Anion binding, aryl-extended cyclotriguaiacylenes and an aryl-bridged cryptophane that provides snapshots of a molecular gating mechanism

Article abstract of DOI:10.1080/10610278.2010.514615

Cyclotriguaiacylene, (±)-CTG, reacted with a series of [(η⁵-C₅H₅)Fe^II- (η⁶-chloroarenes)⁺- to yield a family of aryl-extended, [(η⁵-C₅H₅)Fe ^II-

Full text of DOI:10.1080/10610278.2010.514615

Supramolecular Chemistry
Vol. 22, Nos. 11–12, November–December 2010, 870–890
Anion binding, aryl-extended cyclotriguaiacylenes and an aryl-bridged cryptophane that
provides snapshots of a molecular gating mechanism†
K. Travis Holmana*, Stephen D. Drakea, Jonathan W. Steedb, G. William Orrc and Jerry L. Atwoodc
b
aDepartment of Chemistry, Georgetown University, 37th and O Streets NW, Washington, DC 20057, USA; Department of Chemistry,
Durham University, South Road, Durham DH1 3LE, UK; Department of Chemistry, University of Missouri, 601 S. College Ave.,
Columbia, MO 65211, USA
c
(Received 29 June 2010; final version received 2 August 2010)
Cyclotriguaiacylene, (^)-CTG, reacted with a series of [(h5-C5H5)FeII(h6-chloroarenes)]þ to yield a family of aryl-
extended, [(h5-C5H5)FeII]þ-functionalised cavitands (^)-23þ. Crystal structures of [BF4]2 or mixed [BF4]2/[PF6]2 salts of
these cavitands shed light on their conformational dynamics and anion binding properties. These hosts adopt conformations
that project the [(h5-C5H5)FeII(h6-arene)]þ substituents ‘up’ from the rim of the cavitand, creating deep cavities that are
occupied by a [BF4]2 anion. Hosts (^)-2a–f3þ appear to show some selectivity for the formation of penetrated ion pairs with
[BF4]2 in preference to [PF6]2. The compounds were photochemically demetalated, giving aryl-extended CTGs (^)-3a–d,f.
The crystal structure of (^)-3a reveals that the molecule forms a centrosymmetric, self-included dimer in the solid state.
Cryptophanes syn-5 and (^)-anti-4 were synthesised by the photochemically induced demetalation of the putative metalated
cryptophanes syn-4[PF6]3 and (^)-anti-4[PF6]3, obtained by the ‘capping’ of the cavitand (^)-2d3þ by (^)-CTG. Single
crystal structures of five different solvates of (^)-anti-5 reveal host–guest encapsulated complexes that are interpreted as
revealing ‘snapshots’ of the molecular gating process. The first crystallographically characterised example of an empty
cryptophane, namely (^)-anti-5, reveals that this host can also readily adopt a conformation that fills the bulkofits own cavity.
Keywords: cyclotriveratrylene; cryptophane; encapsulation; gating; penetrated ion pair
Introduction
of container molecule hosts constructed by the joining of
two CTB cavitands – and related higher order container-like
structures derived from the dynamic covalent chemistry (24)
(including metal–ligand self-assembly (25, 26)) of various
CTB derivatives. These and other so-called container
molecules (4, 27) have captured much attention in recent
years related to their ability to selectively encapsulate and/or
organise guests and the relatively high kinetic stabilities of
the corresponding host–guest complexes, allowing com-
plexes of low thermodynamic stability to be sufficiently
long-lived for spectroscopic observation. The lifetimes of
container molecule complexes are attributable to the nearly
closed-surface structures of the hosts: the host imposes a
physical barrier to guest exchange, effectively gating access
to the internal cavity. For very recent reviews concerning
advances in cryptophane chemistry and in the supramole-
cular chemistry of CTBs in general, the reader is referred to
those of Brotin and Dutasta (23), and Hardie (5),
respectively.
[1.1.1]-Orthocyclophane-based macrocycles, i.e. cyclotri-
benzylenes (CTBs, a subset of which are shown in
Scheme 1), have been known for at least 150 years (1).
Present-day interest is commonly associated with the more
or less rigid (2, 3), bowl-shaped – i.e. cavitand (4) –
structure of the CTB scaffold in the context of supramole-
cular chemistry, where these concave hosts can be exploited
for the complexation of various guests. Indeed, CTBs are
among the most studied macrocycles in the field (5). The
cavities of the CTBs have been extensively employed for the
complexation of fullerenes (6), cations (7) and even anions
(8–11). CTB derivatives have also been used as liquid
crystal mesogens (12), to construct monolayers on surfaces
(13) or at the air–solution interface (14), in dendrimers (15),
organic/metallogels (16) and other soft materials (17).
Moreover, like other cavitands, the inability of many CTBs
to close-pack in the solid state has led to their extensive study
in solid-state organic chemistry, forming, for example,
crystalline inclusion compounds (18–20) or polymers of
intrinsic microporosity (21). Also of particular interest are
the cryptophanes (Scheme 1) (22, 23) – Collet’s family
Among the more synthetically useful CTBs that
serve as scaffolds for chemical functionalisation are
the C3v symmetric cyclotriveratrylene (CTV, Scheme 1,
R1 ¼ R2 ¼ OCH3) and cyclotricatechylene (Scheme 1,
*Corresponding author. Email: kth7@georgetown.edu
†This paper is contributed in memory of Prof. Dmitry M. Rudkevich, who positively influenced our thinking in so many ways.
ISSN 1061-0278 print/ISSN 1029-0478 online
q 2010 Taylor & Francis
DOI: 10.1080/10610278.2010.514615
http://www.informaworld.com

Supramolecular Chemistry
871
O
R1
O
R1
R2
O
O
O
R1
R2
O
R1
O
R1
R2
O
O
O
R1
O
R2
R1
R1 R2
R1
R2
bridge
R2
R2
R2
O
CTBs
(±)-CTG
syn
anti
R1 = OH, R2 = OCH3
cryptophanes
Scheme 1. The general structures of cyclotribenzylenes (CTBs) and cryptophanes.
R1 ¼ R2 ¼ OCH3), and the chiral, C3 symmetric cyclo-
trianisidine ((^)-CTA, Scheme 1, R1 ¼ NH2, R2 ¼
OCH3) and cyclotriguaiacylene ((^)-CTG, Scheme 1,
R1 ¼ OH, R2 ¼ OCH3) cavitands (28). The latter, (^)-
CTG, is of particular importance as it constitutes the
structural subunit of most cryptophanes, its chirality giving
rise to syn and anti cryptophane diastereomers. Moreover,
(^)-CTG constitutes the precursor to a variety of chiral,
C3 symmetric CTB derivatives via functionalisation of
the phenolic residues. Our work in the area of anion
complexation (29) by organometallic sandwich com-
pounds of cavitands (10), e.g. [(h6-arene)RuII]2þ or [(h5-
C5Me5)Ir/RhIII]2þ-functionalised CTVs (8), or calix[n ]-
arenes (30), and, more recently, [(h5-C5Me5)RuII]þ-
functionalised cryptophanes (11) (Scheme 2), led us,
some time ago, to pursue deep cavity, upper rim [(h5-
C5H5)Fe(h6-arene)]þ-functionalised CTBs via the nucleo-
philic aromatic substitution (SNAr) of [(h5-C5H5)Fe(h6-
chloroarenes)]þ by (^)-CTG (e.g. (^)-23þ in Scheme 2)
(9). Notably, based upon the premise that limited
conformational flexibility would likely impart anion
selectivity, an important part of the design strategy leading
to cavitands (^)-23þ was the incorporation of aryl ether
moieties, as opposed to, say, benzylic ether moieties at the
upper rim of the CTB cavitand. At the time, (^)-23þ were
the first examples of aryl-functionalised CTGs and a
number of such compounds, also derived from the SNAr
substitution of electron-deficient haloarenes by (^)-CTG,
have since been reported by Pochini and co-workers (31).
Concerning host design, it also occurred to us that unique,
potentially anion-encapsulating cryptophanes such as (^)-
anti-43þ should be available from compounds (^)-23þ
(Scheme 2, X ¼ Cl) by the capping of these cavitands with
another (^)-CTG moiety. The anion hosts in Scheme
2 are comparable to the family of anion binding/sensing
organometallic cavitands developed by Beer and co-
workers (32), except that anion complexation is
not assisted by the presence of traditional hydrogen bond
donor functionalities. Moreover, being constructed almost
entirely of arenes, many of the hosts shown in Scheme 2
are of contemporary relevance in the context of
recent/renewed interest in the interactions between anions
RuCp*
X–
X–
6+
5+
5+
MeO
OMe
MeO
Cp*Ru
MeO
RuCp*
O
OMe OMe
OMe
M(arene)
(arene)M
OH
OH
(arene)M
M(arene)
OH
O–
O
R
MeO
O
(arene)M
X–
M(arene)
n
n
n
M(arene)
X
OMe
O
O
O
OMe
2+
2+
Cp*Ru
X
RuCp*
X
MeO
O
O MeO
O
II
II
MeO
II
Fe
Fe
Fe
RuCp*
II
X–
II
–Fe
II
X
Fe
Fe
MeO
O
O MeO
O
MeO
OMe
O
OMe O
O
OMe
(±)-anti-[4 X–]2+
(±)-[2 X–]2+
Scheme 2. Anion-binding, organometallic cavitands and cryptophanes.

872
K.T. Holman et al.
X
[PF6]3
X
X
X
X
X
FeII
FeII
FeII
X
MeO
OH
HO MeO
OH
[PF6]
MeO
Cl
(i)
(ii)
FeII
O
MeO
+
O MeO
O
MeO
H2
MeO
O
O MeO
O
MeO
H2
H1
H1
X = H
1a[PF6]
(±)-CTG
He
X = o-Cl 1b[PF6]
X = m-Cl 1c[PF6]
X = p-Cl 1d[PF6]
X = p-CH3 1e[PF6]
He
Ha
Ha
X = H
(±)-2a[PF ]3
(±)-2b[PF6]3
(±)-2d[PF6]3
X = H
X = o-Cl
(±)-3a
(±)-3b
X = o-Cl
X = m-Cl (±)-2c[PF6]3
X = m-Cl (±)-3c
X = p-Cl
(±)-3d
X = p-OPh (±)-3f
X = p-Cl
X = p-CH3 (±)-2e[PF6]3
(i)
X = p-OPh (±)-2f[PF 6]3
(i)
C6H5OH
6
(±)-CTG
[PF6]3
[PF6]3
MeO
O
MeO
O
O
MeO
FeII
MeO
O
O
O
O MeO
O MeO
MeO
O
MeO
FeII
MeO
O
O
O
O
O
MeO
MeO
MeO
MeO
MeO
HAr
HAr
(ii)
FeII
FeII
FeII
+
FeII
+
O
O
OMe
OMe
OMe O
O
OMe O
O MeO
O MeO
O
O
O
OMe
OMe
H2
MeO
MeO
H2
H1
H1
He
He
Ha
(±)-anti-5
Ha
syn-5
syn-4[PF6]3
(±)-anti-4[PF6]3
Figure 1. Synthesis of aryl-extended CTGs and aryl-bridged cryptophanes. (i) K2CO3, DMF, room temperature; (ii) hn, CH3CN.
and arenes, as they are early examples of hosts that bind
anions via a charge-assisted form of the so-called anion–p
interaction (33).
procedure reported by Abd-El-Aziz and co-workers for
the synthesis of oligomeric (38) and cyclic (39) aryl ethers
possessing pendant cyclopentadienyliron moieties. Race-
mic cyclotriguaiacylene, (^)-CTG, was reacted at room
temperature with the appropriate [(h5-C5H5)Fe(h6-
chloroarene)][PF6] salt, 1[PF6], in dry dimethyl formamide
(DMF) in the presence of potassium carbonate. Notably, the
reactive chloro substituents at the upper rim of (^)-2d3þ
provide a site for further cavity extension. Thus, extended
(^)-2f[PF6]3 was obtained by reaction of the para-chloro-
functionalised (^)-2d[PF6]3 with phenol under similar
conditions.
Herein, we describe a number of new aryl-extended
CTGs derived from the organometallic compounds (^)-
23þ by the photochemically induced removal of the [(h5-
C5H5)FeII]þ moieties (Figure 1). We also report indirect
evidence for the achievement of [(h5-C5H5)Fe(h6-
arene)]þ-bridged cryptophanes, e.g. (^)-anti-43þ, by
isolating their photochemically demetalated cryptophane
products. New single crystal structures of salts of [(^)-
2]3þ provide insights into their anion complexation
properties and single crystal structures of the correspond-
ing cryptophane solvates reveal host conformations that
can reasonably be interpreted as representing ‘snapshots’
of the molecular gating mechanism in this important family
of supramolecular container compounds. We note that
other deep cavity, C3 symmetric cavitands and containers
have been investigated in the context of their molecular
gating mechanisms (34). Moreover, oligo( p-phenylene
oxides)s are of general interest in supramolecular host–
guest (35) and solid-state organic chemistry (36).
Each new cavitand (^)-2a–f[PF6]3 was characterised
1
by H and 19F NMR spectroscopy, as well as by single
crystal X-ray diffraction whenever possible. The results of
combustion elemental analyses were commonly unsatis-
1
factory (see Experimental section). The H NMR spectra
of these compounds deserve comment as they are often
complicated by anion binding behaviour, which is highly
solvent dependent. The corresponding spectral features
provide a handle for the monitoring of anion binding
behaviour, as discussed in the following section. It is
important to recognise that (^)-2b,c[PF6]3 are likely
obtained as a mixture of all possible stereoisomers that can
arise from the presence of helical chirality at the (^)-CTG
moiety and the introduction of planar chirality at each of
the three metalated, unsymmetrically 1,2-substituted
(2b[PF6]3) or 1,3-substituted (2c[PF6]3) [(h5-C5H5)-
FeII(h6-arene)]þ moieties. Thus, while denoted simply as
(^)-2b,c[PF6]3 in Figure 1, (^)-2b[PF6]3 and (^)-
2c[PF6]3 are in fact a mixture of four diastereomers and
Results and discussion
Synthesis and characterisation
The general synthetic scheme adopted for the work
presented herein is given in Figure 1. For a review on the
use of p-organoiron complexes in arene chemistry, the
reader is referred to that by Astruc (37). Deep cavity, upper
rim [(h5-C5H5)Fe(h6-arene)]þ-functionalised CTGs (^)-
2a–e[PF6] were synthesised by an adaptation of the

