Product-induced distortion of a metalloporphyrin host: Implications for acceleration of Diels-Alder reactions

Article abstract of DOI:10.1021/ja9922227

New cyclic metalloporphyrin hosts, 6 and 7, have been prepared. At 0.33 mM in dichloromethane at 25 °C, they accelerate 65-fold and 840-fold respectively the reaction of diene 1 and dienophile 2 and also bind the hetero Diels-Alder product 3 very strongly. More importantly, small single crystals of solvated 6, 7, and the 6·3 complex were grown and their structures were determined. As the Diels-Alder product resembles the Diels- Alder transition state, the structures of the product-free host 6 and the 6·3 host-product complex allow, for the first time for synthetic receptors, a detailed structural analysis of the geometrical changes imposed on an accelerating agent on binding of a Diels-Alder product. Comparison of these structures reveals that when the Diels-Alder product 3 is bound within the cavity, it induces significant structural changes in 6. This provides the first crystallographic structural evidence that accelerated product formation can be accompanied by substantial host distortion. Desolvation of host and guests emerges as another factor, implying that solvent stabilization is not as significant for the host-accelerated reaction as in the control (host free) reaction.

Full text of DOI:10.1021/ja9922227

5286
J. Am. Chem. Soc. 2000, 122, 5286-5293
Product-Induced Distortion of a Metalloporphyrin Host: Implications
for Acceleration of Diels-Alder Reactions
Moshe Nakash,† Zo1e Clyde-Watson,† Neil Feeder,† John E. Davies,† Simon J. Teat,‡ and
Jeremy K. M. Sanders*,†
Contribution from the Cambridge Centre for Molecular Recognition, UniVersity Chemical Laboratory,
UniVersity of Cambridge, Lensfield Road, Cambridge CB2 1EW, UK, and CLRC Daresbury Laboratory,
Daresbury, Warrington WA4 4AD, UK
ReceiVed June 28, 1999
Abstract: New cyclic metalloporphyrin hosts, 6 and 7, have been prepared. At 0.33 mM in dichloromethane
at 25 °C, they accelerate 65-fold and 840-fold respectively the reaction of diene 1 and dienophile 2 and also
bind the hetero Diels-Alder product 3 very strongly. More importantly, small single crystals of solvated 6, 7,
and the 6.3 complex were grown and their structures were determined. As the Diels-Alder product resembles
the Diels-Alder transition state, the structures of the product-free host 6 and the 6‚3 host-product complex
allow, for the first time for synthetic receptors, a detailed structural analysis of the geometrical changes imposed
on an accelerating agent on binding of a Diels-Alder product. Comparison of these structures reveals that
when the Diels-Alder product 3 is bound within the cavity, it induces significant structural changes in 6. This
provides the first crystallographic structural evidence that accelerated product formation can be accompanied
by substantial host distortion. Desolvation of host and guests emerges as another factor, implying that solvent
stabilization is not as significant for the host-accelerated reaction as in the control (host free) reaction.
Introduction
Supramolecular chemistry offers the prospect that bimolecular
reactions of guests can be efficiently accelerated or catalyzed
within a host cavity, but progress to date has been slow. In part
this is due to the lack of information on the host’s structural
changes resulting from guest binding and reaction.1 The Diels-
Alder reaction is a bimolecular process that proceeds through a
transition state that resembles the Diels-Alder product,2 so
achieving turnover is a challenge for biology and the synthetic
chemist alike.3 The acceleration and catalysis of Diels-Alder
reactions by artificial receptors,4,5,6 antibodies,7 and RNA8 is
† University of Cambridge.
‡ CLRC Daresbury Laboratory.
(1) (a) Kirby, A. J. Angew. Chem., Int. Ed. Engl. 1996, 35, 707. (b)
Sanders, J. K. M. Chem. Eur. J. 1998, 4, 1378.
(2) (a) Sauer, J. Angew. Chem., Int. Ed. Engl. 1967, 6, 16. (b) Houk, K.
N.; Brown, F. K. Tetrahedron Lett. 1984, 25, 4609.
(3) This may be why there is only one current candidate for a natural
Diels-Alderase enzyme: (a) Oikawa, H.; Katayama, K.; Suzuki, Y.;
Ichihara, A. Chem. Commun. 1995, 1321. (b) Oikawa, H.; Kobayashi, T.;
Katayama, K.; Suzuki, Y.; Ichihara, A. J. Org. Chem. 1998, 63, 8748.
(4) Clyde-Watson, Z.; Vidal-Ferran, A.; Twyman, L. J.; Walter, C. J.;
McCallien, D. W. J.; Fanni, S.; Bampos, N.; Wylie, R. S.; Sanders, J. K.
M. New J. Chem. 1998, 22, 493.
Figure 1. (a) Hetero-Diels-Alder reaction of 1 and 2 to give adduct
3. (b) Structure of porphyrins employed in this work. Synthetic details
for 4,4 5,4 7,10and 84 have been given previously.
(5) Marty, M.; Clyde-Watson, Z.; Twyman, L. J.; Nakash, M.; Sanders,
J. K. M. Chem. Commun. 1998, 2265.
well documented but we are not aware of any crystal structures
of synthetic hosts that could shed useful light on the features
responsible for success or failure.9 It is a common view that
host rigidity and preorganization are paramount, but kinetic and
binding results for Diels-Alder reactions accelerated by oli-
goporphyrins suggest that flexibility is important: host respon-
siveness to the geometrical demands of the guests and transition
state is a key factor controlling rate and regioselectivity.4
The regiospecific hetero-Diels-Alder reaction of 4-pyridyl
butadiene 1 with 3-nitroso pyridine 2 to give oxazine 3 (Figure
1a) can be accelerated by various metalloporphyrin trimers, such
as 4 and 5 (Figure 1b), leading to a rate acceleration of 300-
(6) (a) Wang, B.; Sutherland, I. O. J. Chem. Soc., Chem. Commun. 1997,
1495. (b) Rebek, J., Jr.; Kang, J. Nature 1997, 385, 50. (c) Kang, J.;
Santamaria, J.; Hilmersson, G.; Rebek, J., Jr. J. Am. Chem. Soc. 1998, 120,
7389. (d) Ooi, T.; Kondo, Y.; Maruoka, K. Angew. Chem., Int. Ed. Engl.
1998, 37, 3039.
(7) (a) Meekel, A. A. P.; Resmini, M.; Pandit, U. K. J. Chem. Soc., Chem.
