Benzylic imine catenates: Readily accessible octahedral analogues of the sauvage catenates

Article abstract of DOI:10.1002/1521-3773(20010417)40:8<1538::AID-ANIE1538>3.0.CO;2-F

The shape of rings to come: The first members of a new family of simple-to-prepare octahedral analogues of the tetrahedral Sauvage catenates are described (see scheme). M = Mn²⁺, Fe²⁺, Co²⁺, Ni²⁺, Zn²⁺

Full text of DOI:10.1002/1521-3773(20010417)40:8<1538::AID-ANIE1538>3.0.CO;2-F

COMMUNICATIONS
131.2, 121.5, 113.5, 113.2, 71.9, 71.1, 69.5, 59.1, 55.4; FTIR (neat film): nÄ ˆ
[12] Despite the desirability of the approach, combinatorial libraries of
small-molecule fluorescent chemosensors have received limited
attention. See: a) A. W. Czarnik, Chem. Biol. 1995, 2, 423 ± 428;
b) C. T. Chen, H. Wagner, W. C. Still, Science 1998, 279, 851 ± 853;
c) S. E. Schneider, S. N. OꢁNeil, E. V. Anslyn, J. Am. Chem. Soc. 2000,
122, 542 ± 543; d) F. Szurdoki, D. Ren, D. R. Walt, Anal. Chem. 2000,
72, 5250 ± 5257.
2888 (w), 1614 (s), 1445 (s), 1241 (w), 1100 (w), 811 (s) cm 1; HRMS
‡
(MALDI): m/z: calcd for C27H34NO6 [M‡H ] 468.2381; found 468.2370;
TLC Rf (55% EtOAc/Hex) 0.20.
Received: December 18, 2000 [Z16296]
[13] A. F. Littke, C. Y. Dai, G. C. Fu, J. Am. Chem. Soc. 2000, 122, 4020 ±
4028.
[1] a) A. P. de Silva, D. B. Fox, A. J. M. Huxley, T. S. Moody, Coord.
Chem. Rev. 2000, 205, 41 ± 57; b) L. Prodi, F. Bolletta, M. Montalti, N.
Zaccheroni, Coord. Chem. Rev. 2000, 205, 59 ± 83; c) V. Amendola, L.
Fabbrizzi, M. Licchelli, C. Mangano, P. Pallavicini, L. Parodi, A. Poggi,
Coord. Chem. Rev. 2000, 192, 649 ± 669; d) A. P. de Silva, D. B. Fox,
A. J. M. Huxley, N. D. McClenaghan, J. Roiron, Coord. Chem. Rev.
1999, 186, 297 ± 306; e) J. P. Desvergne, A. W. Czarnik, NATO ASI Ser.
Ser. C 1997, 492 ; f) A. W. Czarnik, ACS Symp. Ser. 1993, 538.
[2] For select recent examples, see: a) Y. Kawanishi, K. Kikuchi, H.
Takakusa, S. Mizukami, Y. Urano, T. Higuchi, T. Nagano, Angew.
Chem. 2000, 112, 3580 ± 3582; Angew. Chem. Int. Ed. 2000, 39, 3438 ±
3440; b) S. Deo, H. A. Godwin, J. Am. Chem. Soc. 2000, 122, 174 ± 175;
c) A. P. de Silva, J. Eilers, G. Zlokarnik, Proc. Natl. Acad. Sci. USA
1999 96, 8336 ± 8337.
Benzylic Imine Catenates: Readily Accessible
Octahedral Analogues of the Sauvage
Catenates**
David A. Leigh,* Paul J. Lusby, Simon J. Teat,
Andrew J. Wilson, and Jenny K. Y. Wong
[3] We have recently characterized binding-induced conformational
restriction of biaryl fluorophores as an efficient chemosensor signal-
ling mechanism: S. A. McFarland, N. S. Finney, J. Am. Chem. Soc. in
Historically, one of the triumphs of coordination chemistry
has been its application to the synthesis of mechanically
interlocked molecular architectures, that is catenanes, rotax-
anes, and knots.[1±10] In 1984 Sauvage et al. used the preferred
tetrahedral geometry of CuI to organize appropriately deriv-
atized phenanthroline ligands into a fixed mutually orthogo-
nal orientation, whereupon a double macrocyclization reac-
tion gave the [2]catenate in 27% yield (Scheme 1a).[2]
‡
press. For a previous example in which formation of a 2:2 ligand/K
restricts energy-wasting olefin isomerization in a stilbenoid fluoro-
phore, see: W.-S. Xia, R. H. Schmehl, C.-J. Li, J. Am. Chem. Soc. 1999,
121, 5599 ± 5600.
[4] Quantum yields were determined by standard methods (J. R. Lako-
wicz, Principles of Fluorescence Spectroscopy, 2nd ed., Kluwer, New
York, 1999). The quantum yields of 2 and 3 are 0.027 and 0.024,
respectively, using tryptophan in H2O as a standard (f ˆ 0.13). We
estimate values of f to be accurate to Æ25%. The excitation
maximum for the LE state is 290 nm; the excitation maximum for
the CT band is 345 nm.
O
HO
O
a)
b)
OH
O
O
O
[5] For
a review of fluorescent chemosensors based on induced
N
O
O
O
O
N
N
N
charge transfer, see: B. Valeur, I. Leray, Coord. Chem. Rev. 2000,
205, 3 ± 40.
