C-C
distances between the carbon atoms bearing the methoxy
substituents and the O-O distances for the model compounds
1,4-dimethoxybenzene (1/4DMB), 1,5-dimethoxynaphthalene
(1/5DMN), and 2,7-dimethoxynaphthalene (2/7DMN) are listed
in Figure 4 and were determined by computer molecular
modeling studies. Assuming that these values do not differ
significantly for the dioxy-substituted aromatic rings incorpo-
rated in the macrocyclic polyethers, a parameter XC-C can be
defined for each macrocycle as the sum of the two C-C distance
values for its two aromatic rings. Similarly, XO-O can be
defined as the sum of the two O-O distance values for the two
dioxy-substituted aromatic rings of each macrocycle. A plot
Figure 4. Correlation between the free energy of activation (∆Gq
)
on
associated with the slipping-on processes and the distance parameters
(a) XC-C and (b) XO-O
.
of the ∆Gq values listed in Table 3 for the slipping-on over
on
Table 4. Enthalpic and Entropic Contributions to the Free
the stoppers of the dumbbell-shaped compound Et-D against
XC-C is illustrated in Figure 4a, while Figure 4b shows a plot
Energies (kcal mol-1) of Activation (∆Gq and ∆Gqoff) and of
on
Association (∆G°) for the Slipping of the Macrocyclic Polyether
1/5DN-1/5DNa over the Stoppers of the Dumbbell-Shaped
Compound Et-D in CD3CN at Various Temperatures
of the same ∆Gq values against XO-O
. In both cases,
on
monotonic correlations between the energy barrier ∆Gq and
on
the distance parameters XC-C and XO-O are observed.
parameters
30 °C
40 °C
50 °C
60 °C
In order to investigate the effect of temperature T upon these
processes, the slipping of the macrocycle 1/5DN-1/5DN over
the stoppers of the dumbbell-shaped compound Et-D in CD3CN
was followed (Table 4) at four different temperaturessnamely,
30, 40, 50, and 60 °C. The enthalpic (∆H) and entropic (-T∆S)
contributions (Table 4) to the free energy values were deter-
mined from the straight line plots of the free energies against
T. In the case of the free energy of activation (∆Gqon) for the
slipping-on process, a small difference between enthalpic and
entropic contributions is observed. On the contrary, in the case
of the free energy of activation (∆Gqoff) for the slipping-off
b
∆Gq
20.7
10.1
10.6
24.6
20.7
3.9
3.9
10.0
6.1
21.1
10.1
11.0
24.8
20.7
4.1
3.7
10.0
6.3
21.3
10.1
11.2
24.9
20.7
4.2
3.6
10.0
6.4
21.8
10.1
11.7
25.0
20.7
4.3
3.2
10.0
6.6
on
-∆Hq
c
on
-T∆Sq
d
on
e
∆Gq
off
f
∆Hq
off
-T∆Sq
g
off
-∆G° h
-∆H° i
T∆S° j
a The molar extinction coefficient (ꢀ) values measured for the
[2]rotaxane 1 in CD3CN at 50 °C were employed to calculate the kinetic
and thermodynamic parameters for the slipping of 1/5DN-1/5DN over
the stoppers of Et-D. b Free energy of activation for the slipping-on
process, the enthalpic contribution is predominant (∆Hq
>
off
process (percentage error e 0.03%). c Enthalpic contribution to ∆Gq
-T∆Sqoff). The strong noncovalent bonding interactions be-
tween the π-electron-rich macrocycle and the π-electron-
deficient bipydinium recognition site have to be destroyed in
order to dismember the [2]rotaxane. Thus, the slipping-off
process is enthalpically demanding.
on
(percentage error e 50%). d Entropic contribution to ∆Gqon (percentage
error e 49%). e Free energy of activation for the slipping-off process
(percentage error e 0.07%). Enthalpic contribution to ∆Gqoff (percent-
f
age error e 12%). g Entropic contribution to ∆Gq (percentage error
off
e 62%). h Free energy of association (percentage error e 0.3%).
i Enthalpic contribution to ∆G° (percentage error e 45%). j Entropic
contribution to ∆G° (percentage error e 50%).