Commun. 1995, 571. (b) Hilvert, D.; Hill, K. W.; Nared, K. D.; Auditor,
M.-T M. J. Am. Chem. Soc. 1989, 111, 9261. (c) Schultz, P. G.; Braisted,
A. C. J. Am. Chem. Soc. 1990, 112, 7430. (d) Gouverneur, V. E.; Houk, K.
N.; Pascual-Teresa, B. d.; Beno, B.; Janda, K. D.; Lerner, R. A. Science
1993, 262, 204. (e) Yli-Kauhaluoma, J. T.; Ashley, J. A.; Lo, C.-H.; Tucker,
L.; Wolfe, M. M.; Janda, K. D. J. Am. Chem. Soc. 1995, 117, 7041.
(8) (a) Tarasow, T. M.; Tarasow, S. L.; Tu, C.; Kellogg, E.; Eaton, B.
E. J. Am. Chem. Soc. 1999, 121, 3614. (b) Seelig, B.; Ja¨schke, A. Chem.
Biol. 1999, 6 167.
10.1021/ja9922227 CCC: $19.00 © 2000 American Chemical Society
Published on Web 05/20/2000

Product-Induced Distortion of a Metalloporphyrin Host
J. Am. Chem. Soc., Vol. 122, No. 22, 2000 5287
Alder transition state,2 the structures of the solvated product-
free host 6 and the 6‚3 host-product complex provide insight
into the features responsible for Diels-Alder acceleration by
metalloporphyrin hosts; desolvation of host and guests emerges
as another factor.
Table 1. Host Accelerated Rates for Formation of 3 and Binding
Constant of 3 with Various Hosts, Measured in Dichloromethane at
25 °C
porphyrin
host
rate accelerationa,b
(approx.)
binding constantc
M-1
2 equiv of 8d
2
1030e
280e
65e
6.9 × 103
2.3 × 108e
7.3 ×106e
5.6 × 107e
7.1 × 107e
4d
5d
6
Results and Discussion
Hosts 6 and 7 were synthesized as shown in Scheme 1.
Coupling the porphyrin monomer 9 to succinyl chloride or to
dodecanedioyl dichloride yielded the linear species 10 and 11,
respectively; deprotection and Glaser cyclization gave 6 and 7
in overall yield of 39% and 56%, respectively. All the host
accelerated reactions were carried out using 0.33 mM guests
and host; methods for analysis of kinetics and binding were as
described previously.5 Small crystals of 6, 7, and the 6‚3
complex were grown and the structures were determined.
7
840e
a Increase in initial rate due to the presence of 1 equiv of host
molecule. b The measured rate constant for the control (host free)
reaction is 0.0039 M-1 s-1 5 c The binding constant of oxazine 3 to
.
the trimers corresponds to the initial bidentate interaction of the ligand
with the host observed at low oxazine concentrations. d See ref 5. e To
allow comparison with hosts 6 and 7 one has to consider that trimers
4 and 5 have six constructive binding possibilities for the diene +
dienophile or for oxazine 3, whereas the hosts 6 and 7 have only two
constructive binding possibilities. Therefore, a trimer should inherently
be 6/2 ) 3 times better at accelerating the reaction and binding oxazine
3 than a host molecule that has only two Zn binding sites (as in 6 and
7), and to compare the efficiency of the above host molecules, per two
Zn binding sites, the acceleration rates and binding constants of hosts
6 and 7 should be multiplied by this factor.
Scheme 1. Synthetic Route to Macrocycles 6 and 7
1000 (Table 1).5 We have now prepared10 a series of capped
dimers, including 6 and 7 (Figure 1b), which accelerate 65-
fold and 840-fold respectively the reaction of 1 and 2 and also
bind the product 3 very strongly (Table 1). Addition of 2 equiv
of monomer 8 (Figure 1b) to a reaction mixture of 1 and 2 leads
to an acceleration rate of only 2-fold (Table 1). This suggests
that it is mainly the cyclic structures of hosts 4-7, which hold
the two reactants in close proximity inside the cavity, that lead
to the significant acceleration rates observed with these hosts.11
More importantly, we have succeeded in X-ray structure
determinations of single crystals of methanol-solvated 6 (Figure
2), toluene/methanol-solvated 6 (Figure 3a,b), methanol-solvated
7 (Figure 4), and the 6‚3 complex (Figure 3c,d). These structures
provide the first crystallographic evidence that indeed acceler-
ated product formation can be accompanied by substantial host
distortion. As the Diels-Alder product resembles the Diels-
(9) The X-ray crystal structures for complexes between an antibody and
a transition state analogue and an inhibitor have been recently reported. (a)
Romesberg, F. E.; Spiller, B.; Schultz, P. G.; Stevens, R. C. Science 1998,
279, 1929. (b) Heine, A.; Stura, E. A.; Yli-Kauhaluoma, J. T.; Gao, C.;
Deng, Q.; Beno, B. R.; Houk, K. N.; Janda, K. D.; Wilson, I. A. Science
1998, 279, 1934.
(10) The general route is described in: Twyman, L. J.; Sanders, J. K.
M. Tetrahedron Lett. 1999, 40, 6681.
Two types of solvated 6 crystals could be grown from
different solvent mixtures. From a dichloromethane/hexane/
methanol mixture 6 crystallizes as a methanol-solvate in the
monoclinic space group P21/c, in which two crystallographically
inequivalent cyclic hosts are found in the asymmetric unit
(Figure 2). From a toluene/methanol mixture, 6 crystallizes as
a toluene/methanol-solvate, which has only one host in the
asymmetric unit and in the orthorhombic space group Pbca
(Figure 3a,b). In both solvated forms of 6 (Figures 2 and 3a,b)
and also in the crystal structure of the 6‚3 complex (Figure 3c,d),
the hexyl chains (not shown) are arranged such that they block
the “open” side of the cavity. In the crystal structure of 6 the
two porphyrin moieties are rotated by 7.4°, 3.6°, and 6.8° with
respect to each other (see Figures 2b, 2d, and 3b, respectively),
indicating that the chainlike linkage is somewhat flexible, but
this rotation is insignificant in the 6‚3 complex (see Figure 3d).
Two methanol molecules are bound to the zinc atoms within
the cavity of 6. In addition, six or eight methanol molecules
are trapped in the cavity of 6 (see Figure 2a,b or Figure 2c,d,
respectively,) and two toluene molecules are trapped in the
cavity of 6 (see Figure 3a,b). This is in contrast to the crystal
(11) (a) The very weak acceleration rate observed with monomer 8 and
the fact that addition of 1 equiv of product 3 to the reaction mixture of 1
and 2 in the presence of 5 strongly inhibit the reaction rate (the acceleration
rate is reduced by 20-fold)5 suggest that the accelerated reaction in the
presence of such cyclic-metalloporphyrin hosts occurs inside the cavity of
the host. (b) Free metal cations have also been used to accelerate Diels-
Alder reactions (for recent reviews see: Kagan, H. B.; Riant, O. Chem.