N
N
Cu
Cu
N
N
[6] As detailed in our previous analysis of biphenyl-derived fluorescent
chemosensors,[3] we reason as follows: no changes are observed in the
O
O
X
OH
O
‡
HO
X
UV spectrum upon complexation of Li , suggesting that the excitation
probability (Einstein A coefficient), and thus the emission probability
(Einstein B coefficient) and the rate of radiative decay (kr), remain
constant; the energetic separation of the S1 and S0 potential energy
surfaces is sufficient (>60 kcalmol 1) to preclude internal conversion
(kic ꢀ 0); increased quantum yield must thus arise from decreased
intersystem crossing (ISC); kISC is known to increase with dihedral
angle, and binding-induced conformational restriction thus leads to
X
X
X
N
N
N
N
N
N
N
M N
N
M
N
N
N
X
X
X
Scheme 1. Synthesis of catenates by orthogonalization of coordinated
ligands about metal templates with a) tetrahedral and b) octahedral
reduction in kISC
.
coordination preference. M in (b): Mn2‡, Fe2‡, Co2‡, Ni2‡, Cu2‡, Zn2‡
,
[7] Association constants were determined by nonlinear least-squares
fitting of lg[metal] versus I/I0 plots. Curve fitting was carried out with
the Prism3 software package (Graphpad, Inc., San Diego, CA). The
reported Ka values have 95% confidence limits of Æ0.1 lgKa units. In
all cases reported here curve fitting is consistent with the formation of
1:1 metal/ligand complexes.
Cd2‡, Hg2‡
.
[*] Prof. D. A. Leigh, Dr. P. J. Lusby, A. J. Wilson, J. K. Y. Wong
Centre for Supramolecular and Macromolecular Chemistry
Department of Chemistry, University of Warwick
Coventry CV47AL (UK)
‡
‡
[8] Titration of 2 ± 4 with Na or K does not lead to appreciable change
in fluorescence emission.
Fax : (‡ 44)24-7652-3258
[9] Both 2 and 3 respond slowly to Mg2‡. In the case of 3, the hysteresis is
severe enough to make Ka determination impractical. We tentatively
ascribe the differences in the responses of 2 and 3 to coordination of
E-mail: David.Leigh@Warwick.ac.uk
Dr. S. J. Teat
CCLRC Daresbury Laboratory
Warrington (UK)
‡
Li and Ca2‡ by the distal oxygen atoms of the ether chain, although
details of these interactions remain to be elucidated.
[10] a) R. M. Izatt, K. Pawlak, J. S. Bradshaw, R. L. Bruening, Chem. Rev.
1991, 91, 1721 ± 2085; b) R. M. Izatt, K. Pawlak, J. S. Bradshaw, R. L.
Bruening, Chem. Rev. 1995, 95, 2529 ± 2586.
[11] For recent examples in which metal ion coordination has been used to
ªtuneº fluorescence emission, see: a) H. S. Joshi, R. Jamshidi, Y. Tor,
Angew. Chem. 1999, 111, 2887 ± 2891; Angew. Chem. Int. Ed. 1999, 38,
2722 ± 2725; b) J. C. Loren, J. S. Siegel, Angew Chem. 2001, 113, 776 ±
779; Angew. Int. Ed. Engl. 2001, 40, 754 ± 757.
[**] This work was supported by the EPSRC. D.A.L. is an EPSRC
Advanced Research Fellow (AF/982324). We thank Dr. B. P. Murphy
(Manchester Metropolitan University) for useful discussions and Drs.
T. J. Kidd, S. M. Lacy (University of Warwick), G. Di Orazio and R.
Nasreen (University of Manchester, Institute of Science and Technol-
ogy) for early ligand design.
Supporting information for this article is available on the WWW under
http://www.angewandte.com or from the author.
1538
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
1433-7851/01/4008-1538 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2001, 40, No. 8

COMMUNICATIONS
Subsequent studies on the system boosted yields to near
quantitative levels (by ring closing metathesis, RCM)[3] and
extended the interlocked architectures available to [n]cate-
nates (n ˆ 2 ± 8),[4] catenands (the demetalated, but still
interlocked ligands),[5] rotaxanes,[6] pseudo-rotaxanes,[7] and
knots.[8] Interlocked ligands can have remarkable properties;
catenates are amongst the most stable complexes of CuI that
exist with a neutral ligand,[5] and catenands also stabilize[9] the
ordinarily disfavored tetrahedral geometries of NiI and Cu0.
However, although other metals have been introduced[9] and
alternative trigonal-bipyrimidal ªstationsº employed in re-
dox-switchable systems,[10] the basic catenate assembly system
based on a low oxidation state, tetrahedrally coordinated
metal template and phenanthroline ligands has remained
largely unchanged. Expanding this elegant strategy to include
catenates with higher oxidation state, octahedrally coordi-
nated metal centers is of obvious appeal (Scheme 1b),[11] and
pioneering work in this regard has been carried out by
Schröder et al.,[12] Busch et al.,[13] and others,[1d, 14] yet only two
octahedral catenates[1d, 14] (and two knots[15, 16]) have been
described to date. Catenanes assembled by hydrogen bonds
can exhibit a continuous range of (tunable) dynamic proper-
ties,[17, 18] but the strong coordination bonds present in
catenates lock the macrocyclic components in fixed positions
whereas, in the absence of the metal, these rings rotate with
virtually complete (but uncontrolled) freedom. The difference
in the type of dynamics available to each system led us to seek
a route to catenates of a size and shape that would be both
compatible and interchangeable in molecular devices with
benzylic amide catenanes assembled by hydrogen bonds (e.g. I).
The basic architecture of benzylic amide macrocycles is
rather well suited to adapting to other sorts of assembly
processes (Scheme 2). The rigid framework positions multiple
donor groups such that they converge towards the center of
the cavity, while the isophthalic spacer holds the aromatic
rings in a parallel arrangement at a distance ideal for p
stacking with an orthogonally bound guest. The 1,3-linked
benzylic motif ensures a complete 1808 turn for each fragment
(in comparison to, say, the 1208 turn for a diphenylphenan-
throline unit) holding the endgroups in positions that promote
intracomponent rather than intercomponent cyclizations.[19]
Thus, the only changes necessary to obtain a system based on
metal ion chelation (i.e. arriving at complexes with the
benzylic bis(2,6-diiminopyridine)catenand ligand 1) was to
replace the amide groups with imine groups and, for stability
reasons, the phenolic esters with ethers (Scheme 2).
