Rotaxane or pseudorotaxane? That is the question!

Article abstract of DOI:10.1021/ja9731276

A series of secondary dialkylammonium ions (RCH₂)₂NH₂⁺ have been prepared, and their binding properties toward the macrocyclic polyether dibenzo[24]crown-8 (DB24C8) evaluated. By using this information, a route

Full text of DOI:10.1021/ja9731276

J. Am. Chem. Soc. 1998, 120, 2297-2307
2297
Rotaxane or Pseudorotaxane? That Is the Question!†
Peter R. Ashton,‡ Ian Baxter,§ Matthew C. T. Fyfe,‡ Franc¸isco M. Raymo,‡ Neil Spencer,‡
J. Fraser Stoddart,*,‡ Andrew J. P. White,§ and David J. Williams§
Contribution from The School of Chemistry, The UniVersity of Birmingham,
Edgbaston, Birmingham B15 2TT, UK, and The Chemical Crystallography Laboratory, Department of
Chemistry, Imperial College, South Kensington, London SW7 2AY, UK
ReceiVed September 5, 1997
Abstract: A series of secondary dialkylammonium ions (RCH2)2NH2+ have been prepared, and their binding
properties toward the macrocyclic polyether dibenzo[24]crown-8 (DB24C8) evaluated. By using this
information, a route to a kinetically stable rotaxane-like entitysstabilized by noncovalent bonding interactions
between the DB24C8 macroring and the ammonium centerswas established, in which the crown ether slips
over a dialkylammonium ion’s stopper groups (R). However, we have found that the kinetic stability of this
rotaxane-like entity is extremely dependent on the nature of the solvent in which it is dissolved, suggesting
that pseudorotaxanes lie in the fuzzy domain between two sets of extremes, wherein a beadlike macrocycle
and a dumbbell-like component may either (1) exist as a rotaxane or (2) be completely disassociated from one
another.
Introduction
energy for “dumbbell”-extrusion is too high to be overcome at
ambient temperature. During the past 5 years, we have
developed a novel synthetic strategysnamely, slippage6,7sfor
the construction of rotaxane-like assemblies (Figure 1) that relies
upon (1) the size complementarity between the macrocycle and
the “dumbbell’s” stoppers, in addition to (2) stabilizing nonco-
valent bonding interactions between the macrocycle and the
“dumbbell’s” rod. In this strategy, the macrocycle M and the
“dumbbell” D are synthesized separately, before being heated
together in solution so that the free energy of activation (i.e.,
∆Gqon) for the slippage of M over D’s stoppers can be over-
come. The noncovalent bonding interactions present in the
rotaxane-like structure R make it more stable than its precursors,
viz., M and D, by ∆G°si.e., the free energy of complexationsso
that the free energy of activation for its dissociation (i.e.,
∆Gqoff) becomes insurmountable when the solution is cooled to
ambient temperature. To date,6 we have utilized the supramo-
lecular assistance to synthesis8 provided by, inter alia, aryl-
aryl stacking interactions between π-electron deficient bipyri-
dinium-based “dumbbells” (or macrocycles) and π-electron rich
The quest for nanoscopic device-like systems, based upon
molecular components,1 has led to the rapid growth of a new
field of chemical synthesis involving the production of inter-
locked molecules2 such as rotaxanes.3,4 In these molecular
assemblies,5 a beadlike component (i.e., a macrocycle) is bound
mechanically to a dumbbell-like species so that the activation
* To whom correspondence should be addressed. Current address:
Department of Chemistry and Biochemistry, University of California, Los
Angeles, 405 Hilgard Avenue, Los Angeles, CA 90095.
† Molecular Meccano, Part 33. For Part 32, see: Amabilino, D. B.;
Ashton, P. R.; Stoddart, J. F.; White, A. J. P.; Williams, D. J. Chem. Eur.
J. In press.
‡ University of Birmingham.
§ Imperial College.
(1) (a) de Silva, A. P.; McCoy, C. P. Chem. Ind. (London) 1994, 992-
996. (b) Drexler, K. E. Annu. ReV. Biophys. Biomol. Struct. 1994, 23, 377-
405. (c) Whitesides, G. M. Sci. Am. 1995, 273 (3), 114-117. (d) Ashton,
P. R.; Ballardini, R.; Balzani, V.; Boyd, S. E.; Credi, A.; Gandolfi, M. T.;
Go´mez-Lo´pez, M.; Iqbal, S.; Philp, D.; Preece, J. A.; Prodi, L.; Ricketts,
H. G.; Stoddart, J. F.; Tolley, M. S.; Venturi, M.; White, A. J. P.; Williams,
D. J. Chem. Eur. J. 1997, 3, 152-170 and references therein.
(2) (a) Chambron, J. C.; Dietrich-Buchecker, C. O.; Sauvage, J.-P. Top.
Curr. Chem. 1993, 165, 131-162. (b) Amabilino, D. B.; Stoddart, J. F.
Chem. ReV. 1995, 95, 2725-2828. (c) Ja¨ger, R.; Vo¨gtle, F. Angew. Chem.,
Int. Ed. Engl. 1997, 36, 930-944.
(3) For examples of rotaxane-based device-like systems, see: (a) Bissell,
R. A.; Co´rdova, E.; Kaifer, A. E.; Stoddart, J. F. Nature (London) 1994,
369, 133-137. (b) Anelli, P.-L.; Asakawa, M.; Ashton, P. R.; Bissell, R.
A.; Clavier, G.; Go´rski, R.; Kaifer, A. E.; Langford, S. J.; Mattersteig, G.;
Menzer, S.; Philp, D.; Slawin, A. M. Z.; Spencer, N.; Stoddart, J. F.; Tolley,
M. S.; Williams, D. J. Chem. Eur. J. 1997, 3, 1113-1135. (c) Mart´ınez-
D´ıaz, M.-V.; Spencer, N.; Stoddart, J. F. Angew. Chem., Int. Ed. Engl. 1997,
36, 1904-1907. (d) Murakami, H.; Kawabuchi, A.; Kotoo, K.; Kunitake,
M.; Nakashima, N. J. Am. Chem. Soc. 1997, 119, 7605-7606.
(4) For some recent examples of rotaxanes from other research groups,
see: (a) Li, Z.-T.; Stein, P. C.; Becher, J.; Jensen, D.; Mørk, P.; Svenstrup,
N. Chem. Eur. J. 1996, 2, 624-633. (b) Solladie´, N.; Chambron, J.-C.;
Dietrich-Buchecker, C. O.; Sauvage, J.-P. Angew. Chem., Int. Ed. Engl.
1996, 35, 906-909. (c) Vo¨gtle, F.; Du¨nnwald, T.; Ha¨ndel, M.; Ja¨ger, R.;
Meier, S.; Harder, G. Chem. Eur. J. 1996, 2, 640-643. (d) Anderson, S.;
Anderson, H. L. Angew. Chem., Int. Ed. Engl. 1996, 35, 1956-1959. (e)
Hannak, R. B.; Farber, G.; Konrat, R.; Krautler, B. J. Am. Chem. Soc. 1997,
119, 2313-2314. (f) Leigh, D. A.; Murphy, A.; Smart, J. P.; Slawin, A.
M. Z. Angew. Chem., Int. Ed. Engl. 1997, 36, 728-732.
(5) Schill, G. Catenanes, Rotaxanes, and Knots; Academic Press: New
York, 1971.
(6) (a) Ashton, P. R.; Belohradsky, M.; Philp, D.; Stoddart, J. F. J. Chem.
Soc., Chem. Commun. 1993, 1269-1274. (b) Ashton, P. R.; Belohradsky,
M.; Philp, D.; Spencer, N.; Stoddart, J. F. J. Chem. Soc., Chem. Commun.