The slippage of the macrocycle 1/5DN-1/5DN over the
stoppers of the dumbbell-shaped compound Et-D was followed
in a range of solvents at 30 °C. The resulting free energies of
on going from (CD3)2CO to DMF-d7 (from left to right in Table
5). Interestingly, when the low polar solvents CDCl3 and
CD2Cl2 are employed, the equilibrium is reached immediatelysi.e.
a charge-transfer band of constant intensity with time is detected.
activation (∆Gq and ∆Gqoff) and of asscociation (∆G°) are
on
listed in Table 5. The free energy of activation (∆Gqon) for the
slipping-on processes range over about 1 kcal mol-1, increasing

The Slipping Approach to Self-Assembling [n]Rotaxanes
J. Am. Chem. Soc., Vol. 119, No. 2, 1997 307
Table 5. Free Energies (kcal mol-1) of Activation (∆Gq and ∆Gqoff) and of Association (∆G°) for the Slipping of the Macrocyclic Polyether
on
1/5DN-1/5DN over the Stoppers of the Dumbbell-Shaped Compound Et-D in Various Solvents at 30 °C
a,b
a,b
a
parameters
CDCl3
CD2Cl2
(CD3)2CO
CD3CN
CD3NO2
(CD3)2SOa
THF-d8
DMF-d7
c
∆Gq
<20
<23
2.9
<20
<23
3.6
20.4
25.2
4.8
20.7
24.6
3.9
21.0
24.4
3.4
21.1
22.6
1.5
21.3
26.1
4.8
21.4
23.6
2.2
on
d
∆Gq
off
-∆G° e
a The molar extinction coefficient (ꢀ) values measured for the [2]rotaxane 1 in each of the solvents listed were employed to calculate the kinetic
and thermodynamic parameters for the slipping of 1/5DN-1/5DN over the stoppers of Et-D in these solvents. b In CDCl3 and CD2Cl2, the equilibrium
is reached immediately, i.e., a value constant in time for the absorbance of the charge transfer band associated with the resulting [2]rotaxane 1 is
measured. This result suggests that the values of ∆Gqon in CDCl3 and CD2Cl2 are lower than those measured in any other of the solvents listed in
the table. c Free energy of activation for the slipping-on process (percentage error e 0.06%). d Free energy of activation for the slipping-off process
(percentage error e 0.2%). e Free energy of association (percentage error e 0.8%).
These results suggest that, presumably, the macrocyclic poly-
ether and/or the dumbbell-shaped compound are highly solvated
in a solvent such as DMF-d7. Thus, desolvation is required in
order to slip-on the macrocycle over the stoppers, resulting in
a high energy of activation (∆Gqon). On the contrary, presum-
ably as a result of the low degree of solvation in nonpolar
solvents, low values of the energies of activation (∆Gq and
on
∆Gqoff) are achieved in CDCl3 and CD2Cl2. An approximately
opposite behavior is observed for the free energy of activation
(∆Gqoff) for the slipping-off processes. The lowest value (22.6
kcal mol-1) for ∆Gq is measured in the highly polar solvent
off
(CD3)2SO, while the highest value (26.1 kcal mol-1) for ∆Gq
off
is observed in THF-d8. These results are directly reflected in
the values of the free energy of association (∆G°), which reach
a maximum when either THF-d8 or (CD3)2CO is employed. In
other words, THF-d8 and (CD3)2CO seem to be the best solvents
for the self-assembly of these [2]rotaxanes: a relatively fast
slipping-on process is observed in addition to a high free energy
of associationsthus, leading to a high yield of the [2]rotaxane.
The free energy of association (∆G°) and the free energies of
activation (∆Gq and ∆Gqoff) were plotted against the dipole
on
moments,10 the dielectric constants,10 donor numbers,11 and
Z-values12 of the solvents. However, no significant correlations
between these values were observed.