E
F
a)
D
O
O
O
O
C
I
N
3
3
N
4
N
N
H
G
4
B
ML2
N
A
M N
N
L2
MeOH
N
N
O
O
4
4
[M22]L2
2
PCy3
Cl
CH2Cl2/
C2H2Cl4
Ru
Ph
Cl
PCy3
O
[Mn1](ClO4)2 63
[Fe1]Br2 75 [Zn1](ClO4)2 73
[Co1](ClO4)2 36 [Zn1](BPh4)2 81
[Co1](BPh4)2 41 [Cd1](ClO4)2 70
[Ni1](ClO4)2 32 [Hg1](ClO4)2 05
[Cu1]I2 61
O
N
N
N
M
N
L2
N
N
O
O
[M1]L2
ML2
MeOH
O
b)
H2N
O
H
H
O
O
O
O
O
O
N
O
H
N
O
H
O
N
O
H
N
O
O
O
H
N
I
O
H2N
4
Scheme 2. Synthesis of benzylic imine catenates [M1]L2 by a) double
macrocyclization of a preformed octahedral complex [M22]L2, b) sponta-
neous metal-promoted assembly of nonchelated precursors. The yields (in
%) refer to route (a); the yields from route b are not yet optimized, but are
uniformally higher. The ligand design is based on that of benzylic amide
catenanes, e.g. I.[17b]
ture to the metathesis catalyst for four days equilibrated the
products to give almost exclusively the E,E isomer.
In contrast to the situation with the classic tetrahedral
phenanthroline system, the same catenate [Zn1](ClO4)2 could
also be prepared by assembling the coordinating sites in situ
(Scheme 2b). Treatment of the bis-amine 4[21] with 2,6-
pyridinedicarbaldehyde and Zn(ClO4)2 ´ 6H2O in methanol
resulted in the precipitation of the (E,E)-catenate after one
hour in 53% yield. In this case, the metal center orders the
construction of the catenate about itself, through the rever-
sible formation of four imine bonds from five components, as
the lowest energy means by which it can satisfy its desired
octahedral coordination geometry.
1
The H NMR spectra of free ligand 2, zinc(ii) precursor
complex [Zn22](ClO4)2, and zinc(ii) catenate [Zn{(E,E)-
1}](ClO4)2 are shown in Figure 1a ± c. Shielding of the benzylic
aromatic rings (HE, HF) in both the precursor complex
(Figure 1b) and the catenate (Figure 1c) with respect to the
free ligand (Figure 1a) is indicative of the entwined (in the
case of [Zn22]L2) or interlocked (in the case of [Zn{(E,E)-
1}]L2) architectures. Interestingly, subtle but significant differ-
ences take place in the shifts of all the non-alkyl chain
resonances between [Zn{(E,E)-1}]L2 and [Zn22]L2, indicating
that a certain amount of reorganization of the ligands needs to
occur during catenate formation.
Single crystals suitable for investigation by X-ray crystal-
lography using a synchrotron source were obtained from slow
vapor diffusion of diethyl ether into a solution of [Zn{(E,E)-
1}](ClO4)2 in acetonitrile.[22] The crystal structure (Figure 2)
confirms the interlocked molecular architecture, the octahe-
dral geometry around the coordinated zinc and the E
configuration of the olefinic bonds in both rings. The benzylic
The zinc(ii) perchlorate complex [Zn22](ClO4)2 was isolated
as a precipitate in 83% yield from simple addition of a
solution of the Schiff base ligand 2[20] in dichloromethane to a
methanolic solution of Zn(ClO4)2 ´ 6H2O. Imines are one of
the few classes of substrates not normally compatible with
olefin metathesis (indeed, the free Schiff-base ligand 2 cannot
be converted into the macrocycle by RCM). However, pre-
coordination ties up the imine groups and the tetraolefin
complex [Zn22](ClO4)2 smoothly underwent double macro-
ˆ
cyclization by RCM with Grubbsꢁ catalyst ([Ru( CHPh)-
(PCy3)2Cl2], CH2Cl2, Ar, RT, 4 h) to give the zinc(ii) catenate
[Zn1](ClO4)2 in 73% yield as a mixture of E,E, E,Z, and Z,Z
diastereomers (Scheme 2a). Exposure of the reaction mix-
Angew. Chem. Int. Ed. 2001, 40, No. 8
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
1433-7851/01/4008-1539 $ 17.50+.50/0
1539

COMMUNICATIONS
[Zn22]L2 in halogenated sol-
vents (see NMR data in Fig-
ure 1) and doubtless contrib-
utes to intracomponent rather
than intercomponent cycliza-
tion occurring during RCM.