Supramolecular Chemistry
their enantiomers. For example, if one uses P/M
873
descriptors to ascribe the stereochemistry of the (^)-
CTG moieties and R/S descriptors to ascribe the
stereochemistry of each of the three [(h5-C5H5)FeII(h6-
arene)]þ moieties, then (^)-2b[PF6]3 is truly a mixture
(likely statistical) of the RRRM-2b[PF6]3, RRSM-2b[PF6]3,
RSSM-2b[PF6]3 and SSSM-2b[PF6]3 diastereomers and
their enantiomers – a total of eight unique compounds. No
attempts were made to separate or quantify the relative
amounts of these diastereomers. Consequently, and as
expected, (^)-2b,c[PF6]3 give rise to complicated and/or
1
broad H NMR spectra, as described in the Experimental
section. Their 1H NMR spectra are therefore either
broadened substantially or extremely complicated as most
peaks appear as multiplets.
An advantage of employing [(h5-C5H5)-
FeII(h6-arene)]þ moieties in the synthesis of aryl ethers
is that the [(h5-C5H5)FeII]þ substituents can, in general, be
readily removed from the arene ring (38, 40). Visible light
photolysis is the method of choice. Accordingly,
photolysis of the [(h5-C5H5)Fe(h6-arene)]þ-functionalised
cavitands (^)-2a–d,f[PF6]3 with direct sunlight in
degassed acetonitrile yields the demetalated products
(^)-3a–d,f in good to near-quantitative yields (Figure 1).
Importantly, the near quantitative yields of (^)-3b,c (97%
each) serve to further establish the structures and purity of
the diastereomeric mixtures of (^)-2b,c[PF6]3, which give
Figure 2. The X-ray crystal structure of (^)-3a showing (a) top
and side views of the molecular structure (50% probability
thermal ellipsoids) and (b) the self-included dimer.
Summary data for all X-ray structure determinations are
given in Table 1. The upper rim aryl substituents are all
directed ‘upward’ from the rim of the CTG bowl. For two
of these arene rings, A and B, the mean plane of the arene
ring is nearly orthogonal to the mean plane of the cavitand
upper rim, as defined by the oxygen atoms of the diphenyl
ether moieties. The corresponding interplanar angles,
defined hereafter as f, measure 908 and 898 for rings A and
B, respectively. The third ring, C, is turned significantly
out of plane relative to the cavitand oxygen atoms
(f ¼ 578), but its carbon atoms nonetheless reside entirely
above the upper rim of the cavitand. The arene–arene
dihedral angles of the three diphenyl ether moieties,
involving rings A, B and C, respectively, measure 778, 738
and 728. Perhaps surprisingly, and unlike many other CTB-
based cavitands, (^)-3a does not form an inclusion
compound with the solvent employed for crystallisation, in
this case toluene. Instead, an interesting form of self-
inclusion is observed, in which one of the upper rim arene
rings of one molecule (ring B) is directed into the
molecular cavity of its enantiomer. Two molecules of (^)-
3a thus form a sort of centrosymmetric dimer in the solid
state, with each occupying the molecular cavity of the
other. The cavity-included arene rings are involved in
numerous edge-to-face interactions with the arene rings
that form the walls of the cavity. Lastly, and as is
previously observed in some CTV inclusion compounds
(42), one of the methoxy substituents (indicated by an
arrow in Figure 2) is distorted from the plane of the arene
ring, by 588 in this instance.
1
very complicated H NMR spectra due to their stereo-
chemistry (vide supra). Removal of the [(h5-C5H5)FeII]þ
moieties from (^)-2b,c[PF6]3 reduces the number of
stereocentres from four to only one, the (^)-CTG moiety,
such that (^)-3b,c are easily characterised as a racemic
1
mixture. Compounds (^)-3a–d,f gave expected H and
13C NMR spectra and satisfactory analysis for C and H
elemental composition. For each compound, the upper rim
arene substituents apparently rotate rapidly on the 1H
NMR timescale and arene ring rotation could not be
‘frozen out’, even at low temperatures (2708C) in CDCl3.
This is perhaps not surprising, given that typical barriers to
rotation about the OZC bonds of diphenyl ethers are very
low – of the order of about ,20 kJ mol21 (DH ‡ < 2–
10 kJ mol21) – even at 150 K in a solid matrix (41). It was
initially hoped that, analogous to other CTBs, the new
deep cavity-functionalised CTGs (^)-3a–d,f might be
exploited for the selective binding of fullerenes (6). All
attempts at such applications, however, proved unsuccess-
ful, likely in consequence to the cavity-blocking effect of
the conformationally mobile upper rim aryl rings.
In order to investigate the conformational preferences
of cavitands (^)-3a–d,f – at least in the solid state –
several attempts were made to grow single crystals of these
compounds, but, with the exception of (^)-3a, sufficiently
large single crystals were not obtained. The X-ray crystal
structure of (^)-3a was determined and a thermal
ellipsoid plot of the molecule is depicted in Figure 2.

874
K.T. Holman et al.
Table 1. Summary data for X-ray single crystal structure determinations.
Compound
(^)-2a[BF4][PF6]2·(THF)·(CH3COCH3)a
(^)-2d[BF4]3·(CH3COCH3)
(^)-2e[BF4]3·(CH3COCH3)
Formula
C64H65O8BF16P2Fe3
1506.49
Monoclinic
P21/n
C60H54O7B3F12Cl3Fe3
1421.40
Monoclinic
P21/n
C63H63O7B3F12Fe3
1360.14
Monoclinic
P21/n
Formula weight (g mol21
Crystal system
Space group
)
)
)
Temperature (K)
˚
173
173
173
a (A)
˚
9.5929(5)
21.562(1)
30.579(2)
90
90.769(1)
90
6324.4(6)
4
1.58
9.3918(6)
20.815(1)
31.311(2)
90
90.521(1)
90
6120.9(7)
4
1.54
9.5292(2)
20.541(1)
31.532(2)
90
90.670(1)
90
6171.5(5)
4
1.46
b (A)
˚
c (A)
a (8)
b (8)
g (8)
3
˚
V (A )
Z
rcalc (g cm23
)
m[Mo Ka] (cm21
)
8.30
9.19
7.84
Reflns coll/unique/obs
Parameters/restraints
2umax (8)
20,157/8729/5547
921/474
46
24,732/8431/6093
829/456
46
19,815/8489/5646
839/6
46
R (int)
R1, wR2 [I . 2s(I)]
R1, wR2 (all)
0.0432
0.0425
0.0554
0.091, 0.247
0.126, 0.275
0.82
0.073, 0.203
0.102, 0.225
1.44 (near BF24 )
0.080, 0.187
0.125, 0.211
1.12 (near BF24 )
Residual (e2
Compound
Formula
)
23
˚
A
(^)-2f[BF4]2[PF6]
2g[PF6]
2h[PF6]·DMF
C75H63O9B2F14PFe3
1594.43
Monoclinic
P21/c
C19H19O3F6PFe
496.17
C30H34NO7F6PFe
721.41
Formula weight (g mol21
Crystal system
Space group
Triclinic
P-1
173
Monoclinic
P21/c
173
8.5426(5)
18.396(1)
19.572(1)
90
98.438(1)
90
3042.4(3)
4
1.57
Temperature (K)
˚
173
a (A)
˚
9.2715(5)
20.649(1)
36.446(2)
90
93.424(1)
90
6064.9(6)
4
1.52
8.7171(7)
11.0856(9)
11.5144(9)
107.773(1)
93.041(1)
109.163(1)
986.2(1)
2
b (A)
˚
c (A)
a (8)
b (8)
g (8)
3
˚
V (A )
Z
rcalc (g cm23
)
1.67
9.20
41,20/3370/2901
320/12
50
m[Mo Ka] (cm21
)
7.36
6.22
Reflns coll/unique/obs
Parameters/restraints
2umax (8)
22,194/9551/5803
932/316
46
15,169/5310/3787
467/0
50
R (int)
R1, wR2 [I . 2s(I)]
R1, wR2 (all)
0.0539
0.0115
0.0386
0.091, 0.228
0.147, 0.260
0.53
0.054, 0.154
0.060, 0.159
0.56
0.077, 0.223
0.099, 0.239
0.61
Residual (e2
Compound
Formula
)
23
˚
A
(^)-3a
(^)-anti-5·(CHCl3)4
(^)-anti-5·(C6H5NO2)5
C42H36O6
636.75
Monoclinic
P21/n
C70H58O12Cl12
1516.66
Triclinic
P-1
C96H79O22N5
827.35
Monoclinic
C2/c
Formula weight (g mol21
Crystal system
Space group
Temperature (K)
˚
173
173
173
a (A)
˚
13.142(3)
12.727(3)
19.553(5)
90
94.280(5)
90
12.3672(8)
13.8257(9)
20.075(1)
83.807(1)
87.628(1)
84.960(1)
3397.6(4)
2
31.180(2)
13.930(1)
22.680(3)
90
126.116(1)
90
b (A)
˚
c (A)
a (8)
b (8)
g (8)
V (A )
Z
3
˚
3261(1)
4
7961.5(2)
4

Supramolecular Chemistry
875
Table 1 – continued.
Compound
(^)-3a
(^)-anti-5·(CHCl3)4
(^)-anti-5·(C6H5NO2)5
rcalc (g cm23
m[Mo Ka] (cm21
)
1.30
0.86
1.48
5.51
1.38
0.97
)
Reflns coll/unique/obs
Parameters/restraints
2umax (8)
17,127/5732/2720
436/0
50
10,997/8896/6040
900/102
50
24,810/7014/5479
577/30
56
R (int)
R1, wR2 [I . 2s(I)]
R1, wR2 (all)
0.0918
0.0319
0.0392
0.046, 0.090
0.107, 0.105
0.19
0.099, 0.253
0.137, 0.286
1.33 (near Cl)
0.044, 0.107
0.062, 0.118
0.26
23
Residual (e2
Compound
Formula
A
)
˚
a
(^)-anti-5·(C6H5N)52/3
(^)-anti-5·(C6H5Cl)2.5
(^)-anti-5·(CHCl3)2·(TCE)2
C961/3H821/3O12N52/3
1513.41
Hexagonal
R-3c
C81H66.5O12Cl2.5
1334.55
Triclinic
P-1
C36H30O6Cl7
806.80
Monoclinic
C2/c
173
17.940(1)
18.774(1)
22.528(1)
90
102.580(1)
90
Formula weight (g mol21
Crystal system
Space group
)
Temperature (K)
˚
173
173
a (A)
˚
16.4495(6)
16.4495(6)
49.844(3)
90
90
120
11680.2(9)
6
1.29
12.4658(7)
13.1646(8)
22.426(1)
76.684(1)
79.864(1)
63.664(1)
3198.1(3)
2
b (A)
˚
c (A)
a (8)
b (8)
g (8)
3
˚
V (A )
7405.5(7)
8
1.45
Z
rcalc (g cm23
)
1.39
1.91
m[Mo Ka] (cm21
)
0.85
5.81
Reflns coll/unique/obs
Parameters/restraints
2umax (8)
12,208/1810/1595
160/0
46
10,644/8546/6081
895/0
50
14,649/6480/4512
442/0
50
R (int)
R1, wR2 [I . 2s(I)]
R1, wR2 (all)
0.0283
0.0217
0.0368
0.058, 0.151
0.064, 0.157
0.30
0.104, 0.295
0.134, 0.322
0.75
0.054, 0.133
0.978, 0.145
0.66
23
Residual (e2
A
)
˚
a This structure was treated with SQUEEZE to model highly disordered solvents.
That certain substituted benzyl alcohols can be directly
trimerised to form CTBs led us to synthesise 2g[PF6] and
2h[PF6] by the reaction of excess vanillyl alcohol with
either 1a[PF6] or 1d[PF6], respectively (Figure 3). 2g[PF6]
cyclotrimerisation, and it was conceivable that 2g[PF6] –
a 3-methoxy-4-[(h5-C5H5)FeII(h6-C6H5O-)]þ-substituted
benzyl alcohol – could be directly cyclotrimerised to
(^)-2a[PF6]3. Similarly, it was conceivable that 2h[PF6]
might provide a direct route to metalated cryptophanes syn-
4[PF6]3 and/or (^)-anti-4[PF6]3 in an approach analogous
to Collet’s so-called ‘two-step’ method of cryptophane
synthesis (44). Alternatively, the photodemetalated pro-
ducts of 2g[PF6] or 2h[PF6] – i.e. the corresponding 4-
aryloxy-3-methoxy-substituted benzyl alcohols – were
also plausible precursors for direct cyclotrimerisation to
(^)-3a or cryptophanes syn-5 and/or (^)-anti-5, respect-
ively. Unfortunately, all attempts at the acid-catalysed (e.g.
65% HClO4 or formic acid) cyclisation of 2g[PF6] or
2h[PF6] failed and only starting materials were recovered.
The apparent lack of a reaction might be attributed to a
difficulty in forming the required benzyl cation intermedi-
ate in consequence to the fact that a positive charge is
already present on the molecule in the form of the [(h5-
C5H5)FeII(h6-arene-)]þ moiety. Unfortunately, also,
attempts to cleanly displace the [(h5-C5H5)FeII]þ moieties
1
and 2h[PF6] were characterised by H, 13C and 19F NMR
spectroscopyand bysinglecrystalstructuredeterminations.
ThecationsfromthesestructuresaredepictedinFigure3. In
practice, the cyclotrimerisation of substituted benzyl
alcohols, or the reaction of arenes with a formaldehyde
equivalent, to yield CTBs is governed by the identity of the
arene ring substituents. In general, electron-donating
substituents at the 3,4-positions are necessary for
cyclotrimerisations of benzyl alcohols to proceed in
reasonable yields (28). Typically, the role of the substituent
at the 3-position (commonly ZOCH3) is to activate the
para-position (ortho to the benzylic group) for electrophilic
attack and the role of the substituent at the 4-position (e.g.
OCH2COOH, OCH2CHvCH2, Br, SCH3, NHCOMe) is
largely to inhibit side reactions at this site (43). Substituted
vanillyl alcohols (4-substituted-3-methoxy-benzyl alco-
hols) are good candidates for direct, acid-catalysed

