Rotational oscillation of two interlocked porphyrins in cerium bis(5,15-diarylporphyrinate) double-deckers

Article abstract of DOI:10.1039/a801350k

Dynamic NMR coupled with racemization studies on chiral cerium bis(5,15-diarylporphyrinate) double-deckers demonstrate that the two interlocked porphyrin ligands oscillate rotationally around the metal center.

Full text of DOI:10.1039/a801350k

View Article Online / Journal Homepage / Table of Contents for this issue
Rotational oscillation of two interlocked porphyrins in cerium
bis(5,15-diarylporphyrinate) double-deckers
Kentaro Tashiro, Takashi Fujiwara, Katsuaki Konishi and Takuzo Aida*†
Department of Chemistry and Biotechnology, Graduate School of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku,
Tokyo 113-8656, Japan
(a)
Hb′
Hb
Dynamic NMR coupled with racemization studies on chiral
cerium bis(5,15-diarylporphyrinate) double-deckers demon-
strate that the two interlocked porphyrin ligands oscillate
rotationally around the metal center.
R1
Ha′
R1
Ha
Ha
R2
Ha′
1a R1 = R2
=
=
Me
N
Ce
N
N
N
N
But
But
N
N
Oscillating molecules are interesting in their potential utilities
for sensors, information transfer/storage devices and molecular
machines, and several examples of molecular and supramo-
lecular oscillating systems have been reported to date.1,2
Particularly interesting are oscillators consisting of chromo-
phoric units because of a possible switching of their oscillation
characteristics by light and/or redox. Metal bis(porphyrinate)s
are double-decker complexes, in which a metal atom is
sandwiched by two chromophoric porphyrin ligands.3 Recently,
we have found that the porphyrin ligands in cerium bis(porphyr-
inate)s rotate thermally around the metal center, through studies
on optical resolution and racemization behavior of a chiral
cerium bis(tetraarylporphyrinate) with a symmetry group D2.4
The chirality of the complex originates from the staggered
geometry of the two facing porphyrin ligands, where the relative
rotation of the ligands leads to racemization. Herein we report
that cerium bis(5,15-diarylporphyrinate)s are potential molec-
ular oscillators, in which the two interlocked porphyrin ligands
rotationally oscillate back and forth around the metal center.
We have already reported that the D2-symmetric, chiral
cerium complex 1a, in contrast to a zirconium analogue of the
same porphyrin ligands, shows no sign of optical resolution in
chiral HPLC, as a result of racemization via a fast rotation of the
porphyrin ligands.4 Our motivation to the present study arose
from a preliminary observation that enantiomeric separation of
2a‡ by chiral HPLC was also unsuccessful in spite of the
presence of interferant bulky meta-substituents (But) on the
meso-phenyl groups perpendicular to the porphyrin plane. In
order to compare the rotatability of the porphyrin ligands in 2a
with that of 1a, temperature-dependent NMR signal coales-
cence profile was investigated for exchangeable proton signals
via ligand rotation. At a low temperature such as 220 °C in
CDCl3, 2a showed four doublet signals at d 8.48 (a), 8.58 (aA),
8.76 (b) and 8.84 (bA) [Scheme 1(a)] owing to nonequivalent
pyrrole-b protons,§ where the exchangeable pairs are signals
a/aA and b/bA. Upon elevating the temperature, the paired signals
a and aA, for example, coalesced completely at 0 °C. From this
coalescence profile, the rate constant for the rotation of the
porphyrin ligands in 2a was evaluated at 0 °C to be 0.52 3 102
N
2a R1 = R2
R2
Hb
Hb′
Hb′
Hc
(b)
R1
Ha′
R1
Ha
Hd
R2
Hd′
N
Ce
N
N
N
N
N
N
N
R2
Hc′
Hb
O
1b R1
=
=
Me
O
O
O
R2
=
But
But
2b R1
O
O
Scheme 1
conformation, thereby allowing selective evaluation of the
rotating rate via path II. Prior to dynamic NMR studies, we
investigated the possibility of optical resolution of these two
strapped complexes by chiral HPLC,∑ where 2b was success-
fully separated into enantiomers. The enantiomers exhibited
clear mirror-image circular dichroism (CD) spectra to each
other (Fig. 1), but the spectral intensity at 10 °C gradually
decreased with time, where the rotating rate constant via path II
was evaluated to be 0.39 3 1023 s21.¶ In sharp contrast,
s21,¶ which is almost comparable to that of 1a (0.63 3 102 s21
)
at 1 °C.4
Here one should note two possible rotating paths, I and II, for
the complex to racemize (Scheme 2), where path II is sterically
more demanding than path I, since the former path must
transiently involve an energetically disfavored ‘eclipsed con-
formation’ with respect to the two facing porphyrin ligands. In
order to evaluate the rotating rate via path II, we synthesized
two new chiral double-decker complexes 1b and 2b, which bear
a flexible oligoether strap between the two porphyrin ligands.‡
Their CPK models suggested that path I is unlikely to occur
even when the oligoether strap adopts the most extended
Scheme 2
Chem. Commun., 1998
1121