Spectroscopic and Electrochemical Properties. The ro-
taxanes incorporate a number of chromophoric and electroactive
units, and as a result, they exhibit very interesting spectroscopic
and electrochemical properties. In particular, the [3]rotaxane
3 (Figure 5), which is made of the dumbbell-shaped component
5 and the two macrocycles HQ-HQ and 1/5DN-1/5DN,
incorporates altogether as many as ten chromophoric and
electroactive unitssnamely, two 1,4-dimethoxybenzene-like
units (1/4DMB) and two 1,5-dimethoxynaphthalene-like units
(1/5DMN) in the two macrocyclic components, and two
Figure 5. Macrocycles, dumbbell-shaped compounds, rotaxanes, and
model compounds employed in the photochemical and electrochemical
investigations.
tetraarylmethane-like stoppers (R-S), two methoxybenzene-like
(a) Spectroscopic Properties. In agreement with the be-
units (MB), and two paraquat-like units (PQT) in the dumbbell-
havior exhibited by analogous rotaxanes,3e the [2]rotaxane 1
shaped component 5. It is important to notice that, whereas
and the [3]rotaxane 3 show in the visible region a weak (and
broad in the case of 3) absorption band (Table 6) due to the
charge-transfer (CT) interaction between the electron-donor units
any effect on the absorption spectra, luminescence properties,
the R groups attached to the stoppers play a most important
role in the kinetics of the slippage processes, they do not have
(1/4DMB and/or 1/5DMN) of the macrocycles and the electron-
acceptor units (PQT) present in their dumbbell-shaped com-
ponents. With respect to the [2]rotaxane 4,3e the CT absorption
band of 1 lies at lower energy, since 1/5DN-1/5DN is a better
electron donor than HQ-HQ. In the [3]rotaxane 3 two different
macrocycles are present and therefore two CT bands are
expected. A closer analysis of the broad absorption band of 3
shows, indeed, that there are two components. Subtraction of
the spectrum of previously studied the [2]rotaxane 23e (which
contains two PQT stations and only HQ-HQ macrocycle) from
the spectrum of 3 gives an absorption band almost identical to
that exhibited by the [2]rotaxane 1, where a PQT unit is engaged
with a 1/5DN-1/5DN macrocycle.
and electrochemical behavior.3e This is the reason why the R
groups attached to the stoppers are not indicated in Figure 5
and are not mentioned in the following discussion. The
spectroscopic and electrochemical data of the 1/5DN-1/5DN
macrocycle, [2]rotaxane 1, and [3]rotaxane 3 are listed in Table
6, together with the previously obtained results for the HQ-
HQ macrocycle.3e,7d
(10) Riddick, J. A.; Bunger, W. B. Organic SolVents, Techniques of
Chemistry; Wiley: New York, 1970; Vol. 2.
(11) Reichardt, C. SolVents and SolVent Effects in Organic Chemistry;
VCH: Weinheim, 1988.
(12) (a) Kosower, E. M. J. Am. Chem. Soc. 1958, 80, 3253-3260. (b)
Kosower, E. M. J. Am. Chem. Soc. 1958, 80, 3261-3267. (c) Kosower, E.
M. J. Am. Chem. Soc. 1958, 80, 3267-3270.

308 J. Am. Chem. Soc., Vol. 119, No. 2, 1997
Asakawa et al.
