Nanoporous organic alloys

Article abstract of DOI:10.1002/anie.201105216

Crystal tuning: Organic molecules can be xenophobic, preferring to crystallize with their own kind. Though useful for purification, this precludes the tuning of crystal properties by doping or mixing. Nanoporous steroids provide an exception, as their channels can accept a variety of termini (hexagons and spheres). The steroids can be cocrystallized in any ratio to give a wide range of chiral, potentially porous crystalline materials.

Full text of DOI:10.1002/anie.201105216

Communications
DOI: 10.1002/anie.201105216
Crystal Engineering
Nanoporous Organic Alloys**
Ramalingam Natarajan, Germinal Magro, Lydia N. Bridgland, Anchalee Sirikulkajorn,
Sampriya Narayanan, Lloyd E. Ryan, Mairi F. Haddow, A. Guy Orpen, Jonathan
P. H. Charmant, Andrew J. Hudson, and Anthony P. Davis*
The crystalline state presents many advantages for the design
of functional materials. Components in crystals are ordered
and oriented (usually anisotropically), crystals are generally
more robust than amorphous solids, and the crystal packing
can be arranged to leave gaps to allow absorption of and
interaction with guest species. However crystals are not so
easily tuned as amorphous materials, because compositions
cannot in general be varied freely. Crystal packing generates
an array with symmetry and dimensions which are specific to
the system. An altered component which might moderate a
property (for example, optical extinction or refractive index)
may not fit into the array, and will often be rejected.
Adjustable stoichiometries are fairly common for inor-
ganic materials such as metal alloys, minerals and coordina-
tion polymers, where similarly sized components (e.g. metal
ions) and/or strong directional bonding allows substitutions
within crystal frameworks.[1,2] However for organic molecular
crystals, with diverse structures governed by weak intermo-
lecular forces, such exchanges are more problematic. There
are some examples where solid solutions (“organic alloys”)[3]
are formed from molecules which can be mixed in any ratio
within the same crystal structure. However, this has only been
possible where the molecules concerned are very similar in
size, shape, and nature.[1,3,4] Herein we report a new approach
to the preparation of crystals from continuously variable
mixtures of organic molecules. The strategy allows the
incorporation of multiple and diverse units, including mole-
cules which may not crystallize by themselves. Moreover
these organic alloys can accept guests which are not part of
the crystal array, increasing the prospects for useful function-
ality.
[*] Dr. R. Natarajan, Dr. G. Magro, L. N. Bridgland, A. Sirikulkajorn,
Figure 1. The crystal structure of archetypal channel-forming steroidal
urea 1. a) Formula of 1 and the structure of a single molecule. The
Dr. S. Narayanan, Dr. M. F. Haddow, Prof. A. G. Orpen,
Dr. J. P. H. Charmant, Prof. A. P. Davis
terminal OCH3 and NHPh groups are shown in space-filling mode,
School of Chemistry, University of Bristol
colored magenta and orange, respectively. The steroid is solvated by a
Cantock’s Close, Bristol, BS8 1TS (UK)
molecule of water (shown with thick bonds) which forms hydrogen
E-mail: Anthony.davis@bristol.ac.uk
bonds to all three ureas. b) Packing of 1 viewed down the crystallo-
graphic c axis (along the line of the channels). All the molecules are in
the same environment, but two are highlighted (one blue, one green,
except that OCH3 and NHPh groups are space-filling and retain their
L. E. Ryan, Dr. A. J. Hudson
Department of Chemistry, University of Leicester
Leicester, LE1 7RH (UK)
[**] The synthetic and crystallographic work was supported by the
EPSRC (EP/E021581/1) and the Government of Thailand (Fellow-
ship to A.S.). The fluorescence measurements utilized equipment
provided by the Wellcome Trust. L.B. is a member of the Bristol
Centre for Functional Nanomaterials, an EPSRC-funded Doctoral
Training Centre.
coloring). The channels are of average diameter 1.6 nm with walls
1.3 nm thick and a helix repeat distance of 1.15 nm along the c axis.
c) A single channel sliced in half along the c axis, viewed in space-
filling mode. Terminal OCH3 and NHPh groups retain their coloring,
other atoms near or at the internal surface are shown as light blue.
The channel surface is formed by two parallel helices of steroids with
alternating alignments. Six steroids are required for each turn of each
helix.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201105216.