ReV. 1992, 92, 1007. Pindur, U.; Lutz, G.; Otto, C. Chem. ReV. 1993, 93,
741). However, as the Zn atoms in our systems are coordinated to the pyrrole
rings in the porphyrin units, this is expected to reduce the positive charge
on the Zn atoms and therefore to reduce their Lewis acidity, in comparison
with free Zn2+ cations. Indeed, the charge on the Zn atom in a Zn-porphyrin
unit was calculated to be only +0.4, see: Zerner, M.; Gouterman, M. Theor.
Chim. Acta 1966, 4, 44. Therefore, this is expected to reduce the stabilization
due to acid-base interaction between the pyridine groups and the Zn atoms
in our systems, relative to such a stabilization with free Zn2+ cations, and
could account for the low acceleration rate (of only 2-fold) observed with
monomer 8. (c) As each of the porphyrin units in hosts 4-7 can bind the
reactants inside or outside the cavity, only a small fraction of bound host
is the reactive complex (host + the two reactants inside the cavity) which
leads to accelerated product formation. Therefore, the acceleration rates
shown in Table 1 underestimate the true ability of such hosts to accelerate
the Diels-Alder reaction inside the cavity of the host. The kinetic and
stoichiometric aspects of the Diels-Alder reaction in the presence of cyclic-
metalloporphyrin hosts have been thoroughly discussed elsewhere: Wylie,
R. S.; Sanders, J. K. M. Tetrahedron 1995, 51, 513.

5288 J. Am. Chem. Soc., Vol. 122, No. 22, 2000
Nakash et al.
Figure 2. Two views, a and b (left host) and c and d (right host), of the crystal structure of the “free” cyclic metalloporphyrin host 6, which
contains two inequivalent macrocycles 6 (left side and right side) in the unit cell. Two MeOH molecules are bound to the Zn atoms in the host,
which also contains six or eight MeOH molecules, left or right host, respectively, that are trapped in the cavity. In this crystal, the included
methanol solvent molecules within the cavity of 6 exhibited considerable disorder and therefore the hydrogen atoms for these molecules are not
shown. The hexyl side chains (four per porphyrin unit) have been omitted for clarity. Selected key geometrical features for 6: Zn-Zn distance )
10.792(3) Å (left host) and ) 10.828(3) Å (right host); the two Zn atoms lie at a mean distance of 0.19-0.22 Å above the N4 plane [which are
nearly coplanar, maximum deviation 0.064 Å] toward the oxygen atom; relative rotation of porphyrins ) 7.4° (left host) and ) 3.6° (right host);
Zn-O bond lengths ) 2.22(1), 2.20(1) Å (left host) and 2.21(1), 2.18(1) Å (right host).
structure of the 6‚3 complex (Figure 3c,d) which does not appear
to contain solvent molecules.12a These solvent molecules are
displaced from the cavity on binding of the Diels-Alder
substrates or transition state, and are absent from the product
complex (as seen graphically in Figure 3). This could imply
the following two effects: (a) As the Diels-Alder product is
poorly solvated (if at all) inside the cavity of host 6,12a,b the
host-accelerated reaction is less solvated (if at all) inside the
cavity of the host molecule, suggesting that solvent stabilization
is not as significant for the host-accelerated reaction as in the
solvated control (host free) reaction. This conclusion was
reached previously from solvent effect experiments and binding
results, which showed that the control reaction is faster in CH2-
Cl2 than in toluene while the host-accelerated reaction is faster
in toluene than in CH2Cl2.5 (b) As more than one molecule of
solvent is likely to be displaced from within the cavity of host
6, upon binding of only one ligand 3 (as seen by comparing
Figures 2 or 3a with Figure 3c),12 it is possible that the large
entropic penalty expected for binding a bidentate ligand such
as 3 is partially compensated by entropic gain due to solvent
displacement from within the cavity. While it is in a sense obvi-
ous that some desolvation of hosts and guests is a prerequisite
(12) (a) Because of the poor quality of the diffraction data, it cannot be
proved unequivocally that disordered solvent is not present in the crystal
structure of the 6‚3 complex. (b) However, as 3 occupies a substantial part
of the cavity of 6, clearly seen in Figure 3c and more so in the space-
filling model of the 6‚3 complex (not shown), the Diels-Alder product 3
is poorly solvated (if at all) inside the cavity of 6. (c) In addition, it is
likely that even if solvent is trapped in the space left in this cavity there
will still be fewer molecules of solvent + ligand 3 in the cavity in the case
of the 6‚3 complex (Figure 3c) than in the cavity of the free host 6 (see
Figures 2 and 3a).

Product-Induced Distortion of a Metalloporphyrin Host
J. Am. Chem. Soc., Vol. 122, No. 22, 2000 5289
Figure 3. Two views, a and b (left side) and c and d (right side), of the crystal structures of (on the left side) the “free” cyclic metalloporphyrin
host 6, containing also two molecules of methanol and two toluene molecules, and of (on the right side) the complex between 6 and the hetero-
Diels-Alder product 3. The hexyl side chains (four per porphyrin unit) have been omitted for clarity. Selected key geometrical features for 6:
Zn-Zn distance ) 10.576(5) Å; the two Zn atoms lie at a mean distance of 0.19 Å above the N4 plane [which are nearly coplanar, maximum
deviation 0.001 (left ring) and 0.064 Å (right ring)] toward the oxygen atom; relative rotation of porphyrins ) 6.8°; Zn-O bond lengths ) 2.20(1)
and 2.28(2) Å. Selected key geometrical features for the 6‚3 complex: Zn-Zn distance ) 11.684(3) Å; the two Zn atoms lie at a mean distance
of 0.25 (left ring) and 0.30 Å (right ring) above the N4 planes [which are nearly coplanar, maximum deviation 0.008 (left ring) and 0.019 Å (right
ring)] toward the pyridine nitrogen atom; relative rotation of porphyrins ) 0.1°; NPy-NPy distance ) 7.52(3) Å; Zn-NPy ) 2.20(2) Å (left porphyrin
unit), Zn-NPy ) 2.18(2) Å (right porphyrin unit).