1993, 1274-1277. (c) Amabilino, D. B.; Ashton, P. R.; Belohradsky, M.;
Raymo, F. M.; Stoddart, J. F. J. Chem. Soc., Chem. Commun. 1995, 747-
750. (d) Amabilino, D. B.; Ashton, P. R.; Belohradsky, M.; Raymo, F. M.;
Stoddart, J. F. J. Chem. Soc., Chem. Commun. 1995, 751-753. (e) Asakawa,
M.; Ashton, P. R.; Iqbal, S.; Stoddart, J. F.; Tinker, N. D.; White, A. J. P.;
Williams, D. J. Chem. Commun. 1996, 483-486. (f) 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. (g)
Asakawa, M.; Ashton, P. R.; Ballardini, R.; Balzani, V.; Belohradsky, M.;
Gandolfi, M. T.; Kocian, O.; Prodi, L.; Raymo, F. M.; Stoddart, J. F.;
Venturi, M. J. Am. Chem. Soc. 1997, 119, 302-310. (h) Ashton, P. R.;
Everitt, S. R. L.; Go´mez-Lo´pez, M.; Jayaraman, N.; Stoddart, J. F.
Tetrahedron Lett. 1997, 38, 5691-5694. (i) Raymo, F. M.; Stoddart, J. F.
Pure Appl. Chem. 1997, 69, 1987-1997.
S0002-7863(97)03127-2 CCC: $15.00 © 1998 American Chemical Society
Published on Web 03/03/1998

2298 J. Am. Chem. Soc., Vol. 120, No. 10, 1998
Ashton et al.
Scheme 1. Self-Assembly of the Dibenzylammonium Ion
(1+) with the Macrocyclic Polyether Dibenzo[24]crown-8
(DB24C8) To Form the [2]Pseudorotaxane [DB24C8‚1]+
Figure 1. Schematic representation depicting the self-assembly of
rotaxane-like entities using the slippage approach.
hydroquinone/dioxynaphthalene-based macrocycles (or “dumb-
bells”) for the preparation of rotaxane-like species. We
conjectured that we could self-assemble9 rotaxane-like
entitiessemploying the slippage approachsharnessing another
recognition motif that we have employed for the noncovalent
synthesis of pseudorotaxanes10sspecifically, the interaction of
secondary dialkylammonium ions with the commercially avail-
able macrocyclic polyether dibenzo[24]crown-8 (DB24C8).11,12
By way of illustration, DB24C8 self-assembles with the
dibenzylammonium ion (1+) to form (Scheme 1) a [2]pseu-
dorotaxane superstructure that is stabilized primarily as a result
of [N+-H‚‚‚O] and [C-H‚‚‚O] hydrogen bonds, supplemented
Scheme 2. Slippage-Type Synthesis of Rotaxane-like
Systems by Using the Supramolecular Assistance To
Synthesis Provided by Noncovalent Bonding Interactions
between DB24C8 and Secondary Dialkylammonium Ions
(7) Prior to our research, rotaxane-like species had been synthesized
statistically (i.e., involving methods in which noncovalent bonding interac-
tions were not employed as a “synthetic aid”) with slippage-type approaches.
See: (a) Harrison, I. T. J. Chem. Soc., Chem. Commun. 1972, 231-232.
(b) Schill, G.; Beckmann, W.; Schweickert, N.; Fritz, H. Chem. Ber. 1986,
119, 2647-2655. However, in recent times, other researchers have also
started to apply the slippage methodology using the supramolecular
assistance to synthesis furnished by intermolecular interactions. See: (c)
Macartney, D. H. J. Chem. Soc., Perkin Trans. 2 1996, 2775-2778. (d)
Ha¨ndel, M.; Plevoets, M.; Gestermann, S.; Vo¨gtle, F. Angew. Chem., Int.
Ed. Engl. 1997, 36, 1199-1201.
(8) Fyfe, M. C. T.; Stoddart, J. F. Acc. Chem. Res. 1997, 30, 393-401.
(9) Philp, D.; Stoddart, J. F. Angew. Chem., Int. Ed. Engl. 1996, 35,
1154-1196.
(10) Pseudorotaxanes have been defined (Ashton, P. R.; Philp, D.;
Spencer, N.; Stoddart, J. F. J. Chem. Soc., Chem. Commun. 1991, 1677-
1679) as the noninterlocked counterparts of rotaxanes, in which one or more
linear “filaments” interpenetrate the cavities of one or more macrocyclic
species to form inclusion complexes. Notably, as distinct from rotaxanes,
these supramolecular complexes are free to dissociate into their separate
components since they do not possess bulky stopper groups. For some recent
examples, see: (a) Amabilino, D. B.; Anelli, P.-L.; Ashton, P. R.; Brown,
G. R.; Co´rdova, E.; God´ınez, L. A.; Hayes, W.; Kaifer, A. E.; Philp, D.;
Slawin, A. M. Z.; Spencer, N.; Stoddart, J. F.; Tolley, M. S.; Williams, D.
J. J. Am. Chem. Soc. 1995, 117, 11142-11170. (b) Ashton, P. R.; Langford,
S. J.; Spencer, N.; Stoddart, J. F.; White, A. J. P.; Williams, D. J. Chem.
Commun. 1996, 1387-1388. (c) Baxter, P. N. W.; Sleiman, H.; Lehn, J.-
M.; Rissanen, K. Angew. Chem., Int. Ed. Engl. 1997, 36, 1294-1296. (d)
Mirzoian, A.; Kaifer, A. E. Chem. Eur. J. 1997, 3, 1052-1058. (e) Fyfe,
M. C. T.; Glink, P. T.; Menzer, S.; Stoddart, J. F.; White, A. J. P.; Williams,
D. J. Angew. Chem., Int. Ed. Engl. 1997, 36, 2068-2070.
(11) (a) Ashton, P. R.; Campbell, P. J.; Chrystal, E. J. T.; Glink, P. T.;
Menzer, S.; Philp, D.; Spencer, N.; Stoddart, J. F.; Tasker, P. A.; Williams,
D. J. Angew. Chem., Int. Ed. Engl. 1995, 34, 1865-1869. (b) Ashton, P.
R.; Chrystal, E. J. T.; Glink, P. T.; Menzer, S.; Schiavo, C.; Spencer, N.;
Stoddart, J. F.; Tasker, P. A.; White, A. J. P.; Williams, D. J. Chem. Eur.
J. 1996, 2, 709-728. (c) Glink, P. T.; Schiavo, C.; Stoddart, J. F.; Williams,
D. J. Chem. Commun. 1996, 1483-1490.
by π-π stacking interactions. The 1+ cation has some
difficulty11 threading its way through the DB24C8 macroring
since its phenyl rings are relatively bulky compared to the size
of the macrocycle’s cavity. Hence, it seemed to us that we were
very close to a situation where we could synthesize kinetically
stable rotaxane-like species, via the slippage approach, utilizing
the aforementioned DB24C8-dialkylammonium ion interactions
(Scheme 2), i.e., we envisaged modifying the 1+ cation slightly,
so that we would reach a state where a dumbbell-like dialky-
lammonium ion (D+) would not form a complex with the
beadlike crown ether DB24C8 at ambient temperature, but could
be coaxed into forming a kinetically stable rotaxane-like entity
under the influence of a suitable quantity of thermal energy.
We visualized (Scheme 3) two possible modifications of the
(12) Utilizing this recognition motif, rotaxanes have been synthesized
previously by a “threading followed by stoppering” approach. See: (a)
Kolchinski, A. G.; Busch, D. H.; Alcock, N. W. J. Chem. Soc., Chem.