Table 6. Spectroscopic and Electrochemical Data for the Macrocycles HQ-HQ and 1/5DN-1/5DN, the [2]Rotaxane 1, and the [3]Rotaxane 3
in MeCN at Room Temperature
reductiona
V Vs SCE
absorption
ꢀ (M-1 cm-1
fluorescence
oxidationb
V Vs SCE
compd
λ (nm)
)
λ (nm)
τ (ns)
φ
HQ-HQc
1/5DN-1/5DN
1
290
295
270
313
326
507
269
325
479
5200
17600
33500
11000
9400
320
330
d
2.5
8.5
0.08
0.27
d
+1.28 (1/4DMB)
+1.01 (1/5DMN)
+1.29 (1/5DMN)
-0.52
-0.89
700
3
59000
11500
1100
d
d
-0.46e,f
-0.87e,g
+1.32 (1/5DMN)
+1.41 (1/4DMB)
a Halfwave potential values; unless otherwise noted, the reduction processes are reversible and monoelectronic. b Potential values evaluated from
the DPV peaks; the unit involved in the oxidation process is reported in parentheses. c Literature values (Anelli, P. L.; Ashton, P. R.; Ballardini, R.;
Balzani, V.; Delgado, M.; Gandolfi, M. T.; Goodnow, T. T.; Kaifer, A. E.; Philp, D.; Pietraszkiewicz, M.; Prodi, L.; Reddington, M. V.; Slawin,
A. M. Z.; Spencer, N.; Stoddart, J. F.; Vicent, C.; Williams, D. J. J. Am. Chem. Soc. 1992, 114, 193-218. Ashton, P. R.; Ballardini, R.; Balzani,
V.; Belohradsky, M.; Gandolfi, M. T.; Philp, D.; Prodi, L.; Raymo, F. M.; Reddington, M. V.; Spencer, N.; Stoddart, J. F.; Venturi, M.; Williams,
D. J. J. Am. Chem. Soc. 1996, 118, 4931-4951). d Fluorescence from a very small amount of free 1/5DN-1/5DN precludes the observation of the
very weak stopper fluorescence (see text). e Bielectronic process. f The potential values of the two monoelectronic processes are -0.44 and -0.49
V (see text). g The potential values of the two monoelectronic processes are -0.85 and -0.90 V.
Figure 6. Simplified energy-level diagram of the [3]rotaxane 3. The energies of the CT levels have been estimated from the maxima of the
absorption bands and the energies of the S1 levels have been estimated from the onset of the emission spectra.
The CT interaction between the macrocycles and the PQT
units of the dumbbell-shaped components is also responsible
for the fact that the luminescence properties of 1 and 3 (Table
6) are completely different from those of their separated
components. The CT excited states, in fact, lying below the
fluorescent levels of the HQ-HQ and 1/5DN-1/5DN, introduce
fast radiationless deactivation routes which lead to a complete
quenching of the strong emission of the macrocycles. Whether
or not the very weak fluorescence of the R-S units observed
for the dumbbell-shaped components3e is still present in 1 and
3 cannot be said because of the presence of an interfering
impurity emission. A simplified energy-level diagram of the
[3]rotaxane 3 is illustrated schematically in Figure 6.
obtained3e for dumbbell-shaped components R-D and 5 and the
[2]rotaxanes 2 and 4. In the case of the [3]rotaxane 3, because
of the different electron-donor ability of the two macroclycles,
four monoelectronic reduction processes are expected. It should
also be noted, however, that the reduction of the PQT unit in
1 is only slightly shifted toward more negative potentials
compared to 4 (∆E1/2 ) 50 and 20 mV for the first and second
reduction process, respectively; Figure 7). Therefore, in 3, the
first two and the second two reductions might be expected to
appear as unresolved processes. This expectation is confirmed
by the experimental results, particularly by the differential pulse
voltammogram which exhibits two broad, bielectronic peaks.
With the reasonable assumption that the reduction potentials
of the PQT unit surrounded by HQ-HQ are identical in 2 and
3, the treatment of the differential pulse voltammetric peaks
given by Richardson and Taube13 allows us to obtain E1/2 values
(b) Electrochemical Properties. In Figure 7 the E1/2 values
concerning the reduction of the PQT units of the [2]rotaxane 1
and the [3]rotaxane 3 are correlated with those previously

The Slipping Approach to Self-Assembling [n]Rotaxanes
J. Am. Chem. Soc., Vol. 119, No. 2, 1997 309
macrocycle and the stoppers, has demonstrated that the size
complementarity between the components is the major require-
ment for success in the slippage approach to rotaxanes. Indeed,
only when the macrocyclic compound possesses a cavity large
enough for it to slip over the bulky stoppers covalently attached
at the ends of the dumbbell-shaped compound is rotaxane
formation observed. Furthermore, the kinetic investigation of
slippage, performed over a range of temperatures and in various
solvents, has demonstrated that the enthalpic and entropic
contributions to the free energy of activation associated with
the second-order slipping-on processsnamely the self-assembly
of the rotaxanesare approximately the same, while the opposite
slipping-off processsnamely the dismembering of the rotaxanesis
mainly enthalpically demanding. Also, the polarity of the
solvent is reflected in the rates of both slipping-on and slipping-
off processes suggesting that solvation of the components has
a significant effect on the kinetics of these processes. These
results suggest strongly the possibility of constructing complex
molecular assemblies by carefully selecting appropriate building
blocks and self-assembly approaches. Indeed, a [3]rotaxane,
incorporating a π-electron deficient dumbbell-shaped compo-
nent, comprised of two bipyridinium recognition sites, and
encircled by two constitutionally different macrocyclic poly-
ethers, was self-assembled in good yield. This [3]rotaxane
incorporates altogether as many as ten chromophoric and
electroactive sub-units and its spectroscopic and electrochemical
properties and those of the related [2]rotaxane 1 and model
compounds have been described and discussed. The two
charge-transfer bands originating from the interaction between
the electron-acceptor dumbbell-shaped component and the two
different electron-donor macrocycles have been resolved. Cor-
relations among the redox potentials of the various compounds
have allowed the assignement of the reduction and oxidation
processes to specific electroactive components.