for the chemistry to occur, it is rare to see such a graphic demon-
stration of solvent displacement by a ligand guest molecule.13
The solvated structures of 6 (Figures 2 and 3a,b) have the
same qualitative rectangular structure, in which the two por-
phyrin moieties are virtually planar and almost parallel and have
similar Zn-Zn distances. We chose the toluene solvated
structure (Figure 3a,b) for detailed structural comparison with
the 6‚3 complex (see below); obviously, comparison of the 6‚3
complex with the structures in Figure 2 would lead to the same
qualitative conclusions. As the crystal structure of 6 (Figure
3a,b) does not contain a bidentate ligand that could induce
significant structural distortion, we will refer to it as the free
host 6. The 6‚3 complex (Figure 3c,d) crystallizes from a
dichloromethane/hexane/methanol mixture in the centrosym-
metric monoclinic space group P21/n, with a racemic pair of
enantiomeric oxazine molecules within the structure. The N-N
distance between the two pyridine nitrogens in the bound
oxazine 3 is 7.52(3) Å. 3 is a viscous oil, and attempts to obtain
crystals for an unbound oxazine 3 from different combinations
of solvents and temperatures were unsuccessful. Force field
calculations for 3 show that the calculated oxazine ring has a
similar structure to that found in the 6‚3 complex but that the
two pyridine rings are rotated to a different extent (23° and 35°
(13) In some systems, displacement of solvent molecules has been
suggested as the driving force for binding and acceleration: (a) Kang, J.;
Rebek, J., Jr. Nature 1996, 382, 239. (b) See also ref 6c.

5290 J. Am. Chem. Soc., Vol. 122, No. 22, 2000
Nakash et al.
for the 3-pyridine ring and for the 4-pyridine ring, respectively)
with respect to the oxazine ring.14 This leads to a calculated
N-N distance between the two pyridine nitrogens in 3 of 8.1
Å, some 0.6 Å larger than found in the crystal. The Zn-N
intermolecular distances between the two zinc atoms and the
3-pyridine and the 4-pyridine nitrogens in the 6‚3 complex are
2.20(2) and 2.18(2) Å, respectively; these are consistent with
other zinc porphyrin-pyridine complexes.15
The dihedral angles between the pyridine ring plane and the
N4 plane are 61.7(5)° and 59.0(6)° for the 3-pyridine and the
4-pyridine, respectively. These features are unusual for mono-
meric pyridine.zinc porphyrin complexes, where the pyridine
ring is normally almost perpendicular to the porphyrin ring,15
but do appear to be characteristic of oligomers that contain
pyridylporphyrin-metalloporphyrin linkages.16 In the latter
cases, the lack of orthogonality has been rationalized by crystal
packing effects, but this explanation is not tenable in the 6‚3
complex where the distortion is an inevitable result of the
geometrical mismatch between host and guest (see below). The
tilted position of the pyridine rings in the 6‚3 complex implies
that the presumed perpendicular binding of the two substrates
does not lead directly to the transition state geometry.
Comparison of the solvated structure of the free host 6 (Figure
3a) with that of the 6‚3 complex (Figure 3c) reveals that when
the Diels-Alder product 3 is bound within the cavity, it induces
structural changes in 6: the two porphyrin moieties are pushed
apart, increasing the Zn-Zn distance by 1.1 Å (see Figure 3);
the right-hand porphyrin is forced to lean away from the cavity
while remaining virtually planar; and the left-hand porphyrin
adopts a strikingly domed shape and a slight lean toward the
cavity. The degree of leaning toward or away from the cavity
(of the porphyrin units) can be estimated by the angle between
three meso porphyrin carbons, of which two are connected to
the butadiyne linker and one is connected to the diester linker.17a
For the left-hand porphyrin unit in the 6‚3 complex (Figure 3c)
this angle is 86°, indicating that this porphyrin unit leans slightly
toward the cavity, while for the right-hand porphyrin unit in
the 6‚3 complex this angle is 103°, indicating that this porphyrin
unit leans away from the cavity. For comparison, these angles
for both porphyrin units in the qualitatively rectangular molec-
ular structure of 6 (Figure 3a) are 94°. These changes result
from the orientation of the pyridine rings and lone pairs in 3
relative to the Zn sites in 6. In particular, the Npy-Zn-Zn angles
are 5.2(6)° and 21.3(4)° for the 3- and 4-pyridine, respectively.
These structural changes to accommodate the product 3 are
facilitated by the flexible diester chain linkage, which is
significantly stretched in the product complex. The ability of
host 6 to change its structure is responsible for the observation
that it binds 3 very strongly and is able to accelerate the reaction
of 1 and 2 (see Table 1), although the Zn-Zn distance in the
free host 6 (10.576(5) Å, Figure 3a) is much shorter than that
in the 6‚3 complex (11.684(3) Å, Figure 3c), and even more so
in the transition state complex (in which this distance is expected
to be ∼13.5 Å, see below). Therefore, if 6 would not be able to
distort its structure as seen in Figure 3, it would be less effective
in stabilizing the transition state complex and in accelerating
the reaction in Figure 1a. Obviously, host molecules with a
longer Zn-Zn distance, closer to that expected in the transition
state complex, should have better potential in accelerating the
reaction in Figure 1a. There are other examples where flexibility
has also been invoked as a key feature of catalytic success.1b,18
Although the two crystal structures of 6 (Figures 2 and 3a,b)
have quite different packing and cavity solvation, they retain
the same qualitative rectangular structure, in which the two
porphyrin moieties are virtually planar and almost parallel and
have similar Zn-Zn distances (see above). This implies that
for the 6‚3 complex (Figure 3c,d) we are observing fundamental
structural changes rather than purely environmental effects.
The bicyclo[2.2.2.]octene system, such as 12, is commonly
regarded as a transition state analogue for the Diels-Alder
reaction.7a-d,9 The ethano bridge in such systems locks the cyclo-
hexene ring (or the oxazine ring in 12) in a conformation that
resembles the proposed pericyclic transition state2 for the Diels-
Alder reaction of a cisoid diene and dienophile.7a-d The rigid
bicyclo[2.2.2.]octene system in 12 imposes a somewhat different
geometrical arrangement on the oxazine ring in this system, in
comparison with the favored conformation found for the oxazine
ring in the crystal structure of the 6‚3 complex (see Figure 3c).
