Comparative solvolytic stabilities of copper(II) nanoballs and dinuclear Cu(II) paddle wheel units

Article abstract of DOI:10.1016/j.ica.2010.06.022

The self-assembly of copper(II) ions and 5-(2-(2-hydroxyethoxy)ethoxy) benzene-1,3-dicarboxylic acid (2) leads to hollow nanoballs in which 12 dinuclear copper(II) paddle wheel units are interconnected via 24 ligands, as determined by single crystal X-ray

Full text of DOI:10.1016/j.ica.2010.06.022

Inorganica Chimica Acta 363 (2010) 4220–4229
Contents lists available at ScienceDirect
Inorganica Chimica Acta
journal homepage: www.elsevier.com/locate/ica
Comparative solvolytic stabilities of copper(II) nanoballs and dinuclear Cu(II)
paddle wheel units
Markus Tonigold, Dirk Volkmer *
Ulm University, Institute of Inorganic Chemistry 2, Materials and Catalysis, Albert-Einstein-Allee 11, D-89081 Ulm, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 19 June 2010
The self-assembly of copper(II) ions and 5-(2-(2-hydroxyethoxy)ethoxy)benzene-1,3-dicarboxylic acid
(
2) leads to hollow nanoballs in which 12 dinuclear copper(II) paddle wheel units are interconnected
via 24 ligands, as determined by single crystal X-ray structure analysis. The nanoball dissociates in aque-
ous solutions, and in the presence of an excess of ligand it transforms into a three-dimensional network,
but is stable in organic solvents. The thermodynamic stability of the nanoball against dissociation in
aqueous solution is studied and compared to simple copper(II) paddle wheel complexes. The results
reveal enhanced thermodynamic stability of the nanoball as compared to discrete copper(II) paddle
wheel complexes due to chelate effects and positive cooperativity.
Intended as Full Paper for Inorg. Chim. Acta.
Special Issue dedicated to Prof. Dr. H.C. Mult.
Achim Müller on the occasion of his 72nd
birthday.
Keywords:
MOP
Nanoball
Ó 2010 Elsevier B.V. All rights reserved.
Thermodynamic stability
Solvolysis
Chelate effect
Positive cooperativity
1
. Introduction
A broad range of existing and future applications for nanome-
zene-1,3-dicarboxylate (bdc) or its derivatives [4–7]. By changing
the substituent R in the 5-position of bdc, nanoballs which are
highly soluble in polar solvents (e.g. alcohols or N,N-dimethylform-
ter-sized hollow molecules is currently under discussion in litera-
ture: they can serve for instance as nanoreactors, which are a
simple mimic of enzymes for a diverse range of thermal and pho-
tochemical reactions [1]. Further present applications for (drug)
delivery, storage, and separation technology [2] involve molecular
recognition, host–guest chemistry and chiral recognition, while
long-term goals are focused on the design of new materials with
desirable properties and the manufacturing of nanoscale devices
and molecular machinery [3,4].
Tracking these aims, metal-organic polyhedra (MOPs, also
named nanoballs, nanocontainers or nanocages) [4–7], have come
to the foreground in the last few years, in addition to the well-
established organic molecular containers [8]. MOPs are built
through the interconnection of metal ions or clusters with simple,
polyfunctional organic ligands (so-called linkers, mostly containing
nitrogen or carboxylate groups as complex builders). One way to
obtain nanoballs in high yield is to interconnect copper(II) paddle
wheel complexes at a 120° angle with the bifunctional ligand ben-
amide (DMF) for R = OH [5a], DMF or dimethyl sulfoxide (DMSO)
for R = OC
romethane or THF for R = OC12
2
H
5
OH [6]), or in nonpolar solvents (e.g. toluene, dichlo-
25 [5d]) are obtained.
H
Especially for the applications mentioned above, knowledge
about kinetic and thermodynamic stability of these molecular con-
tainers is crucial. Studies on the disassembly kinetics of smaller
supramolecular architectures (such as helices [9] and organic
self-assemblies) [10] have been intensively performed, but to date
only a few studies focus on the kinetic stability of larger, nanome-
ter-sized metal-organic polyhedra [6,7]. To the best of our knowl-
edge, no studies concerning the thermodynamic stability of metal-
organic polyhedra or similar compounds have been carried out in
detail yet.
We have already shown earlier that nanoballs based on inter-
connected copper(II) paddle wheels are kinetically far more stable
in solvents like DMF or DMSO than the (kinetically labile) dinuclear
copper(II) paddle wheel complexes such as copper acetate or ben-
zoate, which is presumably due to their more rigid structure (as
2 5
shown for R = OC H OH in 5-position of bdc by spectrophotome-
try) [6].
Since the kinetic stability compared to simple paddle wheel
*
Corresponding author. Tel.: +49 (0)731 50 23921; fax: +49 (0)731 50 23039.
E-mail address: dirk.volkmer@uni-ulm.de (D. Volkmer).
units is increased due to the structural features of the nanoball,
0
020-1693/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2010.06.022