Commun. 1995, 1289-1291. (b) Ashton, P. R.; Glink, P. T.; Stoddart, J.
F.; Tasker, P. A.; White, A. J. P.; Williams, D. J. Chem. Eur. J. 1996, 2,
729-736. (c) Reference 3c.

Rotaxane or Pseudorotaxane?
J. Am. Chem. Soc., Vol. 120, No. 10, 1998 2299
Scheme 3. Modifications to the 1+ Cation Which Might
Lead to Useful “Dumbbells” for a Slippage Synthesis of
Rotaxane-like Species Involving DB24C8 and (RCH2)2NH2
Ions
Scheme 4. Syntheses of the Secondary Dialkylammonium
Salts 2‚PF6-6‚PF6
+
1+ cation for the synthesis of rotaxane-like structures using this
approach. Either (1) bulky stopper groups could be placed
at the termini of the cation 1+, as in 2+ or 3+, whose phenyl
rings are para-substituted with i-Pr and t-Bu groups, respectively,
or (2) the 1+ cation’s phenyl rings could be replaced by
fully saturated, more flexiblesand hence, sterically more
encumberingscycloalkyl rings, as in 4+, 5+, or 6+, where the
phenyl rings have been exchanged for c-C5H9, c-C6H11, or
c-C7H13 rings, respectively. This article describes a systematic
study of the complexation behavior of DB24C8 with the cations
2+-6+ and reports the discovery of a 1:1 complexsformed
between DB24C8 and the 5+ cationswhich is quite stable in
nonpolar solvents that are reluctant to donate their nonbonding
electrons into noncovalent bonds, i.e., those endowed with low
Gutmann donor numbers (DN), but dissociates in polar solvents
that are inclined to form such intermolecular bonds, i.e., those
possessing a high DN.13 Thus, we pose the following questionsIs
the [DB24C8‚5]+ complex a rotaxane, or is it a pseudorotaxane?
We provide a fuzzy14 answer which states that all pseudoro-
taxanes possess, to a certain extent, some rotaxane-like character.
alkyl]amine with LiAlH4.15 The secondary diamines obtained
by these methods were then converted into the salts
2‚PF6-6‚PF6 by heating with hydrochloric acid, followed by
counterion exchange from chloride to hexafluorophosphate.
Pseudorotaxane Synthesis. We chose to monitor 1:1
complex formation by 1H NMR spectroscopy, since the
complexes formed between DB24C8 and secondary dialkylam-
1
monium ions tend to show11 substantially different H NMR
spectra compared to those of the sum of their uncomplexed
components. A CDCl3/CD3CN (3:1) solvent system (herein-
after, for the sake of convenience, this mixed solvent system is
referred to simply as MSS) was chosen to study the syntheses
of the complexes: the CDCl3 ensures11 that the complexes will
form with high stability constants (Ka), while the CD3CN confers
solubility upon the dialkylammonium salts.
A 1:1 mixture of DB24C8 and 2‚PF6 was dissolved in the
1
MSS. After 5 min, the room temperature H NMR spectrum
Results and Discussion
of this solution revealed (Figure 2a) no discernible complex
formation. However, after a short period of time, the formation
of additional peaksscorresponding to the [DB24C8‚2][PF6]
speciesswas noticed (Figure 2b). As time progressed, these
peaks grew in intensity (Figure 2, parts c and d) so that the
[DB24C8‚2]+ complex eventually developed into the dominant
species present in solution. After ca. 9.5 days, equilibrium was
attained, i.e., the rates of complexation and decomplexation
became identical. Moreover, the same equilibrium position
could be reached when a single crystal of the complex was
dissolved in the MSS at 20 °C. Since the decomplexation
process was observed at ambient temperature, the [DB24C8‚2]+
complex appears to be a pseudorotaxane and not a rotaxane.
That is, the ∆Gqoff barrier is not large enough to prevent
extrusion of the 2+ ion from the [DB24C8‚2]+ species at 20 °C
in the MSS. On the other hand, the p-t-BuC6H4 stopper groups
of the dialkylammonium ion 3+ are too bulky to permit the
slippage of the DB24C8 macroring over them at various
Dumbbell Synthesis. The ammonium ions used in this
investigation were prepared as their hexafluorophosphate salts
to (1) augment their solubilities in organic solvents and (2)
weaken ion pairing, i.e., to destabilize the reactants’ ground-
state energy compared to that of their complexes, thus making
∆G° more negative. The requisite secondary diamine precursors
of the salts 2‚PF6-6‚PF6 were prepared (Scheme 4) by two
methods: specifically, by Route Asvia a reductive amination
sequence, in which the appropriate aldehyde and amine were
condensed to produce an imine that was subsequently reduced
with NaBH4sand by Route Bsvia the condensation of the
appropriate aldehyde with NH3 and benzotriazole (BtH), fol-
lowed by reduction of the resultant bis[1-(benzotriazol-1-yl)-
(13) Previously,11 we have noted that secondary dialkylammonium ions
form the strongest complexes with DB24C8 in solvents possessing the lowest
DNs, i.e., those lacking the ability to donate their electrons into noncovalent
bonds.
(14) For excellent overviews describing the utility of fuzzy logic in
chemistry, see: (a) Rouvray, D. H. Chem. Br. 1995, 31, 544-546. (b)
Rouvray, D. H. Chem. Ind. (London) 1997, 60-62.
(15) Katritzky, A. R.; Zhao, X.; Hitchings, G. J. Synthesis 1991, 703-
708.

2300 J. Am. Chem. Soc., Vol. 120, No. 10, 1998
Ashton et al.
Figure 3. 1H NMR spectrum (MSS, 400.1 MHz, 20 °C) of a 1:1
mixture of DB24C8 and 4‚PF6 recorded 90 min after dissolution. The
system has reached equilibrium, i.e., the relative intensities of the peaks
for both the uncomplexed (uc) and complexed (c) species remain
constant after this point. The initial concentrations of the macrocyclic
polyether and the salt were both 1.0 × 10-2 M. Chemical shift data
for the compounds and complexes are cataloged in Table 1.
At 20 °C, the equilibrium between the [DB24C8‚4]+ complex
Figure 2. 1H NMR spectra (MSS, 300.1 MHz, 20 °C) of a 1:1 mixture
of DB24C8 and 2‚PF6, illustrating the slow formation of the [DB24C8‚2]-
[PF6] complex. The spectra portrayed were recorded after (a) 5 min,
(b) 2.25 h, (c) 29 h, and (d) 9.5 days, respectively. Peaks associated
with the uncomplexed crown ether and dialkylammonium ion are
denoted by the descriptor uc, while those associated with the complex
are denoted by the descriptor c. The initial concentrations of the crown
ether and the salt were both 1.0 × 10-2 M. Chemical shift data for the
compounds and complexes are listed in Table 1.
1
and its components was established by H NMR spectroscopy
(Figure 3) approximately 90 min after dissolving an equimolar
mixture of DB24C8 and 4‚PF6 in the MSS. Indeed, the slippage
of the DB24C8 macroring over the c-C5H9 rings of the 4+ cation
1
is so rapid that we were unable to obtain a H NMR spectrum
where only the free crown ether and the free salt could be
observed at room temperature. Conversely, the 1H NMR
spectrum of a 1:1 solution (MSS) of DB24C8 and 5‚PF6 at 20
°C showed (Figure 4a) peaks for the uncomplexed crown ether
and salt only. Thus, there is clearly not enough thermal energy
present at ambient temperature to permit the slippage of the
DB24C8 macroring over the 5+ cation’s c-C6H11 groups such
that the ∆Gqon barrier can be surmounted. Encouragingly in
this case, however, this barrier could be overcome at elevated
temperatures. After the solution had been heated at 40 °C for
temperatures. In other words, DB24C8 and 3‚PF6 coexist as
two distinct chemical entities in solution, since only peaks for
1
the uncomplexed polyether and salt were visible16 in the H
NMR spectrum of a 1:1 mixture of DB24C8 and 3‚PF6 in the
MSS at 20 °C. Furthermore, this spectrum was not altered in
the slightest when the solution was heated to 50 °C, i.e., even
at elevated temperatures, the large ∆Gqon barrier is insur-
mountable. This observation suggests that this system is
extremely sensitive to steric effects:17 the difference between
the formation of a rotaxane-like species being detected or not
corresponds to a simple change of stopper groups from p-i-
PrC6H4 to p-t-BuC6H4. The sensitivity of this system may be
ascribed to the (relatively) small and semirigid cavity of the
DB24C8 macroring.
