Narcissistic self-sorting in self-assembled cages of rare earth metals and rigid ligands

Article abstract of DOI:10.1002/anie.201500400

Highly selective, narcissistic self-sorting can be achieved in the formation of self-assembled cages of rare earth metals with multianionic salicylhydrazone ligands. The assembly process is highly sensitive to the length of the ligand and the coordination geometry. Most surprisingly, high-fidelity sorting is possible between ligands of identical coordination angle and geometry, differing only in a single functional group on the ligand core, which is not involved in the coordination. Supramolecular effects allow discrimination between pendant functions as similar as carbonyl or methylene groups in a complex assembly process. Equals among equals: The formation of self-assembled cages of rare earth metals and ligands proceeds through highly selective, narcissistic self-sorting. Pendant functionalities as similar as carbonyl and methylene groups are discriminated in this complex assembly process.

Full text of DOI:10.1002/anie.201500400

Angewandte
Chemie
DOI: 10.1002/anie.201500400
Self-Assembly
Narcissistic Self-Sorting in Self-Assembled Cages of Rare Earth Metals
and Rigid Ligands**
Amber M. Johnson, Calvin A. Wiley, Michael C. Young, Xing Zhang, Yana Lyon,
Ryan R. Julian, and Richard J. Hooley*
Abstract: Highly selective, narcissistic self-sorting can be
achieved in the formation of self-assembled cages of rare
earth metals with multianionic salicylhydrazone ligands. The
assembly process is highly sensitive to the length of the ligand
and the coordination geometry. Most surprisingly, high-fidelity
sorting is possible between ligands of identical coordination
angle and geometry, differing only in a single functional group
on the ligand core, which is not involved in the coordination.
Supramolecular effects allow discrimination between pendant
functions as similar as carbonyl or methylene groups in
a complex assembly process.
more difficult, and geometrical differences are often
exploited for good selectivity.
[4]
Narcissistic self-sorting is less common, and generally
observed in hydrogen-bonded systems. Isaacs and Wu showed
that a mixture of ten components can narcissistically self-sort,
despite the fact that each component contains complementary
[
2a]
hydrogen-bonding motifs. There are also capsular aggre-
gates that display this effect, and these examples take
[5]
advantage of geometrical differences in the components.
Narcissistic sorting in metal–ligand assemblies is rarer still.
Discrimination between identical coordinating motifs is
challenging, unless geometric constraints are added. The
pioneering example showed narcissistic self-sorting of bipyr-
idyl ligands of varied lengths in the formation of self-
S
elf-sorting and the controlled organization of self-assem-
bled systems are vital for the controlled formation of macro-
molecular constructs. The selective assembly of DNA is the
classical example, but the selective folding of proteins could
also be thought of as “self-sorting”, in that some individual,
highly similar hydrogen-bond interactions are favored over
others as part of a greater self-assembly process. On the
whole, the discrimination between individual hydrogen-
bonding interactions is strong, but each interaction by itself
is weak. Self-sorting can be achieved in the formation of
assembled M L3 helices; other groups have applied this
2
[6]
concept to bis(catecholate) coordinators. Other tactics that
confer some type of selectivity in sorting (not necessarily
narcissistic) between identical coordinating groups are to vary
[7]
the coordination angle or distance, employ stereoinduc-
[
8]
[9]
tion, or provide steric hindrance in the internal cavity.
This leads to a question: how similar can individual
components be, while still allowing narcissistic self-sorting?
In complexes with very similar geometry and the same
coordinating motif, mixing is most commonly observed, and
[
1]
synthetic supramolecular assemblies. There are two possible
types of self-sorting upon multicomponent self-assembly:
narcissistic self-sorting, whereby the individual component
forms an assembly solely with itself, or social self-sorting, in
which mixed assemblies (heterocomplexes) are favored.
[10]
highly controlled narcissistic selectivity is rare. We recently
showed that multianionic salicylhydrazone ligands display
subtle selectivity for differently sized rare earth metals, an
[
2]
[3]
[11]
Social self-sorting is relatively common, and multiple differ-
ent components can be engineered to form impressively
complex structures. Social self-sorting in metal–ligand assem-
blies can exploit different coordination motifs in the forma-
effect that is maximized by supramolecular cooperativity.
Other groups have investigated self-assembly exploiting
[
12]
lanthanide coordinators,
and this also focuses on the
[
13]
selectivity between different metal ions.