Metal coordination necessarily
buries the polar imine groups
at the center of the molecular
structure with the alkyl chains
to the outside; a similar overall
co-conformation to that ob-
served in the solid state for
amphiphilic benzylic amide
catenanes.[17a]
The bite angle of the ligand
in [Zn{(E,E)-1}](ClO4)2 is
slightly smaller (1508) than in
nonmacrocyclic bis(2,6-diimi-
nopyridine) complexes with
ZnII[13] (1518) and NiII[12, 13]
(151 ± 1568), indicating that
the ligand is able to adapt to
the demands of its environ-
ment. Encouraged, we investi-
gated the tolerance of the
octahedral catenate assembly
system by extending the RCM
approach to metals both across
and down the periodic table
with respect to zinc (i.e. Mn
Zn and Zn !Hg). The results
(see Scheme 2) were unifor-
mally satisfying with the pre-
cursor complexes and cate-
nates obtained in each case,
often in good yields despite the
Figure 1. 1H NMR spectra (400 MHz, C2D2Cl4, 298 K) of a) free Schiff-base ligand 2, b) acyclic tetraolefin
precursor complex [Zn22](ClO4)2, c) benzylic imine catenate [Zn{(E,E)-1}](ClO4)2, d) free benzylic amine
catenand (E,E)-3, and e) benzylic amine catenate [Zn{(E,E)-3}](ClO4)2. The assignments correspond to the
lettering shown in Scheme 2 for compound 2.
mixture of diastereomers complicating the purification pro-
cess.[23] Preliminary studies show that the same catenates are
also readily produced by the direct imine bond formation
route (Scheme 2b). The crystal structures of two of these new
catenates, [Cu{(E,E)-1}](ClO4)2 and [Co{(E,E)-1}]I2, were
obtained by using a synchrotron radiation source (Figure 3).
The bite angles in these catenates proved larger than in the
zinc system (152.1 and 152.78 (Cu), 152.9 and 154.58 (Co)) in
line with observed trends with acyclic ligands, confirming the
geometrical flexibility of the benzylic bis(2,6-diiminopyridi-
ne)catenand ligand system. Interestingly, all the catenates are
much more thermally stable than the corresponding precursor
complexes. Whilst the acyclic coordination compounds
[M22]L2 typically possess sharp melting points in the range
221 ± 2528C, the analogous [M1]L2 catenates tend to gradually
lose color and decompose at temperatures in excess of 3508C.
The properties of the zinc(ii) catenate [Zn{(E,E)-1}](ClO4)2
were investigated in more detail (Scheme 3). By single-phase
or biphasic extraction of [Zn{(E,E)-1}](ClO4)2 in various
organic solvents (DMF, CHCl3/H2O, etc.) with up to
100 equivalents of EDTA, disodium salt, no signs of demeta-
lating the catenate were obtained, suggesting that the
Figure 2. Structure of [Zn{(E,E)-1}](ClO4)2 as determined by X-ray
crystallography.[22] Carbon atoms of one macrocycle are shown in light
blue and those of the other in yellow; oxygen atoms are red, nitrogen dark
blue, chlorine green, hydrogen white, and zinc silver. Non-olefinic hydro-
gen atoms and a molecule of acetonitrile are omitted for clarity. Selected
bond lengths [Š]: Zn-N5 2.289, Zn-N8 2.048, Zn-N11 2.211, Zn-N205 2.264,
Zn-N208 2.045, Zn-N211 2.276; other selected interatomic distances [Š]:
N5-N11 4.348, N8-N208 4.081, N205-N211 4.388; ligand bite angles [8]: N5-
Zn-N11 150.14, N205-Zn-N211 150.34.
groups of each macrocycle p-stack with the 2,6-diiminopyr-
idine groups of the other interlocked ring, an interaction
which is also clearly present in the precursor complex
1540
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
1433-7851/01/4008-1540 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2001, 40, No. 8

COMMUNICATIONS
metallation proceded smoothly to give the corresponding,
clearly more labile, benzylic amine catenate [Zn{(E,E)-
1
3}](ClO4)2. The H NMR spectra of (E,E)-3 and [Zn{(E,E)-
3}](ClO4)2 are shown in Figure 1d and e, respectively. The
compact nature of the free ligand (E,E)-3 is convincingly
demonstrated by the shielding of several resonances com-
pared to the acyclic ligand 2. Protons HE and HF, for example,
which intrinsically should be relatively unaffected by imine
reduction, are shifted in (E,E)-3 by a similar magnitude to
those in the [Zn({E,E)-1}](ClO4)2 catenate and the
[Zn22](ClO4)2 precursor complex despite no directed inter-
component interactions being present between the macro-
cycles in the catenand. The broad appearance of several
resonances in the spectrum of the catenand is probably due to
co-conformational exchange processes that are not fast on the
NMR time scale at room temperature. The broadened peaks
in the spectrum of the amine catenate [Zn{(E,E)-3}]L2,
however, may be a result of multiple asymmetric centers
being produced by coordination of the four nitrogen atoms
which are prochiral in the free ligand 3.
In conclusion, a simple, versatile and effective route exists
to octahedrally coordinated analogues of the Sauvage cate-
nates. This route should be extendable to the synthesis of
other metal-based interlocked architectures (such as rotax-
anes, shuttles, and knots), and the octahedrally coordinated
catenates should be interchangeable in possible molecular
devices with catenanes assembled by hydrogen bonds. Cate-
nates are not only a means of stabilizing unusual oxidation
states and normally disfavored geometries, but could also find
uses in areas where complexes with particularly high kinetic
stabilities are required, such as radiotherapy[24] and magnetic
resonance imaging,[25] because of the wide range of metals that
can adopt octahedral coordination patterns. Finally, we are
well aware that metal complexes of 2,6-diiminopyridine
ligands also form an important new generation
Figure 3. Structures of a) [Co{(E,E)-1}]I2 and b) [Cu{(E,E)-1}](ClO4)2 as
determined by X-ray crystallography.[22] Color code as in Figure 2, cobalt
red, copper green, iodide purple. Non-olefinic hydrogen atoms are omitted
for clarity. Ligand bite angles [8]: N5-Co-N11 152.9, N205-Co-N211 154.5;
N5-Cu-N11 152.1, N205-Cu-N211 152.7.
of non-metallocene olefin polymerization cata-
O
O
O
lysts[26] and that intertwined versions of such
O
EDTA
N
N
N
N
N
O
disodium salt
N
structures exhibit novel forms of columnar liquid
crystalline behavior.[27] The possible application of
translationally switchable mechanically inter-
locked architectures (containing unsaturated met-
al centers in the case of potential polymerization
catalysts) to these areas should be fascinating.