1
ca. 16 weeks, a H NMR spectrum was obtained (Figure 4b)
that revealed the formation of a new complex. Inspired by this
result, we sought a method by which this complex could be
produced on a preparative scale. It appeared to us that CH2Cl2
was the solvent of choice, since it boils at 40 °C (so that
temperature control of the reaction may be achieved simply by
heating the reaction mixture under reflux) and has a low DN,
thus permitting11 the formation of strong rotaxane-like com-
plexes. An indication that the ammonium-containing “dumb-
bell” 5+ was forming a complex with DB24C8, using this
solvent as the reaction medium, was provided by the fact that
the salt 5‚PF6swhich, under normal circumstances, is fairly
(16) See the Supporting Information for details.
(17) Hitherto, we have discovered6a,f-g slippage systems that are much
less susceptible to steric effects. For instance, a much broader range of
stopper groups can be used for slippage processes involving the larger and
more flexible crown ether bis-p-phenylene[34]crown-10.

Rotaxane or Pseudorotaxane?
J. Am. Chem. Soc., Vol. 120, No. 10, 1998 2301
Table 1. 1H NMR Data [δ Values] for Compounds and Complexes in the MSSa
-
DB24C8
(RCH2)2NH2+ PF6
b
b
b
c
compd/complex
R-OCH2
â-OCH2
γ-OCH2
C6H4
6.74
CH2
c-CnH2n-1
p-C6H4-i-Pr
DB24C8
3.97
3.72
3.63
2‚PF6
3.98
4.42
1.09, 2.78, 7.09-7.21
[DB24C8‚2][PF6]
3.92
4.04
4.02
3.59
3.74
3.73
3.30
3.57
3.57
6.57-6.66,
6.68-6.77
1.02, 2.66, 6.88, 7.05
4‚PF6
2.72
3.10
2.63
2.98
0.98-1.16, 1.41-1.64,
1.69-1.81, 1.85-2.12
0.80-0.96, 1.16-1.33,
1.41-1.64, 1.75-1.90
0.71-0.94, 0.97-1.21,
1.44-1.67
[DB24C8‚4][PF6]
5‚PF6
6.77
[DB24C8‚5][PF6]
6.76
0.47-0.63, 0.65-0.90,
1.08-1.32, 1.33-1.48
a The spectra were obtained at 20 °C on a Bruker AC300 spectrometer (at 300.1 MHz) or on a Bruker AMX400 spectrometer (at 400.1 MHz)
with the deuterated solvent as the lock and the residual solvent as internal reference. b The descriptors R, â, and γ are defined in Scheme 2.
c Methylene protons adjacent to the ammonium center.
Table 2. Conversion of 5‚PF6 into [DB24C8‚5][PF6] upon Heating
with a 3-Fold Excess of DB24C8 in CH2Cl2 at 40 °Ca
time (days)
5‚PF6 (%)b
[DB24C8‚5][PF6] (%)c
0
4
8
12
16
20
24
28
32
36
100.0
63.5
43.5
30.6
18.6
11.0
7.5
4.6
3.3
2.2
0.0
36.5
56.5
69.4
81.4
89.0
92.5
95.4
96.7
97.8
a Reaction monitoring. A small portion of the reaction mixture was
removed at the time stated. Thereupon, the solvent was evaporated
off under reduced pressure, then the 1H NMR spectrum of the residue
was recorded in the MSS. The relative intensities of the peaks
associated with the NH2+CH2 protons for the uncomplexed and
complexed 5‚PF6 saltswhich were observed at δ 2.63 and 2.98,
respectivelyswere then evaluated. b Percentage amount of the uncom-
plexed 5‚PF6 salt in the reaction mixture. c Percentage amount of the
complexed 5‚PF6 salt in the reaction mixture.
cyclic polyether for 4 days. Moreover, the 1H NMR spectrum
of the residuesproduced after the solvent had been evaporated
off from a small portion of the reaction mixturesrecorded in
the MSS, revealed the formation of the same species that had
1
been generated in our earlier experiments. We used H NMR
spectroscopy to study the disappearance of the uncomplexed
5+ cation and the appearance of its complex with DB24C8,
which was shown (vide infra) to be the rotaxane-like complex
[DB24C8‚5]+ by X-ray crystallography. By examining the
relative intensities of the doublet18 located at δ 2.63scorre-
sponding to the NH2+CH2 protons in the uncomplexed
“dumbbell”sand the multiplet18 situated at δ 2.98sarising from
the NH2+CH2 protons in the rotaxane-like complexswe were
able to monitor (Table 2) the gradual buildup of the complex
over a long period of time. Indeed, this experiment revealed
that, after 36 days, almost 98% of the “dumbbell” had been
converted19 into its complex with DB24C8. Moreover, the
complex could be easily isolated in 90% yield following a
simple recrystallization procedure that employed a vapor
diffusion technique. The 1H NMR spectrum (Figure 4c) of the
pure
Figure 4. 1H NMR spectra (MSS, 300.1 MHz, 20 °C) of (a) a 1:1
mixture of DB24C8 and 5‚PF6 (both 1.0 × 10-2 M), (b) the same
mixture after being heated at 40 °C for 16 weeks, and (c) the pure
[DB24C8‚5][PF6] complex (1.0 × 10-2 M). Peaks associated with
uncomplexed DB24C8 and 5+ are indicated by the descriptor uc, while
those associated with the [DB24C8‚5]+ complex are denoted by the
descriptor c. Chemical shift data for the compounds and complexes
are listed in Table 1.
(18) In the free 5+ cation, there is no coupling between the vicinal NH2
+
and CH2 protons since the NH2+ protons exchange with adventitious H2O
present in the solvent. On account of this exchange process, the signal for
the CH2 protons resonates as a doublet in the 1H NMR spectrum of the
uncomplexed cation. However, the crown ether’s presence in the [DB24C8‚5]+
complex results in a cessation of the exchange process, so that the signal
for the complex’s NH2+CH2 protons appears as a second-order multiplet
presumably as a result of vicinal coupling between the two sets of protons.
insoluble in CH2Cl2sdissolved completely when it was heated
under reflux in the presence of a 3-fold excess of the macro-

2302 J. Am. Chem. Soc., Vol. 120, No. 10, 1998
Ashton et al.