876
K.T. Holman et al.
Figure 3. Synthesis of potential CTB and cryptophane precursors 2g[PF6] and 2h[PF6]: (i) K2CO3, DMF, excess vanillyl alcohol.
Thermal ellipsoid plots of the cations from the X-ray crystal structures of (a) 2g[PF6] and (b) 2h[PF6]·DMF. The [(h5-C5H5)FeII]þ moiety
of 2hþ is disordered about the two faces of the coordinated arene. Only the major occupancy portion (80%) is shown.
from 2g[PF6] or 2h[PF6] by photochemical means
(sunlight, CH3CN) were equally unsuccessful.
only one doublet corresponding to the [PF6]2 counterions
and thus, contrary to our expectations, the putative syn-
4[PF6]3 and/or (^)-anti-4[PF6]3 cryptophanes seemingly do
not incarcerate [PF6]2 anions upon shell closure. Presumed,
nonetheless, to contain the desired syn-4[PF6]3 and (^)-
anti-4[PF6]3, and lacking a suitable means for separation of
these cryptophanes from the poly/oligomeric material, the
product mixture was subjected to sunlight photolysis in
degassed CH3CN in order to remove the [(h5-C5H5)FeII]þ
moieties. Work-up by column chromatography yielded
mixtures of the two cryptophane diastereomers syn-5 and
(^)-anti-5, which were further separated by fractional
crystallisation, giving unoptimised yields of 3 and 7%,
respectively – yields that are somewhat typical of other
cryptophane syntheses. By inference, the successful
isolation of syn-5 and (^)-anti-5 unequivocally establishes
the existence of syn-4[PF6]3 and (^)-anti-4[PF6]3, though
we have not as yet been successful in isolating these
cryptophanes.
Of the now hundreds of known cryptophanes, most are
synthesised by one of two methods (22, 23): (i) the so-called
‘template’ method whereby three appropriately substituted
benzyl alcohol moieties are first appended to the upper rim
of a pre-formed CTB cavitand and are then cyclotrimerised,
the pre-formed CTB serving as a template for cyclisation of
the second; or (ii) the ‘two-step’ method whereby two
appropriately substituted benzyl alcohol moieties (usually
vanillyl alcohols) are first covalently linked via some bridge
and are then simultaneously cyclotrimerised. A third, less
conventional approach involves the ‘capping’ of a pre-
formed CTB cavitand with another pre-formed CTB. Cram
et al. (45) first used the capping approach successfully in the
copper-catalysed coupling of two terminal acetylene-
functionalised CTBs to give the corresponding crypto-
phanes. More recently, the smallest cryptophane, so-called
cryptophane-1.1.1, was synthesised by a capping method
(46). Xu and Warmuth (24) have also recently exploited
dynamic covalent chemistry to adjoin two pre-formed
CTBs, resulting in amine-bridged cryptophanes. Shinkai
et al. (25) have used a related capping approach, taking
advantage of the transition metal coordination chemistry of
CTB-based ligands.
1
Syn-5 and (^)-anti-5 were characterised by H and
13C NMR spectroscopy, elemental analysis ((^)-anti-5
only) and single crystal X-ray diffraction (vide infra).
Though readily discernable in most solvents, the 1H NMR
spectra of syn-5 and (^)-anti-5 (Figure 4) are nearly
identical, reflecting the identical symmetry-unique por-
tions (i.e. fundamental regions) of the two molecules and
the equivalent order (6) of their respective C3h and D3
point group symmetries. Notably, protons of the meso
form, syn-5, resonate downfield from (^)-anti-5 in most
solvents. Also notable is the observation that the protons of
the 1,4-dioxylaryl bridge of both diastereomers appear as a
singlet in the 1H NMR spectrum, indicating, analogous to
the aryl-functionalised CTGs (^)-3, fast rotation of the
arene bridges on the 1H NMR timescale, even at
temperatures as low as 2658C. Unambiguous assignment
of the diastereomers was easily accomplished by crystal
Our inability to cyclise 2g[PF6] or 2h[PF6] necessitated
a capping approach for the synthesis of the corresponding
cryptophanes syn-4[PF6]3, (^)-anti-4[PF6]3 and ultimately
syn-5 and (^)-anti-5. The 1,4-arylether-bridged crypto-
phanes were synthesised by the capping of the 4-chloro
terminal (^)-3d[PF6]3 with an additional equivalent of (^)-
CTG (Figure 1). The reaction was accomplished at low
concentrations (3.0mM) in basic DMF, yielding after work-
up, a yellow solid that, by 1H NMR spectroscopy, appeared
to consist mostly of poly/oligomeric material. Somewhat
surprisingly, the 19F NMR spectrum of this material showed

Supramolecular Chemistry
877
Figure 4. 1H NMR spectrum (400 MHz, CDCl3) of an
approximate 5:1 mixture of syn-5 and (^)-anti-5.
Figure 5. The 1H NMR spectrum of (^)-2d[PF6]3 in (a)
NO2Me-d3, (b) NO2Me-d3 with excess added [NBu4]Br and (c)
acetone-d6, highlighting the peak splitting that occurs upon anion
binding.
structure determinations. Vapour diffusion of n-hexane
into a solution of (^)-anti-5 in CHCl3 results in well-
formed single crystals of composition (^)-anti-5·4CHCl3.
X-ray analysis of (^)-anti-5·4CHCl3 and several other
solvates (Table 1, vide infra) confirmed the anti
stereochemistry of the host. Moreover, a low quality
crystal structure of syn-5, crystallising in the space group
P63/m, established the stereochemistry and C3h point
group symmetry of the syn diastereomer.
[(h5-C5H5)FeII(h6-arene)]þ moieties is slowed, presum-
ably as a consequence of halide binding within the
extended CTB cavity. This behaviour is likely attributed to
the formation of a ‘penetrated ion pair’ (47) where the
halide ion is bound, reversibly, within the (^)-2d3þ host
cavity. Control experiments show that the addition of
excess [NBu4][PF6] has little to no effect on the 1H NMR
spectrum, suggesting that [PF6]2 ions are at best weakly
bound in NO2Me-d3.
Anion binding by cavitands (6)-2a–f31
The anion binding properties of cavitands (^)-2a,d,e3þ
,
1
as revealed by H NMR spectroscopic properties of (^)-
2d3þ and the X-ray single crystal structure of (^)-
2d[PF6]3·1.5Et2O·H2O, were reported in an earlier
communication (9). These studies will not be repeated
here but a summary of the findings is in order.
Inacetone-d6 (1 ¼ 20), theupper rim[(h5-C5H5)FeII(h6-
arene)]þ moieties of (^)-2d[PF6]3 experience inhibited
rotation in the absence of added halide, probably indicating
the intracavity binding of a [PF6]2 anion in this less polar
solvent (Figure 5(c)). The actual host–guest binding
geometry is likely very similar to that observed in the
crystal structure of this salt, which was earlier reported (9).
The 19F NMR spectrum, however, exhibits only one doublet
for the [PF6]2 anion and therefore bound-to-free anion
exchange occurs in the fast exchange domain with respect to
the 19F NMR timescale, whereas [(h5-C5H5)FeII(h6-
The solution anion binding properties of cavitands
1
(^)-2a,d,e3þ become apparent upon examining the H
NMR spectra of these hosts. The spectrum of (^)-
2d[PF6]3 is particularly diagnostic in that the para-
substitution pattern of metal-coordinated arene rings
provides a good and simple spectroscopic handle to
monitor anion complexation in solution. In NO2Me-d3
(1 ¼ 38), the para-substituted arenes of the upper rim
[(h5-C5H5)FeII(h6-arene)]þ moieties appear as a typical
AB pattern, indicative of fast rotation of these groups on
the NMR timescale such that the protons on the ‘inside’ of
the cavitand (H_i) are in fast exchange with those on the
‘outside’ of the cavitand (H_o) (Figure 5). Addition of an
excess of halide ions to the solution, in the form of
[NBu4]Br, results in the concomitant splitting of the AB
pattern into two separate patterns as the rotation of the
1
arene)]þ rotation is slow on the H NMR timescale. [(h5-
C5H5)FeII(h6-arene)]þ rotation is therefore not completely
coincident with intracavity anion exchange as the 1H and 19
F
timescales are similar. It is therefore likely that some degree
of external [PF6]2 complexation also occurs in this solvent,
contributing to slow rotation of the [(h5-C5H5)FeII(h6-
arene)]þ moieties. This sort of partial ion paring is consistent
with a recent study concerning ion pairs of [(h5-C5Me5)-
RuII(h6-arene)][PF6] salts in acetone-d6 as examined by

878
K.T. Holman et al.
19F–1H HOESY and PGSE diffusion NMR experiments
(48). Halide (X ¼ Cl2, Br2, I2) binding by (^)-2d3þ in
acetone-d6 was previously studied by 1H NMR titrations and
the results suggested that the host has the greatest affinity for
iodide (9).
presumably because these compounds are present as a
mixture of four diastereomers (vide supra).
The crystal structures of (^)-2d[BF4]3·CH3COCH3
and (^)-2e[BF4]3·CH3COCH3 are isostructural, exempli-
fying the chloro-methyl exchange rule (49). Their unit cells
and crystal packing are also very similar to those of the
mixed salts (^)-2a[BF4][PF6]2·THF·CH3COCH3 and (^)-
2e[BF4]2[PF6]3, despite the difference in counterion
compositions, upper rim para-substituents, and included
solvents. The most notable feature in each of these four
crystal structures is the presence of a cavity-bound [BF4]2
anion. The anions rest central to three ‘upward’ projected
[(h5-C5H5)FeII(h6-arene)]þ moieties of the cavitand
(Figure 6) and are deeply buried, with boron positions that
lie beneaththe mean planes of theironatoms by, onaverage,
The X-ray crystal structures of salts of (^)-23þ further
demonstrate their ability to complex anions deep within the
cavitand bowls. The structure of (^)-2d[PF6]3·1.5Et2-
O·H2O was reported earlier (9) and demonstrates a
centrally embedded [PF6]2 ion, with the host adopting a
pseudo-C3 symmetric conformation similar to that
observed for the aforementioned (^)-3a with all [(h5-
C5H5)FeII(h6-arene)]þ moieties directed ‘upward’ from
the CTG bowl. In general, however, cations (^)-23þ are
difficult to crystallise as their [PF6]2 or halide salts.
Addition of excess [NBu4][BF4] to solutions of (^)-2a–
f[PF6]3 in acetone, however, results in the precipitation
of tetrafluoroborate salts of these hosts. Thus, [PF6]2
anions can be exchanged for [BF4]2 using this procedure.
Single crystals of the [BF4]2 salts of (^)-2d,e can be
obtained when exactly three equivalents of [NBu4][BF4]
are used. Alternatively, mixed [PF6]x[BF4]32x salts of (^)-
23þ can be prepared using H[BF4] and/or Na[BF4] in
the synthetic work-up and single crystals can be grown by
the diffusion of ethers (e.g. THF, Et2O) into acetone
solutions of the mixed salts. By these methods, single
crystals of (^)-2a[BF4][PF6]2·THF·CH3COCH3, (^)-
2d[BF4]3·CH3COCH3, (^)-2e[BF4]3·CH3COCH3 and
(^)-2e[BF4]2[PF6]3 were obtained and their X-ray
structures were determined. Summary crystallographic
details for these salts are provided in Table 1. Notably,
single crystals of salts of (^)-2b,c3þ could not be obtained,
˚
0.4(1) A. One comparative measure of the space provided
for the anions by the [(h5-C5H5)FeII(h6-arene)]þ moieties
of cavitands (^)-2a,d–f3þ is the Fe· · ·Fe distance. In (^)-
2d[PF6]3·1.5Et2O·H2O, the [(h5-C5H5)FeII(h6-arene)]þ
moieties adopt positions such that the iron atoms essentially
form an isosceles triangle about the guest, with Fe· · ·Fe
˚
distances of 8.56, 8.65 and 9.23 A, defining an area of
2
33.5 A . These values highlight the slight deviation of the
˚
host from ideal C3 symmetry. In the four structures that
exhibit cavity-bound [BF4]2 anions, the Fe· · ·Fe distances
are, on average, slightly shorter, defining triangular areas of
2
28.2, 30.2, 29.4 and 29.5 A for (^)-2a,d–f3þ, respect-
˚
ively. The smaller cavity areas, as defined by the iron atoms,
reflect the ability of the host to accommodate the smaller
anionic guest via a slight contraction.
Coupled with the fact that the [BF4]2 salts precipitate
from acetone, the observation that [BF4]2 anions are found
within the host cavities of the two mixed [BF4]2/[PF6]
Figure 6. Single crystal X-ray structures of cations (^)-23þ and their cavity-bound anions. (a) The [2d , (PF6)]2þ species (9),
illustrating the planes used to describe the cation conformations; the planes are defined by the arene rings of substituents A, B and C (red),
and the oxygen atoms at the upper rim of the cavitand (tan). (b)–(e) ‘Top’ and ‘side’ views of the [2a,d,e,f , (BF4)]2þ species,
respectively. Only the major occupancy portions of disordered species are shown.