The delicate
[
4]
tion of polygons, polyhedra, catenanes, and knots. Obvi-
ously, each of these sorting outcomes can be favored by
maximizing the difference between the coordinating ligands.
The greater the difference in geometry, size, coordination
angle, and coordination denticity, the easier the self-sorting
becomes. When using ligands with highly similar, or even
identical coordination motifs, discrimination becomes far
discrimination we observed between metals of very similar
charge, size, and coordination sphere suggested that ligand-
based self-sorting behavior might be possible. Here we show
that highly selective narcissistic self-sorting can be achieved in
the lanthanide-mediated self-assembly of highly similar
bis(salicylhydrazone) ligands. Most remarkably, changing
a single function (from CH to C=O) that is not involved in
2
the coordination process can allow ligand discrimination.
A number of bis(salicylhydrazone)-based ligands were
synthesized to investigate the effect of variable coordination
geometry/angle, length, and pendant substitution on the
narcissistic self-sorting behavior (Figure 1). Each ligand
displayed two identical salicylhydrazone coordination
motifs, a tridentate coordinator known to form polyhedral
[
*] A. M. Johnson, C. A. Wiley, M. C. Young, X. Zhang, Y. Lyon,
Prof. R. R. Julian, Prof. R. J. Hooley
University of California Riverside, Department of Chemistry
Riverside, CA, 92521 (USA)
E-mail: richard.hooley@ucr.edu
[
14]
self-assemblies upon exposure to lanthanide salts.
Fluo-
[
**] The National Science Foundation (CHE-1151773 to R.J.H., CHE-
0
747481 to R.R.J.) is acknowledged for support.
rene-based ligand cores are good candidates for controlling
[15]
self-assembly properties.
Ligands 1–4 display identical
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201500400.
coordination lengths and angles, and differ only in the
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
and 6. After treatment with sodium hydride and isolation as
the tetrasodium salts, pairs of these ligands were combined
3
+
with Ln salts in [D ]DMSO at a total M:L ratio of 2:3 (i.e.
6
3
+
1
molar equivalent Ln , 0.75 equivalents of each ligand), and
the resulting assemblies were analyzed by H NMR spectros-
copy (Figure 2). While complexes could be assembled from
1
Figure 1. a) Ligands with identical shape and variable functions;
b) ligands with variable lengths and coordination angles; c) represen-
6
À
tative self-assembly process forming [1 Y ]
.
3
2
group on the central carbon atom. These ligands were
synthesized from commercially available 9-fluorenone-2,7-
dicarboxylic acid (see the Supporting Information for syn-
thetic procedures and characterization). Four different inter-
nal groups were employed, with slightly different sizes and
polarities: fluorenone ligand 1, fluorene ligand 2, which
displays a smaller CH group, and the larger ligands oxime 3,
2
or benzyl oxime 4. Other ligands were synthesized to study
Figure 2. Narcissistic self-sorting in ligands of different geometry:
[14]
[11]
6À
6À
12À
6À
larger changes. Known ligands 5
and 6
vary in both
a) [1
3
Y
]
2
; b) [6
Y
3
2
]
; c) [5
6
Y
4
]
; d) [8 Y ] ; e) 1+6+Y(OTf) ;
3 3
2
f) 1+5+Y(OTf) ; g) 5+6+Y(OTf) ; h) 1+8+Y(OTf) (400 MHz,
coordination length and angle. The length of the ligand can be
varied without changing the coordination angle by extending
the core fluorenone motif with a rigid linear spacer. These
longer ligands required an alternate synthetic route, and were
accessed by Suzuki coupling between 2,7-dibromofluorene
and 4-ethoxycarbonylphenylboronic acid followed by deriva-
tization. The extended ligand with an unsubstituted fluorene
core proved susceptible to oxidation, so we focused on the
extended fluorenone ligand 7 and a,a-dimethylfluorene
ligand 8. Finally, control ligands 9a–c, which contain only
a single salicylhydrazone motif, were synthesized to evaluate
the selectivity of the coordinating motif itself, without any
supramolecular considerations.
3
3
3
[
D ]DMSO, 298 K).