[DB24C8‚5][PF6] complex revealed no dissociation in the
MSSseven upon standing for several weeks at 20 °C. Nev-
ertheless, the complex was not stable in all solvents at room
temperature. When the [DB24C8‚5][PF6] complex was dis-
solved in (CD3)2SOsan electron-donating solvent that disrupts
hydrogen bonding interactions and thus inhibits pseudorotaxane
formation between DB24C8 and dialkylammonium ions11sat
25 °C, we noticed (Figure 5) that the signals associated with
its NH2+CH2 and OCH2 protons progressively diminished in
size with respect to the corresponding peaks identifiable with
the uncomplexed cation and polyether, thus demonstrating that
the complex can dissociate into its componentssviz., DB24C8
and 5‚PF6sunder the appropriate conditions. After 18 h,
complete extrusion of the 5+ “dumbbell” from the complex was
observed (Figure 5d).20 This “schizophrenic” type of behavior
raises the questionsIs the [DB24C8‚5][PF6] complex a rotaxane,
or is it a pseudorotaxane? At 20 °C, the complex is kinetically
inert in the MSS, but it is kinetically labile in (CD3)2SO since
the free energy of activation for extrusion of the cation 5+
becomes surmountable, in addition to the complex becoming
less stable than its components. Thus, the [DB24C8‚5][PF6]
complex appears to be a rotaxane when it is studied in nonpolar
solvent systems, but manifests itself as either a pseudorotaxane
or two distinct chemical species in polar solvents with high
DNs.21 By definition,5 rotaxanes are compounds in which
“bulky end groups prevent the extrusion of a threaded chain
from a macrocycle”. Therefore, the [DB24C8‚5][PF6] complex
is a pseudorotaxane, and not a rotaxane, since its c-C6H11
stoppers are not bulky enough to maintain its integrity under
all circumstances. Where do we draw the line between
rotaxanes and pseudorotaxanes, i.e., when does a complex stop
being a pseudorotaxane and become a rotaxane? The best
answer is obtained by employing a fuzzy14 description of these
systems, i.e., rather than impose an arbitrary and complicated
cutoff parameter which distinguishes rotaxanes from pseudoro-
taxanes, it is best to say that the transition is a gradual one22 in
which pseudorotaxanes progressively acquire more rotaxane-
like qualities until they are completely interlocked, as is observed
in genuine rotaxanated structures. For instance, it can be seen
that, as the cations’ stopper groups are changed from c-C5H9 to
1
Figure 5. Partial H NMR spectra ((CD3)2SO, 400.1 MHz, 25 °C)
portraying the dissociation of the [DB24C8‚5][PF6] complex (initial
concentration 1.0 × 10-2 M). The spectra illustrated were recorded
after (a) 3 min, (b) 48 min, (c) 4 h, and (d) 18 h, respectively. Peaks
associated with the complex are denoted by the descriptor c, while
those associated with the uncomplexed crown ether and dialkylammo-
nium ion are denoted by the descriptor uc.
(19) Further experimentation showed that the temperature and the
presence of a large excess of DB24C8 were necessary to maximize the
conversion of 5‚PF6 into the [DB24C8‚5][PF6] complex. We observed only
a 30% conversion of 5‚PF6 into the complex when it was heated under
reflux with DB24C8 (3 mol equiv) in CHCl3sa solvent possessing a similar
DN to CH2Cl2 but with a higher boiling point (61 °C). This observation is
readily explained by Le Chatelier’s principle, which states that the
equilibrium is driven to the side of the reactionsin this case, the
reactantssthat absorbs energy upon heating. Moreover, this principle can
also be used to explain the fact that we observed only an 88% conversion
of the salt 5‚PF6 into its DB24C8 complex when an equimolar mixture of
the two components was heated under reflux in CH2Cl2 for 92 days.
(20) The first-order rate constantscalculated by using the integrated form
of the first-order rate law and the 1H NMR spectroscopic datasfor the
extrusion of the 5+ cation from the [DB24C8‚5][PF6] complex was found
to be 1.2 × 10-4 s-1 in (CD3)2SO at 25 °C. This specific reaction rate
corresponds to the complex having a half-life (t1/2) of ca. 100 min in
(CD3)2SO at 25 °C.
(21) In many respects, systems such as these may be likened to
“molecular locks” (Fujita, M.; Ibukuro, F.; Yamaguchi, K.; Ogura, K. J.
Am. Chem. Soc. 1995, 117, 4175-4176) which have been described for
systems that exploit the reversibility of coordinate covalent bonds between
metals and ligands. In this instance, the overall bond (equivalent to the
sum of the noncovalent and mechanical bonding interactions present)
“locking” the DB24C8 macroring and the “dumbbell” 5+ together is stable
in low donicity solventsshowever, this “lock” is “released” in high donicity
solvents.
c-C6H11, the rotaxane-like properties of the DB24C8-containing
pseudorotaxanes are enhancedsat 20 °C, the [DB24C8‚4]+
complex dissociates in the MSS, while its c-C6H11-stoppered
congener [DB24C8‚5]+ does not. Ultimately, to see if we could
synthesize a system that possessed even more rotaxane-like
character, we examined the DB24C8-6‚PF6 system. As we
had observed with several of the systems studied earlier, the
1
room temperature H NMR spectrum16 of a 1:1 mixture of
DB24C8 and 6‚PF6 exhibited signals which could only be
attributed to the uncomplexed cation and crown ether. Similar
to what had been observed previously with an equimolar mixture
1
of DB24C8 and 3‚PF6 (vide supra), the H NMR spectrum of
this solution was not altered, even upon heating to 50 °C,
indicating that the c-C7H13 stoppers are far too bulky to permit
slippage of the DB24C8 macroring over them to form a
rotaxane-like structure in this case. As before, this observation
indicates that this system is highly susceptible to steric
(22) It should be noted that the demeanor of this transitional phase will
vary with solvent and temperature, i.e., the degree of rotaxane-like
characteristics that a particular species is endowed with depends strongly
upon the external conditions it experiences.

Rotaxane or Pseudorotaxane?
J. Am. Chem. Soc., Vol. 120, No. 10, 1998 2303
Table 3. Kinetic and Thermodynamic Data for the Synthesesa of the [2]Pseudorotaxanes [DB24C8‚2][PF6], [DB24C8‚4][PF6] and
[DB24C8‚5][PF6], in the MSS at Various Temperatures
temp
(°C)
Ka
-∆G°
kon
∆Gqon
koff
)
∆Gqoff
probe
b
c
d
f
e
f
complex
(M-1
)
(kcal mol-1
)
(M-1 s-1
)
(kcal mol-1
)
(s-1
(kcal mol-1
)
protonsg
[DB24C8‚2][PF6]
[DB24C8‚2][PF6]
[DB24C8‚4][PF6]
[DB24C8‚4][PF6]
[DB24C8‚5][PF6]
20
40
20
40
40
2870
2470
290
110
110
4.6
4.9
3.3
2.9
2.9
1.1 × 10-3
3.2 × 10-3
8.2 × 10-2
1.4 × 10-1
2.9 × 10-5
21.1
21.9
18.6
19.6
24.9
3.9 × 10-7
1.3 × 10-6
2.8 × 10-4
1.3 × 10-3
2.6 × 10-7
25.7
26.8
21.9
22.5
27.8
CHMe2
CHMe2
CH2NH2
CH2NH2
CH2NH2
+
+
+
a The reactions were followed with 1H NMR spectroscopy by monitoring the changes in the relative intensities of the signals associated with the
probe protons in the complexed and uncomplexed ammonium ions. b The Ka values were obtained from single-point measurements of the concentrations
of the complexed and uncomplexed cations, in the relevant 1H NMR spectrum, by using the expression Ka ) [DB24C8‚cation‚PF6]/
[DB24C8][cation‚PF6]. c The free energies of association (∆G°) were calculated from the Ka values by using the expression ∆G° ) - RT ln Ka.
d The second-order rate constants for the slippage processes (kon) were calculated employing eq 1. e The first-order rate constants for the extrusion
processes (koff) were calculated from the kon and Ka values by using the expression koff ) kon/Ka. f The free energies of activation for the slippage
(∆Gqon) and extrusion (∆Goqff) processes were calculated by using the relationships ∆Gqon ) -RT ln(konh/kT) and ∆Gq ) -RT ln(koffh/kT),
respectively, where R, h and k correspond separately to the gas, Planck and Boltzmann constants. g Probe protons locatedofofn the ammonium ion.