As a result, the distance between the two pyridine nitrogens in
12 is calculated to be 9.1 Å, around 1 Å longer than that
calculated for 3.14 This, in addition to the two Zn-N distances
in the transition state complex, approximated to the value as in
the 6‚3 complex (∼2.2 Å, see above), provides an estimated
distance of ∼13.5 Å between the two zinc atoms in the transition
state complex, rather than 11.7 Å which was found in the crystal
structure for the 6‚3 complex (Figure 3c). This implies that the
transition state complex experiences maximum strain energy,
the two porphyrin moieties being pushed even further apart than
in the 6‚3 complex.19 The ability of 6 to change its structure,
as manifested in the 6‚3 complex (Figure 3c) and even more
dramatically expected in the transition state complex, enables
6 to accelerate the reaction of 1 and 2 (see Table 1), although
the Zn-Zn distance in the free host 6 (10.576(5) Å, Figure 3a)
is much shorter than that in these complexes. The substantial
geometrical distortions imposed on 6 (seen in the 6‚3 complex
and expected in the transition state complex) could account for
the modest rate acceleration displayed by 6 compared with
trimer 4 (Table 1), in which the Zn-Zn distances across the
acetylene linkages are calculated to be 14 Å, similar to that
expected in the transition state complex. In support of this idea,
trimer 5, which has three rigid butadiyne linkages, leading to
Zn-Zn distances of ∼16 Å and that are much longer than that
expected in the transition state complex, is less effective than 4
at accelerating the reaction in Figure 1a.
(14) (a) The possible geometries of 3 were explored by a Monte-Carlo
conformational search and were optimized. The most stable structure for 3
has the same conformation for the oxazine ring and the same type of
inversion about the nitrogen atom in this ring, as was found in the 6‚3
crystal structure. (b) The computational method used in this work is
described in the Experimental and Computational Procedures section.
(15) (a) Anderson, H. L.; Bashall, A.; Henrick, K.; McPartlin, M.;
Sanders, J. K. M. Angew. Chem., Int. Ed. Engl. 1994, 33, 429. (b) Cullen,
D. L., Jr.; Meyer, E. F. Acta Crystallogr. 1976, B32, 2259. (c) Collins, D.
M.; Hoard, J. L. J. Am. Chem. Soc. 1970, 92, 3761.
(16) Fleischer, E. B.; Shachter, A. M. Inorg. Chem. 1991, 30, 3763.
(17) (a) As the two porphyrin units in 6 and in the 6‚3 complex are
qualitatively not offset relative to each other (see Figure 3b,d), this angle
is truly describing the degree of leaning toward or away from the cavity of
the two porphyrin units in the host. (b) The same is also true for 7 (Figure
4).
As the distance between the two pyridine nitrogens in the
transition state is expected to be longer than that in 3 (see above)
(18) Menger, F. M.; Ding, J.; Barragan, V. J. Org. Chem. 1998, 63, 7578.
(19) (a) This could be seen by examining molecular models. (b) Due to
the lack of high-quality Zn‚‚‚N parameters in force field packages that we
know of, we do not present here a calculated structure for the 6‚12 complex.
Obviously, these type of complexes are too large for high-level molecular
orbital calculations with current computational resources available to us.

Product-Induced Distortion of a Metalloporphyrin Host
J. Am. Chem. Soc., Vol. 122, No. 22, 2000 5291
Figure 4. Two views, a and b, of the crystal structure of the cyclic metalloporphyrin 7. Two MeOH molecules are bound to the Zn atoms in the
host, which also contains two additional MeOH molecules that are each hydrogen bonded to the bound methanols. The hexyl side chains (four per
porphyrin unit) have been omitted for clarity. Selected key geometrical features for 7: Zn-Zn distance ) 14.625(3) Å; the two Zn atoms lie at a
mean distance of 0.14 Å above the N4 plane [which are nearly coplanar, maximum deviation 0.006 Å (left and right rings)] toward the oxygen
atom; relative rotation of porphyrins ) 15.1°; Zn-O bond lengths ) 2.34(1) Å.
and as the structural changes found on going from the free host
6 to the 6‚3 complex are facilitated by the flexible ester linkage,
hosts with larger and more flexible linkages, such as -(CH2)10-,
as in 7 (see Figure 4), instead of -(CH2)2- (as in 6) should
better stabilize the transition state complex and lead to improved
acceleration rate. Indeed, host 7 is much more efficient in
accelerating the Diels-Alder reaction (in Figure 1a) than 6,
giving 840-fold acceleration rather than 65-fold, respectively
(see Table 1). 7 crystallizes from a dichloromethane/hexane/
methanol mixture as a methanol-solvate in the monoclinic space
group C2/c (Figure 4). Two hexyl chains (not shown) on each
porphyrin unit interpenetrate the cavity of an adjacent dimer,
filling the cavity space with four hexyl chains from two adjacent
dimers. The other two hexyl chains fill the gaps between
porphyrin units of adjacent dimers. The two porphyrin units in
7 are qualitatively planar and have identical structure, as 7 has
C2 symmetry. Two methanol molecules are bound to the zinc
atoms within the cavity of 7. In addition, two methanol
molecules are trapped in the cavity of 7 and each is hydrogen
bonded to the bound methanol molecules (see Figure 4). The
intermolecular O‚‚‚O distance, 2.55(2) Å, and the O‚‚‚H distance
in each of these pairs of hydrogen bonded MeOH molecules
are in the range known for O-H‚‚‚O type hydrogen bonding.20,21
The two porphyrin moieties in the crystal structure of 7 are
rotated by 15.1° with respect to each other (see Figure 4b) and
to a higher degree than in 6 (7.4°, 3.6°, and 6.8°, see above
and Figures 2b, 2d, and 3b, respectively). This is probably due
to the longer and more flexible -(CH2)10- linkage in 7 than
the -(CH2)2- linkage in 6.
the angles between three meso porphyrin carbons, of which two
are connected to the butadiyne linker and one is connected to
the ester linker,17 are 115° for both porphyrin units in 7 (Figure
4a), while in 6 these angles are only 94° (see above and Figure
3a). The long and flexible -(CH2)10- linker should make it
easier for the porphyrin units in 7 to adopt the “right” Zn-Zn
distance and leaning angles to accommodate 3 or the transition
state for the reaction in Figure 1a.22 This might be the reason
for the higher binding constant found for the 7‚3 complex in
comparison with the 6‚3 complex (see Table 1), although a
higher entropic penalty is expected to be paid upon binding of
3 by the more flexible host 7 and for which the Zn-Zn distance
is longer by ∼3 Å than in the 6‚3 complex.23 Since the N-N
distance between the two pyridine nitrogen atoms in the
transition state is expected to be longer than in 3 (see above),
this effect should be even more dramatic in the transition state
to form the Diels-Alder product 3 than in the case of binding
3. This would account for the significantly higher acceleration
rate observed with 7 than with 6 (see Table 1). We are now
preparing a series of different but closely related hosts, with a
view to exploring the structure-activity relationships kinetically,
thermodynamically, and structurally.