11386
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 11386 –11390

To obtain an organic alloy, it is generally necessary that
the crystals of the individual molecules should possess almost
identical structures with the same space group and closely
similar unit cell parameters. Usually, this implies that the
molecular structures are almost identical, perhaps differing by
a few atoms. However, in principle this need not be true if the
crystal contains substantial open space. Effectively, the
molecule can expand into the voids without interfering with
the crystal packing. A variety of compounds can therefore
crystallize in the same format, part of the structure remaining
constant and controlling the packing while the remainder
contains the variability. If the spaces are sufficiently large, and
the canonical structure robust, it should be possible to form a
wide range of continuously tunable materials. Organic
crystals with large voids are quite rare,[5] but we have recently
discovered a family of compounds which allow the realization
of this concept. Exemplified by 1, these nanoporous steroidal
ureas (NPSUs) form crystals with hexagonal P61 symmetry in
which the molecules form helices surrounding broad channels
(up to 1.6 nm average diameter, 30% of crystal by volume).[6]
As illustrated in Figure 1, the central core of the steroid
provides the scaffolding for the material while groups at
either end point into the channels. The channels are chiral and
can accept guest molecules, but can also accommodate a
variety of end groups. Indeed, we have found that a number of
molecules of general structure 2 crystallize in the same
manner as 1 with almost identical unit cell parameters.[6,7]
Groups R1 and R2 form part of the channel surface, and may
be used to control the internal diameter, optical properties,
and chemical nature of the pores.
Given the isostructural nature of NPSUs, there seemed no
reason why two or more should not crystallize together in any
ratio, to form continuously variable nanoporous materials. As
illustrated in Figure 2, the phenomenon could be exploited in
various ways. Firstly, binary mixtures could be formed in any
ratio to achieve desired macroscopic physical properties such
as density, dielectric constant, refractive index, etc. (Fig-
ure 2b). Secondly, steroids with exceptionally large and
complex termini could be included in the mixture (Figure 2c).
Such molecules could not form the nanoporous structure by
themselves, because the channels could not accommodate
such large units attached to every molecule (6 for each turn of
the helix). However, if they are mixed in low ratio with
smaller host steroids, the problem disappears. Thirdly, any
Figure 2. Schematic depictions of NPSUs. a) Crystal formed from a
single component, with spheres and hexagonal prisms representing
the terminal groups (OMe and NHPh for 1; cf. Figure 1c). Guest
molecules may enter the channels. b) Alloy formation with a second
component (blue) allows tuning of macroscopic properties. c) Larger
groups can be positioned in the channel if incorporated in low ratios.
Depending on the size of the group and the level of doping, guest
molecules may also be accommodated. d) Multicomponent alloy with
diverse terminal groups acting in concert on absorbed guest mole-
cules.
Angew. Chem. Int. Ed. 2011, 50, 11386 –11390
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 11387

Communications
number of steroids could be mixed together in principle,
confirm the presence of all components. For mixtures with
colored components, optical microscopy provided further
evidence that the chromophores were distributed throughout
the sample. Overall the results suggested that, in a given
mixture, some components were more readily included than
others. Thus the compositions of the crystals deviated some-
what from the initial ratios, while some micrographs revealed
variations in color intensity. The latter may be explained by
unequal rates of incorporation—towards the end of a
crystallization, the composition in solution could differ
markedly from the initial ratio, leading to inhomogeneity.
However, differences seem relatively small, and no evidence
was obtained for the segregation of pure components, either
in different crystals or in domains within individual crystals.
Having demonstrated the formation of binary (Figure 2b)
and multicomponent (Figure 2d) alloys, we investigated the
possibility of incorporating steroids with large terminal units
as dopants in a matrix of simpler hosts (Figure 2c). Peryle-
nediimide (PDI, see 3) is a widely used chromophore which is
valued for its high fluorescence quantum yield, thermal and
giving chiral crystalline nanoporous materials with highly
diverse interior microenvironments (Figure 2d). These mate-
rials could be capable of sophisticated functions such as
enzyme-like catalytic activity on bound guests, especially if a
way could be found to organize the different components.
To explore these possibilities, we first employed prototype
NPSU 1 and the closely similar analogue 2a, bearing an
iodine atom as a crystallographic probe. 2a was prepared
from the inexpensive steroid cholic acid, by a procedure
analogous to that used for 1.[6,8] As expected, 2a crystallized
from methyl acetate to give needles similar in appearance to
those formed by 1. Single crystal X-ray diffraction (SCXRD)
confirmed that 2a was isostructural with 1, with slightly
smaller cell dimensions (ca. 0.6%). A 1:1 mixture of 1 and 2a
was then crystallized under similar conditions. The resulting
needles were collected and analyzed by 1H NMR spectrosco-
py to reveal a 65:35 ratio of 1 to 2a.[8] To establish whether
both molecules occurred in the same crystal, one of the
needles was subjected to SCXRD.[8] Iodine atoms were
detected in the appropriate position at roughly 30% occu-
pancy, consistent with formation of an organic alloy.