Table 4. LSIMS Dataa for DB24C8-Secondary
Dialkylammonium Salt (D‚PF6) Mixtures and Crystals of Their
Associated Pseudorotaxanes
effects17sslippage of the DB24C8 macroring over the stopper
units of the dialkylammonium ions rapidly becomes more
difficult as more CH2 groups are inserted into the cations’
cycloalkyl termini.
mixture or complex
m/z D+ d
m/z [DB24C8‚D]+ e(%)f
b
DB24C8 + 2‚PF6
282
282
310
182
182
210
210
238
730 (10)
730 (667)
758 (6)
630 (145)
630 (149)
658 (7)
The chemical shift data for the pseudorotaxanes and their
components are presented in Table 1. In general, downfield
shifts were observed for the resonances associated with the
ammonium ions’ protons upon complexation, the largest shifts
being exhibited by the methylene protons adjacent to the
ammonium center. For the most part, the proton resonances
associated with the DB24C8 macroring were shifted to a lesser
extent than their counterparts on the ammonium ions.
Thermodynamics and Kinetics of Pseudorotaxane Forma-
tion. The absolute concentrations of the uncomplexed crown
ether/salts and their associated pseudorotaxane complexes can
be calculated11 from the known original concentrations of
DB24C8 and the ammonium salts, in addition to the relative
intensities of the 1H NMR signals associated with each species.
Hence, at equilibrium, the observation of both complexed and
uncomplexed species in solution provides a suitable single-point
determination23 of the complexes’ Ka values (Table 3) and their
derived free energies of complexation (∆G°). Not surprisingly,
the [DB24C8‚2][PF6] complex possesses the highest Ka values,
probably as a result of the greater acidity of the benzylic CH2
protons with respect to their cycloalkyl CH2 counterparts, in
addition to the presence of supplementary stabilizing aryl-aryl
stacking interactions. Noticeably, in all cases, the Ka values
decrease at higher temperatures, an effect that may be rational-
ized utilizing Le Chatelier’s principle, i.e., at higher tempera-
tures, the equilibrium is shifted toward the species that absorbs
energy.
[DB24C8‚2][PF6]c
b
DB24C8 + 3‚PF6
DB24C8 + 4‚PF6
b
[DB24C8‚4][PF6]c
b
DB24C8 + 5‚PF6
[DB24C8‚5][PF6]c
658 (250)
687 (3)
b
DB24C8 + 6‚PF6
a LSIMS spectra were obtained by using a VG Zabspec mass
spectrometer equipped with a cesium ion source and utilizing a
m-nitrobenzyl alcohol matrix. b Equimolar mixture of DB24C8 and
dialkylammonium salt. c Crystals of pseudorotaxane. d (M - PF6)+ peak
corresponding to the uncomplexed dialkylammonium ion D+. e (M -
PF6)+ peak corresponding to a DB24C8-dialkylammonium ion com-
plex [DB24C8‚D]+. f Relative intensity of the [DB24C8‚D]+ peak
compared to the D+ peak expressed as a percentage.
the pseudorotaxanes at equilibrium and at time, t, respectively.
Nonlinear curve fitting of the plot of Pt against t furnished values
of kon from which estimations of the respective free energies of
activation for slippage (∆Gqon) and extrusion (∆Gqoff) could be
determined (Table 3). As anticipated from previous experi-
ments, the free energy of activation for the slippage of the
DB24C8 macroring over the dialkylammonium ions’ stopper
groups, viz., ∆Goqn, increases in the series 4+ < 2+ < 5+,
indicating that it is relatively easy for DB24C8 to slip over
c-C5H9 rings, while it is much more difficult for it to slip over
c-C6H11 rings. Moreover, the free energy of extrusion, viz.,
∆Gqoff, also increases in the same series. Since this value is a
measure of the kinetic stability of the pseudorotaxane super-
molecules in the MSSsi.e., it represents a barrier that must be
overcome for complex dismemberment to occurswe can say
that the [DB24C8‚5]+ species has more rotaxane-like character
than the [DB24C8‚2]+ species, which, in turn, has more
rotaxane-like character than the [DB24C8‚4]+ system.
To examine the kinetic parameters associated with these
1
systems, pseudorotaxane formation was studied by H NMR
spectroscopy at 20 and 40 °C in the MSS. By monitoring the
decrease in the relative intensities of the 1H NMR signals
associated with various probe protons in the free dialkylam-
monium ions with respect to their concomitant appearance in
the complexes with time, we could calculate (Table 3) the rate
constants for the slippage processes, kon, using eq 1:24
Mass Spectrometry. Data acquired from the liquid second-
ary ion mass spectra (LSIMS) of 1:1 mixtures of DB24C8 and
the secondary dialkylammonium salts (D‚PF6), in addition to
the LSIMS data for cocrystals of some of the DB24C8-
dialkylammonium salt complexes, are displayed in Table 4. In
all instances, the peaks observed for the uncomplexed macro-
cyclic polyether were very small and insignificant25 compared
2
2
2
2
Pt ) [D20Pee(k t(D -P )/P ) - D20Pe]/[D02e(k t(D -P )/P ) - P2e]
(1)
on
0
e
e
on
0
e
e
where D0 corresponds to the initial concentration of the
uncomplexed dialkylammonium salts or macrocyclic polyether,
while Pe and Pt are commensurate with the concentration of
(24) This equation was derived16 from the integrated rate equation for
an equilibrating reaction between two reactantssin this instance, DB24C8
and the dialkylammonium ionsand a single productshere, the 1:1
pseudorotaxane complex formed between DB24C8 and the dialkylammo-
nium ion.
(23) Adrian, J. C., Jr.; Wilcox, C. S. J. Am. Chem. Soc. 1991, 113, 678-
680 and references therein.

2304 J. Am. Chem. Soc., Vol. 120, No. 10, 1998
Ashton et al.
to the peak for the uncomplexed ammonium ions 2+-6+.
However, several important features became evident when we
examined the relative intensities of the peaks associated with
the uncomplexed ammonium ions and their DB24C8 complexes.
Only in the case of the [DB24C8‚4]+ complex were peaks of
approximately the same intensity detected for (1) a 1:1
DB24C8-secondary dialkylammonium salt mixture26 and (2)
cocrystals26 of the corresponding pseudorotaxane, indicating that
the “gas phase” rates of complexation and decomplexation are
both relatively fast for this system. On the other hand, vast
differences were noted between (1) the LSIMS of 1:1 mixtures26
of DB24C8 with the salts 2‚PF6 and 5‚PF6 and (2) the LSIMS
of crystals26 of the corresponding 1:1 complexes, viz.,
[DB24C8‚2][PF6] and [DB24C8‚5][PF6]. The [DB24C8‚2]+
and [DB24C8‚5]+ peaks observed in the LSIMS spectra of 1:1
mixtures26 of DB24C8 with either 2‚PF6 or 5‚PF6si.e., systems
in which no 1:1 complexes were present initiallyswere
infinitesimal compared to the signals retrieved from their
associated cocrystals.26 This observation indicates that these
systems’ complexation/decomplexation rates are exceptionally
slow during their flight through the mass spectrometer. Equimo-
lar mixtures26 of DB24C8 with the salts 3‚PF6 and 6‚PF6ssalts
which were shown (vide supra) not to form complexes with
DB24C8 under any circumstances in solutionsexhibit LSIMS
which display very weak signals for the 1:1 DB24C8-
dialkylammonium ion complexes. These very weak signals
presumably arise as a result of a small amount of “face-to-face”
complexation between the ammonium ions and the macrocyclic
polyether in the “gas phase”. Interestingly, these signals were
smaller than those detected in the LSIMS of 1:1 mixtures of
DB24C8 with either 2‚PF6 or 5‚PF6, perhaps indicating that there
is a certain degree of formation of [DB24C8‚2]+ or [DB24C8‚5]+
pseudorotaxane superstructures in the “gas phase”.