By comparison with conventional small molecule crystal-
lography, the R-factors associated with the structures described
here are rather large. This is hardly surprising: the large unit
cell and lack of heavy atoms ensure that the X-ray scattering
(22) Unfortunately, many attempts to obtain single crystals suitable for
X-ray crystallography for the 7‚3 complex were not successful.
(23) Obviously, entropic factors due to different desolvation processes
for binding 3 to hosts 6 vs 7 could also partially contribute to the binding
constants measured between these hosts and the Diels-Alder product 3:
It is likely that in solution host 7 would contain more solvent molecules in
its cavity than the smaller host 6. Therefore, this could lead to a different
entropic gain due to solvent release upon binding of 3 in the cavity of 6 vs
7 (see text above). However, this does not have to be always the case, as
having more solvent molecules in a larger cavity does not necessarily ensure
that more solvent molecules will be displaced by the ligand. The desolvation
process upon binding of ligands inside the cavity of host molecules is a
complex phenomenon and, therefore, only some possible qualitative aspects
of this problem are discussed here.
The long and flexible -(CH2)10- linker in 7 should make it
possible for the two porphyrin units in 7 to lean toward or away
from the cavity and to a much higher extent than in 6; indeed,
(20) For the geometry of hydrogen bonding see: (a) Allen, F. H.;
Motherwell, W. D. S.; Raithby, P. R.; Shields, G. P.; Taylor, R. New J.
Chem. 1999, 23, 25. (b) Jeffrey, G. A.; Saenger, W. Hydrogen Bonding in
Biological Structures; Springer-Verlag: Berlin, 1991.
(21) As the precision in the location of hydrogen atoms is less than that
for oxygen atoms, we present the O‚‚‚O distances.

5292 J. Am. Chem. Soc., Vol. 122, No. 22, 2000
Nakash et al.
1H NMR (250 MHz, CDCl3) δ 0.91-1.02 (24H, m, hexyl Me), 1.30-
1.54 (32H, s, hexyl CH2), 1.74 (16H, m, hexyl CH2), 2.09-2.55 (16H,
m, hexyl CH2 + 24H, ring Me + 4H, CH2 cap), 3.19 (2H, s, CtC-
H), 3.70-4.26 (16H, s, CH2-Por), 5.37, 5.38 (4H, 2 × s, benzyl CH2),
7.56-8.33 (16H, m, aryl-H), 9.93, 10.08, 10.21 (4H, 3 × s, meso-H);
13C NMR (62.5 MHz, CDCl3, APT) δ 13.94, 14.67, 15.15, 15.62 (Me),
19.72, 20.58, 22.77, 23.96, 26.73, 29.32, 30.08, 31.99, 33.35 (CH2),
66.78 (CH2O-aryl), 83.94 (C≡C-Ar), 97.49 (meso), 117.80, 118.54,
121.37, 134.92, 137.60, 137.71, 141.50, 143.34, 143.66, 143.92, 144.11,
146.18, 146.29, 147.41 (quaternary pyrrole and aryl carbons), 127.57,
127.58, 127.59, 127.61, 131.82, 132.81, 133.09, 133.74, 136.77 (aryl-
H), 172.09 (ester CdO); FABMS (C130H162N8O4Si2Zn2) calcd 2027.130,
found 2027.5.
power of the crystals will be very low,24 while the four
solubilizing hexyl chains per porphyrin unit tend to be disor-
dered; furthermore, solvent molecules are often included within
and between molecules in the lattice. In general, therefore, we
have needed recourse to synchrotron sources of X-rays. It is
important to note that although the large R-factors associated
with this kind of system inevitably limit the precision with which
one can determine bond lengths and angles, they do not detract
from our ability to draw conclusions about intermolecular
interactions or large scale molecular distortions.
Conclusions
Host 6: The deprotected 10 (118 mg, 0.058 mmol) was dissolved
in dry dichoromethane (340 mL) and subjected to Glaser-Hay coupling
cyclization conditions; freshly prepared copper(I) chloride (402 mg,
4.06 mmol, 70 equiv) and TMEDA (0.62 mL, 4.06 mmol, 70 equiv)
were added and the solution stirred for 15 h at room temperature under
a drying tube. The mixture was washed with water (6 × 200 mL), the
organics were dried (MgSO4) and filtered, and the solvent was removed
by evaporation to give a red solid. The cyclic product was purified by
chromatography on silica, eluting with hexane/ethyl acetate (4:1 v/v),
and recrystallized from chloroform/methanol (88 mg, 75%). 1H NMR
(250 MHz, CDCl3) δ 0.76-0.81 (24H, m, hexyl Me), 1.14-1.39 (32H,
m, hexyl CH2), 1.51-1.59 (16H, m, hexyl CH2 ), 1.95 (16H, m, hexyl
CH2), 2.20, 2.26 (24H, 2 × s, ring Me), 2.29 (4H, s, CH2 cap), 3.71
(16H, m, CH2-Por), 5.21 (4H, s, benzyl CH2), 7.42 (2H, s, aryl-H2),
7.45 (2H, s, aryl-H2′), 7.54-7.72 (8H, m, aryl-H), 8.16 (2H, d, J )
6.8 Hz, aryl-H), 8.35 (2H, d, J ) 6.8 Hz, aryl-H), 9.83 (4H, s, meso-
H); 13C NMR (62.5 MHz, CDCl3, APT) δ 13.90, 15.07, 15.48 (Me),
22.57, 26.53, 28.27, 29.85, 31.78, 33.14 (hexyl CH2), 66.36 (benzyl
CH2), 74.39, 78.08 (CtC), 97.03 (meso), 112.63, 116.74, 119.81,
120.05, 120.78, 134.71, 137.02, 137.33, 143.10, 143.28, 144.44, 146.02,
147.22, 147.30 (quaternary pyrrole and aryl carbons), 126.99, 127.11,
127.38, 129.27, 132.43, 132.85, 133.13, 139.74 (aryl-H), 172.34 (ester
CdO); MALDI-TOF (C130H160N8O4Zn2) calcd 2025.114, found 2025.1.