N
N
Zn
N
X
N
N
O
O
N
O
[Zn{(E,E)-1}]2+
(E,E)-1
1 NaBH4, EtOH, 1h
2 EDTA, disodium salt
O
O
O
NH
O
O
H
H
Experimental Section
N
N
HN
HN
N
Zn(ClO4)2
MeOH
N
Zn N
N
NH
Route (a) of Scheme 2 (RCM): Grubbsꢁ metathesis catalyst
O
N
N
ˆ
H
H
[Ru( CHPh)(PCy3)2Cl2] (0.02 g, 0.0243 mmol, 20 mol%)
O
O
was placed in a sealed, argon-purged, flame-dried Schlenk
tube, and subjected to a constant stream of argon for ten
(E,E)-3
[Zn{(E,E)-3}]2+
minutes.
A degassed and argon-purged solution of the
Scheme 3. The chemistry of the octahedral metal catenate [Zn{(E,E)-1}](ClO4)2. Direct
demetalation is unsuccessful, but reduction of the imine groups allows extraction of the
zinc to yield the free benzylic amine catenand (E,E)-3 (78%). Remetalation affords the
benzylic amine catenate [Zn{(E,E)-3}](ClO4)2 (ca. 100%).
[M22]L2 complex (0.122 mmol) in anhydrous dichlorome-
thane (500 mL) was transferred to the Schlenk tube by
injection over ten minutes. Reaction was allowed to continue
until all the starting material was consumed as evidenced by
TLC (typically 4 h), or, in order to maximize the amount of
E,E diastereomer, for 4 days. The resulting solution was evaporated to
dryness and purified by repeated chromatography (silica gel, 1 ± 5%
MeOH in CH2Cl2 as eluent) to give the catenates [M1]L2 as variously
colored solids. Selected data for [Zn{(E,E)-1}](ClO4)2: yield: 112 mg, 73%;
m.p. 3508C (decomp); 1H NMR (400 MHz, [D6]DMSO): d ˆ 1.58 (m, 8H,
combination of the directional, planar diiminopyridyl coordi-
nating motif and the tight, encapsulated architecture provides
exceptional kinetic stability. However, reduction of the imine
groups followed by washing with aqueous EDTA, disodium
salt, resulted in the free benzylic amine catenand (E,E)-3. Re-
ˆ
CH2), 1.74 (m, 8H, CH2), 2.12 (m, 8H, CH2CH CH), 3.85 (t, 8H, J ˆ
Angew. Chem. Int. Ed. 2001, 40, No. 8
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
1433-7851/01/4008-1541 $ 17.50+.50/0
1541

COMMUNICATIONS
6.5 Hz, HG), 4.49 (brs, 8H, HD), 5.61 (t, 4H, J ˆ 3.0 Hz, HH), 6.36 (AA'BB'
system, 8H, J ˆ 8.0 Hz, HE), 6.53 (AA'BB' system, 8H, J ˆ 8.0 Hz, HF), 7.70
(d, 4H, J ˆ 8.0 Hz, HB), 8.42 (t, 2H, J ˆ 8.0 Hz, HA), 8.50 (s, 4H, HC);
13C NMR (100 MHz, [D6]DMSO): d ˆ 25.89, 28.05, 31.50, 60.76, 67.17,
113.91, 127.02, 129.34, 129.97, 130.62, 144.37, 145.25, 158.24, 160.27, FAB-MS
on metal templates with octahedral coordination preference, see:
D. H. Busch, J. Incl. Phenom. 1992, 12, 389 ± 395; N. V. Gerbeleu, V. B.
Arion, J. Burgess, Template Synthesis of Macrocyclic Compounds,
Wiley-VCH, Weinheim, 1999.
[12] A. J. Blake, A. J. Lavery, T. I. Hyde, M. J. Schröder, J. Chem. Soc.
Dalton Trans. 1989, 965 ± 970.
‡
(mBNA matrix): m/z: 1126 [M ClO4 ] ; anal. calcd for:
C62H70O12N6Cl2Zn (1225): C 60.73, H 5.71, N 6.86%, found: C 60.46, H
5.97, N 6.67%.
[13] A. L. Vance, N. W. Alcock, J. A. Heppert, D. H. Busch, Inorg. Chem.
1998, 37, 6912 ± 6920.
[14] J.-P. Sauvage, M. Ward, Inorg. Chem. 1991, 30, 3869 ± 3874.
[15] G. Rapenne, C. Dietrich-Buchecker, J.-P. Sauvage, J. Am. Chem. Soc.
1999, 121, 994 ± 1001.
Route (b) of Scheme 2: A solution of 2,6-pyridinedicarbaldehyde (0.97 g,
7.2 mmol) in methanol (10 mL) was added dropwise over 30 min to a
solution of bis-amine 4 (7.2 mmol) and metal salt (ML2, 3.6 mmol) in
methanol (10 mL). The reaction mixture was stirred at room temperature
for 1 h after which time the [M1]L2 catenate could either be isolated as a
precipitate by filtration, or the solvent removed under reduced pressure
and the resulting solid purified as for the RCM procedure.
[16] C. A. Hunter, Chem. Br. 1998, 34(5), 17.
[17] a) D. A. Leigh, K. Moody, J. P. Smart, K. J. Watson, A. M. Z. Slawin,
Angew. Chem. 1996, 108, 326 ± 321; Angew. Chem. Int. Ed. Engl. 1996,
35, 306 ± 310; b) T. J. Kidd, D. A. Leigh, A. J. Wilson, J. Am. Chem.