6
a range of lanthanide salts, assembly was primarily studied
3
+
3+
with diamagnetic La and lanthanide surrogate Y for ease
of NMR analysis. These metals represent variation in the
lanthanide series (Y is similar in size to Ho) and should
provide representative data for both larger and smaller
metals. The initial ligand combinations test the self-sorting
behavior of species with relatively large changes in coordina-
tion angle. Even though the coordinators are identical, one
might expect narcissistic sorting to be easily achieved. The
lengths of the ligands are obviously different, but not vastly
[11]
Ligands 1–8 are all capable of multicomponent assembly
3
+
[14]
[11]
6À
with Ln ions. The assembly properties of ligands 5, 6,
and 9a
so. Modeling indicates that the M–M distances in [1 Y ] ,
[6 Y ] , and [5 Y ]
3
2
[11]
6À
12À
4
are known and the complexes were previously
are 14.41 ꢀ, 16.44 ꢀ, and 16.61 ꢀ,
3
2
6
characterized. Upon treatment with a strong base (e.g., NaH)
and variable Ln salts, three salicylhydrazone motifs form
respectively.
3
+
As expected, only single species were observed in the
H NMR spectra of the various combinations of 1, 5, 6, and
1
a complex with a single metal ion. The V-shaped ligand 6
6
À
1
6À
2
forms an M L helicate complex (e.g. [6 La ] ), while the
Y(OTf) . For example, H NMR analysis of the [(1 + 6) Y ]
2
3
3
2
3
3
rigid, linear ligand 5 is incapable of this and an M L assembly
mixture only showed resonances corresponding to those of
4
6
[
14]
6À
6À
is formed in this case.
Obviously, ligand 9a can only
[1 Y ] and [6 Y ] , and this selectivity was also observed for
3 2 3
2
3
À
complex one metal ion and forms an [ML3] complex.
Ligands 1–4, 7, and 8 show a slightly V-shaped ligand
geometry and M L assembly would be expected, consistent
the other combinations (Figure 2). Each mixture contained
signals for the homocomplexes alone. While some resonances
overlap (most notably from the protons in ortho position of
the salicyl groups), the NMR spectra were quite distinguish-
able, and no resonance broadening or extra resonances for
heterocomplexes were observed. The selectivity appeared to
be both kinetically and thermodynamically favorable. The
2
3
with the vast majority of self-assembled rare-earth-metal
[
12]
complexes.
The initial tests of the sorting properties were the
simplest, and involved geometrically distinct ligands 1, 5,
2
www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
homocomplexes were formed immediately and no equilibra-
tion to other species occurred. When the samples were heated
for up to 5 days at 608C, no other products were observed. It
has been previously shown that stirring for 6–12 h at ambient
temperature gave complete exchange between different
[11]
lanthanide ions in M L3 complexes of 6.
Additionally,
2
ESI-MS analysis in positive mode confirmed the stoichiom-
+
etry of the process. Only M ions of the M L complexes were
2
3
observed for ligands 1–4 and 7 (see the Supporting Informa-
+
tion for spectra). The observed M ions were assigned as the
2
+
[
M L -4H] ion in each case. Other fragments were observed,
2 3
+
2
corresponding to the free ligand and [M L -4H] species. MS/
MS analysis of the [M L -4H] ion showed these ions
2
2
2
+
2
3
originated from fragmentations of the M L complex.
2
3
Ligands 1, 5, and 6 are quite different in both length and
geometry, so this selectivity is encouraging, but not surprising.
A more challenging test is the combination of ligands 1 and 7/
8. The fluorene cores are rigid, and so the introduction of
additional phenyl groups between the core and the salicylhy-
drazone coordinator confers a change in the length, but not in
6
À
the angle (the M–M distance in [8 Y ] is 23.08 ꢀ, by
3
2
modeling). As can be seen in Figure 2h (and in the Support-
ing Information), combination of the short and long fluore-
none ligands 1 and 8 with Y(OTf) led to the observation of
3
1
two sets of signals in the H NMR spectrum, corresponding to
6
À
6À
self-sorted homocomplexes [1 Y ]
and [8 Y ] . Even
3
2
3
2
though the ligands vary only in length, no mixing is observed.
These results are also encouraging, but hardly unprece-
dented, as they corroborate results seen with more flexible
[
6]
bis(catecholate) and bipyridyl ligands. The most exciting
possibility would be to achieve sorting between ligands of
identical coordination angle, length, and coordinating motif.
The fluorenone cores provided this. By simple alteration of
the central carbonyl group, very subtle changes can be made
to the ligands without appreciably altering the coordination
geometry or the length. Fluorenone-derived ligands 1–4 are of
the same size and shape and contain the same coordination
motif, and they differ only in their endohedral functionality.