Figure 7. Correlations of the halfwave reduction potentials of the
rotaxanes and their dumbbell-shaped components. Argon-purged MeCN
solution. Bielectronic process (bold concentric circles); calculated values
of the monoelectronic processes (shaded circle).
Experimental Section
of -0.44, -0,49, -0,85, and -0,90 V Vs SCE for the four one-
electron processes of 3 (Table 6 and Figure 7). Looking at
Figure 7, one can also note quantitative (considering the
experimental error) correlations among all the reduction proc-
esses, particularly for the first reduction of each PQT unit. For
example, 1/5DN-1/5DN has the same effect on the first
reduction of the PQT unit in 1 and 3, i.e. shifts of 150 mV
toward more negative potential in going from R-D to 1 and
160 mV in going from 5 to 3 are observed.
Although oxidations are not fully reversible, a correlation can
be attempted, at least for the potentials of the first observed
processes. On the basis of the results obtained for the
[2]rotaxanes 2 and 4,3e in 1 the first process, that occurs at less
positive potential than in 2 and 4, can be assigned to the
oxidation of the 1/5DN-1/5DN macrocycle, easier to oxidize
than HQ-HQ (Table 6). In 3, where both the 1/5DN-1/5DN
and HQ-HQ macrocycles are present, the first and second
processes can be assigned to the oxidation of 1/5DN-1/5DN
and HQ-HQ, respectively, by comparison with the potential
values of 1 and 4 (or 2).
General Methods. Chemicals were purchased from Aldrich and
used as received. Solvents were dried [MeCN (from P2O5), DMF (from
CaH2)] according to literature procedures.14 The macrocyclic polyethers
HQ-HQ,7d HQ-1/5DN,7e HQ-2/7DN,7e 1/5DN-1/5DN,7b 1/5DN-2/
7DN,7e and 2/7DN-2/7DN,7e the bis(hexafluorophosphate) salt
[BPYPYXY][PF6]2,7d the tetrarylmethane-based stopper i-Pr-S, and
the dumbbell-shaped compounds Et-D,3e i-Pr-D,3e cyclohexyl-D,3e and
t-Bu-D3e were prepared according to published procedures. Reactions
requiring ultrahigh pressures were carried out in Teflon vessels using
a custom-built ultrahigh pressure reactor manufactured by PSIKA
Pressure Systems Limited of Glossop, UK. Thin layer chromatography
(TLC) was carried out on aluminium sheets coated with silica gel 60
(Merck 5554). Column chromatography was performed on silica gel
60 (Merck 9385, 230-400 mesh). Melting points were determined
on an Electrothermal 9200 melting point apparatus and are uncorrected.
Liquid secondary ion mass spectrometry (LSIMS) in conjunction with
a 3-nitrobenzyl alcohol or 2-nitrophenyl octyl ether matrix was
performed on a VG Prospec instrument. Electrospray positive-ion mass
spectra (ESMS) were also measured on the VG Prospec mass
spectrometer. UV-visible spectra were obtained and reaction kinetics
followed by means of a computer-controlled Perkin-Elmer Lambda 2
spectrophotometer fitted with a thermostatic temperature controller
((0.2 °C). 1H-NMR spectra were recorded on a Bruker AC300 (300
MHz) spectrometer. 13C-NMR spectra were recorded on a Bruker
AC300 (75 MHz) spectrometer.