Soc. 1999, 121, 1599 ± 1600.
[18] D. A. Leigh, A. Murphy, J. P. Smart, M. S. Deleuze, F. Zerbetto, J. Am.
Chem. Soc. 1998, 120, 6458 ± 6467.
Received: August 14, 2000
Revised: February 5, 2001 [Z15632]
[19] The importance of these structural features for promoting catenate
formation was exemplified by a very recent study which showed that it
is not possible to produce octahedral catenates from terpy ligands
which were not well preorganized for intracomponent cyclization (N.
Belfrekh, C. Dietrich-Buchecker, J.-P. Sauvage, Inorg. Chem. 2000, 39,
5169 ± 5172).
[1] a) Molecular Catenanes, Rotaxanes and Knots (Eds.: J.-P. Sauvage, C.
Dietrich-Buchecker), Wiley-VCH, Weinheim, 1999; for catenanes
based on coordination bonds other than the Sauvage catenates see:
b) G.-J. M. Gruter, F. J. J. de Kanter, P. R. Markies, T. Nomoto, O. S.
Akkerman, F. Bicklehaupt, J. Am. Chem. Soc. 1993, 115, 12179 ±
12180; c) M. Fujita, F. Ibukuro, H. Hagihara, K. Ogura, Nature
1994, 367, 720 ± 723; d) C. Piguet, G. Bernardinelli, A. F. Williams, B.
Bocquet, Angew. Chem. 1995, 107, 618 ± 621; Angew. Chem. Int. Ed.
Engl. 1995, 34, 582 ± 584; e) D. M. P. Mingos, J. Yau, S. Menzer, D. J.
Williams, Angew. Chem. 1995, 107, 2045 ± 2047; Angew. Chem. Int. Ed.
Engl. 1995, 34, 1894 ± 1895; f) A. C. Try, M. M. Harding, D. G.
Hamilton, J. K. M. Sanders, Chem. Commun. 1998, 723 ± 724; g) D.
Whang, K.-M. Park, J. Heo, P. Ashton, K. Kim, J. Am. Chem. Soc.
1998, 120, 4899 ± 4900; h) C. P. McArdle, M. J. Irwin, M. C. Jennings,
R. J. Puddephatt, Angew. Chem. 1999, 111, 3571 ± 3573; Angew. Chem.
Int. Ed. 1999, 38, 3376 ± 3378; i) M. Fujita, Acc. Chem. Res. 1999, 32,
53 ± 61; j) H. W. Gibson, S.-H. Lee, Can. J. Chem. 2000, 78, 347 ± 355;
for catenane-like coordination interpenetrating networks see: k) S. R.
Batten, R. Robson, Angew. Chem. 1998, 110, 1558 ± 1595; Angew.
Chem. Int. Ed. 1998, 37, 1460 ± 1494; l) A. J. Blake, N. R. Champness,
H. Hubberstey, W.-S. Li, M. A. Withersby, M. Schröder, Coord. Chem.
Rev. 1999, 183, 117 ± 138.
[20] Compound 2 was conveniently prepared in three steps from 4-hy-
ˆ
droxybenzonitrile: 1) HO(CH2)4CH CH2, Ph3P, DEAD, THF, 08C,
59%; 2) LiAlH4, THF, 788C !reflux, 94%; 3) 2,6-pyridinedicarb-
aldehyde, MeOH, 92%.
[21] Compound
4 was conveniently prepared in three steps from
5-hexenyloxybenzylamine: 1) tBoc2O, NEt3, MeOH, 87%;
ˆ
2) [Ru( CHPh)(PCy3)2Cl2], CH2Cl2, Ar, RT, 24 h, 78%; 3) CF3CO2H,
then NEt3, CH2Cl2, 93%.
[22] [Zn{(E,E)-1}](ClO4)2: C62H70Cl2N6O12Zn ´ 0.5(C2H3N), Mr ˆ 1248.04,
crystal size 0.14  0.08  0.06 mm, monoclinic P21/c, a ˆ 10.6258(5),
b ˆ 17.4158(9), c ˆ 33.4757(18) Š, b ˆ 97.232(2)8, Vˆ 6145.6(5) Š3,
Z ˆ 4, 1calcd ˆ 1.349 Mgm 3; synchrotron radiation (CCLRC Dares-
bury Laboratory Station 9.8, silicon monochromator, l ˆ 0.69280 Š),
m ˆ 0.553 mm 1, Tˆ 150(2) K. 22797 data (7081 unique, Rint ˆ 0.0489,
1.65 < q < 21.008) were collected on
a Siemens SMART CCD
diffractometer using narrow frames (0.28 in w), and were corrected
semiempirically for absorption and incident beam decay (transmission
0.83 ± 1.00). The structure was solved by direct methods and refined by
full-matrix least-squares methods on F 2 values of all data (G. M.
Sheldrick, SHELXTL Manual, Version 5, Siemens Analytical X-ray
Instruments, Madison, WI, 1994) to give wR ˆ {S[w(Fo2
[2] C. O. Dietrich-Buchecker, J.-P. Sauvage, J.-M. Kern, J. Am. Chem.
Soc. 1984, 106, 3043 ± 3045; for the earliest CuI catenate synthesis,
involving a threaded intermediate complex, see: C. O. Dietrich-
Buchecker, J.-P. Sauvage, J.-P. Kintzinger, Tetrahedron Lett. 1983, 24,
5095 ± 5098.
F 2†2] S{w(F2†2]}1/2 ˆ 0.2984, conventional R ˆ 0.1087 for F values of
/
c
o
7081 reflections with Fo2 > 2s(Fo2†, S ˆ 1.066 for 761 parameters.
3
Residual electron density extremes were 0.974 and 0.563 eŠ
.