Supramolecular Chemistry
879
salts seems to suggest a preference of the cavitands (^)-
2a–f2þ for this anion. Addition of [NBu4][BF4] to
acetone-d6 solutions of (^)-2d[PF6]3 results in only very
slight changes in the 1H NMR spectrum before the onset of
precipitation. The lack of significant changes in the
spectrum ought not necessarily be interpreted as a lack of
binding, however, since the environment of the protons
inside the cavity would be expected to be quite similar in
substituents (red planes in Figure 6(a)) and the mean
planes of the cavitand upper rims, as defined by the oxygen
atoms of the diphenyl ether moieties (tan plane in
Figure 6(a), Table 2). Of the five available crystal
structures of (^)-2a,d–f3þ [BF4]2/[PF6]2 salts, including
the previously reported (^)-2d[PF6]3·1.5Et2O·H2O, 14 of
the 15 [(h5-C5H5)FeII(h6-arene)]þ substituents (two
disordered) are observed in the ‘anticlockwise’ confor-
mation. The arenes of the anticlockwise substituents are all
nearly normal to the cavitand bowl, with mean f angles of
86(4)8. Each is turned such that the aryl and cyclopenta-
dienyl CZH bonds are directed more-or-less towards the
centre of the cavity, allowing the cavity-bound anion to
nestle between the planes of the arenes and exposing it to
the iron atom. There are four instances (including
disordered positions) of substituents in the ‘clockwise’
rotameric position (see Figure 6(e)). These substituents
exhibit mean f angles of 56(9)8 reflecting the sterics
between the [(h5-C5H5)FeII]þ moieties and the ortho-
positioned methoxy substituent in this conformation. In all
observed cavitand conformations, the dihedral angles
between the phenyl rings of the diphenyl ether moieties are
within the normal range for such compounds (50).
comparing
a
[2d , (PF26 )]2þ complex with
a
[2d , (BF24 )]2þ complex, as is borne out by the crystal-
lography. It is possible that these hosts do indeed display a
greater affinity for the [BF24 ] ion, which would
presumably be attributed to the greater charge density of
the smaller anion. Notably, and somewhat surprisingly, the
cavity-bound [BF4]2 anions of these structures all exhibit
considerably greater orientational disorder than the
corresponding exterior positioned anions in the solid
state, even at 21008C. Exactly the opposite effect was
observed for the [PF6]2 salt of (^)-2d3þ, which exhibited
a highly ordered, cavity-bound [PF6]2 anion. Speculation
allows one to suppose that a difference in solution binding
anion affinity might also be in part attributable to the
entropic differences between binding a smaller, orienta-
tionally disordered [BF4]2 anion relative to the larger,
more rigidly held [PF6]2 anion.
The five crystal structures of (^)-2a,d–f3þ as
[BF4]2/[PF6]2 salts allows one to generate statistics with
regard to the cation· · ·anion interactions. These inter-
actions are most easily summarised by examining the non-
bonded distances between the iron atoms of the [(h5-
C5H5)FeII(h6-arene)]þ substituents, which bear the great-
est positive charge, and the central atom of the spheroidal
anions. Table 2 lists the closest three to four Fe· · ·B or
Fe· · ·P distances for each of the [BF4]2 or [PF6]2 anions
It is convenient to designate the three [(h5-C5H5)-
FeII(h6-arene)]þ substituents of hosts (^)-2a,d–f3þ with
descriptors A–C as shown in Figure 6. In (^)-2d[BF4]3·
CH3COCH3 and (^)-2e[BF4]3·CH3COCH3, the host
cation adopts a pseudo-C3 conformation such that all
three [(h5-C5H5)FeII]þ moieties of the A–C substituents
project away from the ortho-positioned methoxy sub-
stituent, defining an anticlockwise turn with respect to the
cavitand enantiomer depicted in Figure 6. In (^)-
2e[BF4]2[PF6]3, however, one of the three substituents, A
(designated with an arrow in Figure 6(e)), is projected
‘clockwise’ with respect to the cavitand, representing a
rotamer of the ‘anticlockwise’ position. The two
substituent conformations differ by essentially a 1808
rotation about the [(h5-C5H5)FeII(h6-arene)]þZoxygen
bond and a smaller rotation about the CTBZoxygen bond.
The two rotameric conformations reflect the confor-
mational dynamics of the [(h5-C5H5)FeII(h6-arene)]þ
substituents and are not unique to the structure of
2e[BF4]2[PF6]3. In (^)-2a[BF4][PF6]2·THF·CH3COCH3,
two of the substituents (A and C) are disordered such that
both rotamers of [(h5-C5H5)FeII]þ moieties are observed
within the crystal (approximately 1:1 and 7:3 antic-
lockwise:clockwise occupancies, respectively). The dis-
ordered situation is not depicted in Figure 6(b); only the
anticlockwise positions are shown, which for substituent C
is the major occupancy position.
found in the five crystal structures of (^)-2a,d–f3þ
.
Where disorder of iron or boron atom positions is
observed, an average value is listed. The most notable
features are as follows. First, the cavity-bound anions are
all found to be nestled between the arene rings of the three
surrounding [(h5-C5H5)FeII(h6-arene)]þ sandwich com-
pounds, being simultaneously exposed to the CZH groups
of the arenes and the central iron atoms bearing the bulk of
the positive charge. For all structures, the anions within the
cavities exhibit the closest set of three Fe· · ·B or Fe· · ·P
contacts. For instance, the three closest iron atom distances
2
˚
for the cavity-included [BF4] anions average 4.8(1) A for
the four examples, whereas the three closest iron atom
distances of the exterior-positioned [BF4]2 anions (seven
˚
examples) measure 5.6(3) A. Similarly, the cavity-bound
[PF6]2 anion of (^)-2d[PF6]3·1.5Et2O·H2O, exhibits the
˚
closest set of three Fe· · ·P contacts (5.13(4) A) as
compared to the exterior-located [PF6]2 anions in these
˚
structures (5.6(3) A, five examples). Moreover, in the solid
state, only the cavity-bound anions show close contacts to
more than one iron atom of the same host molecule.
Collectively, these observations strongly support the
contention that the central cavity of hosts (^)-2a–f3þ
The conformations of the hosts are more quantitatively
summarised by measures of the dihedral angles (f)
between the mean planes of the arene rings of the A–C

880
K.T. Holman et al.
is essentially the only significant supramolecular binding
site for anions. That is, in the observed host conformations,
there are no sites, other than the cavitand cavity, where an
anion forms a contact ion pair with more than one [(h5-
C5H5)FeII(h6-arene)]þ moiety of the host.
Table 2. Cavitand conformations and closest anion–cation
contacts in crystal structures of (^)23þ saltsa.
(^)-2a[BF4][PF6]2·(THF)·(CH3COCH3)
[(C5H5)Fe(arene)] þconformationsb
Substituent
A (50%), A0 (50%)
B
f (8)
84, 59
90
Host–guest chemistry of cryptophane (6)-anti-5
C (70%), C0 (30%)
Cavity-bound [BF4] 2
B1· · · FeA, FeB, FeC
Exterior [PF6] 2
P1· · ·FeA, FeA00, FeA000, FeB
Exterior [PF6] 2
76, 46
Distance (A)
4.80, 4.76, 4.50
˚
Cryptophanes commonly form kinetically stable com-
plexes (millisecond or longer lifetimes) with both small
neutral molecules and small organic cations in lipophilic
solvents (22). Thus, a variety of spectroscopic studies were
performed to determine whether the new cryptophanes
syn-5 or (^)-anti-5 exhibit similar binding properties. In
order to reduce competition from the solvent, the studies
were performed in deuterated 1,1,2,2-tetrachloroethane
(TCE), which was deemed to be too large to be effectively
complexed by the host cavity. The crystal structure of (^)-
anti-5, obtained from this solvent, supports this contention
as the host does not encapsulate the TCE (vide infra).
Addition of potential small molecule guests such as
CH2Cl2, CHCl3, C6H6, C6H5CN, NO2C6H6 results in no
changes to the room temperature spectrum of (^)-anti-5,
suggesting that these guests are either not bound, or that
they exchange between the solvent and host cavity rapidly
on the NMR timescale. Notably, however, each of these
molecules (except CH2Cl2) forms a solid-state inclusion
compound with (^)-anti-5 where the guests are contained
within the host cavity (vide infra).
0
5.25, 6.15, 5.85, 5.66
6.48, 6.15, 5.41, 5.81
P2· · · FeB00, FeB000, FeC, FeC
000
(^)-2d[PF6]3·(Et2O)1.5·H2Oc
[(C5H5)Fe(arene)] þconformations b
Substituent
f (8)
87, 89, 88
Distance (A)
5.14, 5.08, 5.16
5.73, 5.88, 5.09
5.12, 5.36, 5.95
A, B, C
Cavity-bound [PF6] 2
P1· · · FeA, FeB FeC
Exterior [PF6] 2
˚
0
P2· · · FeA, FeB00, FeC
Exterior [PF6] 2
0
P3· · ·FeA00, FeB , FeC
(^)-2d[BF4]3·(CH3COCH3)
[(C5H5)Fe(arene)] þconformationsb
Substituent
f (8)
82, 87, 89
Distance (A)
4.85, 4.99, 4.88
A, B, C
Cavity-bound [BF4] 2
B1· · ·FeA, FeB, FeC
Exterior [BF4] 2
˚
Typically, in lipophilic solvents, cryptophane com-
plexes of small cations are more thermodynamically stable
than complexes of neutral guests. Other arene-bridged
cryptophanes, e.g. those possessing para- or ortho-xylyl
bridges, are known to form kinetically stable complexes
with small alkyl ammonium cations at low temperatures
(51). Specifically, the anti diastereomer of the six-carbon,
p-xylyl-bridged cryptophane (Scheme 1, R1 ¼ R2 ¼
OCH3), forms a more stable [DG8(210 K) ¼ 22.9 -
kcal/mol] complex with NEt3Meþ than with smaller or
larger cations. The related four-carbon, o-xylyl-bridged
cryptophane forms a mildly stable complex with
NMeþ4 (DG8 (210 K) ¼ 20.98 kcal/mol). Under identical
conditions (210 K, 9:1 CD2Cl2:CD3OD), however, (^)-
anti-5 shows no evidence for complex formation with either
[NMe4]þ or [NHMe3]þ as their picrate salts.
0
B2· · ·FeB000, FeC, FeC
5.93, 5.58, 5.64
Exterior [B0F4] 2
B3· · · FeA , FeA00, FeA000, FeB
5.39, 5.37, 5.47, 4.97
(^)-2e[BF4]3·(CH3COCH3)
[(C5H5)Fe(arene)] þconformationsb
Substituent
f (8)
88, 89, 84
Distance (A)
4.88, 4.83, 4.89
A, B, C
Cavity-bound [BF4] 2
B1· · · FeA, FeB, FeC
Exterior [BF4] 2
˚
0
00
B2· · · FeB , FeB000, FeC, FeC
6.08, 6.45, 5.70, 5.63
5.50, 5.37, 5.50, 5.08
Exterior [BF4] 2
0
B3· · ·FeA, FeA , FeA00, FeB
(^)-2f[BF4]2[PF6]
[(C5H5)Fe(arene)] þconformationsb
Substituent
f (8)
64.0, 89, 80
Distance (A)
A, B, C
Snapshots of molecular gating
Cavity-bound [BF4] 2
B1· · ·FeA, FeB, FeC
Exterior [BF4] 2
˚
The inability to observe host–guest complexes of (^)-
anti-5 in solution does not imply that the host cavity of this
new cryptophane is not capable of encapsulating guests. In
fact, a number of crystal structures of (^)-anti-5, obtained
from various solvents of crystallisation, have been
determined which shed light on its host–guest chemistry.
X-ray crystal structure data for five of these solvates are
4.70, 4.95, 5.05
5.92, 5.69, 6.51, 5.59
5.82, 5.21, 5.74
0
B2· · · FeB00, FeB000, FeC, FeC
Exterior [PF6] 2
0
00
P1· · · FeA , FeB, FeC
a Each prime (0) mark indicates an interaction distance involving a
different cavitand.

Supramolecular Chemistry
881
given in Table 1. In addition to providing a definitive
assignment to 1H NMR spectra of diastereomers syn-5 and
(^)-anti-5, these structures provide general experimental
insights into the conformational properties and molecular
gating mechanisms of this and other cryptophane
container molecules.
be summarised by a number of simple geometric parameters
(Figure 7(f)). One is the ‘twist angle’, u, representing the
relative turn of the northern and southern CTB caps. Host–
guest complexes with different twist angles are reminiscent
of the (hemi)carcerplex ‘twistomers’ observed by Chapman
and Sherman (55). The twist angle is formally defined by the
average dihedral angles subtended by the oxygen atoms of
the threebridges with respect to the C3 (or pseudo-C3) axis of
the host. A zero twist angle therefore represents a
conformation wherein the arene bridges lie perfectly normal
to both CTB caps, maximising the length (or height) of the
cryptophane cavity, h. The approximate cavity height can be
conveniently measured by the distance between the centres
of the arene rings of the two CTB caps. (^)-Anti-5 , CHCl3
exhibits a minimal (58) twist angle and the height of the
The constrictive binding of guests by Cram’s
hemicarcerand container molecules – i.e. the large DG ‡
associated with the guest complexation process – has been
shown to be attributable to the energy costs associated with
the conformational fluctuations necessary to increase the
portal size sufficiently for possible guest entrance/escape
(52). Obviously, analogous, albeit energetically less
restricted, processes must occur in the gating of guest
molecules by the cryptophane class of container
molecules. For cryptophane hosts, two distinct types of
conformational processes can be envisaged which would
dramatically affect the size and shape of the portals
regulating access to the molecular cavity. Firstly, out-of-
plane distortions of the methoxy (or other) substituents
would provide a larger portal diameter and allow easier
exchange of the guest. In keeping with Houk’s description
of the gating mechanisms of hemicarcerands (52), such a
process could be described as a ‘french door’ gate whereby
the methoxy groups open outward to allow access to the
cavity. The barrier to rotation of the methoxy group in
anisole has been the subject of debate, but the coplanar
conformation is undoubtedly the minimum energy
configuration, implying that there exists a reasonable
barrier (a few kcal/mol) to the out-of-plane distortion of
each methoxy substituent (53). Similarly, other confor-
mational motions of the cryptophanes, such as contortions
of the orthocyclophane moieties, conformational adjust-
ments of the bridges or twisting of the two CTB caps about
the C3 axis, might be considered to be analogous to Houk’s
‘sliding door’ gating mechanism described for hemi-
carcerands. McCammon and co-workers (54) have studied
the structural fluctuations of a water-soluble hexa acid
derivative (anti; R1 ¼ R2 ¼ OCH2COOH; bridge ¼
O(CH2)3O) of cryptophane-E by molecular dynamics
simulations. The results have been interpreted in terms of a
probability controlled, i.e. stochastic, gating process
whereby fluctuations of the molecular portals dramatically
influence the kinetics of guest binding.
˚
cavity (8.4A) is nearly maximised in order to accommodate
the encapsulated guest. Notably, unlike the CHCl3 complex
of cryptophane-E (Scheme 1, anti, R1 ¼ R2 ¼ OCH3,
bridge ¼ ZO(CH2)3OZ), the guest chloroform molecule
does not align its CZH bond with the C3 axis of the host,
presumablybecauseofthelongerandbulkierbridgesof(^)-
anti-5, which cause the long axis of the chloroform
molecules to align with the long axis of the host such that
two of its chlorine atoms are directed into the two CTB caps.
A closer examination of the host conformation in (^)-
anti-5 , CHCl3 reveals some other remarkable features. Of
the six methoxy moieties which help define the portals of the
host cavity, four are situated either coplanar with the arene
rings to which they are attached or are directed slightly
inwardsuchthatthe diameters oftwo ofthe threehostportals
are minimised. As highlighted in Figure 7(a), however, the
two methoxy substituents associated with one of the portals
aredirectedoutoftheplaneofthearene(by118and308)soas
to significantly enlarge the portal diameter. A useful
parameter for establishing the size of a portal is the distance
between the two methoxy carbon atoms that define it. These
˚
distances measure 4.7 and 5.2 A for the closed portals and
˚
7.8 A for the ‘open’ portal in this structure, indicated by the
dashed line. Inspection of this portal region reveals the
presenceofanotherchloroformmoleculeessentiallyperched
on the doorstep of the partially opened molecular gate. Thus,
it seems that one chloroform molecule is not large enough to
effectively fill the cavity volume and allow for efficient
packing in the crystal. The host, therefore, gets caught in a
‘french door’ type of gating conformation with the methoxy
groups of one portal stuck slightly open to allow another
chloroform to fill the excess volume. Interestingly, the
second chloroform molecule directs its CZH towards the
face of the bridging arene ring and is seemingly involved in a
‘Open gate’ structures
Crystallisation of (^)-anti-5 by diffusion of hexane into a
chloroform solution of the host yields a racemic tetrachloro-
form solvate (^)-anti-5·(CHCl3)4. The X-ray crystal
structure establishes the anti diastereomeric configuration
of the host, but, more importantly, shows that one of the
chloroform molecules of crystallisation is found encaged
within the cryptophane cavity. The (^)-anti-5 , CHCl3
complex is shown in Figure 7(a). The host conformation can
˚
CZH· · ·p interaction [C(H)·· ·(centroid) ¼ 3.50A]. This
type of interaction has been of significant interest in
supramolecular chemistry and has been calculated to be
energetically favourable by ,3 kcal/mol with
a
˚
C(H)· · ·(centroid) distance of 3.3 A (56).