View Article Online
1.5
1.0
Notes and References
† E-mail: aida@macro.t.u-tokyo.ac.jp
‡ According to the method reported in ref. 4, 2a was obtained as brown
powder in 11% yield. 1H NMR (270 MHz, CDCl3, 220 °C) d 9.57 (s, 4 H,
o-endo-H), 9.14 (s, 4 H, meso), 8.84, 8.76, 8.58, 8.48 (d 3 4, J 4 Hz, 16 H,
pyrrole-b), 7.78 (s, 4 H, p-H), 6.34 (s, 4 H, o-exo-H), 2.06 (s, 36 H, endo-
But), 1.14 (s, 36 H, exo-But). FAB-HRMS (m/z); calc. for M + H+
(C96H105CeN8): 1509.7517, found: 1509.7510. For the synthesis of 1b and
2b, the ether linkages of 5-(4A-methoxyphenyl)-15-(4A-tolyl)porphine
and 5-(4A-methoxyphenyl)-15-(3A,5A-di-tert-butylphenyl)porphine were
cleaved, respectively, by BR3 and the products were reacted with the
corresponding a,w-tosylated oligoether in the presence of Cs2CO3 to give
bridged free bases, which were metallated with Ce(acac)3·nH2O in refluxing
1,2,4-trichlorobenzene for 3 h under Ar. The crude products were
chromatographed on alumina and recrystallized from CH2Cl2–hexane. The
isolated yields were 5.6 (1b) and 17 (2b) %, respectively. 1b: 1H NMR (270
MHz, CDCl3, 0 °C) d 9.68 (d 3 2, 4 H, o-endo-H in C6H4Me and C6H4O),
9.09, 9.08 (s 3 2, 4 H, meso), 8.82, 8.81, 8.72, 8.68, 8.60, 8.54, 8.38, 8.30
(d 3 8, J 4 Hz, 16 H, pyrrole-b), 8.12 (d, J 7 Hz, 2 H, m-endo-H in C6H4Me),
7.89 (br s, 2 H, m-endo-H in C6H4O), 7.17–7.08 (m, 6 H, m-exo-H in
C6H4Me and C6H4O2), 6.98 (br s, 2 H, m-exo-H in C6H4O), 6.61 (br s, 2 H,
o-exo-H in C6H4O), 6.35 (d, J 7 Hz, 2 H, o-exo-H in C6H4Me), 4.80–4.26
(m, 16 H, CH2), 2.80 (s, 6 H, CH3). FAB-HRMS calc. (m/z): for M+
0.5
0.0
–0.5
–1.0
–1.5
300
400
500
l / nm
600
700
Fig. 1 Circular dichroism (CD) spectra of the enantiomers of 2b in hexane–
EtOH (1/1) at 10 °C
complex 1b showed no sign of optical resolution, indicating that
1b is more subject than 2b to the ligand rotation via path II.
In order to evaluate the rotation rate of 1b, we conducted 1H
NMR spectroscopy in CDCl3 at varying temperatures. If the
relative rotation of the porphyrin ligands is slower than the
NMR chemical shift timescale, 1b should display eight non-
equivalent pyrrole-b signals because of the low symmetry of the
molecular structure. In fact, we observed at 0 °C eight doublets
at d 8.30 (a), 8.38 (d), 8.54 (dA), 8.60 (aA), 8.68 (b), 8.72 (c), 8.81
(bA) and 8.82 (cA) [Scheme 1(b)]§, where exchangeable pairs via
ligand rotation are a/aA, b/bA, c/cA and d/dA, as determined from
the cross-peaks in the EXSY spectra. However, even upon
elevating the temperature to 55 °C, none of the paired signals
coalesced, indicating that the site exchange in 1b is much slower
than that in the non-strapped analogue 1a. Thus, the saturation
transfer method5,6 was applied, which is informative of
exchange phenomena much slower than the NMR chemical
shift timescale: When the pyrrole-b signal dA was irradiated at,
e.g., 25 °C (d 8.50), the paired signal d (d 8.35) decreased to
one-third in intensity. As the temperature was lowered, the
saturation transfer was less pronounced, and virtually dis-
appeared at 0 °C. From the degree of saturation transfer from the
signals dA to d at 10 °C, the rate constant for the rotation of the
porphyrin ligands (path II) in 1b was evaluated to be 0.37 s21,**
which is almost three orders of magnitude larger than that of
2b.
The observed rotation rates for the strapped complexes (1b
and 2b), compared with those for non-strapped 1a and 2a, allow