The alkyl chains were modeled by using both geometrical and
displacement parameter restraints. Hydrogen atoms were added in
calculated positions and constrained to a riding model. Data for the
copper catenate [Cu{(E,E)-1}](ClO4)2 was collected and solved as
above except: C62H71CuN6O12, Mr ˆ 1226.69, crystal size 0.14  0.10 Â
0.01 mm, monoclinic P21/c, a ˆ 10.2948(6), b ˆ 17.4273(10), c ˆ
[3] a) B. Mohr, M. Weck, J.-P. Sauvage, R. H. Grubbs, Angew. Chem. 1997,
109, 1365 ± 1367; Angew. Chem. Int. Ed. Engl. 1997, 36, 1308 ± 1310;
b) M. Weck, B. Mohr, J.-P. Sauvage, R. H. Grubbs, J. Org. Chem. 1999,
64, 5463 ± 5471.
Â
[4] F. Bitsch, C. O. Dietrich-Buchecker, A.-K. Khemiss, J.-P. Sauvage,
A. V. Dosselaer, J. Am. Chem. Soc. 1991, 113, 4023 ± 4025.
[5] C. O. Dietrich-Buchecker, J.-P. Sauvage, J.-M. Kern, J. Am. Chem.
Soc. 1989, 111, 7791 ± 7800.
[6] a) C. Wu, P. R. Lecavalier, Y. X. Shen, H. W. Gibson, Chem. Mater.
1991, 3, 569 ± 572; b) J.-C. Chambron, V. Heitz, J.-P. Sauvage, J. Chem.
Soc. Chem. Commun. 1992, 1131 ± 1133.
33.1511(18) Š, b ˆ 96.120(2)8, Vˆ 5913.8(6) Š3, Z ˆ 4, 1calcd
ˆ
1.378 Mgm 3; synchrotron radiation (CCLRC Daresbury Laboratory
1
Station 9.8, silicon monochromator, l ˆ 0.68950 Š), m ˆ 0.528 mm
,
Tˆ 150(2) K. 10342 data (3460 unique, Rint ˆ 0.0784, 1.93 < q <
16.508), correction for absorption and incident beam decay (trans-
mission 0.9298 ± 0.9947) wR ˆ {S[w(Fo2 F 2†2] S[w(F 2†2]}1/2 ˆ 0.3840,
[7] J.-P. Collin, P. GavinaÄ, J.-P. Sauvage, J. Chem. Soc. Chem. Commun.
1996, 2005 ± 2006.
/
c
o
conventional R ˆ 0.1556 for F values of 3460 reflections with Fo2
>
2s(Fo2†, S ˆ 2.349 for 339 parameters. Residual electron density
extremes were 0.795 and 0.785 eŠ 3. Data for the cobalt catenate
[Co{(E,E)-1}]I2 was collected and solved as the others except:
C63H70CoI2N6.5O4, Mr ˆ 1294.99, crystal size 0.14  0.10  0.01 mm,
monoclinic P21/c, a ˆ 10.2948(6), b ˆ 17.4273(10), c ˆ 33.1511(18) Š,
b ˆ 96.120(2)8. Vˆ 5913.8(6) Š3, Z ˆ 4, 1calcd ˆ 1.454 Mgm 3; synchro-
tron radiation (CCLRC Daresbury Laboratory Station 9.8, silicon
monochromator, l ˆ 0.68950 Š), m ˆ 1.386 mm 1, Tˆ 150(2) K. 28253
data (5997 unique, Rint ˆ 0.1119, 1.93 < q < 20.008), correction for
absorption and incident beam decay (transmission 0.8296 ± 0.9863),
[8] C. O. Dietrich-Buchecker, J. P. Sauvage, Angew. Chem. 1989, 101,
192 ± 195; Angew. Chem. Int. Ed. Engl. 1989, 28, 189 ± 192.
[9] N. Armaroli, L. De Cola, V. Balzani, J.-P. Sauvage, C. O. Dietrich-
Buchecker, J.-M. Kern, A. Bailal, J. Chem. Soc. Dalton Trans. 1993,
3241 ± 3247.
Â
[10] D. J. Cardenas, A. Livoreil, J.-P. Sauvage, J. Am. Chem. Soc. 1996, 118,
11980 ± 11981.
[11] The idea of using octahedral ligand geometries in catenate formation
was actually proposed a decade before the Sauvage tetrahedral system
was introduced (V. I. Sokolov, Russ. Chem. Rev. 1973, 42, 452 ± 463).
For more recent discussions of possible strategies to catenates based
wR ˆ {S[w(Fo2 F 2†2] S[w(F 2†2]}1/2 ˆ 0.2662, conventional R ˆ 0.1246
/
c
o
1542
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
1433-7851/01/4008-1542 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2001, 40, No. 8

COMMUNICATIONS
for F values of 5997 reflections with Fo2 > 2s(Fo2†, S ˆ 1.215 for 689
parameters. Residual electron density extremes were 0.955 and
1.662 eŠ 3. Crystallographic data (excluding structure factors) for
the structures reported in this paper have been deposited with the
Cambridge Crystallographic Data Centre as supplementary publica-
tion no. CCDC-147870 ([Zn{(E,E)-1}](ClO4)2), CCDC-152441
([Cu{(E,E)-1}](ClO4)2), and CCDC-152442 ([Co{(E,E)-1}]I2). Copies
of the data can be obtained free of charge on application to CCDC, 12
Union Road, Cambridge CB21EZ, UK (fax: (‡44)1223-336-033;
e-mail: deposit@ccdc.cam.ac.uk).
complexes offer a number of prospects for novel materials.[4±6]
A large effort has been invested in finding efficient methods
for the preparation and purification of these compounds.[1, 2, 7]
So far, endohedral complexes have been formed with
lanthanide or alkaline earth metals as charge transfer
species,[1, 2] and with the rare gases[8, 9] or atomic nitrogen[10]
as neutral complexes. Processes using the evaporation of
graphite ± metal oxide composites, as well as high-pressure
and high-temperature or even high-energy plasma insertions
into pure fullerenes, have been reported.[1±7] These methods
constitute remarkable achievements and have already
brought forth important insight into the properties of
endohedral fullerene complexes. Nevertheless, they are still
limited in scope and, more significantly, tend to give small
amounts of material or meager incorporation fractions.