882
K.T. Holman et al.
Figure 7. Spacefill and thermal ellipsoid representations of the solvent-encapsulated complexes (a) (^)-anti-5 , CHCl3·CHCl3, with a
3
˚
molecule of CHCl3 perched at the cavity portal, (b) (^)-anti-5 , NO2Ph, (c) (^)-anti-5 , C5H5N·3C5H5N (the guest-occupied 116 A
central cavity is depicted in orange) and (d) (^)-anti-5 , C6H5Cl, as determined from the crystal structures of the respective solvates. (e)
3
The empty form of (^)-anti-5 from the crystal structure of (^)-anti-5·(CHCl3)2·(TCE)2. The empty 52 A cavity is depicted in orange.
˚
Principal structural features are highlighted. (f) A summary of the conformational dynamics of (^)-anti-5.
A similar open ‘french door’ situation is observed
when the host is crystallised with a guest that is just
slightly too large for the host cavity. The crystal structure
of (^)-anti-5·(NO2C6H5)5 displays a host conformation
similar to that observed in the tetrachloroform solvate with
the same twist angle of approximately u ¼ 58 (Figure
7(b)). In this structure, however, the host adopts crystal-
lographically imposed C2 symmetry. One cavity-bound
molecule of nitrobenzene is situated centrally within the
host cavity with its plane approximately parallel to the
cryptophane’s pseudo-C3 axis. The extra bulk of the nitro
group precludes its complete encapsulation, however, as
this portion of the guest is directed out from the core of the
cavity, again through a partially open ‘french door’ type
gate. The two methoxy moieties that define the portal are
distorted out of the plane of the arene ring by 148 and the
which all of the methoxy groups lie in the plane of the
arene to which they are attached. Thus, (^)-anti-5 was
crystallised by ether-induced precipitation from benzene
and pyridine to form the isostructural pyridine and
benzene solvates. Full details of the pyridine solvate are
reported here. The crystals pack in a trigonal crystal
system (space group R-3c) and the high symmetry of the
structure exacerbates an extreme disorder of the included
solvents. The (^)-anti-5 molecule, however, is well
ordered and conforms to ideal, crystallographically
imposed D3 symmetry. Despite the small volume of
pyridine, the host adopts its fully extended conformation,
displaying a twist angle of only 278 (Figure 7(c)) and
giving rise to the most elongated possible cavity
˚
(h ¼ 8.6 A). Interestingly, in this extended host confor-
mation, the two orthocyclophane caps of (^)-anti-5 are
sufficiently separated to allow three pyridine (or benzene)
guests to intercalate between the methoxy groups of each
portal and partially penetrate the host cavity. Notably,
though, all of the methoxy moieties remain coplanar with
their respective arenes (^38) and the ‘french doors’ must
be considered closed. The portals must nonetheless be
considered to be ‘open’ to the pyridines, as they are
˚
distance between these two carbon atoms is 7.2 A. The
other four methoxy substituents are roughly coplanar with
the arene to which they are attached and the C· · ·C
˚
distances of these portals are both only 5.3 A.
From the crystal structure of (^)-anti-5·(NO2C6H5)5,
one might expect that a slightly smaller guest would allow
the host to adopt a ‘closed gate’ type conformation in

Supramolecular Chemistry
883
wedged in between the methoxy substituents and are
clearly partially within the cryptophane cavity. The C· · ·C
˚
distances of the portals are 7.25 A. This structure can
1:1 TCE:CHCl3. Hexachloroacetone and bromoform were
chosen because they were deemed to be too large to
possibly reside within the cryptophane cavity. Indeed, in
each of these solvates, the cryptophane crystallises with an
empty ‘cavity’, the cryptophane conformation being
essentially identical in structure. Only the highest quality
of these structure determinations, that of (^)-anti-
5·(CHCl3)2·(TCE)2, is reported here. It is a curiosity that
the 1:1 mixture of CHCl3 and TCE gives rise to crystals of
an ‘empty’ cryptophane, when the former of these guests
can clearly fit within the cavity of the host (Figure 7(a)).
The crystallographic observation of an empty cryptophane
cavity is unprecedented. This behaviour is in stark contrast
to the three-carbon-bridged cryptophane-E, which will
efficiently scavenge CHCl3 molecules from a solution of
TCE. This observation points to a low energetic penalty for
the cryptophane being empty vs. occupied, possibly
explaining why stable host–guest complexes of (^)-anti-5
are not observed in solution. We (3) and others (2b,c) have
found that other cryptophane container molecules tend to
conformationally ‘implode’ in response to the emptying of
their cavities. In short, the low energetic penalty for finding
(^)-anti-5 to be empty is likely a consequence of the
molecules’ ability to adopt a conformation that effectively
fills most of its own cavity. Indeed, in the crystal structure
of (^)-anti-5·(CHCl3)2·(TCE)2 (and the other solvates
mentioned above), the cryptophane adopts a fully twisted
conformation (u ¼ 378) in order to minimise the end-to-
therefore be regarded as a snapshot of a purely ‘sliding
door’ type of gating of guests to (^)-anti-5, where the
portal diameter has been increased, large enough to allow
passage of a guest, strictly through twisting of the northern
and southern caps.
The portal-situated pyridine positions of (^)-anti-
5·(C5H5N)52/3 are the only solvent locations that could be
identified with any degree of crystallographic/chemical
certainty. Interstitial sites in the crystal are occupied by what
appear to be extremely disordered pyridine molecules and
the cryptophane central cavity is occupied by extremely
disordered, ill-defined species. SQUEEZE (57) estimates an
internal cavity bound by the cryptophane and the three
3
˚
portal-situated pyridines of 116 A with an associated
electron count of 21 electrons, a value corresponding exactly
to either one half of a pyridine (or diethyl ether). Therefore,
eitheronly50% ofthe cryptophanecavitiesare occupiedbya
solvent molecule or there exists a different species within the
cavity, for example, two water molecules (20 electrons). In
any case, the ascribed material composition of (^)-anti-
5·(C5H5N)52/3 is not likely to be formally correct, but, in the
absence of more information, is sufficient to give a
reasonable approximation for the density of the material.
A ‘closed gate’ structure
Presumably, an optimally sized guest would allow the
cryptophane to encapsulate a guest so as to adopt a
minimum energy conformation where all of the methoxy
substituents lie in the plane of their respective arene rings
while the portals remain closed. Such a situation is realised
in many of the known crystal structures of cryptophanes,
and it was worthwhile to establish whether (^)-anti-5 is
capable of doing the same. Co-crystallisation of (^)-anti-
5 with a guest intermediate in size between NO2Ph and
benzene (i.e. chlorobenzene) leads to a structure exhibiting
a ‘closed gate’ host conformation, namely (^)-anti-
5·(C6H5Cl)2.5. Like the other structures of (^)-anti-5, the
guest is located within the host cavity, but the cryptophane
host is significantly more twisted (u ¼ 178) and the more
appropriately sized guest is found to be completely
encapsulated by the host (Figure 7(d)). The closed portals
are exemplified by the spacefill representation of this
complex, which makes it difficult to see the encapsulated
guest, and the close C· · ·C separations between the
˚
end length and volume of its internal cavity (h ¼ 7.7 A,
Figure 7(e)). The 1,4-diphenoxy bridges are turned inward
so as to most effectively fill the void. The symmetry-
unique dihedral angles of the bridging arene planes relative
to the upper rim of the cavitand measure f ¼ 558, 578 and
618. Each of the methoxy groups lies roughly coplanar to
its respective arene and the C· · ·C separations that define
˚
the portal size have been reduced to 3.98–4.08 A.
SQUEEZE was used to examine the structure of (^)-
anti-5·(CHCl3)2·(TCE)2 in order to determine the volume
of empty space within the (^)-anti-5 cavity, and to
establish that it is in fact devoid of any guest, including the
possibility of highly disordered, ill-defined species.
SQUEEZE finds discrete, solvent-accessible voids centred
3
˚
within the cryptophane cavities measuring 52 A and finds
no electron density (only two electrons) that can be
attributed to any putative species that might be inside. To
be sure, the cavities of (^)-anti-5 in crystals of (^)-anti-
5·(CHCl3)2·(TCE)2 are in fact empty. It is quite rare to
observe such large volumes of contiguous empty space in
crystalline organic solids owing to their propensity to close
pack, and include solvents of crystallisation where
necessary in order to fill interstitial voids (58).
˚
methoxy groups across the portal (4.2, 4.3 and 4.6 A).
An empty cryptophane container
Single crystals of solvated forms of (^)-anti-5 can also be
obtained by the slow diffusion of hexane into solutions of
the host in 1,1,2,2-TCE, bromoform, hexachloroacetone or

884
K.T. Holman et al.
at a heating rate of 108C/min. The melting range is given
Conclusions
as the full width at half maximum of peak corresponding
to the melting endotherm. Microanalyses were performed
by M-H-W Laboratories (Phoenix, AZ, USA).
In summary, aryl-extended CTGs (^)-3a–d,f were
synthesised from their respective [(h5-C5H5)FeII(h6-
arene)]þ-functionalised derivatives (^)-23þ, some of
which had been reported previously (9). New crystal
structures of [BF4]2 or mixed [BF4]2/[PF6]2 salts of (^)-
2a,d–f3þ shed light on the conformations and anion
binding properties of these deep cavity cavitands. In
general, these molecules adopt a conformation that
projects all of the [(h5-C5H5)FeII(h6-arene)]þ substituents
‘up’ from the upper rim of the cavitand, creating a deep
cavity that is occupied by a [BF4]2 anion. The host cavities
of 23þ appear to show some selectivity for the inclusion of
[BF4]2 in preference to available [PF6]2 counterions. The
crystal structure of (^)-3a was also obtained and reveals
that, in the solid state, the molecule forms a self-included
dimer wherein the cavitand conformation is similar to the
The [(h5-C5H5)FeII(h6-chloroarene)][PF6] starting
materials 1a–e[PF6] were prepared by ligand substitution
of ferrocene (Sigma-Aldrich, St. Louis, MO, USA)
according to the method of Nesmeyanov (59). For all
compounds, purification was achieved by passing through a
short column of neutral alumina using acetone as an eluent,
followed by precipitation with diethyl ether. A typical
preparative procedure, that of 1a[PF6], is given below.
Racemic CTG ((^)-CTG) was synthesised according to
published procedures (60). All reactions, unless otherwise
stated, were performed under nitrogen atmosphere and in the
minimal presence of visible light, but work-up procedures
were conducted on the benchtop in the presence of air. All
other reagents and solvents were obtained from the usual
commercial sources and were used as received. Compounds
(^)-2a,d,e[PF6]3 were reported previously (9), though
complete synthetic details are reported herein. Compounds
(^)-2a–h[PF6]3 are difficult to obtain in high purity and
commonlygave disagreeablecombustionanalysesdue to: (i)
their photoinstability, (ii) their propensity to include solvents
of crystallisation and (iii) the presence of small amounts
(,5%) ofimpurities that are not completely removed during
the chromatographic step. Compounds (^)-2b,c[PF6]3 were
presumed to be obtained as a mixture of the four expected
diastereomers, e.g. RRRM-2b[PF6]3, RRSM-2b[PF6]3,
RSSM-2b[PF6]3 and SSSM-2b[PF6]3 and their enantiomers
due to the introduction of planar chirality at each of the three
1,2- or 1,3-unsymmetrically substituted [(h5-C5H5)FeII(h6
-arene)]þ moieties (R/S descriptors in the above notation)
and the racemic nature of the helically chiral (^)-CTG
moieties (M/P descriptors in the above notation). Conse-
quently, and as expected, (^)-2b,c[PF6]3 give rise to
conformations observed in structures of 2a,d–f3þ
.
Cryptophanes syn-5 and (^)-anti-5 were synthesised
by the photochemically induced demetalation of the
putative-metalated cryptophanes syn-4[PF6]3 and (^)-
anti-4[PF6]3, obtained by the synthetic ‘capping’ of (^)-
2d3þ by another molecule of (^)-CTG. Single crystal
structures of five different solvates of (^)-anti-5 reveal
host–guest encapsulated complexes displaying host
conformations that can be fairly interpreted as being
snapshots of the host involved in the process of molecular
gating. The structures provide valuable insights into the
guest-gating conformational processes of cryptophane
container molecules and can be described in the context of
Houk’s ‘french/sliding’ door mechanisms (52), originally
delineated for hemicarcerand containers. The first crystal-
lographically characterised example of an empty, that is,
non-imploded (3) cryptophane, that of (^)-anti-5, reveals
that this host can also readily adopt a conformation that
fills the bulk of its own cavity, perhaps explaining why this
host does not appear to form kinetically stable host–guest
encapsulated complexes in solution.
1
complicated and/or broad H NMR spectra, as described
below. Moreover, the regions of the 1H NMR spectra
corresponding to the [(h5-C5H5)FeII(h6-arene)]þ moieties of
(^)-2a–d,f[PF6]3 are often further complicated (in, for
example, acetone-d6) by the effects of anion complexation,
which serves to freeze out rotation of these groups on the
NMR timescale. The near-purity of (^)-2a-d,f[PF6]3 is
perhaps best evidenced by the typically high yield recovery
and complete characterisation of the corresponding photo-
chemically demetalated cavitands (^)-3a–d,f obtained
in pure form.
Experimental
Materials and methods
NMR spectra were recorded on a Bruker ARX-250
spectrometer operating at 250.1 MHz (1H), 235.3 MHz
(19F), or a Varian Inova 400 MHz spectrometer operating
at 400 MHz (1H) or 100.5 MHz (13C). All NMR spectra
were collected at room temperature unless otherwise
noted and signals were indirectly referenced to tetra-
methylsilane using residual solvent signals as internal
standards. All [(h5-C5H5)FeII(h6-arene)][PF6] salts gave
expected 19F NMR resonances as a doublet at or near d
X-ray crystallography
Single crystal X-ray diffraction data are summarised in
Table 1. Crystals were mounted with epoxy on the end of a
thin glass fibre, attached to a copper pin and centred in the
X-ray beam (Mo Ka) of a Siemens SMART CCD
diffractometer equipped with either a 1K or an APEX II
1
270.68 (d, JP-F ¼ 708 Hz, [PF6]2). Melting points were
determined by differential scanning calorimetry using a
TA Instruments Q20 operating under a purge of nitrogen