In this study we have demonstrated, for the first time for
synthetic receptors, a detailed structural analysis of the geo-
metrical changes imposed on an accelerating agent on binding
of a Diels-Alder product. We have confirmed the importance
of host flexibility, for acceleration of the hetero-Diels-Alder
within the cavity of a cyclic metalloporphyrin receptor. This
opens the possibility of exploring in detail the influence of host
geometry changes on acceleration rates and will direct us in
the design of new and better accelerating metalloporphyrin
receptors. The absence of solvent molecules from the host-
guest 6‚3 complex (Figure 3c,d), in contrast to the solvated
crystal structures of the free host 6 (Figures 2 and 3a,b), implies
that solvent stabilization is not as significant for the host-
accelerated reaction as in the control (host free) reaction, a
conclusion that had previously been reached from kinetic and
binding results.5
Experimental and Computational Procedures
General. 1H NMR spectra (250 MHz) was recorded on Bruker AC-
250 spectrometers. 13C NMR spectra were obtained on a Bruker AC-
250 operating at 62.5 MHz. All NMR measurements were carried out
at room temperature in deuteriochloroform. Fast atom bombardment
mass spectra (FAB MS) were recorded on a Kratos MS-50 mass
spectrometer. MALDI-TOF mass spectra were recorded on a Kratos
Analytical Ltd., Kompact MALDI IV mass spectrometer. A nitrogen
laser (337 nm, 85 kW peak laser power, 3 ns pulse width) was used to
desorb the sample ions, and the instrument was operated in linear time-
of-flight mode with an accelerating potential of 20 kV. Results from
50 laser shots were signal averaged to give one spectrum. An aliquot
(1 µL) of a saturated solution of the matrix (sinapinic acid) was
deposited on the sample plate surface. Before the matrix completely
dried, a small volume (1 µL) of analytes (dissolved in dichloromethane/
chloroform at 1 mg/mL) was layered on the top of the matrix and
allowed to air-dry.
Deprotected 10: Monomer 9 (232 mg, 0.223 mmol, 2.1 equiv),
DMAP (48 mg, 0.39 mmol), and triethylamine (33 µL, 0.23 mmol)
were added to a stirring solution of succinyl dichloride (12 µL, 0.106
mmol, 1 equiv) in tetrahydrofuran (20 mL, freshly distilled over LiAlH4/
CaH2) at 0 °C. After 2 h the mixture was allowed to warm to room
temperature after which stirring was continued for a further 12 h. The
solvent was removed by evaporation to give a red solid before
redissolving in dichloromethane (100 mL). The mixture was washed
with saturated sodium bicarbonate solution (3 × 100 mL) and water
(3 × 100 mL), dried over MgSO4, and evaporated to dryness. The
product was purified by column chromatography on silica, eluting with
hexane/ethyl acetate (3:1 v/v), and the solvent was removed by
evaporation to give a red solid. This product was redissolved in
dichloromethane (150 mL) and excess TBAF added (300 µL of a 1.1
M solution in tetrahydrofuran, 0.33 mmol). The mixture was stirred
under a drying tube for 10 min, after which no starting material could
be detected by TLC (3:1 v/v ethyl acetate/hexane). Excess calcium
chloride was added to quench any remaining TBAF (4-5 spatula tips)
and the mixture washed with water (2 × 100 mL), dried (MgSO4),
filtered, and evaporated to give 118 mg of the deprotected 10 (52%).
Host 7: The preparation and spectroscopic data for 7 are described
elsewhere.10
Molecular Modeling. Molecular modeling was carried out on a
Silicon Graphics Indy workstation using CERIUS2 version 3.0 (BIO-
SYM/Molecular Simulations) with the UNIVERSAL 1.02 force field.25
Atom and bond parameters were taken directly from the program
database. In comparison with the 6‚3 crystal structure, the program
severely underestimates the pyramidal angle around the nitrogen atom
in the oxazine ring. Therefore, we fixed that pyramidal angle to the
value found in the 6‚3 crystal structure. The MM2 and MM3 force
fields calculate a similar degree of pyramidalization about the oxazine
nitrogen for a model of 3 (in which the pyridine rings were replaced
with Me substituents) as in the 6‚3 complex. However, the MM2 and
MM3 force fields lack the parameters for a pyridine substituted 3 and
therefore were not used. Fixing the pyramidal angle around the oxazine
nitrogen is further justified, as ab initio molecular orbital calculations
(at the HF/6-31G* level) show a nearly identical pyramidal angle about
the nitrogen in N-Me-oxazine (used as a model for 3) to that found in
the 6‚3 crystal structure.
Single-Crystal Structure Determinations. After much effort, only
very small crystals of 6 (MeOH) solvate, 6‚3, and 7 (MeOH) solvate
could be grown. These proved to be only weakly diffracting. To
determine these structures it was necessary to exploit the high intensity
of a synchrotron radiation source. Data were collected at the Daresbury
SRS (UK), Station 9.8, using a Bruker AXS Smart CCD area-detector
diffractometer, in narrow frame mode.26,27 Intensities were integrated28
(25) Rappe´, A. K.; Casewit, C. J.; Colwell, K. S.; Goddard, W. A., III;
Skiff, W. M. J. Am. Chem. Soc. 1992, 114, 10024.
(26) Cernik, R. J.; Clegg, W.; Catlow, C. R. A.; Bushnell-Wye, G.;
Flaherty, J. V.; Greaves, G. N.; Hamichi, M.; Borrows, I. D.; Taylor, D. J.;
Teat, S. J. J. Synchrotron Radiat. 1997, 4, 279.
(27) Clegg, W.; Elsegood, M. R. J.; Teat, S. J.; Redshaw C.; Gibson, V.
C. J. Chem. Soc., Dalton Trans. 1998, 3037.
(28) SMART (control) and SAINT (integration) software, version 4;
Bruker AXS Inc.: Madison, WI 1994.
(24) Harding, M. M. J. Synchrotron Radiat. 1996, 3 (No. Pt6), 250.

Product-Induced Distortion of a Metalloporphyrin Host
J. Am. Chem. Soc., Vol. 122, No. 22, 2000 5293
from several series of exposures. For 6 (which was grown from a
dichloromethane/hexane/methanol mixture, Figure 2) each exposure
covered 0.5° in ω, with an exposure time of 2 s, and the total data set
was more than a hemisphere. For 6‚3 each exposure covered 0.6° in
ω, with an exposure time of 1 s, and the total data set was more than
a hemisphere. For 7 each exposure covered 0.3° in ω, with an exposure
time of 10 s, and the total data set was more than a hemisphere. Data
were corrected for absorption and incident beam decay.29 The unit cell
parameters were refined using LSCELL.30 Crystals grown for 6 (toluene/
MeOH) solvate were of a sufficient quality to use for structure
determination with a conventional laboratory X-ray source. Measure-
ments were made using a Rigaku R-Axis IIc imaging plate system.
Integration and data reduction was performed within the R-AXIS
PROCESS31 software.