Conclusions
We have developed a synthetic approachsnamely that of the
slippagesto self-assembling [n]rotaxanes incorporating π-electron-
deficient bipyridinium-based dumbbell-shaped components and
π-electron-rich hydroquinone-, 1,5-dioxynaphthalene-, and/or
2,7-dioxynaphthalene-based macrocyclic polyether components.
A systematic investigation of the kinetics of these self-assembly
processes, performed while varying the size of both the
[2]Rotaxane 1. A solution of i-Pr-D (108.0 mg, 0.06 mmol) and
1/5DN-1/5DN (21.0 mg, 0.03 mmol) in dry MeCN (3 mL) was stirred
at 50 °C for 24 h. The solvent was removed in Vacuo and the residue
was purified by column chromatography (SiO2, MeOH/CH2Cl2/
MeNO2/2 M NH4Cl, 6:2:1:0.2). The resulting purple solid was
(14) Furniss, B. S.; Hannaford, A. J.; Smith, P. W. G. A.; Tatchell, R.
(13) Richardson, D. E.; Taube, H. Inorg. Chem. 1981, 20, 1278-1285.
Practical Organic Chemistry, Longman: New York, 1989.

310 J. Am. Chem. Soc., Vol. 119, No. 2, 1997
Asakawa et al.
dissolved in H2O/(Me)2CO and a saturated aqueous solution of NH4PF6
was added. After the evaporation of the (Me)2CO under reduced
pressure, a purple solid precipitated. The solid was filtered and washed
with H2O to give 2 (41.1 mg, 57%): mp 146 °C dec; LSIMS m/z 2272
polyether were dissolved in one of the deuterated solvents, preheated
at temperature T, listed in Table 5 to afford a solution of known
concentration C0. The solution was mantained at temperature T by
means of a thermostatic water bath over a period of time (t) and the
absorbance (A) of the developing charge-transfer band associated with
the resulting [2]rotaxane was monitored by absorption UV-visible
spectroscopy. Heating was continued further until no change in the
absorbance was detectedsi.e. until the equilibrium was reached. At
1
[M - PF6]+, 2128 [M - 2PF6]+; ESMS: m/z 2272 [M - PF6]+; H-
NMR (CD3COCD3) δ 9.02 (4H, d, J ) 7 Hz), 7.88 (4H, d, J ) 9 Hz),
7.38 (4H, d, J ) 7 Hz), 7.30-7.19 (12H, m), 7.14-7.04 (16H, m),
7.01 (4H, d, J ) 9 Hz), 6.72 (4H, d, J ) 9 Hz), 6.70-6.50 (8H, m),
6.45 (4H, d, J ) 7 Hz), 6.02 (4H, s), 4.23-4.18 (4H, m), 4.04-3.91
(32H, m), 3.87-3.75 (12H, m), 2.90-2.81 (2H, m), 1.28 (36H, s),
1.20 (12H, d, J ) 7 Hz); 13C-NMR (CD3CN) δ 161.3, 157.5, 154.0,
149.3, 147.1, 145.8, 145.4, 144.5, 140.5, 132.5, 132.2, 131.4, 131.1,
126.6, 126.4, 126.4, 126.0, 125.3, 124.9, 116.5, 114.1, 114.0, 106.6,
106.1, 71.9, 71.7, 71.4, 70.8, 70.3, 70.1, 68.9, 68.2, 65.3, 63.9, 34.8,
34.1, 31.5, 24.1. Anal. Calcd for C140H162F12N2O16P2‚H2O: C, 69.01;
H, 6.78; N, 1.15. Found: C, 68.99; H, 7.09; N, 1.15.