Supramolecular Chemistry
885
detector and a Siemens LT-2 or an Oxford Cryosystems
700 series variable temperature apparatus regulating the
temperature with nitrogen gas. All structures were
collected at 173(2) K, with the exception of (^)-anti-
5·5NO2Ph, which was collected at 100(2) K. Full hemi-
sphere data were collected and accurate unit cells were
determined using reflections harvested from the entire data
set. Frames were integrated using the program SAINT.
Data were corrected for the effects of absorption
(SADABS, simulated c-scans) and crystal decay where
appropriate. All structures were solved (direct methods)
and refined using the SHELX suite of software (61).
Refinement was proceeded using conventional alternating
cycles of least-squares refinement (SHELXL-97) on F 2
and difference Fourier synthesis. Whenever possible, all
non-hydrogen atoms were refined with anisotropic
displacement parameters, while hydrogen atoms were
fixed in idealised positions and given displacement
parameters in accord to the atom to which they are
attached. X-SEED (62) was used as a graphical interface to
manipulate the model between refinements and to generate
figures. Many of the structures exhibited considerable
disorder of the anions, including solvents, cyclopentadie-
nyl rings and/or [(h5-C5H5)Fe(h6-arene)]þ moieties. In
such instances, efforts were made to implement a
refinement model consistent with known molecular
geometries by applying reasonable geometric restraints.
Occasionally, thermal parameters were also restrained. In
general, these refinement restraints lead to higher R-
factors, but give rise to more chemically reasonable
molecular geometries. In the interest of full disclosure, the
instruction files for the final SHELXL refinements are
included in the crystallographic information files. (^)-
anti-5·(C6H5N)52/3 and (^)-2a[BF4][PF6]2·(THF)·(CH3-
COCH3) were treated with the SQUEEZE subroutine of
Platon (57) in order to model highly disordered solvents.
The structure of (^)-2d[PF6]3·(Et2O)1.5·(H2O) has been
published previously (9). CCDC 782040–782051 contain
the supplementary crystallographic data for this paper.
These data can be obtained free of charge via
ww.ccdc.cam.ac.uk/data_request/cif, by emailing
data_request@ccdc.cam.ac.uk or by contacting The Cam-
bridge Crystallographic Data Centre, 12, Union Road,
Cambridge CB2 1EZ, UK; Fax: þ44-1223-336033.
3–5 h. The reaction mixture was allowed to cool and was
then quenched with ca. 100 g of ice (CAUTION!).
The bilayered solution was then filtered through Celite and
the aqueous layer was separated and repeatedly washed
with diethyl ether until the washings were clear. Slow
addition of excess [NH4][PF6](aq.) to the solution
_resulted in the precipitation of 1a[PF6] as a yellow
solid. The product was removed by filtration, dried under
vacuum and passed through a short column of neutral
alumina using acetone as an eluent. The pure product
(2.43 g, 40%) was precipitated from the acetone by the
addition of diethyl ether. The compound gave clean,
expected 1H, 13C and 19F NMR spectra and was not
characterised further. Compounds 1b–e[PF6] were
obtained by an identical procedure, though the yields of
the other [(h5-C5H5)FeII(h6-chloroarene)][PF6] salts are
significantly arene dependent.
(^)-2a[PF6]3
Anhydrous DMF (10 ml) was quickly added to a round
bottom flask containing 800 mg (2.11 mmol) of [CpFe(C6-
H5Cl)][PF6] (1a[PF6]), 247 mg (0.605 mmol) of (^)-CTG
and 584 mg (4.22 mmol) of K2CO3 powder. The reaction
was stirred overnight at room temperature, under nitrogen
and in the absence of visible light, after which it was
quenched to slightly acidic pH by the careful addition of
2 M HCl(aq.). To this mixture was slowly added a
concentrated aqueous solution of 0.5 equivalents
[NH4][PF6]. At this point, the solution was transparent.
Precipitation of the crude product was effected by addition
of ca. 100 ml of distilled water. This yellow solid was
filtered and dried under vacuum (room temperature).
The material was further purified by passage through a
short column of neutral alumina with acetone, followed by
precipitation with diethyl ether. Storage in the absence of
light is necessary to avoid decomposition. Yield 711 mg
(0.496 mmol), 82%. 1H NMR (acteone-d6, J/Hz): d 7.73 (s,
3H), 7.53 (s, 3H), 6.20–6.39 (m, 15H, OC6H5), 5.18 (s,
2
15H, C5H5), 5.10 (d, 3H, J ¼ 13.7 Hz, Ha), 3.93 (d, 3H,
2J ¼ 13.7 Hz, He), 3.79 (s, 9H, OCH3). The mixed
[BF24 ]/[PF62]2 salt, for which the crystal structure is
reported, was obtained using H[BF4](aq.) and Na[BF4] in
the place of HCl(aq.) and [NH4][PF6] in the above
procedure.
Syntheses
[(h5-C5H5)FeII(h6-Chloroarene)][PF6] salts
[(h5-C5H5)FeII(h6-C6H5Cl)][PF6], 1a[PF6]
(^)-2b[PF6]3
As for (^)-2a[PF6]3: 1.00 g (2.42 mmol) [CpFe(o-
C6H4Cl2)][PF6] (1b[PF6]), 300 mg (0.734 mmol) (^)-
CTG, 670 mg (4.85 mmol) K2CO3. Yield 858 mg
(0.558 mmol), 76%. 1H NMR (CD3CN, J/Hz): d
7.49–7.50 (m, 3H), 7.25–7.30 (m, 3H), 6.25–6.65 (m,
3H), 5.95–6.06 (m, 6H), 5.80–5.91 (m, 3H), 4.91–5.15
To 250 ml of a stirred, room temperature, continuously
degassed (N2) solution of chlorobenzene containing 3.00 g
(16.1 mmol) of ferrocene and 0.453 g (16.1 mmol) of
aluminium powder was added 8.60 g (64.5 mmol) of
AlCl3. A vigorous reaction immediately ensued and the
temperature was then raised to near reflux conditions for
2
(m, 15H, C5H5 and Ha), 3.87 (d, 3H, J ¼ 14.5 Hz, He),

886
K.T. Holman et al.
3.70–3.82 (m, 9H, OCH3). Note: The spectrum is
complicated due to the presence of diastereomers (see
text). The spectrum in acetone-d6 is more complicated.
and 110 mg (0.796 mmol) K2CO3 powder was used. Yield
151 mg (0.088 mmol), 68%. 1H NMR (acteone-d6, J/Hz): d
7.66 (s, 3H, H1), 7.54 (t, 6H, 3J ¼ 7.5 Hz, H2), 7.50 (s, 3H),
3
7.35 (t, 3H, J ¼ 7.5 Hz), 7.28 (d, 6H), 6.22 ( , q, 12H),
2
5.23 (s, 15H, C5H5), 5.06 (d, 3H, J ¼ 13.5 Hz, Ha), 3.88
(^)-2c[PF6]3
(d, 3H, 2J ¼ 13.5 Hz, He), 3.80 (s, 9H, OCH3). The mixed
[BF4]2/[PF6] salt, for which the crystal structure is
reported, was obtained using H[BF4](aq.) and Na[BF4]
in place of HCl(aq.) and [NH4][PF6] in the work-up.
As for (^)-2a[PF6]3: 2.00 g (4.84 mmol) [CpFe(m-
C6H4Cl2)][PF6] (1c[PF6]), 565 mg (1.38 mmol) (^)-
CTG, 1.34 g (9.70 mmol) K2CO3. Yield 1.68 g
1
1
(1.09 mmol), 79%. H NMR (CD3CN, J/Hz): H NMR
(CD3CN, 250 MHz, J/Hz): d 7.44 (s, 3H, H1), 7.26 (s, 3H,
H2), 6.43 (d, 3H), 6.35 (d, 3H), 6.23 (t, 3H), 5.99 (t, 3H),
2g[PF6]
2
5.09 (s, 15H, C5H5), 4.95 (d, 3H, J ¼ 13.9 Hz, Ha), 3.83
As for (^)-2a[PF6]3, except with 600 mg (1.59 mmol) of
1a[PF6], 232 mg (1.50 mmol) 4-hydroxy-3-methoxyben-
zyl alcohol (vanillyl alcohol), 325 mg (2.35 mmol) K2CO3
powder. Yield 411 mg (0.828 mmol), 55.2%. 1H NMR
(d, 3H, 2J ¼ 13.9 Hz, He), 3.76 (s, 9H, OCH3). Note: The
spectrum is complicated due to the presence of
diastereomers (see text). The spectrum in acetone-d6 is
more complicated.
3
(acteone-d6, J/Hz): d 7.33 (d, 1H, J ¼ 8.1 Hz, Hf), 7.29
3
(s, 1H, Hd), 7.09 (d, 1H, J ¼ 8.1 Hz, He), 6.39–6.44 (m,
2H, Hb), 6.25–6.29 (m, 3H, Ha and Hc), 5.21 (s, 5H,
C5H5), 4.70 (d, 2H, 3J ¼ 5.1 Hz, CH2), 4.31 (t, 1H,
3J ¼ 5.1 Hz, OH), 3.86 (s, 3H, OCH3). 13C NMR
(acteone-d6): Aromatic region: d 151.1; d 143.0; d
139.3; d 133.9; d 122.3; d 119.2; d 111.5; d 86.7; d 84.5; d
(^)-2d[PF6]3
As for (^)-2a[PF6]3: 1.00 g (2.42 mmol) [CpFe( p-
C6H4Cl2)][PF6] (1d[PF6]), 282 mg (0.691 mmol) (^)-
CTG, 670 mg (4.85 mmol) K2CO3. Yield 829 mg
1
1
(0.539 mmol), 78%. H NMR (acetone-d6, J/Hz): d 7.72
3
77.0; d 75.7. Aliphatic region: d 63.0; d 55.3. H NMR
(s, 3H, H1), 7.52 (s, 3H, H2), 6.74 (dd, 3H, J ¼ 6.9 Hz,
peak assignments are according to Figure 3. Single
crystals of 2g[PF6] were grown by the vapour diffusion of
diethyl ether into a dimethylformamide solution of the
compound.
4J ¼ 1.9 Hz, Hto), 6.68 (dd, 3H, 3J ¼ 6.9 Hz, 4J ¼ 1.9 Hz,
3
4
Hti), 6.46 (dd, 3H, J ¼ 6.9 Hz, J ¼ 1.9 Hz, Hbo), 6.31
(dd, 3H, 3J ¼ 6.9 Hz, 4J ¼ 1.9 Hz, Hbi), 5.31 (s, 15H,
C5H5), 5.10 (d, 3H, 2J ¼ 13.8 Hz, Ha), 3.91 (d, 3H,
2J ¼ 13.8 Hz, He), 3.78 (s, 9H, OCH3). Note: In acetone-
d6, rotation about the aryl ether bonds is slow relative to
the 1H NMR timescale and the protons on the inside of the
cavitand cavity (H_i) can be distinguished from those on
the outside of the cavity (H_o) by their response to the
addition of halide ions. Aryl protons at the top of the
cavitand (Ht_), adjacent to the 4-chloro substituent, are
distinguished from those at the bottom of the cavitand
(Hb_), adjacent to the ethereal oxygen, by 2D COSY
experiments.
2h[PF6]
As for (^)-2a[PF6]3 except with 250 mg (0.605 mmol) of
1d[PF6], 233 mg (1.51 mmol) 4-hydroxy-3-methoxyben-
zyl alcohol (vanillyl alcohol), 522 mg (3.78 mmol) K2CO3
1
powder. Yield 291 mg (0.0449 mmol), 74.2%. H NMR
(acteone-d6, J/Hz): d 7.30 (d, 2H, 3J ¼ 8.2 Hz, Hd), 7.28 (s,
3
2H, Hb), 7.07 (d, 2H, J ¼ 8.2 Hz, Hc), 6.20 (s, 4H, Ha),
5.24 (s, 5H, C5H5), 4.69 (d, 4H, 3J ¼ 5.6 Hz, CH2), 4.33 (t,
3
2H, J ¼ 5.6 Hz, OH), 3.88 (s, 6H, OCH3). 13C NMR
(acteone-d6): Aromatic region: d 151.9; d 142.9; d 139.8; d
131.2; d 122.2; d 119.2; d 111.5; d 77.8; d 73.3. Aliphatic
(^)-2e[PF6]3
1
region: d 63.1; d 55.4. H NMR peak assignments are
As for (^)-2a[PF6]3: 200 mg (0.510 mmol) [CpFe( p-
C6H4ClMe)][PF6] (1e[PF6]), 63 mg (0.154 mmol) (^)-
CTG, 128 mg (0.926 mmol) K2CO3. Yield 170 mg
according to Figure 3. Single crystals of 2h[PF6]·DMF
were grown by the vapour diffusion of diethyl ether into a
dimethylformamide solution of the compound.
1
(0.115 mmol), 75%. H NMR (acetone-d6, J/Hz): d 7.71
(s, 3H, H1), 7.52 (s, 3H, H2), 6.10–6.31 (m, 12H,
OC6H4CH3), 5.14 (s, 15H, C5H5), 5.09 (d, 3H,
(^)-3a
2
2J ¼ 13.8 Hz, Ha), 3.92 (d, 3H, J ¼ 13.8 Hz, He), 3.78
In a round bottom flask, 600 mg (0.418 mmol) of (^)-
2a[PF6]3 was dissolved in dry, degassed (N2) acetonitrile
(300 ml) and the solution was exposed to direct sunlight.
The initially yellow solution turned purple/brown almost
immediately, indicating the intermediate formation of
[CpFe(CH3CN)3]þ (63), and the reaction was allowed to
continue with stirring for 3–4 h. Evaporation of the
(s, 9H, OCH3), 2.44 (s, 9H, OC6H4CH3).
(^)-2f[PF6]3
As for (^)-2a[PF6]3 except that 200 mg (0.130 mmol) of
(^)-2d[PF6]3 was used instead of 1a[PF6], 55 mg
(0.58 mmol) of phenol was used instead of (^)-CTG,