The study was then expanded to involve NPSUs 2b–i,[9]
with further variations in both R1 and R2. Systems bearing
azobenzene chromophores (2c–e) were included to facilitate
analysis by optical microscopy. The synthesis of 2e was
reported previously,[7] while the remainder were prepared by
analogy with established procedures.[8] A range of binary,
tertiary and quaternary mixtures were co-crystallized as
exemplified in Table 1.[8] In all but a very few cases, the
mixtures gave needles with similar appearance to those
formed by 1. Crystallographic analyses confirmed the reten-
tion of the P61 nanoporous structure, with minor variations in
cell parameters (see Table 1). Ratios of components in the
bulk crystalline samples were determined by 1H NMR
integration, while individual crystals were analyzed by
SCXRD or electrospray mass spectrometry (ESMS) to
photochemical stability, and ability to act as an n-type
semiconductor.[10] Modeling suggested that, as present in 3,
it is also sufficiently bulky to exhaust the free space in the P61
nanoporous structure.[8] Steroid 3 was synthesized from 1 by
ester hydrolysis followed by coupling to the appropriate N-
hydroxyethyl perylenediimide.[8] As expected, 3 failed to
crystallize by itself, despite extensive efforts. How-
ever, when mixed in 1:1 ratio with 1 it yielded red
needles of composition 3:1 = 40:60 (1H NMR spec-
troscopy) (Figure 3a). SCXRD confirmed the P61
structure but could not locate the perylene units,
which were presumably disordered within the
channels.
Table 1: Organic alloys formed from steroidal ureas 1 and 2.
Components Initial
ratio
Ratio in
crystals
(bulk)[a]
Homogeneity[b] Crystallographic unit
cell parameters [ꢀ][c]
a,b
c
1, 2a
1, 2b
1, 2c
1, 2d
1, 2e
1, 2 f
1, 2g
50:50
50:50
50:50
50:50
50:50
50:50
50:50
65:35
50:50
65:35
55:45
55:45
45:55
50:50
SCXRD
ESMS
29.0453(4) 11.4639(2)
With PDI derivative 3 in hand, we were
positioned to illustrate the potential of NPSUs for
crystal tuning. The optical properties of PDI-based
materials are strongly affected by the relationship
between individual chromophores. If the PDI units
are spread throughout the material at low density
they will behave independently, as if in dilute
solution. However, if they are closer together, they
can interfere with each other resulting in quenching
and other phenomena.[11] In particular, anisotropic
arrays of PDI systems can serve as optical wave-
guides in which light-energy is absorbed, re-emitted
at longer wavelengths and transmitted over long
distances.[12] Much effort has been devoted to
controlling the arrangement of PDI units, but
28.986(1)
28.800(3)
28.918(6)
28.911(8)
28.834(9)
28.91(4)
11.456(5)
11.470(14)
11.429(3)
11.425(5)
11.426(5)
11.450(14)
11.409(11)
ESMS[d]
ESMS[d]
ESMS[d]
ESMS
ESMS
1, 2h
50:50
50:50
ESMS
28.85(3)
2a, 2e
1, 2b, 2e
1, 2c, 2g
1, 2d, 2i
50:50
45:55
SCXRD
ESMS[d]
ESMS[d]
ESMS[d]
29.1427(19) 11.4342(15)
33:33:33
33:33:33
33:33:33
35:35:30
35:35:30
40:25:35
28.880(8)
28.984(8)
28.942(5)
11.425(3)
11.462(3)
11.448(2)
1, 2b, 2c, 2h 25:25:25:25 25:25:30:20 ESMS[d]
28.899(11) 11.439(5)
[a] Determined by 1H NMR integration. [b] Method used to show that all components
were present in a single crystal. [c] Standard deviations are given in parentheses, and
are smaller for those cases where a full structural determination was performed.
[d] Supported by optical microscopy.