The sorting experiments were performed as before. Ligands
Figure 3. Narcissistic self-sorting between ligands of identical angle
and length, differing only in endohedral functionality: a) cartoon of
6À
6À
6À
6À
narcissistic self-sorting; b) [1 Y ] ; c) [2 Y ] ; d) [3 Y ] ; e) [4 Y ]
;
3
2
3
2
3
2
3
2
f) 1+2+Y(OTf) ; g) 1+3+Y(OTf) ; h) 2+3+Y(OTf) ; i) 3+4+Y-
3
3
3
(
OTf) (400 MHz, [D ]DMSO, 298 K). Minimized structures of
3
6
6
À
6À
2
j) [1 Y ] ; k) [7 Y ] ; l) space-filling representation of the central core
3
2
3
6
À
of [1 Y ] . (SPARTAN, AM1 force field).
3 2
1
–4 were treated with sodium hydride, isolated and combined
the internal groups has no effect on the self-assembly. Even
though benzyl oxime 4 is significantly larger than the other
internal groups, the bulk of the benzyl group can be oriented
with Y(OTf) , and the results are shown in Figure 3.
Remarkably, complete narcissistic self-sorting was observed
in all cases when mixing these highly similar ligands.
Complete discrimination was observed immediately after
the addition of the metal ions to the mixture of the ligands.
Spectra were acquired periodically, and no evidence of mixed
complexes was observed, even after several days. Figure 3
3
[9b]
away from the central cavity, thus allowing self-assembly.
1
Of particular note in the H NMR spectra are the singlets near
9.0 and 5.5 ppm (Figure 3d,e,i). If mixed complexes were
[9]
forming, multiplicity would be observed in these resonances.
1
shows the H NMR spectra of various combinations of
endohedrally derivatized fluorenone ligands 1–4. As with
the other mixing experiments, only single species were
1
observed in the H NMR spectra, and can be identified by
comparison with the spectra of the individual complexes
(
Figure 3b–e). While ligands 1–4 have the same length, shape,
and coordination motif, there are size differences in the
endohedral functionality. Ligands 1 and 2 are symmetrical,
and display simple spectra upon complexation. The oxime
ligands 3 and 4 are desymmetrized and display more complex
spectral patterns, but the species are obviously assignable and
present in the assembled complexes (Figure 4g–i). The size of
Figure 4. Heteroleptic complex formation with singly coordinating
3
À
3À
3À
control ligands: a) [9a Y] ; b) [9b Y] ; c) [(9a) (9b) Y] (400 MHz,
3
3
x
y
[D ]DMSO, 298 K).
6
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
While there is a large difference between 2/4 and 3/4 with
regard to the size of the internal functional groups, the sorting
selectivity is remarkable. Similar variation in internal steric
hindrance in self-assembled Pd–pyridine and Fe–iminopyr-
idine complexes shows little control of sorting behavior, and
no narcissistic sorting whatsoever. Most interestingly, sort-
ing was also observed with combinations of ligands that are
much more similar in size, such as fluorenone 1 and oxime 3
This level of self-sorting is extremely surprising, and the
explanation is unclear. An analysis of the minimized molec-
ular structures of the complexes does shed some light on the
process. The almost linear nature of the fluorene-based cores
forms an M L assembly with virtually no internal cavity at all.
2
3
[
9]
Figure 3l shows a space-filling model of the central core of
6
À
[1 Y ] , which shows the close-packed nature of the C=O
3
2
groups. The slight kink in the ligand ensures that the pendant
functions cannot orient “outward”, and so the central groups
are forced into close proximity upon assembly. The assembly
is therefore sensitive to the steric environment inside the
cavity, and close packing of functional groups appears to
disfavor mismatches. The interaction is such that even small
changes in the size of internal groups are “felt” by the other
ligands in the assembly, promoting the narcissistic sorting that
is seen. The control ligands 9a–c cannot exploit this effect,
and so heteroleptic complexes are formed that show minimal
narcissistic self-sorting.
(
Figure 3g). The most striking result is the combination of
fluorenone 1 and fluorene 2 (Figure 3 f), with which narcis-
sistic self-sorting is observed in ligands that differ in the
addition of only one atom. As with the previously described
experiments, the samples from Figure 3 were exposed to
[
11]
equilibrating conditions in an attempt to induce exchange,
1
but no change in the composition was observed by H NMR
spectroscopy. MS analysis was consistent with these results.