Supramolecular Chemistry
solvent, followed by chromatography on silica gel with 1:2 (^)-3d
887
hexane:CHCl3 (Rf ¼ 0.48) yielded crude, colourless (^)-
3a, which was recrystallised by vapour diffusion of hexane
into a concentrated toluene solution of the compound.
As for (^)-3a: 380 mg (0.247mmol) (^)-2d[PF6]3 yielded
176 mg (0.237mmol) (^)-3d, 96%. Mp ¼ 195–1988C. 1H
NMR (CD3CN, J/Hz): d 7.27 (dt, 6H, 3J ¼ 6.8 Hz,
4J ¼ 2.2 Hz), 7.15 (s, 3H, H1), 6.99 (s, 3H, H2), 6.79 (dt,
1
Yield 192 mg (0.301 mmol), 72%. Mp ¼ 216–2188C. H
NMR (CD3CN, J/Hz): d 7.29 (t, 6H, 3J ¼ 7.4 Hz), 7.14 (s,
3
4
2
6H, J ¼ 6.8 Hz, J ¼ 2.2 Hz), 4.77 (d, 3H, J ¼ 13.6Hz,
Ha), 3.65 (s, 9H, OCH3), 3.60 (d, 3H, 2J ¼ 13.6Hz, He). 1H
NMR (CDCl3): d 7.21 (d, 3J ¼ 8.6 Hz, 6H); d 6.91 (s, 3H); d
6.85 (d, 3J ¼ 8.6 Hz, 6H); d 6.74 (s, 3H); d 4.72 (d,
2J ¼ 14.0Hz, 3H, Ha); d 3.69 (s, 9H, OCH3); d 3.53 (d,
2J ¼ 14.0Hz, 3H, He). 13C NMR (CDCl3): Aromatic region:
d156.5;d149.8;d143.5;d136.1;d131.9;d129.4;d127.5;d
121.9; d 118.6; d 114.1. Aliphatic region: d 55.9; d 36.3.
Anal. calc. for C42H33O6Cl36: C, 68.2%; H, 4.5%. Found: C,
68.2%; H, 4.1%.
3
3H, H1), 7.03 (t, 3H, J ¼ 7.4 Hz), 6.96 (s, 3H, H2), 6.82
(d, 6H, 3J ¼ 7.7 Hz), 4.78 (d, 3H, 2J ¼ 13.6 Hz, Ha), 3.64
2
1
(s, 9H, OCH3), 3.60 (d, 3H, J ¼ 13.6 Hz, He). H NMR
3
(CDCl3, J/Hz): d 7.28 (d, J ¼ 7.6 Hz, 6H); d 7.03 (t,
3
3J ¼ 7.6 Hz, 3H); d 6.94 (d, J ¼ 8.4 Hz, 6H); d 6.90 (s,
2
3H); d 6.72 (s, 3H); d 4.71 (d, J ¼ 13.8 Hz, 3H, Ha); d
3.68 (s, 9H, OCH3); d 3.51 (d, 2J ¼ 13.8 Hz, 3H, He). 13
C
NMR (CDCl3): Aromatic region: d 157.7; d 149.7; d
143.9; d 135.5; d 131.8; d 129.5; d 122.7; d 121.4; d 117.7;
d 114.0. Aliphatic region: d 56.0; d 36.3. Anal. calc. for
C42H36O6: C, 79.2%; H, 5.70%. Found: C, 78.8%; H,
5.6%. Single crystals of (^)-3a were grown by slow
evaporation of a toluene solution of the compound.
(^)-3f
As for (^)-3a: 89 mg (0.052 mmol) (^)-2f[PF6]3 yielded
1
33 mg (0.036 mmol) 4.11c, 69%. Mp ¼ 136–1408C. H
3
NMR (acetone-d6, J/Hz): d 7.33 (dd, 6H, J ¼ 7.6 Hz,
3J ¼ 7.6 Hz); d 7.23 (s, 3H, H1); d 7.17 (s, 3H, H2); d 7.06
(^)-3b
(t, 3H, 3J ¼ 7.6 Hz); d 6.95 (d, 6H, 3J ¼ 8.4 Hz); d 6.94 (d,
As for (^)-3a: 800 mg (0.520 mmol) of (^)-2b[PF6]3
yielded 373 mg (0.504 mmol) (^)-3b, 97%. Mp ¼ 207–
2098C. 1H NMR (acetone-d6, J/Hz): d 7.48 (dd, 3H,
3
3
6H, J ¼ 7.6 Hz); d 6.87 (d, 6H, J ¼ 8.4 Hz); d 4.89 (d,
2
3H, J ¼ 13.6 Hz, Ha); d 3.69 (s, 9H, OCH3); d 3.69 (d,
2
4
3H, J ¼ 13.6 Hz, He). 13C NMR (acetone-d6): Aromatic
3J ¼ 7.9 Hz, J ¼ 1.6 Hz), 7.24 (s, 3H, H1), 7.23 (s, 3H,
region: d 158.2; d 154.4; d 151.5; d 150.2; d 143.6; d
136.7; d 132.5; d 129.7; d 122.7; d 122.4; d 120.4; d 117.9;
d 117.8; d 114.8. Aliphatic region: d 55.5; d 35.4. Anal.
calc. for C60H48O9: C, 78.9%; H, 5.3%. Found: C, 78.7%;
H, 5.4%.
H2), 7.19 (td, 3H, 3J ¼ 7.5 Hz, 4J ¼ 1.6 Hz), 7.05, (td, 3H,
3J ¼ 7.5 Hz, 4J ¼ 1.6 Hz), 6.74 (dd, 3H, 3J ¼ 7.5 Hz,
4J ¼ 1.6 Hz), 4.93 (d, 3H, 2J ¼ 13.5 Hz, Ha), 3.74 (d, 3H,
2J ¼ 13.5 Hz, He), 3.69 (s, 9H, OCH3). 13C NMR
(acetone-d6): Aromatic region: d 205.5; d 153.8; d 149.9;
d 142.7; d 137.2; d 132.6; d 130.3; d 128.0; d 123.3; d
122.2; d 117.3; d 114.9. Aliphatic region: d 55.6; d 35.4.
Anal. calc. for C42H33O6Cl3: C, 68.2%; H, 4.5%. Found:
C, 67.9%; H, 4.7%.
Syn-5 and (^)-anti-5
Two separate solutions of (^)-CTG and (^)-2d[PF6]3
(50.0 ml, 6.5 mM, 0.325 mmol each) in DMF were
concurrently added via syringe pump over 24 h to a room
temperature solution of excess K2CO3 in dry, degassed
DMF. The mixture was allowed to react for a further 24 h in
the dark and was subsequently neutralised with 6 M
HCl(aq.), treated with one half of an equivalent of
NH4PF6(aq.), rotary evaporated (,658C) to ,35 ml and
filtered. Addition of water to the filtrate led to the
precipitation of a yellow solid (720 mg) which appeared by
1H NMR spectroscopy to contain mostly polymeric
material. The presence of cryptophane products syn-
4[PF6]3 and (^)-anti-4[PF6]3 in this material is demon-
strated by the isolation of syn-5 and (^)-anti-5 from its
photolysed reaction product (500 ml degassed CH3CN,
sunlight, 5 h; as for (^)-3a). Evaporation of the solvent
followed by chromatography of the crude, photolysed
product on silica gel with CHCl3 yielded crude (^)-anti-5
as a broad, slightly yellow band (Rf < 0.36). Recrystallisa-
tion of this crude material by the partial evaporation of
acetone/CHCl3 solution of the compound yielded colour-
(^)-3c
As for (^)-3a: 1.20 g (0.780 mmol) (^)-2c[PF6]3 yielded
560 mg (0.757 mmol) (^)-3c, 97%. Mp ¼ 168–1718C. 1H
NMR (CD3CN, J/Hz): d 7.25 (t, 3H, 3J ¼ 8.1 Hz), 7.22 (s,
3H, H1), 7.01–7.06 (dd and s, 6H, H2 and para-H), 6.81 (t,
4
3
4
3H, J ¼ 1.5 Hz), 6.75 (dd, 3H, J ¼ 8.1 Hz, J ¼ 1.6 Hz,
2
Hh), 4.80 (d, 3H, J ¼ 13.6 Hz, Ha), 3.65 (s, 9H, OCH3),
3.66 (d, 3H, J ¼ 13.6 Hz, He). 1H NMR (CDCl3): d 7.17
2
(t, 3J ¼ 8.4 Hz, 3H); d 6.99 (d, 3J ¼ 8.4 Hz, 3H); d 6.97 (s,
3H); d 6.91 (s, 3H); d 6.80 (d, 3J ¼ 8.4 Hz, 3H); d 6.79 (s,
2
3H); d 4.75 (d, J ¼ 14.5 Hz, 3H, Ha); d 3.69 (s, 9H,
OCH3); d 3.57 (d, 2J ¼ 14.5 Hz, 3H, He). 13C NMR
(CDCl3): Aromatic region: d 158.8; d 150.1; d 142.8; d
136.4; d 134.9; d 131.9; d 130.2; d 122.6; d 122.4; d 117.4;
d 115.3; d 114.3. Aliphatic region: d 56.1; d 36.4. Anal.
calc. for C42H33O6Cl3: C, 68.2%; H, 4.5%. Found: C,
68.0%; H, 4.4%.

888
K.T. Holman et al.
(2) (a) Even sterically unrestricted CTBs can exist in a nearly
isoenergetic, so-called ‘saddle-twist’ conformation. See, for
example: (a) Zimmerman, H.; Tolstoy, P.; Limbach, H.-H.;
Puopko, R.; Luz, Z. J. Phys. Chem. B. 2004, 108,
18772–18778. (b) Garcia, C.; Aubry, A.; Collet, A. Bull.
Soc. Chim. Fr. 1996, 133, 853–867. (c) Huber, G.; Brotin,
T.; Dubois, L.; Desvaux, H.; Dutasta, J.-P.; Berthault, P.
J. Am. Chem. Soc. 2006, 128, 6239–6246.
(3) Mough, S.T.; Goeltz, J.C.; Holman, K.T. Angew. Chem. Int.
Ed. 2004, 43, 5631–5635.
(4) Cram, D.J.; Cram, J.M. Container Molecules and Their
Guests; Ch. 5: Cavitands, Royal Society of Chemistry:
Cambridge, 1994.
(5) Hardie, M.J. Chem. Soc. Rev. 2010, 39, 516–527.
(6) (a) Matsubara, H.; Oguri, S.; Asano, K.; Yamamoto, Y.
Chem. Lett. 1999, 431–432. (b) Eckert, J.-F.; Byrne, D.;
Nicoud, J.-F.; Oswald, L.; Niergarten, J.-F.; Namuta, M.;
Ikeda, A.; Shinkai, S.; Armaroli, N. New. J. Chem. 2000, 24,
749–758. (c) Hardie, M.J.; Raston, C.L. Chem. Commun.
1999, 1153–1163. (d) Bond, A.M.; Miao, W.; Raston, C.L.;
Ness, T.J.; Barnes, M.J.; Atwood, J.L. J. Phys. Chem. B.
2001, 105, 1687–1695. (e) Huerta, E.; Cequier, E.; de
Mendoza, J. Chem. Commun. 2007, 5016. (f) Huerta, E.;
Metselaar, G.A.; Fragoso, A.; Santos, E.; Bo, C.; de
Mendoza, J. Angew. Chem. Int. Ed. 2007, 46, 202–205.
(7) Holman, K.T.; Steed, J.W.; Atwood, J.L. Angew. Chem. Int.
Ed. 1997, 36, 1736–1738.
(8) (a) Holman, K.T.; Halihan, M.M.; Steed, J.W.;
Jurisson, S.S.; Atwood, J.L. J. Am. Chem. Soc. 1995, 117,
7848–7849. (b) Holman, K.T.; Halihan, M.M.;
Mitchell, A.R.; Burkhalter, R.S.; Steed, J.W;
Jurisson, S.S.; Atwood, J.L. J. Am. Chem. Soc. 1996, 118,
9567–9576.
(9) (a) Holman, K.T.; Orr, G.W.; Steed, J.W.; Atwood, J.L.
Chem. Commun. 1998, 2109–2110. (b) Gawenis, J.A.;
Holman, K.T.; Atwood, J.L.; Jurisson, S.S. Inorg. Chem.
2002, 41, 6028–6031.
(10) (a) Steed, J.W.; Holman, K.T.; Atwood, J.L. Chem.
Commun. 1996, 1401–1407. (b) Lenthall, J.T.; Steed,
J.W. Coord. Chem. Rev. 2007, 251, 1747–1760.
(11) Fairchild, R.M.; Holman, K.T. J. Am. Chem. Soc. 2005,
127, 16364–16365.
(12) Collet, A.; Dutasta, J.P.; Lozach, B.; Canceill, J. Top. Curr.
Chem. 1993, 165, 103–129.
(13) Zhang, S.; Palker, A.; Fragoso, A.; Prados, P.; de Mendoza,
J.; Echegoyen, L. Chem. Mater. 2005, 17, 2063–2068.
(14) Gambut, L.; Chauvet, J.-P.; Garcia, C.; Berge, B.; Renault,
A.; Riviere, S.; Meunier, J.; Collet, A. Langmuir. 1996, 12,
5407–5412.
less prisms of (^)-anti-5·nsolvent, which was heated to
1408C under a nitrogen atmosphere to remove the solvent.
1
Yield 25 mg (7.4%). Mp < 2358C (decomp.). H NMR
(CDCl3, J/Hz) d 6.94 (s, 6H, H1), 6.63 (s, 6H, H2), 6.33 (s,
2
12H, HAr), 4.61 (d, 6H, J ¼ 13.8 Hz, Ha), 3.47 (d, 6H,
2J ¼ 13.8 Hz, He), 3.39 (s, 18H, HOMe). 1H NMR (DMSO-
d6): d 7.26 (s, 6H, H1); d 7.07 (s, 6H, H2); d 6.37 (s, 12H,
2
HAr); d 4.71 (d, 6H, J ¼ 13.4 Hz, Ha); d 3.55 (d, 6H,
2J ¼ 13.4 Hz, He) d 3.35 (s, 18H, OCH3). 13C NMR
(DMSO-d6): Aromatic region: d 195.8; d 151.0; d 142.4; d
142.1; d 135.7; d 130.3; d123.7; d116.0. Aliphatic region: d
56.2; d 29.1. The material did not give agreeable elemental
analyses, presumably related to its high propensity to form
thermally unstable solvates of varying composition.
To obtain samples of syn-5, CH2Cl2 was used during the
chromatography step, instead of CHCl3, to yield a crude
1
mixture (,2.5:1 by H NMR spectroscopy; Rf < 0.64 for
syn-5 and Rf < 0.48 for (^)-anti-5) of syn-5 and (^)-anti-5.
These diastereomers could be separated by fractional
crystallisation. Diffusion of hexane into NO2C6H5 solutions
of the crude material yields macroscopic single crystals of
(^)-anti-5·5NO2C6H5, for which a single crystal structure
was obtained, and microcrystalline syn-5·nNO2C6H5 as a
fine powder. Repeated recrystallisations yielded syn-5 as a
pure NO2C6H5 solvate. The nitrobenzene was removed by
heating the sample to 1408C under a nitrogen atmosphere.
Yield 10 mg (3.0%) of pure material. Mp < 2408C
(decomp.). Although not reported here due to the low
resolution of the data, an X-ray crystal structure (syn-
˚
5·nsolvent; hexagonal, P63/m, a ¼ 16.8054(13)A,
˚
c ¼ 15.5160(16)A) was obtained and unequivocally estab-
lished the identity of the compound as the syn-5
1
diastereomer. H NMR (CDCl3, J/Hz): d 7.02 (s, 6H, H1),
6.89 (s, 6H, H2), 6.53 (s, 12H, HAr), 4.70 (d, 6H,
J 2 ¼ 13.8Hz, Ha), 3.50 (d, 6H, J 2 ¼ 13.8Hz, He), 3.48 (s,
18H, HOMe). 1H NMR (DMSO-d6): d 7.30 (s, 6H, H1); d 7.18
(s, 6H, H2); d 6.40 (s, 12H, HAr); d 4.75 (d, 2J ¼ 13.4Hz, 6H,
Ha); d 3.56 (d, 2J ¼ 13.4Hz, 6H, He) d 3.40 (s, 18H, OCH3).
13C NMR (DMSO-d6): Aromatic region: d 153.5; d 150.6; d
142.4; d 138.1; d 133.1; d 124.5; d 116.1; d 116.0. Aliphatic
region: d 56.3; d 35.5. Elemental analysis was not performed
due to the paucity of material.
(15) (a) Percec, V.; Imam, M.R.; Peterca, M.; Wilson, D.A.;
Heiney, P.A. J. Am. Chem. Soc. 2009, 131, 1294–1304.
(b) Peterca, M.; Percec, V.; Imam, M.R.; Leowanawat, P.;
Morimitsu, K.; Heiney, P.A. J. Am. Chem. Soc. 2008, 130,
14840–14852.
(16) lder, D.; Heinrich, B.; Guillon, D.; Nicoud, J.-F.;
Nierengarten, J.-F. Chem. Eur. J. 2000, 6, 3501–3507.
(17) (a) Bardelang, D.; Camerel, F.; Ziessel, R.; Schmutz, M.;
Hannon, M.J. J. Mater. Chem. 2008, 18, 489–494. (b) Kubo,
Y.; Yoshizumi, W.; Minami, T. Chem. Lett. 2008, 37, 1238–
1239. (c) McKeown, N. B.; Gahnem, B.; Msayib, K.J.;
Budd, P.M.; Tattershall, C.E.; Mahmood, K.; Tan, S.; Book,
D.; Langmi, W.; Walton, A. Angew. Chem. Int. Ed. 2006, 45,
1804–1807. (d) Westcott, A.; Sumby, C.J.; Walshaw, R.D.;
Hardie, M.J. New J. Chem. 2009, 33, 902–912.
Acknowledgements
This work was supported in part by the US National Science
Foundation (DMR-0349316), the University of Missouri and
Georgetown University. We thank Prof. Christian Wolf (George-
town University) for useful advice regarding the notation of
stereochemistry for compounds 2b,c3þ
.
References
(1) (a) Grealis, J.P.; Muller-Bunz, H.; Casey M.; McGlinchey,
M.J. Tet. Lett. 2008, 9, 1527–1530. (b) Cannizzaro, S. Ann.
Chem. Pharm. 1854, 90, 252–254. (c) Cannizzaro, S. Ann.
Chem. Pharm. 1854, 92, 113–117.