1
this juncture, an H-NMR spectrum of the solution was recorded at
the same temperature T. By integrating the resonances appearing in
1
the H-NMR spectrum associated with the [2]rotaxane and the free
macrocyclic and dumbbell-shaped components, the concentration of
the [2]rotaxane at equilibrium Ce was evaluated.9 As a result, the molar
extinction coefficient ꢀ of the [2]rotaxane was calculated from the
expression Ae ) ꢀlCe, where Ae is the absorbance measured at
equilibrium and l the optical path. By employing a nonlinear curve-
fitting program, the plot of A against time (t) was fitted by eq 18 thus,
affording the value of the rate constant kon for the slipping-on process.
[2]Rotaxane 2. A solution of i-Pr-S (300.0 mg, 0.43 mmol), HQ-
HQ (124.0 mg, 0.23 mmol), and [BPYPYXY][PF6]2 (150.0 mg, 0.20
mmol) in dry DMF (10 mL) was subjected to a pressure of 12 kbar at
30 °C for 36 h. The solvent was removed in vacuo and the resulting
residue was washed with Et2O (60 mL) and then dissolved in H2O/
(Me)2CO (200 mL). An aqueous solution of NH4PF6 was added and,
after the evaporation of (Me)2CO, a red solid precipitated. The solid
was filtered off and purified by column chromatography (SiO2, MeOH/
CH2Cl2/MeNO2/2 M NH4Cl, 6:2:1:0.2) to give, after counterion
exchange [H2O/(Me)2CO, saturated aqueous solution of NH4PF6] 2 as
a red solid (111.2 mg, 19%): mp 220 °C dec; LSIMS m/z 2721 [M -
(C02 - Ce )
2
2
ont
ꢀ C0 Ce 1 - ek
(
)
[
]
Ce
A )
(1)
(C02 - Ce )
2
Ce2 - C0 ek
2
ont
Ce
Absorption and Luminescence Measurements. Room temperature
experiments were carried out in MeCN solutions. Electronic absorption
spectra were recorded with a Perkin-Elmer λ6 spectrophotometer.
Emission spectra were obtained with a Perkin-Elmer LS50 spectro-
fluorimeter. Emission spectra in butyronitrile rigid matrix at 77 K were
recorded using quartz tubes immersed in a quartz Dewar filled with
liquid nitrogen. Fluorescence quantum yields were determined using
naphthalene in degassed cyclohexane as a standard (Φ ) 0.23).15
Nanosecond lifetime measurements were performed with a previously
described Edinburgh single-photon counting equipment.16 Experimental
errors: absorption maxima, (2 nm; emission maxima, (2 nm; excited
state lifetimes, (10%; fluorescence quantum yields, (20%.
Electrochemical Measurements. Electrochemical experiments
were carried out in argon-purged MeCN solution with a Princeton
Applied Research 273 multipurpose instrument interfaced to a personal
computer, using cyclic voltammetry (CV) and differential pulse
voltammetry (DPV) techniques. The exact configuration and proce-
dures for these measurements have been previously reported.3e
1
PF6]+, 2576 [M - 2PF6]+; ESMS m/z 1290 [M - 2PF6]2+; H-NMR
(CD3COCD3) δ 9.32 (4H, d, J ) 7 Hz), 9.24 (4H, d, J ) 7 Hz), 8.47
(4H, d, J ) 7 Hz), 8.44 (4H, d, J ) 7 Hz), 7.91 (4H, s), 7.68 (4H, d,
J ) 9 Hz), 7.30 (8H, d, J ) 9 Hz), 7.18-7.05 (24H, m), 6.82 (4H, d,
J ) 9 Hz), 6.19 (4H, s), 6.05 (12H, s), 4.21 (4H, t, J ) 5 Hz), 4.12
(4H, t, J ) 5 Hz), 3.92-3.85 (8H, m), 3.81-3.70 (24H, m), 3.64-
3.61 (8H, m), 2.93-2.80 (2H, m), 1.29 (36 H, s), 1.22 (12 H, d, J )
7 Hz); 13C-NMR (CD3CN) δ 161.3, 157.8, 153.2, 149.6, 149.0, 147.4,
146.7, 146.4, 146.0, 145.6, 140.9, 135.8, 132.8, 132.6, 131.6, 131.4,
127.5, 127.3, 126.6, 125.7, 125.5, 116.6, 115.8, 114.45, 71.5, 71.3,
70.8, 70.6, 70.4, 68.9, 68.6, 68.5, 65.3, 65.0, 64.2, 35.1, 34.3, 31.7,
24.3. Anal. Calcd for C150H174F24N4O16P4: C, 62.80; H, 6.11; N, 1.95.
Found: C, 62.95; H, 6.20; N, 1.95.