us to estimate that the rotation via sterically more demanding
path II takes place much less frequently at every 102 to 105 times
the rotation via path I occurs. Therefore, cerium bis(5,15-
diarylporphyrinate) double-deckers may be regarded as molec-
ular oscillators, where the two facing porphyrin ligands
rotationally oscillate around the metal center, because of the
interlocking by the meso-aryl groups. Studies on stimuli-
responsive oscillations of metal bis(porphyrinate) double-
deckers are the subject worthy of further investigation.
We thank Dr N. Morisaki for HRMS measurement, and K. T.
thanks JSPS Research Fellowships for Young Scientists.
1
(C80H62CeN8O6): 1370.3847, found: 1370.3876. 2b: H NMR (270 MHz,
CDCl3, 21 °C) d 9.64 (br s, 2 H, o-endo-H in C6H4O), 9.52 (t, J 2 Hz, 2 H,
o-endo-H in C6H3), 9.06, 9.02 (s 3 2, 4 H, meso), 8.81, 8.80, 8.76, 8.65,
8.64, 8.56, 8.51, 8.28 (d 3 8, J 4 Hz, 16 H, pyrrole-b), 7.85 (br s, 2 H,
m-endo-H in C6H4O), 7.76 (t, J 2 Hz, 2 H, p-H in C6H3), 7.14–7.03 (m, 4
H, C6H4O2), 6.95 (br s, 2 H, m-exo-H in C6H4O), 6.59 (br s, 2 H, o-exo-H
in C6H4O), 6.31 (t, J 2 Hz, 2 H, o-exo-H in C6H3), 4.76–4.22 (m, 16 Hz,
CH2), 2.01 (s, 18 H, endo-But), 1.11 (s, 18 H, exo-But). FAB-HRMS (m/z):
calc. for M + H+ (C94H91CeN8O6): 1567.6117, found: 1567.6068.
§ Assignments were made by 1H–1H COSY and EXSY spectroscopies upon
consideration of magnetic effects of proximate meso-aryl substituents.
¶ Calculated by the method reported in ref. 4.
∑ With a Chiralcel OD-H (Daicel) column using hexane–EtOH (1/1) as
eluent at a flow rate of 0.5 ml min21; retention times: 35.9 and 85.8 min.
** According to the method described in ref. 6, the rate constant of
21
exchange (k) was calculated by the equation: k = (I0 2 Iirrad)/Iirrad T1
,
where I0 and Iirrad represent intensity of the pyrrole-b signal d and that upon
irradiation to saturate the exchangeable proton singal dA, respectively, and
T1 represents the spin–lattice relaxation time for the signal dA.
1 M. Postel, M. Pfeffer and J. G. Riess, J. Am. Chem. Soc., 1977, 99, 5623;
J. T. B. H. Jastrzebski, G. van Koten, M. Konijn and C. H. Stam, J. Am.
Chem. Soc., 1982, 104, 5490; E. W. Abel, N. J. Long, K. G. Orrell, A. G.
Osborne, H. M. Pain and V. Sik, J. Chem. Soc., Chem. Commun., 1992,
303; H. Go¨rls, B. Neumu¨ller, A. Scholz and J. Scholz, Angew. Chem., Int.
Ed. Engl., 1995, 34, 673.
2 D. B. Amabilino and J. F. Stoddart, Chem. Rev., 1995, 95, 2725.
3 J. W. Buchler, K. Elsa¨sser, M. Kihn-Botulinsky and B. Scharbert, Angew.
Chem., Int. Ed. Engl., 1986, 25, 286; J. W. Buchler and B. Scharbert,
J. Am. Chem. Soc., 1988, 110, 4272; G. S. Girolami, S. N. Milam and K.
S. Suslik, J. Am. Chem. Soc., 1988, 110, 2011; K. M. Kadish, G. Moninot,
Y. Hu, D. Dubois, A. Ibnfassi, J.-M. Barbe and R. Guilard, J. Am. Chem.
Soc., 1993, 115, 8153; D. Y. Dawson, H. Brand and J. Arnold, J. Am.
Chem. Soc., 1994, 116, 9797; G. S. Girolami, C. L. Hein and K. S. Suslik,
Angew. Chem., Int. Ed. Engl., 1996, 35, 1223; D. K. P. Ng and J. Jiang,
Chem. Soc. Rev., 1997, 26, 433.
4 K. Tashiro, K. Konishi and T. Aida, Angew. Chem. Int. Ed. Engl., 1997,
36, 856.
5 C. L. Perrin and T. J. Dwyer, Chem. Rev., 1990, 90, 935.
6 J. K. M. Sanders and B. K. Hunter, Modern NMR Spectroscopy, Oxford
University, Oxford, 1987.
Received in Cambridge, UK, 17th February 1998; 8/01350K
1122
Chem. Commun., 1998