Figure 6. The solid-state superstructure of the [2]pseudorotaxane
[DB24C8‚2][PF6]. The hydrogen bonding geometries are (a) [N+‚‚‚O],
[H‚‚‚O] 2.96, 2.11 Å, [N+-H‚‚‚O] 158° and (b) [C‚‚‚O], [H‚‚‚O], 3.26,
2.43 Å, [C-H‚‚‚O] 145°.
X-ray Crystallography. The X-ray analysis of the 1:1
pseudorotaxane complex formed between the 2+ cation and
DB24C8 reveals (Figure 6) a structure that has a co-conforma-
tion27 very similar to that observed (cf. Scheme 1) for the related
[DB24C8‚1]+ complex.11 The DB24C8 macrocycle adopts an
extended, approximately Ci-symmetric conformation reminiscent
of that observed28 in its uncomplexed state, i.e., its two catechol
rings retain coplanarity with their adjoining OCH2CH2 frag-
ments. One of the phenylene rings of the cation is positioned
over one of the catechol rings of the DB24C8 macroring (mean
interplanar separation 3.56 Å, centroid-centroid distance 4.38
Å, the rings being inclined by ca. 14° to one another), while
the other phenylene ring is oriented approximately axially to
Figure 7. The solid-state superstructure of one of the two crystallo-
graphically independent [2]pseudorotaxanes present in the crystals of
[DB24C8‚4][PF6].
this macrocycle.29 Complex stabilization is achieved via π-π
stacking interactions (vide supra) and a combination of both
[N+-H‚‚‚O] and [C-H‚‚‚O] hydrogen bonds (Figure 6).
Inspection of the packing of the [DB24C8‚2]+ pseudorotaxane
supermolecules does not reveal any additional intercomplex
interactions in the crystal.
(25) This feature of the LSIMS arises since the dialkylammonium ions
are already charged and do not need to be ionized before they are desorbed
into the “gas phase” from the matrix. On the other hand, the neutral crown
ether molecules are not desorbed to such a great extent since they must be
ionized beforehand.
(26) In solvent-free 1:1 mixtures of DB24C8 and the (RCH2)2NH2‚PF6
salts, no pseudorotaxane complexes are present originallysthe only chance
that the compounds have to form pseudorotaxanes, prior to desorption,
occurs when they are dissolved in the matrix. Consequently, the pseudoro-
taxane content of samples produced from DB24C8-(RCH2)2NH2‚PF6
mixtures may be considered to be negligible before desorption. On the other
hand, cocrystals of DB24C8 with the saltssisolated after the components
had been allowed to equilibrate completelysmay be considered to consist
of pure pseudorotaxane, as verified by X-ray crystallographic analysis (vide
supra).
(27) Strictly speaking, the term “conformation” refers only to the three-
dimensional spatial organization of atoms in discrete molecular species that
are interconvertible merely by free rotation about chemical bonds. Accord-
ingly, we have adopted10e the expression “co-conformation” to designate
the three-dimensional spatial arrangement of the atoms in supramolecular
systems that can be interconverted by bond rotation.
The solid state superstructure of the 1:1 pseudorotaxane
complex is markedly altered on changing the cation from one
+
containing substituted benzyl groups, attached to the NH2
center, to one carrying cyclopentylmethyl substituents. In this
instance (Figure 7), the DB24C8 macroring adopts a folded,
V-shaped conformation, with the cation inserted approximately
axially through its center. The crystals contain two crystallo-
graphically independent [DB24C8‚4]+ pseudorotaxanes, both
with very similar co-conformations.27 The “cleft angles”
between the two catechol rings in the two complexes are 47
and 50°, respectively. The cations’ CHCH2NH2+CH2CH
backbones both have all-anti geometries and are slipped such
that, in each case, the nitrogen atoms lie ca. 0.6 Å from the
centroid of the eight polyether oxygen atoms. The centroid-
(28) Hanson, I. R.; Hughes, D. L.; Truter, M. R. J. Chem. Soc., Perkin
Trans. 2 1976, 972-978.
(29) Stronger π-π stacking interactions between the two aromatic
systems are probably precluded by the bulky i-Pr substituent.

Rotaxane or Pseudorotaxane?
J. Am. Chem. Soc., Vol. 120, No. 10, 1998 2305
Figure 8. The two crystallographically independent 1:1 complexes [DB24C8‚5][PF6]. In complex A, the hydrogen bonding geometries are (a)
[N+‚‚‚O], [H‚‚‚O] 3.05, 2.20 Å, [N+-H‚‚‚O] 156° and (b) [C‚‚‚O], [H‚‚‚O] 3.30, 2.35 Å, and [C-H‚‚‚O] 170°. In complex B, the hydrogen
bonding geometries are (c) [N+‚‚‚O], [H‚‚‚O] 2.89, 2.01 Å, [N+-H‚‚‚O] 165°, (d) 2.87, 2.24 Å, 126°, and (e) [C‚‚‚O], [H‚‚‚O] 3.14, 2.27 Å
[C-H‚‚‚O] 151°.
Table 5. Crystal Data, Data Collection, and Refinement Parametersa
data
[DB24C8‚2][PF6]
[DB24C8‚4][PF6]
C36H56NO8‚PF6
[DB24C8‚5][PF6]
C38H60NO8‚PF6
formula
solvent
C44H60NO8‚PF6
0.25 C6H14
897.4
formula wt
color, habit
crystal size/mm
lattice type
space group symbol, no.
cell dimensions
a/Å
775.8
803.8
clear prisms
0.77 × 0.70 × 0.60
monoclinic
P21/n, 14
clear needles
1.00 × 0.40 × 0.17
orthorhombic
Pbca, 61
clear prisms
0.47 × 0.43 × 0.40
orthorhombic
Pbca, 61
12.202(3)
10.749(1)
40.325(4)
98.60(1)
5230(1)
4
16.264(1)
30.838(2)
32.173(4)
16.539(1)
31.722(3)
32.130(2)
b/Å
c/Å
â/deg
V/Å3
16137(3)
16b
16857(2)
16b
Z
Dc/g cm-3
F(000)
1.140
1.277
6592
1.26
1.267
6848
1.23
1906
µ/mm-1
1.04
θ range/deg
no. of unique reflcns
measd
2.2-55.0
2.8-55.0
2.8-60.0
6490
10142
12308
obsd, |Fo| > 4σ(|Fo|)
2589
4803
4832
no. of variables
638
618
641
c
R1
0.114
0.146
0.197
d
wR2
0.309
0.376
0.539
weighting factors a, be
0.246, 0.000
0.73, -0.27
0.215, 71.165
0.68, -0.45
0.415, 25.238
0.78, -0.57
largest difference peak, hole/eÅ-3
a Details in common: graphite monochromated Cu KR radiation, ω-scans, Siemens P4 diffractometer, 293 K, refinement based on F2. b There
2
2
are two crystallographically independent molecules in the asymmetric unit. c R1 ) Σ||Fo| - |Fc||/Σ|Fo|. d wR2 ) {Σ[w(Fo - Fc2)2]/Σ[w(Fo )2]}.1/2
e w-1 ) σ2(Fo ) + (aP)2 + bP.