[3]Rotaxane 3. A solution of 2 (45.0 mg, 0.02 mmol) and 1/5DN-
1/5DN (40.0 mg, 0.06 mmol) in dry MeCN (3 mL) was stirred at 50
°C for 48 h. The solvent was removed in vacuo and the red residue
was purified by column chromatography (SiO2, MeOH/CH2Cl2/
MeNO2/2 M NH4Cl, 6:2:1:0.2) to give, after counterion exchange [H2O/
(Me)2CO, saturated aqueous solution of NH4PF6], 3 as a red solid (27.0
mg, 49%): mp 134 °C dec; LSIMS m/z 3360 [M - PF6]+, 3215 [M -
2PF6]+; ESMS m/z 1609 [M - 2PF6]2+, 1024 [M - 3PF6]3+; 1H-NMR
(CD3COCD3) δ 9.16 (2H, d, J ) 7 Hz), 9.09 (2H, d, J ) 7 Hz), 9.05
(2H, d, J ) 7 Hz), 9.00 (2H, d, J ) 7 Hz), 8.28-8.15 (4H, m), 8.14-
8.07 (4H, m), 7.91 (2H, d, J ) 9 Hz), 7.70 (2H, d, J ) 9 Hz), 7.39
(2H, d, J )7 Hz), 7.35-7.22 (12H, m), 7.19-7.03 (22H, m), 6.87-
6.76 (4H, m), 6.72-6.63 (4H, m), 6.60-6.48 (8H, m), 6.22 (2H, s),
6.18 (2H, s), 6.09 (2H, s), 6.02-5.96 (10H, m), 4.24-4.17 (4H, m),
4.14-3.54 (76H, m), 2.93-2.79 (2H, m), 1.29 (18H, s), 1.28 (18H, s),
1.22 (6H, d, J ) 7 Hz), 1.21 (6H, d, J )7 Hz); 13C-NMR (CD3CN) δ
154.1, 152.9, 149.5, 147.3, 146.4, 145.5, 145.0, 140.8, 136.1, 132.8,
132.6, 132.5, 132.2, 132.0, 131.5, 131.3, 126.8, 126.5, 126.1, 125.4,
125.2, 125.0, 118.3, 116.6, 116.4, 115.6, 114.4, 114.0, 106.3, 72.0,
71.8, 71.3, 71.2, 70.9, 70.7, 70.5, 70.3, 69.0, 68.8, 68.5, 68.4, 64.1,
35.0, 34.2, 31.6, 24.3. Anal. Calcd for C186H218F24N4O26P4: C, 63.73;
H, 6.27; N, 1.60. Found: C, 63.57; H, 6.40; N, 1.62.
Acknowledgment. This research was supported by the
Engineering and Physical Sciences Research Council and by
the European Community Human Capital and Mobility Pro-
gramme, Italian MURST and CNR (Progetto Strategico Tec-
nologie Chimiche Innovative).
Supporting Information Available: Derivation of eq 1 and
equations employed to calculate the free energies of activation
and of association, and tables listing the kinetic and thermo-
dynamic parameters for the slipping processes (8 pages). See
any current masthead page for ordering and Internet access
instructions.
JA961817O
(15) Berlman, Handbook of Fluorescence Spectra of Aromatic Com-
pounds, Academic Press: London, 1965.
(16) Armaroli, N.; Balzani, V.; Barigelletti, F.; De Cola, L.; Flamigni,
L.; Sauvage, J.-P.; Hemmert, C. J. Am. Chem. Soc. 1994, 116, 5211-5217.
General Method for the Determination of the Rate Constants.
Equimolar amounts of a dumbbell-shaped compound and a macrocyclic