No ions for heterocomplexes were observed when the
solutions from mixing experiments were subjected to ESI
analysis. Unfortunately, the low ionization efficiency of the
complexes led to preferential ionization of single species in
the mixture, and accurate MS analysis of the mixtures was not
possible, even under a variety of ionization conditions.
The extended fluorene-based ligands 7 and 8 were also
tested for their sorting behavior, and the results were identical
to the shorter fluorene-based ligands 1–4 (see the Supporting
In conclusion, we have shown that highly selective
narcissistic self-sorting is possible in the formation of rigid
M L3 rare-earth helicates with identical salicylhydrazone
2
coordinating motifs. Variations in ligand length and coordi-
nation geometry are easily discriminated, as might be
expected. However, when ligands derived from fluorenone
cores are applied, complete narcissistic self-sorting between
species that vary only by one atom, which is not involved in
the coordination process, is possible. The sorting process is
a supramolecular effect, that is, the application of singly
coordinating control ligands leads to heteroleptic complexes
with little discrimination. The ability of self-assembled
complexes to show self-sorting between ligands that are so
similar in structure is reminiscent of the selectivity observed
in nature. Further investigations of this phenomenon in the
formation of selective photoactive lanthanide assemblies are
underway in our laboratory.
4
À
Information for spectra). Combination of deprotonated 7
4
À
and 8 with Y(OTf) gave rise to only self-sorted homocom-
plexes [7 Y ] and [8 Y ] , with no new signals representing
6
À
6À
2
3
2
3
the formation of mixed complexes. These larger assemblies
were substantially less soluble than those from 1–4, and far
more challenging to characterize by MS methods, so we
limited our study to the 7/8 combination alone.
One obvious question is whether this is a supramolecular
effect, or a quirk of the coordinating ligand itself. To test this,
we applied control ligands 9a–c, which contain single
salicylhydrazone coordinating motifs. Ligands 9a–c have
very slight electronic differences because of the substituent
Received: January 15, 2015
Revised: February 23, 2015
Published online: && &&, &&&&
in para position of the aryl ring (Br, OMe, and NO for 9a, b,
2
and c respectively) and can be easily distinguished in the
1
H NMR spectra. Figure 4 shows the results for (9a) (9b) Y.
x
y
2
À
2À
Keywords: coordination modes · helical structures ·
Upon immediate addition of 9a and 9b to a solution of
.
lanthanides · self-assembly · supramolecular chemistry
Y(OTf) in [D ]DMSO, a single species was initially observed,
3
6
but it was quickly replaced by a mixture of different species.
Complete formation of heteroleptic complexes was ensured
by heating at 608C overnight. As can be seen in Figure 4,
while resonances that correspond to homoleptic complexes
[1] a) S. Ghosh, A. Wu, J. C. Fettinger, P. Y. Zavalij, L. Isaacs, J. Org.
Chem. 2008, 73, 5915; b) Q. Shi, K.-E. Bergquist, R. Huo, J. Li,
M. Lund, R. Vꢁcha, A. Sundin, E. Butkus, E. Orentas, K.
Wꢂrnmark, J. Am. Chem. Soc. 2013, 135, 15263.
3
À
3À
[
(9b) Y] and [(9a) Y] are observed, a statistical mixture
3
3
1
of species is present. The new resonances in the H NMR
spectrum (labeled with triangles) do not correspond to either
of the component complexes, and can be assigned as the
[
2] a) A. Wu, L. Isaacs, J. Am. Chem. Soc. 2003, 125, 4831; b) K.
Acharyya, S. Mukherjee, P. S. Mukherjee, J. Am. Chem. Soc.
2013, 135, 554; c) A. S. Singh, S.-S. Sun, Chem. Commun. 2012,
3
À
3À
[
(9a) (9b) Y] or [(9a) (9b) Y] heterocomplexes. ESI-MS
2 1 1 2
48, 7392; d) A. Jimꢃnez, R. A. Bilbeisi, T. A. Ronson, S. Zarra,
C. Woodhead, J. R. Nitschke, Angew. Chem. Int. Ed. 2014, 53,
4556; Angew. Chem. 2014, 126, 4644; e) W. Wang, Y. Zhang, B.
Sun, L.-J. Chen, X.-D. Xu, M. Wang, X. Li, Y. Yu, W. Jiang, H.-B.