2
centroid distances between the c-C5H9 rings and their associated
“sandwiching” catechol units are 4.26 and 4.67 Å in one
complex, while they are 4.37 and 4.74 Å in the other. There is
a marked absence of any strong [N+-H‚‚‚O] or [C-H‚‚‚O]
hydrogen bonding interactions. The shortest respective [N‚‚‚O]
distances are 3.04 and 3.08 Å in the two complexes, but the
closest approaches of any of the NH2+ hydrogen donors to any
potential DB24C8 oxygen acceptors are all greater than 2.3 Å.
Similarly, the shortest contact from either of the cations’
NH2+CH2 hydrogen atoms to the polyether oxygen atoms is
2.44 Å. As was the case with the [DB24C8‚2]+ complex, there
are no intercomplex interactions of any note.
In the solid-state superstructure of the [2]pseudorotaxane
[DB24C8‚5]+, there are also two crystallographically indepen-
dent 1:1 complexes in the asymmetric unit. Although the
DB24C8 macrocycle adopts a folded, V-shaped conformation
in both cases, with respective “cleft angles” of 57 and 55°, the
co-conformations27 and geometries of the binding of the cations
differ markedly in each case. In one of the complexes, the
geometries about the two C-N bonds within the cation are anti,
whereas in the other, the geometry about one of these bonds is
gauche. In both independent complexes, one of the c-C6H11
rings is sandwiched between the catechol rings of the DB24C8
macrocycle with ring-centroid-ring-centroid separations of 4.39
and 4.69 Å in one complex and 4.49 and 4.70 Å in the other.
In both pseudorotaxane complexes, there are distinct [N+-
H‚‚‚O] and [C-H‚‚‚O] stabilizing interactions (Figure 8). As
with the previous two complexes, there are no notable interp-
seudorotaxane interactions.
The crystal data, data collection and refinement parameters
for the crystal structures reported in this article are summarized
in Table 5.
Molecular Modeling. To gain further insight into the size
complementarity requirements associated with the slippage
processes, the passage of the macrocyclic polyether over the
terminal aryl or cycloalkyl groups of the dumbbell-like com-

2306 J. Am. Chem. Soc., Vol. 120, No. 10, 1998
Ashton et al.
Figure 10. Schematic illustration, indicating that pseudorotaxanes are
represented by the intersection set [R ∩ I] of the sets of rotaxanes (R)
and isolated chemical entities (I), i.e., the fuzzy domain between these
two sets. Consequently, pseudorotaxanes are species that are endowed
simultaneously with characteristics associated with species belonging
to both R and I. The numbers 1 and 0 have been assigned arbitrarily
to the respective species which belong either totally or not at all to the
sets R and I.
the energy barrier associated with the passage of the macrocyclic
component over one of the terminal cycloalkyl groups increases
(Figure 9b) with the number of methylene groups incorporated
within the cycloalkyl residues. In conclusion, for both the aryl-
and cycloalkyl-terminated cations 1+-3+ and 4+-6+, respec-
tively, the slippage of the macrocyclic component over the
terminal groups becomes increasingly difficult as the size of
these groups is enlarged. This trend is consistent (vide supra)
with the experimental observations, both in solution and in the
“gas phase”. Nonetheless, the energy barriers derived from the
substituted benzyl-terminated dumbbell-like cations 2+ and 3+
are significantly higher than those determined for their cy-
cloalkyl-containing congeners 5+ and 6+. These findings are
apparently in contrast with the experimental observations, which
reveal (vide supra) that the macrocyclic polyether is able to pass
over the p-i-PrC6H4 groups of 2+, but not over the c-C7H13
groups of 6+si.e., the free energy of activation associated with
the slippage process, viz., ∆Gqon, is higher in the case of 6+
than that of 2+. This apparent contradiction is presumably a
result of entropic effects, which are expected to be significant
for the rather flexible cycloalkyl groups under the experimental
conditions. However, the entropic term is not taken into account
by the molecular mechanics calculations, since the calculated
∆E values can be equated, at best, with enthalpy differences at
0 K and not with free energy differences. Thus, the energy
barriers obtained from these slippage simulations must be
considered as relatiVe measures, comparable only when similar
dumbbell-like cations are being considered.
Figure 9. Schematic diagram illustrating the protocol utilized for the
computational simulations delineated in this article, in addition to energy
profiles associated with the passage of the macrocyclic polyether
DB24C8 over the (a) terminal aryl groups of the cationic “dumbbells”
1+-3+ and (b) cycloalkyl termini of the dumbbell-like cations 4+-
6+.
ponents 1+-6+ was simulated computationally. In particular,
the preformed complexessconstructed within the input mode
of Macromodel30 5.0swere employed as the simulations’ initial
geometries. In each case, the distance, D (Figure 9), separating
the carbon atom in position 4 of one of the macrocyclic
polyether’s two catechol rings and a reference “dummy”
atomslocated at a distance of 60.0 Å from the nitrogen
atomswas reduced stepwise from 56.5 to 35.0 Å. As a result,
the macrocycle was forced to move along the dumbbell-like
component, eventually passing over one of its two termini. An
energy minimization was performed at each stepsthis step’s
initial geometry coincided with the previous step’s final
geometry. The energies obtained for each step were “normal-
ized” by subtracting the concluding step’s energyscorresponding
to the energy associated with the well-separated macrocyclic
and dumbbell-like componentssfrom each value. The resulting
energy differences ∆E were plotted against the values of D to
furnish the energy profiles illustrated in Figure 9. In the case
of the dumbbell-like cation 1+ (Figure 9a)si.e., when R is equal
to Phsthe energy profile is remarkably “flat”. However, by
increasing the size of Rsi.e., by replacing Ph for p-i-PrC6H4
and p-t-BuC6H4 in 2+ and 3+, respectivelyspronounced energy
barriers were observed at D values of 44.0 and 43.0 Å, for 2+
and 3+, individually. Similarly, in the case of the dumbbell-
like cations 4+-6+, incorporating terminal cycloalkyl groups,
Conclusions
By modifying the termini of the cation 1+, we have
demonstrated that it is possible to synthesize rotaxane-like
speciesscomposed of DB24C8 and secondary dialkylammo-
nium ionsswhich are kinetically stable under the appropriate
conditions. In particular, we have shown that the [DB24C8‚5]-
[PF6] complex is stable at ambient temperature in a solvent
endowed with a low DN. “Under the appropriate conditions”
is the key phrase in this instance, for the other rotaxane-like
complexes which we have studied in this article will undoubt-
edly be kinetically stable at lower temperatures. These examples
(30) Mahamadi, F.; Richards, N. G. K.; Guida, W. C.; Liskamp, R.;
Lipton, M.; Caufield, D.; Chang, G.; Hendrickson, T.; Still, W. C. J. Comput.
Chem. 1990, 11, 440-467.

Rotaxane or Pseudorotaxane?
J. Am. Chem. Soc., Vol. 120, No. 10, 1998 2307
serve to highlight the fact that the transformation from a
pseudorotaxane into a rotaxane is a progressive one, and implies
that some pseudorotaxanes possess more rotaxane-like character
than others. Moreover, it shows that the concept of a pseu-
dorotaxane is one that is inherently vague and is best described
using fuzzy sets,14 in which the set of pseudorotaxanes (P)
belongs (Figure 10), to some extent, to both the set of
mechanically interlocked rotaxanes (R) and the set of two
isolated chemical entities (I), i.e., P ) [R ∩ I], a description
that provides much more meaning than one in which a definite
cutoff is made between rotaxanes and pseudorotaxanes.
Acknowledgment. This research was sponsored in the U.K.
by an Engineering and Physical Sciences Research Council
CASE Award (to M.C.T.F.).
Supporting Information Available: Experimental proce-
duressincluding the derivation of eq 1scrystallographic data,
1
and H NMR spectra (43 pages). See any current masthead
page for ordering information and Web access instructions.
JA9731276