A stimulating prospect for the formation of these com-
pounds can be described as a ªmolecular surgeryº approach,
which consists of the chemical creation of an opening within
the fullerene cage (Figure 1a).[4] The opened species would
allow the introduction of an atomic or molecular species,
which could be followed by the ªsuturingº of broken bonds
back onto the original framework. Each of these steps
presents unique nontrivial challenges. The development of
an effective method to open up the framework of C60 by a one-
pot reaction with bisazide 2, to afford bislactam 1,[11] provides
an unprecedented opportunity for the preparation of endo-
hedral complexes. We now report the successful insertion of
two small neutral gases, helium and molecular hydrogen, into
the fullerene bislactam derivative 1 which results in the
highest incorporation fractions for any gas to date in a direct
insertion process.
[23] The yields listed in Scheme 2 are yields after chromatography or
recrystallization. Sometimes the mixture of olefin diastereomers
complicates purification and, in particular, is a factor in the modest
yields of [Co1]L2 and [Ni1]L2. The very low yield of [Hg1]L2 by
route (a) is caused by the lability of the ligands in the precursor
complex [Hg22]L2, which liberates free amine or imine in the reaction
and prevents RCM occurring. All compounds gave satisfactory mass
spectra, elemental analysis, IR, UV/Vis, and, with the exception of the
1
paramagnetic catenates, 13C and H NMR data.
[24] S. Jurisson, D. Berning, W. Jia, D. Ma, Chem. Rev. 1993, 93, 1137 ±
1156.
[25] D. Parker, Chem. Br. 1994, 818 ± 822.
[26] G. J. P. Britovsek, V. C. Gibson, D. F. Wass, Angew. Chem. 1999, 111,
448 ± 468; Angew. Chem. Int. Ed. 1999, 38, 428 ± 447.
[27] L. Douce, A. El-Ghayoury, A. Skoulios, R. Ziessel, Chem. Commun.
1999, 2033 ± 2034.
Insertion of Helium and Molecular Hydrogen
Through the Orifice of an Open Fullerene**
Yves Rubin,* Thibaut Jarrosson, Guan-Wu Wang,
Michael D. Bartberger, K. N. Houk,* Georg Schick,
Martin Saunders,* and R. James Cross*
Bislactam 1 has currently the largest orifice formed in the
shell of a fullerene with an inviting open-mouth shape.[4, 11]
The amount of energy needed to push atoms or molecules
through the orifice of 1 was first calculated using hybrid
density functional theory. Activation barriers to insertion
obtained as a function of guest size are listed in Table 1. The
ground-state structures for empty bislactam 1, its inclusion
complexes of the inert gases He, H2, Ne, N2, and Ar, and the
transition structures for their encapsulation were fully opti-
mized at the B3LYP/3-21G level of theory, and energies were
computed from these geometries using the 6-31G** basis
set.[12]
Dedicated to Professor Fred Wudl
on the occasion of his 60th birthday
One of the most exciting features of C60 and the higher
fullerenes is that their carbon cages have inner cavities large
enough to hold any atom and even small molecules.[1, 2] The
physical and chemical properties of these caged compounds
(endohedral complexes) are determined by the degree of
interaction established with the p-electron shell and the
overall oxidation state of the complex.[3] As such, endohedral
The overall insertion processes for the smallest species
(He, H2) are predicted to be slightly endothermic (0.2 ±
1.4 kcalmol , Table 1). Patchkovskii and Thiel have inves-
[*] Prof. Y. Rubin, Prof. K. N. Houk, T. Jarrosson, Dr. M. D. Bartberger,
Dr. G. Schick
1
Department of Chemistry and Biochemistry
University of California, Los Angeles
Los Angeles, CA 90095-1569 (USA)
Fax: (‡1)310-206-7649
tigated the encapsulation of helium by C60 using MP2
1
calculations,[13] and found a favorable binding of 2.0 kcalmol
as a result of van der Waals interactions of all sixty carbon
atoms with the helium atom. These authors conclude that
DFT calculations underestimate the exothermicity of He
E-mail: rubin@chem.ucla.edu, houk@chem.ucla.edu
Prof. M. Saunders, Prof. R. J. Cross, Dr. G.-W. Wang[‡]
Department of Chemistry
Yale University
1
encapsulation within C60 by about 3 kcalmol compared to
New Haven, CT 06520 (USA)
the MP2 results, and thus the encapsulation of H2 and He
within the fullerene framework of 1 can be expected to be
thermoneutral or slightly exothermic.
The predicted barrier for insertion of H2 is nearly double
that of helium, even though both species have similar
van der Waals radii (1.20 and 1.22 Š, respectively).[14] The
single imaginary vibrational modes at the transition states for
Fax: (‡1)203-432-6144
E-mail: ms@gaus90.chem.yale.edu, james.cross@yale.edu
‡
[ ] Present address: Department of Chemistry
University of Science and Technology of China
Hefei, Anhui 230026 (PR China)
[**] This work was supported by grants from the National Science
Foundation.
Angew. Chem. Int. Ed. 2001, 40, No. 8
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
1433-7851/01/4008-1543 $ 17.50+.50/0
1543