Yang, Chem. Sci. 2014, 5, 4554; f) J.-F. Ayme, J. E. Beves, C. J.
Campbell, D. A. Leigh, Angew. Chem. Int. Ed. 2014, 53, 7823;
Angew. Chem. 2014, 126, 7957.
analysis confirmed the ML stoichiometry of this process, and
3
the presence of heteroleptic complexes. The same results
were obtained when the other control ligands were mixed (see
the Supporting Information for other spectra). Small varia-
tions in donor ability of singly coordinating ligands are not
sufficient to allow self-sorting, and it can be deduced that the
[
3] a) M. Yamanaka, Y. Yamada, Y. Sei, K. Yamaguchi, K.
Kobayashi, J. Am. Chem. Soc. 2006, 128, 1531; b) P. Mukhopad-
hyay, A. Wu, L. Isaacs, J. Org. Chem. 2004, 69, 6157; c) Q.-F. Sun,
supramolecular nature of the M L assembly is essential for
2
3
successful sorting.
4
www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
S. Sato, M. Fujita, Angew. Chem. Int. Ed. 2014, 53, 13510; Angew.
Chem. 2014, 126, 13728; d) M. Lal Saha, M. Schmittel, J. Am.
Chem. Soc. 2013, 135, 17743; e) M. M. J. Smulders, A. Jimꢃnez,
J. R. Nitschke, Angew. Chem. Int. Ed. 2012, 51, 6681; Angew.
Chem. 2012, 124, 6785; f) M. L. Saha, S. Pramanik, M. Schmittel,
Chem. Commun. 2012, 48, 9459; g) L.-P. Cao, J.-G. Wang, J.-Y.
Ding, A.-X. Wu, L. Isaacs, Chem. Commun. 2011, 47, 8548;
h) M. L. Pellizzaro, K. A. Houton, A. J. Wilson, Chem. Sci. 2013,
[8] a) C. Gꢄtz, R. Hovorka, G. Schnakenburg, A. Lꢄtzen, Chem.
Eur. J. 2013, 19, 10890; b) S. W. Sisco, J. S. Moore, Chem. Sci.
2014, 5, 81; c) T. Kamada, N. Aratani, T. Ikeda, N. Shibata, Y.
Higuchi, A. Wakamiya, S. Yamaguchi, K. S. Kim, Z. S. Yoon, K.
Dongho, A. Osuka, J. Am. Chem. Soc. 2006, 128, 7670; d) O.
Gidron, M.-O. Ebert, N. Trapp, F. Diederich, Angew. Chem. Int.
Ed. 2014, 53, 13614; Angew. Chem. 2014, 126, 13833.
[9] a) A. M. Johnson, R. J. Hooley, Inorg. Chem. 2011, 50, 4671;
b) M. C. Young, A. M. Johnson, A. S. Gamboa, R. J. Hooley,
Chem. Commun. 2013, 49, 1627.
4, 1825.
[
4] a) M. M. Safont-Sempere, G. Fernꢁndez, F. Wꢄrther, Chem. Rev.
2
2
2
011, 111, 5784; b) M. D. Ward, P. R. Raithby, Chem. Soc. Rev.
013, 42, 1619; c) M. L. Saha, M. Schmittel, Org. Biomol. Chem.
012, 10, 4651; d) M. L. Saha, S. De, S. Pramanik, M. Schmittel,
[10] a) A. M. Johnson, O. Moshe, A. S. Gamboa, B. W. Langloss,
J. F. K. Limtiaco, C. K. Larive, R. J. Hooley, Inorg. Chem. 2011,
50, 9430; b) M. Frank, L. Krause, R. Herbst-Irmer, D. Stalke,
G. H. Clever, Dalton Trans. 2014, 43, 4587; c) R. A. Bilbeisi, J. K.
Clegg, R. Elgrishi, X. de Hatten, M. Devillard, B. Breiner, P.
Mal, J. R. Nitschke, J. Am. Chem. Soc. 2012, 134, 5110.
[11] A. M. Johnson, M. C. Young, X. Zhang, R. R. Julian, R. J.
Hooley, J. Am. Chem. Soc. 2013, 135, 17723.
