A R T I C L E S
Dominguez et al.
phenylene rings and one with “facile” flipping motion.29 The
Scheme 3
“
mobile” polymorph was shown to belong to the space group
C2 with two independent half molecules per asymmetric unit.
As expected from its crystal symmetry, two dynamic environ-
ments were indeed observed. The most mobile phenylene was
6
-1
shown to undergo flipping motions at a rate of 5 × 10 s at
3
05 K. Phenylene motion in the static polymorph was slower
2
-1
2
than 10 s , as suggested by the H NMR data.
In the case of more complex polymeric samples, rotational
rate constants and activation energies reflect the structural and
dynamic heterogeneity of each material. Despite a range of
dynamics properties, a 2-fold flipping model has been proven
suitable in most cases. Deuterium line shape analyses with poly-
crystalline penicillin V dynamics by X-ray and 13C CPMAS
NMR, Wendeler and co-workers documented the rate of phenyl
flipping in various crystalline salts. Crystals of the free acid,
23
+
+
+
+
+
H , and with Li , Na , Rb , and Cs were shown to influence
the rate of phenyl rotation over 8 orders of magnitude at ambient
(
butylene therephthalate) (PBT) at 70 °C revealed a crystalline
2
-1
phase with very slow motion (krot < 10 s ), an amorphous
+
temperature. The rates of rotation for the Li salt were the
5
-1
phase with fast flips at a rate of krot ≈ 5.5 × 10 s , and a
-
1
+
slowest (∼0.1 s at 300 K), and the rates in the Rb salt were
5
-1
slower interphase with flipping rates of krot ≈ 1.2 × 10 s .
An activation energy of only 5.9 kcal/mol calculated for rotation
in the amorphous region suggests a fluidlike environment.
Dynamic studies with phenyl deuterated PET were interpreted
in terms of distributions of rate constants and narrow distribution
8
-1
the fastest (5 × 10 s ). Differences in activation energy for
phenyl flip in these crystals ranged by a factor of 2, between
1
2.3 and 26.5 kcal/mol. Interestingly, differences in flipping
rates in these cases arise primarily from differences in the
calculated preexponential factors which varied by as much as
3
-1
of flip angles giving rotational rate constants of <10 s at
ambient temperature. Rates of phenylene flip in the case of
14
18 -1 24
5
orders of magnitude (ca. 10 -10 s ). In another example
30
of phenyl flipping in crystals reported by Casarini and co-
bisphenol-containing polymers have also been measured in the
2
3 -1
workers, rates of rotation of 3.5 × 10 and 1.2 × 10 s were
calculated at ambient temperature for the two formally enan-
tiotopic phenyl groups of diphenyl sulfoxide.25
4
6
-1
31
1
0 -10 s range in the amorphous regions. Activation
energies measured in the amorphous phase of various polymers
range between 6 and 20 kcal/mol.
If a single point of attachment reduces the rates of rotation
of phenyl groups in crystals, one may expect that the linkage
of a benzene ring along its 1,4-axis may restrict its rotational
3. Conclusions
1
3
An ideal crystalline geared lattice would be characterized by
the concerted rotation of individual molecular rotors, such that
they can influence each other over long distances by mechanisms
that transfer angular momentum from one to another in a
structurally predetermined manner. Extended correlated motions
in crystals would require precise rotational symmetry with
clockwise and anticlockwise rotors acting on each other without
interference. Parity rules that determine the motion of molecular
gears in a chain have been analyzed by Mislow,32 and the
transfer of angular momentum at the molecular level in
molecular gear trains of triptycyl ethers has been analyzed in
solution by Iwamura and co-workers. An ideal crystalline
geared lattice would “operate” under thermal equilibrium by
exchanging energy between molecular modes and extended
phonons. Unless activation energies and preexponential factors
are identical for different rotors, geared lattices would only
operate under a limited temperature range when their rotational
motions become isochronous or commensurate. Regarding the
benzene clathrate of compound 3, it is clear that in-plane rotation
of the benzene molecules and 2-fold phenylene flipping occur
with very different rates and very different activation energies.
dynamics even further. However, coalescence analysis by
CPMAS NMR in the case of 3 revealed rates of rotation of ca.