Supramolecular Chemistry
889
(18) (a) Structural Aspects of Inclusion Compounds Formed by
Organic Host Lattices. In Inclusion Compounds; Atwood,
J.L., Davies, J.E.D., MacNicol, D.D., Eds.; Academic Press:
London, 1984; Vol. 2; pp 91–121. (b) Hardie, M.J.;
Nichols, P.J.; Raston, C.L. Supramol. Chem. 2002, 8, 1–41.
(19) (a) Mough, S.T.; Holman, K.T. Chem. Commun. 2008,
1407–1409. (b) Sumby, C.J.; Fisher, J.; Prior, T.J.; Hardie,
M.J. Chem. Eur. J. 2006, 12, 2945–2959.
(20) (a) Clark, T.E.; Makha, M.; Sobolev, A.N.; Raston, C.L.
J. Chem. Soc., Dalton Trans. 2008, 36, 4855–4859.
(b) Hardie, M.J.; Raston, C.L. Angew. Chem. Int. Ed. 2000,
39, 3835–3839.
(21) (a) McKeown, N.B.; Gahnem, B.; Msayib, K.J.; Budd,
P.M.; Tattershall, C.E.; Mahmood, K.; Tan, S.; Book, D.;
Langmi, H. W.; Walton, A. Angew. Chem. Int. Ed. 2006, 45,
1804–1807. (b) Budd, P.M.; Butler, A.; Selbie, J.;
Mahmood, K.; McKeown, N.B.; Ghanem, B.; Msayib, K.;
Book, D.; Walton, A. Phys. Chem. Chem. Phys. 2007, 9,
1802–1808.
(22) (a) Holman, K.T. in Encyclopedia of Supramolecular
Chemistry, Atwood, J.L., Steed J.W., Eds.; Dekker: New
York, 2004; Vol. 1; pp 340–348. (b) Collet, A.
In Comprehensive Supramolecular Chemistry, Atwood,
J.L., Davies, J.E.D., MacNicol, D.D., Vogtle, F., Eds.;
Permagon: Oxford, 1996; Vol. 2; pp 325–365.
(23) Brotin, T.; Dutasta, J.-P. Chem. Rev. 2009, 109, 88–130.
(24) Xu, D.; Warmuth, R. J. Am. Chem. Soc. 2008, 130,
7520–7521.
(25) Zhong, Z.; Ikeda, A.; Shinkai, S.; Sakamoto, S.;
Yamaguchi, K. Org. Lett. 2001, 3, 1085–1087.
(26) (a) Ronson, T.K.; Fisher, J.; Harding, L.P.; Rizkallah, P.J.;
Pierre, J.; Warren, J.E.; Hardie, M.K. Nat. Chem. 2009, 1,
212–216. (b) Ronson, T.K.; Carruthers, C.; Fisher, J.;
Brotin, T.; Harding, L.P.; Rizkallah, P.J.; Hardie, M.J. Inorg.
Chem. 2010, 49, 675–685. (c) Abrahams, B.F.;
FitzGerald, N.J.; Robson, R. Angew. Chem. Int. Ed. 2010,
49, 2896–2899.
(27) (a) Jasat, S.; Sherman, J.C. Chem. Rev. 1999, 99, 931–968.
(b) Conn, M.M.; Rebek, J. Chem. Rev. 1997, 97,
1647–1668.
(36) (a) Das, D.; Barbour, L.J. Chem. Commun. 2008,
5110–5112. (b) Das, D.; Barbour, L.J. J. Am. Chem. Soc.
2008, 130, 14032–14033.
(37) Astruc, D. Top. Curr. Chem. 1991, 160, 47–95.
(38) Abd-El-Aziz, A.S.; de Denus, C.R.; Zaworotko, M.J.;
MacGillivray, L.R. J. Chem. Soc. Dalton Trans. 1995,
3375–3393.
(39) Abd-El-Aziz, A.; de Denus, C.R.; Zaworotko, M.J.;
Sharma, C.V.K. Chem. Commun. 1998, 265–266.
(40) (a) Gill, T.P; Mann, K.R. Inorg. Chem. 1980, 19, 3007–
3010. (b) Gill, T.P.; Mann, K.R. Inorg. Chem. 1983, 22,
1986–1991. (c) McNair, A.M.; Schrenk, J.L.; Mann, K.R.
Inorg. Chem. 1984, 23, 2633–2640. (d) Schrenk, J.L.;
McNair, A.; M.; McCormick, F.B.; Mann, K.R. Inorg.
Chem. 1986, 25, 3501–3504.
(41) Chao, J.; Desando, M.A.; Gourlay, D.L.; Orr, D.E.; Walker,
S. J. Phys. Chem. 1984, 88, 711–716.
(42) Steed, J.W.; Zhang, H.M.; Atwood, J.L. Supramol. Chem.
1996, 7, 37–45.
¨
(43) (a) Frensch, K.; Vogtle, F. Liebigs Ann. Chem. 1979, 2121–
2123. (b) Canceill, J.; Collet, A. Nouv. J. Chem. 1986, 10,
17. (c) Garcia, C.; Andraud, C.; Collet, A. Supramol. Chem.
ˆ
1992, 1, 31–45. (d) Garcia, C.; Malthete, J.; Collet, A. Bull.
Soc. Chim. Fr. 1993, 130, 93–95.
(44) Canceill, J.; Collet, A. J. Chem. Soc., Chem. Commun.
1988, 582.
(45) Cram, D.J.; Tanner, M.E.; Keipert, S.J.; Knobler, C.B.
J. Am. Chem. Soc. 1991, 113, 8909–8916.
(46) (a) Fogarty, H.A.; Berthault, P.; Brotin, T.; Huber, G.;
Desvaux, H.; Dutasta, J.-P. J. Am. Chem. Soc. 2007, 129,
10332–10333. (b) Traore, T.; Delacour, L.; Garcia-Argote,
S.; Berthault, P.; Cintrat, J.-C.; Rousseau, B. Org. Lett.
2010, 12, 960–962.
(47) (a) Boche, G. Angew. Chem. Int. Ed. Engl. 1992, 31, 731–
732. (b) Abbott, A.P.; Schiffrin, D.J. J. Chem. Soc., Faraday
Trans. 1990, 86, 1453–1459.
(48) Moreno, A.; Pregosin, P.S.; Veiros, L.F.; Albinati, A.;
Rizzato, S. Chem. Eur. J. 2008, 14, 5617–5629.
(49) Desiraju, G.R.; Sarma, J.A.R.P. Proc. Indian Acad. Sci.
(Chem. Sci.) 1986, 96, 599–605.
(50) Straßner, T. Can. J. Chem. 1997, 75, 1011–1022.
(51) Miura, M.; Yuzawa, S.; Takeda, M.; Takeda, M.;
Habata, Y.; Tanase, T.; Akabori, S. Supramol. Chem.
1996, 8, 53–66.
(52) (a) Nakamura, K.; Houk. K.J. Am. Chem. Soc. 1995, 117,
1853–1854. (b) Houk, K.N.; Nakamura, K.; Sheu, C.;
Keating, A.E. Science 1996, 273, 627–629. (c) Rouhi, M.
Chem. Eng. News. 1996, 74, 4. (d) Sheu, C.; Houk, K. J. Am.
Chem. Soc. 1996, 118, 8056–8070.
(53) (a) Vincent, M.A.; Hillier, I.H. Chem. Phys. 1990, 140,
35–40. (b) Spellmeyer, D.C.; Grootenhuis, P.D.J.; Miller,
M.D.; Kuyper, L.F.; Kollman, P.A. J. Phys. Chem. 1990, 94,
4483–4491. (c) Schaefer, T; Sebastian, R. Can. J. Chem.
1989, 67, 1148–1152.
(28) Collet, A. Cyclotriveratrylene and Related Hosts.
In Comprehensive Supramolecular Chemistry; Atwood,
J.L., Davies, J.E.D., MacNicol, D.D., Vogtle, F., Eds.;
Permagon: Oxford, 1996; Vol. 9, pp 281–302.
(29) (a) Sessler, J.L.; Gale, P.A.; Cho, W.-S. Anion Receptor
Chemistry; Royal Society of Chemistry: Cambridge, 2006.
(b) Steed, J.W. Chem. Soc. Rev. 2009, 38, 506–519.
(30) Staffilani, M.; Hancock, K.S.B.; Steed, J.W.; Holman, K.T.;
Atwood, J.L.; Juneja, R.K.; Burkhalter, R.S. J. Am. Chem.
Soc. 1997, 119, 6324–6335.
(31) Arduini, A.; Calzavacca, F.; Demuru, D.; Pochini, A.;
Secchi, A. J. Org. Chem. 2004, 69, 1386–1388.
(32) (a) Beer, P. D.; Cadman, J. Coord. Chem. Rev. 2000, 205,
131–155. (b) Beer, P.D.; Hayes, E.J. Coord. Chem. Rev.
2003, 240, 167–189.
(54) (a) Kirchhoff, P.D.; Bass, M.B.; Hanks, B.A.; Collet, A.;
McCammon, J.A. J. Am. Chem. Soc. 1996, 118,
3237–3246. (b) Kirchhoff, P.D.; Dutasta, J.-P.; Collet, A.;
McCammon, J.A. J. Am. Chem. Soc. 1997, 119,
8015–8022.
(55) Chapman, R.G.; Sherman, J.C. J. Am. Chem. Soc. 1999,
121, 1962–1963.
(56) Jorgensen, W.L.; Severence, D.L. J. Am. Chem. Soc. 1990,
112, 4768–4774.
(33) Hay, B.P.; Bryantsev, V.S. Chem. Commun. 2008,
2417–2428.
(34) (a) Reith, S.; Bao, X.; Wang, B.-Y.; Hadad, C.M.; Badjic,
J.D. J. Am. Chem. Soc. 2010, 132, 773–773. (b) Rieth, S.;
Wang, B.-Y.; Bao, X.; Badjic, J.D. Org. Lett. 2009, 11,
2495–2498. (c) Wang, B.-Y.; Rieth, S.; Badjic, J.D. J. Am.
Chem. Soc. 2009, 131, 7250–7252.
(35) Takeuchi, D.; Asano, I.; Osakada, K. J. Org. Chem. 2006,
71, 8614–8617.
(57) (a) Spek, A. J. Appl. Crystallogr. 2003, 36, 7. (b) Spek, A.L.
Acta Cryst. A. 1990, A46, C34.

890
K.T. Holman et al.
(58) (a) Atwood, J.L.; Barbour, L.J. Science 2002, 296,
2367–2369. (b) Barbour, L.J. Chem. Commun. 2006,
1163–1168.
Shilovtseva, L.S. Dokl. Akad. Nauk SSSR 1965, 160,
1327–1332.
(60) (a) Canceill, J.; Gabard, J.; Collet, A. J. Chem. Soc., Chem.
Commun. 1983, 122–123. (b) Canceill, J.; Collet, A.;
Gotarelli, G. J. Am. Chem. Soc. 1984, 106, 5997–6003.
(61) Sheldrick, G.M. Acta Cryst. A. 2008, 64, 112–122.
(62) (a) Barbour, L.J. Supramol. Chem. 2001, 1, 189–191.
(b) http://x-seed.net.
(59) (a) Nesmeyanov, A.N.; Vol’kenau, N.A.; Bolesova, I.N.
Tetrahedron Lett. 1963, 1725–1729. (b) Nesmeyanov,
A.N.; Vol’kenau, N.A.; Bolesova, I.N. Dokl. Akad. Nauk
SSSR 1963, 149, 615. (c) Nesmeyanov, A.N.; Vol’kenau,
N.A.; Bolesova, I.N. Dokl. Akad. Nauk SSSR 1966, 166,
607–611. (d) Nesmeyanov, A.N.; Vol’kenau, N.A.;
(63) Freedman, D.A.; Gill, T.P.; Blough, A.M.; Koefod, R.S.;
Mann, K.R. Inorg. Chem. 1997, 36, 95–102.

Copyright of Supramolecular Chemistry is the property of Taylor & Francis Ltd and its content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.