Chem. Soc. Rev. 2013, 42, 6860; e) D. Schultz, J. R. Nitschke,
Proc. Natl. Acad. Sci. USA 2005, 102, 11191; f) V. E. Campbell,
X. de Hatten, N. Delsuc, B. Kauffmann, I. Huc, J. R. Nitschke,
Nat. Chem. 2010, 2, 684; g) J. R. Nitschke, J.-M. Lehn, Proc. Natl.
Acad. Sci. USA 2003, 100, 11970.
[
[
[
5] a) D. Ajami, J.-L. Hou, T. J. Dale, E.; Barrett, J. Rebek, Jr., Proc.
Natl. Acad. Sci. USA 2009, 106, 10430; Barrett, J. Rebek, Jr.,
Proc. Natl. Acad. Sci. USA 2009, 106, 10430; b) A. Lledꢅ, S.
Kamioka, A. C. Sather, J. Rebek, Jr., Angew. Chem. Int. Ed.
[12] a) C. Piguet, G. Bernardinelli, G. Hopfgartners, Chem. Rev. 1997,
97, 2005; b) J. Shi, Y. Hou, W. Chu, X. Shi, H. Gu, B. Wang, Z.
Sun, Inorg. Chem. 2013, 52, 5013; c) P. E. Ryan, L. Guꢃnꢃe, C.
Piguet, Dalton Trans. 2013, 42, 11047.
[13] a) T. B. Jensen, R. Scopelliti, J.-C. G. Bꢄnzli, Dalton Trans. 2008,
1027; b) T. B. Jensen, R. Scopelliti, J.-C. G. Bꢄnzli, Chem. Eur. J.
2007, 13, 8404; c) L. Aboshyan-Sorgho, H. Nozary, A. Aebischer,
J.-C. G. Bꢄnzli, P.-Y. Morgantini, K. R. Kittilstved, J. Am. Chem.
Soc. 2012, 134, 12675; d) M. Albrecht, O. Osetska, J.-C. G.
Bꢄnzli, F. Gumy, R. Frçhlich, Chem. Eur. J. 2009, 15, 8791; e) T.
Riis-Johannessen, G. Bernardinelli, Y. Filinchuk, S. Clifford,
N. D. Favera, C. Piguet, Inorg. Chem. 2009, 48, 5512.
[14] a) Y. Liu, Z. Lin, C. He, L. Zhao, C. Duan, Dalton Trans. 2010,
39, 11122; b) X. Zhu, C. He, D. Dong, Y. Liu, C. Duan, Dalton
Trans. 2010, 39, 10051.
[15] T. Schulte, M. Krick, C. I. Asche, S. Freye, G. H. Clever, RSC
Adv. 2014, 4, 29724.
2011, 50, 1299; Angew. Chem. 2011, 123, 1335.
6] a) R. Krꢂmer, J.-M. Lehn, A. Marquis-Rigault, Proc. Natl. Acad.
Sci. USA 1993, 90, 5394; b) D. L. Caulder, K. N. Raymond,
Angew. Chem. Int. Ed. Engl. 1997, 36, 1440; Angew. Chem. 1997,
109, 1508; c) M. Albrecht, M. Schneider, H. Rçttele, Angew.
Chem. Int. Ed. 1999, 38, 557; Angew. Chem. 1999, 111, 512.
7] a) Q.-F. Sun, J. Iwasa, D. Ogawa, Y. Ishido, S. Sato, T. Ozeki, Y.
Sei, K. Yamaguchi, M. Fujita, Science 2010, 328, 1144; b) R.
Pinalli, V. Cristini, V. Sottili, S. Geremia, M. Campagnolo, A.
Caneschi, E. Dalcanale, J. Am. Chem. Soc. 2004, 126, 6516; c) D.
Samanta, P. S. Mukherjee, Chem. Eur. J. 2014, 20, 12483; d) J. J.
Henkelis, J. Fisher, S. L. Warriner, M. J. Hardie, Chem. Eur. J.
2014, 20, 4117.
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
Communications
Self-Assembly
A. M. Johnson, C. A. Wiley, M. C. Young,
X. Zhang, Y. Lyon, R. R. Julian,
R. J. Hooley*
&&&& — &&&&
Narcissistic Self-Sorting in Self-
Assembled Cages of Rare Earth Metals
and Rigid Ligands
Equals among equals: The formation of
self-assembled cages of rare earth metals
and ligands proceeds through highly
selective, narcissistic self-sorting. Pend-
ant functionalities as similar as carbonyl
and methylene groups are discriminated
in this complex assembly process.
6
www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
These are not the final page numbers!