C
3
-1
1
0 s at ambient temperature, and an activation energy of only
2.8 kcal/mol was deduced. Although this analysis is of limited
1
precision, it is remarkable that rotation of the phenylene group
in desolvated samples, determined by deuterium line shape
analyses over a wider temperature range (Figure 13), has a
remarkably similar activation energy of only 14.6 kcal/mol.
Despite limitations in our data, a review of the literature
reveals that the phenylene group in compound 3 performs as
an excellent molecular rotor. There are fewer examples of
phenylene rotation in solids reported, and most of them relate
to phenylene-containing polymers. A few examples of phenylene
rotation in crystalline samples reveal a very slow process. In
one case, measurements carried out with drawn crystalline fibers
of poly(ethyleneterephthalate) (PET, R ) CH2-CH2-, Scheme
3
3
1
3
3
) measured by C exchange NMR give 2-fold flipping rate
-
1
26
constants of 1.5 s at 293 K. Phenylene flip in crystals of
,4-diphenoxy benzene, a model for poly(phenyl-O(-phenylene)
PPE), was shown to occur at a rate of 1.3 s-1 at 300 K.27
Similarly, studies with crystalline bisphenol phenyl carbonate
1
(
8
-1
While a lower limit of krot > 10 s was deduced for in-plane
(
BPPC, R ) -CO2Ph), a model for poly(bisphenyl carbonate)
28
(PBPC), revealed two crystalline forms, one with “immobile”
(
28) Enrichs, P. M.; Luss, H. R. Macromolecules 1988, 21, 860.
29) Enrichs, P. M.; Luss, H. R. Macromolecules 1988, 22, 2731.
(
(
23) Wendeler, M.; Fattah, J.; Twyman, J. M.; Edwards, A. J.; Dobson, C. M.;
Heyes, S. J.; Prout, K. J. Am. Chem. Soc. 1997, 119, 9793-9803.
(30) (a) Horii, F.; Kaji, H.; Ishida, H.; Kuwabara, K.; Masuda, K.; Tai, T. J.
Mol. Struct. 1998, 441, 303-311. (b) Kawaguchi, T.; Mamada, A.;
Hosokawa, Y.; Horii, F. Polymer 1998, 39, 2725-2732.
(24) Some of the suggested preexponential factors may be in error as they exceed
1
5
the limit of rotational inertia (Φ
0
) for a 2-fold-flip of ca. Φ
0
) 2.4 × 10
(31) (a) Shi, J.-F. I., P. T.; Jones, A. A.; Meadows, M. D. Macromolecules
1996, 29, 605-609. (b) Klug, C. A. W.; Jinhuang Xiao, C.; Yee, A. F.;
Schaefer, J. Macromolecules 1997, 30, 6301-6306.
-1
s
.
(
25) Casarini, D.; Lunazzi, L.; Mazzanti, A. Angew. Chem., Int. Ed. 2001, 40,
2
536-2540.
(32) Mislow, K. Chemtracts: Org. Chem. 1988, 2, 151-174.
(33) (a) Kawada, Y.; Iwamura, H. J. Am. Chem. Soc. 1983, 105, 1449-1459.
(b) Koga, N.; Kawada, Y.; Iwamura, H. Tetrahedron 1986, 42, 1679-
1686.
(
(
26) Wilhelm, M. S., H. W. Macromolecules 1996, 29, 1088-1090.
27) Reichert, D.; Hempel, G.; Zimmermann, H.; Schneider, H.; Luz, Z. Solid
State Nucl. Magn. Reson. 2000, 18, 17-36.
7726 J. AM. CHEM. SOC.
9
VOL. 124, NO. 26, 2002