23547 measured reflections of which 12688 were independent [R(int)
) 0.097]; final residuals (12688 included reflections, 508 parameters)
R1[I > 2σ(I)] ) 0.1626, wR2 ) 0.4010, S ) 1.155 (w ) 1/[σ2Fo
+
2
(0.2P)2], where P ) (Fo + 2Fc2)/3)). Hydrogen atoms were fixed
geometrically, riding on the relevant heavy atom and refined with
isotropic temperature factors. The largest peak and hole in the final
difference map are 1.160 and -0.706 eÅ-3, respectively. The structure
was solved by direct methods using SIR-9232 and refined with
SHELXL-97.33
2
Crystal data for the 6‚3 complex: C144H169N11O5Zn2, Mr ) 2264.64,
red microcrystals from a dichloromethane/hexane/methanol mixture
(crystal dimension 0.20 × 0.20 × 0.02 mm3); monoclinic, space group
P21/n (no. 14), a ) 16.174(2) Å, b ) 29.754(5) Å, c ) 28.970(4) Å,
â ) 102.63(10)°, V ) 13604(3) Å3, Z ) 4, Fcalc ) 1.106 Mg m-3, µ )
0.408 mm-1; 2θmax ) 40.00°, λ ) 0.6919 Å (synchrotron), T ) 150(2)
K, F(000) ) 4840; 31413 measured reflections of which 12891 were
independent [R(int) ) 0.1197]; final residuals (12891 included reflections,
574 parameters) R1[I > 2σ(I)] ) 0.1899, wR2 ) 0.4217, S ) 1.407
(w ) 1/[σ2Fo2 + (0.2P)2], where P ) (Fo2 + 2Fc2)/3). Hydrogen atoms
were fixed geometrically, riding on the relevant heavy atom and refined
with isotropic temperature factors. The largest peak and hole in the
final difference map are 1.064 and -0.834 eÅ-3, respectively. The
structure was solved by direct methods using SIR-9232 and refined with
SHELXL-97.33
Even using a synchrotron radiation source, these crystals were found
to be weakly diffracting because of extensive disorder of the n-hexyl
chains. This resulted in relatively high R1 values. The carbon atoms in
these disordered groups were refined with isotropic temperature factors
and restrained to give chains with a reasonable geometry.
Crystal data for 6 (MeOH solvate, Figure 2): C136H180N8O10Zn2,
Mr 2217.62, red crystals from a dichloromethane/hexane/methanol
mixture (crystal dimensions 0.28 × 0.16 × 0.16 mm3); monoclinic,
space group P21/c(14), a ) 29.750(4) Å, b ) 29.301(4) Å, c ) 29.914-
(4) Å â ) 90.670(10)°, V ) 26074(6) Å3, Z ) 8, Fcalc ) 1.130 Mg
m-3, µ ) 0.426 mm-1; θmax ) 22.56° (Mo KR), λ ) 0.6887 Å, T )
150(2) K; 81486 measured reflections of which 31773 were independent
[R(int) ) 0.1044]; final residuals (31773 included reflections, 1278
parameters) R1[I > 2σ(I)] ) 0.1454, wR2 ) 0.3291, S ) 1.077 (w )
Crystal data for 7 (MeOH solvate, Figure 4): C142H188N8O8Zn2,
Mr ) 2265.74, red crystals from a dichloromethane/hexane/methanol
mixture (crystal dimensions 0.12 × 0.10 × 0.03 mm3); monoclinic,
space group C2/c (no. 15) a ) 8.038(1) Å, b ) 32.731(4) Å, c )
47.780(6) Å, â ) 91.410(10)°, V ) 12567(3) Å3, Z ) 4, Fcalc ) 1.198
Mg m-3, µ ) 0.442 mm-1; θmax ) 22.50° (Mo KR), λ ) 0.6890 Å, T
) 150(2) K; 27697 measured reflections of which 8971 were
independent [R(int) ) 0.0852]; final residuals (8971included reflections,
655 parameters) R1[I > 2σ(I)] ) 0.1469, wR2 ) 0.3448, S ) 1.153
1/[σ2Fo + (0.1008P)2 + 439.7665P], where P ) (Fo + 2Fc2)/3).
Hydrogen atoms were fixed geometrically, riding on the relevant atom
and refined with isotropic temperature factors. The largest peak and
hole in the difference map are 1.61 and -1.008 eÅ-3, respectively.
The structure was solved by direct methods using SIR-9232 and refined
with SHELXL-97.33
2
2
2
2
(w ) 1/[σ2Fo + (0.1342P)2 + 177.0312P], where P ) (Fo + 2Fc2)/
3). Hydrogen atoms were fixed geometrically, riding on the relevant
atom and refined with isotropic temperature factors. The largest peak
and hole in the difference map are 1.141and -0.708 eÅ-3, respectively.
The structure was solved by direct methods using SIR-9232 and refined
with SHELXL-97.33
Crystal data for 6 (toluene/MeOH solvate, Figure 3a,b): C146H180
-
N8O6Zn2, Mr ) 2273.72, red crystals from a toluene/methanol mixture
(crystal dimensions: 0.20 × 0.20 × 0.20 mm3); orthorhombic, space
group Pbca (no. 61) a ) 30.686(10) Å, b ) 30.124(10) Å, c ) 28.941-
(10) Å, V ) 26753(16) Å3, Z ) 8, Fcalc ) 1.129 Mg m-3, µ ) 0.415
mm-1; 2θmax ) 41.46°, λ ) 0.71069 Å, T ) 180(2) K, F(000) ) 9760;
Acknowledgment. This work was supported by the British
Council, Israel Academy and Ministry of Science, and B′nai
B′rith.
(29) G. M. Sheldrick, SADABS, program for scaling and correction of
area detector data; University of Go¨ttingen, 1997 (based on the method of
Blessing: Blessing, R. H. Acta Crystallogr. Sect A 1995, 51, 33).
(30) Clegg, W.; LSCELL, program for refinement of cell parameters
from SMART data; University of Newcastle upon Tyne, 1995.
(31) For R-AXIS PROCESS: Molecular Structure Corporation, 1995.
PROCESS Version 3.2. MSC, 3200 Research Forest Drive, The Woodlands,
TX 77381, 1997.
(32) Altomare, A.; Cascarano, G.; Giacavazzo, C.; Guagliardi, A.; Burla,
M. C.; Polidori, G.; Camalli, M. C. J. Appl. Crystallogr. 1994, 27, 435.
(33) G. M. Sheldrick, 1997; SHELXL97, Program for the Refinement
of Crystal Structures; University of Go¨ttingen, Germany.
Supporting Information Available: Tables of crystal data,
structure solution and refinement, atomic coordinates, bond
lengths and angles, and anisotropic thermal parameters for
solvated 6 and 7 and for the 6‚3 complex (PDF). This material
is available free of charge via the Internet at http://pubs.acs.org.
JA9922227