11388 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 11386 –11390

thinly throughout the crystal [sample (C)], emitted light was
observed from all regions as expected. However, for sample
(A), with the highest density of PDI units, light was observed
only from the ends (Figure 3d). When illumination was
restricted to a small region in the middle of the crystal no
emission was observed (Supporting Information, Fig-
ure S63b), but when one end was illuminated both ends
emitted brightly (Figure 3e). It is known that the long axis of
the needles coincides with the channel axis.[6] The results
therefore imply that fluorescence is strongly quenched in the
bulk of the material, or at surfaces parallel to the channels,
but that emission occurs readily at (or near) surfaces which
cross channels. Moreover, this emission is waveguided along
the line of the channels with remarkable efficiency, with
almost no loss of light from internal scattering in the bulk of
the crystal (Figure S63c). The properties of the crystal are
clearly anisotropic, highlighting the contrast between this
organic alloy and a simple, amorphous mixture of compounds.
The distribution of the components in NPSU-based
organic alloys remains an open question. The evidence thus
far is consistent with random dispersions, but small-scale
segregation or ordering may occur in some cases. The ability
to control organization would further increase the potential
utility, and it is interesting to speculate whether a guest
molecule might serve as a template to impose order during
crystallization. Alternatively, specific non-covalent interac-
tions could be used to induce pairing, or steroidal components
could be linked covalently before crystallization. Even with
random ordering, there is broad scope for the development of
applications through tuning of physical properties (by choice
of components and adjustment of ratios), and/or the doping of
effector units (e.g. for sensing, catalysis or flow control) within
the channels.
Figure 3. a) Color bright-field image of a single needle obtained from
crystallization of 3:1=1:1. b) Fluorescence excitation spectra, recorded
at a fixed emission wavelength of 620Æ5 nm, for 3 dissolved in
chloroform (black) and for solid solution single needles crystallized
from 3:1=1:1, (blue), 1:10 (green), and 1:100 (red). c) Fluorescence
emission spectra recorded with excitation at 487 nm. Color coding as
for (b). d) Fragments from single needles of solid solutions crystallized
from 3:1=1:100 and 1:1, visualized by total internal reflection
fluorescence (TIRF) microscopy using a laser wavelength of 488 nm.
e) The crystal from 3:1=1:1 visualized by TIRF using an illumination
field restricted to the marked circular region. The TIRF image was
merged with a separate bright-field image to illustrate the location of
the sample.
Received: July 25, 2011
Published online: October 4, 2011
tuning by continuous variation within a crystalline environ-
ment has not previously been possible.
We therefore prepared crystals from a range of mixtures
of 3 + 1, and studied their optical behavior using fluorescence
spectroscopy.[8] Figure 3b and c show selected results from
three samples, obtained by crystallization of 3:1 in the ratio
A) 1:1 (as discussed above), B) 1:10, and C) 1:100. The
excitation (Figure 3b) and emission spectra (Figure 3c)
show clearly the effect of chromophore proximity within the
crystal. For the most dilute sample (C), the spectra are similar
to those of 3 dissolved in chloroform, showing that there is
little interaction between the chromophores (and also imply-
ing that the PDI units are evenly distributed within the solid).
In contrast, the spectra for crystals (A) and (B) are very
different to that of 3 in solution. The excitation spectra
(Figure 3b) show substantial blue-shifts, consistent with
interacting PDI chromophores. The emission spectra are
red-shifted, implying the involvement of excimers (also
supported by fluorescence lifetime measurements, see Sup-
porting Information). The excitation and emission spectra
from sample (A) are closely similar to literature absorption
and emission data for a helical polyisocyanide polymer
densely substituted with PDI units.[13]
Keywords: crystal engineering · doping ·
fluorescence microscopy · microporous materials ·
supramolecular chemistry
.
[1] A. I. Kitaigorodsky, Mixed Crystals, Springer, Berlin, 1984.
[2] T. Fukushima, S. Horike, Y. Inubushi, K. Nakagawa, Y. Kubota,
M. Takata, S. Kitagawa, Angew. Chem. 2010, 122, 4930; Angew.
Chem. Int. Ed. 2010, 49, 4820; C. J. Adams, M. F. Haddow, M.
Lusi, A. G. Orpen, Proc. Natl. Acad. Sci. USA 2010, 107, 16033;
H. X. Deng, C. J. Doonan, H. Furukawa, R. B. Ferreira, J.
Towne, C. B. Knobler, B. Wang, O. M. Yaghi, Science 2010, 327,
846; R. Custelcean, D. E. Jiang, B. P. Hay, W. S. Luo, B. H. Gu,
Cryst. Growth Des. 2008, 8, 1909.
[3] K. Sada, K. Inoue, T. Tanaka, A. Epergyes, A. Tanaka, N.
Tohnai, A. Matsumoto, M. Miyata, Angew. Chem. 2005, 117,
7221; Angew. Chem. Int. Ed. 2005, 44, 7059; M. Dabros, P. R.
Emery, V. R. Thalladi, Angew. Chem. 2007, 119, 4210; Angew.
Chem. Int. Ed. 2007, 46, 4132.
[4] M. Morimoto, S. Kobatake, M. Irie, J. Am. Chem. Soc. 2003, 125,
11080; S. Takami, L. Kuroki, M. Irie, J. Am. Chem. Soc. 2007,
129, 7319.
[5] For leading references see: G. Couderc, J. Hulliger, Chem. Soc.
Rev. 2010, 39, 1545; J. D. Wuest, Chem. Commun. 2005, 5830;
Major differences were also revealed through fluores-
cence microscopy.[8] When PDI chromophores were spread
Angew. Chem. Int. Ed. 2011, 50, 11386 –11390
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 11389

Communications
S. J. Dalgarno, P. K. Thallapally, L. J. Barbour, J. L. Atwood,
[10] F. Wꢀrthner, Chem. Commun. 2004, 1564; T. Weil, T. Vosch, J.
Hofkens, K. Peneva, K. Mꢀllen, Angew. Chem. 2010, 122, 9252;
Angew. Chem. Int. Ed. 2010, 49, 9068; X. W. Zhan, A. Facchetti,
S. Barlow, T. J. Marks, M. A. Ratner, M. R. Wasielewski, S. R.
Marder, Adv. Mater. 2011, 23, 268.
[11] G. Klebe, F. Graser, E. Hadicke, J. Berndt, Acta Crystallogr. Sect.
B 1989, 45, 69; P. M. Kazmaier, R. Hoffmann, J. Am. Chem. Soc.
1994, 116, 9684; H. Langhals, O. Krotz, K. Polborn, P. Mayer,
Angew. Chem. 2005, 117, 2479; Angew. Chem. Int. Ed. 2005, 44,
2427; T. E. Kaiser, H. Wang, V. Stepanenko, F. Wꢀrthner,
Angew. Chem. 2007, 119, 5637; Angew. Chem. Int. Ed. 2007, 46,
5541.
Chem. Soc. Rev. 2007, 36, 236; C. H. Gçrbitz, Chem. Eur. J. 2007,
13, 1022; A. J. Cruz-Cabeza, G. M. Day, W. Jones, Chem. Eur. J.
2009, 15, 13033; N. B. McKeown, J. Mater. Chem. 2010, 20,
10588; A. Comotti, S. Bracco, G. Distefano, P. Sozzani, Chem.
Commun. 2009, 284; J. T. A. Jones, T. Hasell, X. Wu, J. Bacsa,
K. E. Jelfs, M. Schmidtmann, S. Y. Chong, D. J. Adams, A.
Trewin, F. Schiffman, F. Cora, B. Slater, A. Steiner, G. M. Day,
A. I. Cooper, Nature 2011, 474, 367.
[6] A. L. Sisson, V. del Amo Sanchez, G. Magro, A. M. E. Griffin, S.
Shah, J. P. H. Charmant, A. P. Davis, Angew. Chem. 2005, 117,
7038; Angew. Chem. Int. Ed. 2005, 44, 6878.
[7] R. Natarajan, J. P. H. Charmant, A. G. Orpen, A. P. Davis,
Angew. Chem. 2010, 122, 5251; Angew. Chem. Int. Ed. 2010, 49,
5125.
[12] Y. K. Che, X. M. Yang, K. Balakrishnan, J. M. Zuo, L. Zang,
Chem. Mater. 2009, 21, 2930; Q. L. Bao, B. M. Goh, B. Yan, T.
Yu, Z. A. Shen, K. P. Loh, Adv. Mater. 2010, 22, 3661.
[13] J. Hernando, P. A. J. de Witte, E. van Dijk, J. Korterik, R. J. M.
Nolte, A. E. Rowan, M. F. Garcia-Parajo, N. F. van Hulst,
Angew. Chem. 2004, 116, 4137; Angew. Chem. Int. Ed. 2004,
43, 4045.
[8] For details, see the Supporting Information.
[9] Compounds 2b–i crystallized individually to give P61 nano-
porous structures. The structure of 2e is reported in Ref. [7]. For
the remainder, details will be given in a separate publication.
11390 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 11386 –11390