M. Tonigold, D. Volkmer / Inorganica Chimica Acta 363 (2010) 4220–4229
4221
O
O
OH
O
OH
O
OH
a
b
MeO C
CO Me
MeO C
CO Me
HO C
CO H
2
2
2
2
2
2
1
2
O
O
OH
c
24
+ 24 Cu(NO ) *3 H O
[Cu24(2)24]•S
3
2
2
HO C
CO H
2
2
2
3
2 3 3 2
Scheme 1. Synthetic route for linker 2 and nanoball 3. a: KI, K CO , CH CN, reflux, 24 h; b: (1) KOH, EtOH/H O (1:1), reflux, 6 h (2) HCl; c: 2 equivalents 2,6-dimethylpyridine,
methanol, room temperature, 1 min. S: solvent molecules.
the question about the influence on the thermodynamic stability
arises. Dinuclear copper(II) carboxylates (to which we refer as a
reference standard) are remarkably dissociated in aqueous solution
2. Results and discussion
The linker 5-(2-(2-hydroxyethoxy)ethoxy)benzene-1,3-dicar-
boxylic acid 2 was synthesized from dimethyl-5-hydroxyben-
zene-1,3-dicarboxylate and 2-(2-chloroethoxy)ethanol to give the
intermediate 1, which was subsequently saponified and afterwards
protonated to yield 2 in an overall yield of 77% (Scheme 1).
From an equimolar solution of 2 and Cu(NO ) ꢀ3H O in metha-
[
11], whereas the dissociation is strongly decreased for less polar
solvents [12]. For hydroxylated nanoballs (R = OH in 5-position of
bdc) in methanol solution, spectroscopic studies even showed no
detectable dissociation at all [13], qualitatively showing the in-
creased thermodynamic stability above the underlying dinuclear
copper(II) carboxylates. Since the thermodynamic stability of these
nanoballs seems to be most interesting for aqueous solutions, we
thus prepared a new nanoball based on a bis(ethylene glycol)-func-
tionalized linker which is soluble in H O:DMF (1:1) to quantify the
2
thermodynamic stability, as existing functionalised nanoballs [5,6]
in our hands showed poor solubility in aqueous solvents.
3
2
2
2
4
12 12 7 24
nol, the nanoball [Cu (C H
O ) ]ꢀ24MeOH 3 immediately pre-
1
cipitates upon adding two equivalents of 2,6-dimethylpyridine.
Recrystallisation from dimethylsulfoxide (DMSO) and butanol affor-
ded 3b as green crystals suitable for single crystal X-ray diffraction
studies.
To evaluate the dissociation stability of 3, it is necessary to facil-
itate the dissociation of the complexes by adding an acid, since the
dissociation of the nanoball is kinetically hindered. For this pur-
pose, it might seem reasonable to add the diprotonated ligand 2
to a solution of the nanoball 3. Consequently, a closed spherical
structure starts to compete with open polymeric structures if the
ratio in solution after dissociation mismatches the stochiometry
of the spherical structure: compared to open polymeric structures,
an increased number of bonds between the constituents is possible
in a cyclic or spherical arrangement, and will be favoured from the
enthalpic point of view (since the donor and acceptor sites at each
end of the polymer will always remain uncoordinated). However,
when the ligand-to-metal ratio in solution mismatches the stochi-
ometry of the cyclic or spherical supramolecular assembly, the
enthalpic advantage diminishes due to the present excess of donor
or acceptor sites [1,20–22]. Accordingly, when the nanoball 3 is
Whether the thermodynamic stability is increased, or even de-
creased, depends on the cooperativity of the system. Cooperativity
[
14] arises from the interplay of two or more interactions, so that
the system as a whole behaves differently from expectations based
on the properties of the individual interactions acting in isolation.
Coupling of interactions can lead to positive or negative coopera-
tivity, depending on whether one interaction favours or disfavours
another. Positive cooperativity implies a low concentration of
intermediates. Positive cooperativity leads to a low peak concen-
tration of intermediates and a sharp transition from unbound to
bound. In other words, as the system approaches the limit of strong
positive cooperativity, only the extreme states are significantly
populated. Cooperativity exhibited in the folding of proteins and
supramolecular self-assemblies is driven by the difference in
strength between the intermolecular and intramolecular interac-
tions, and is a consequence of the molecular architecture. This phe-
nomenon gives rise to the chelate effect [15], and is thus also called
chelate cooperativity.
dissolved in H O:DMF (1:1) together with three equivalents of
2
the ligand 2 per copper atom, the nanoball slowly transforms
throughout several weeks into single crystals of the three-dimen-
Already small organic assemblies [16], inorganic cylindrical
architectures [17] or double helicates [9a–c] show increased ther-
modynamic stabilities due to positive cooperative effects, and an
even more pronounced influence on the nanoball built from 48
subunits might be expected.
sional network [Cu(C12H12O )]ꢀ0.5 H Oꢀ0.5 DMF 4. This course from
7
2
least stable to more stable (as well as less dense to more dense)
polymorph of Cu(C12H12O ) follows the classical Ostwald–Volmer
7
rule [23]. Only slow dissociation and no transformation is observed
after the addition of acetic acid, emphasising the strong influence
of the ligand-to-metal ratio for the stability of the nanoball over
open polymeric structures. Therefore, the dissociation constant of
3 can be determined after the addition of acetic acid (see below).
Noteworthly, no transformation of 3 into a three-dimensional
network is observed for solutions in anhydrous DMF or DMSO.
These solutions have remained unchanged for up till now more
than 12 months – even when up to 500 equivalents of the diacid
ligand 2 per copper atom are present in these solutions. These re-
sults show, in analogy to the strongly increased stability against
Furthermore, the formation of insoluble coordination poly-
mers is always given as a possible concurrence to the dissocia-
tion of a supramolecular self-assembly into its constituents: by
linking copper(II) paddle wheel complexes with functionalised
benzene-1,3-dicarboxylate (bdc), two- or three-dimensional net-
works [18,19] might be obtained instead of nanoballs. Which of
these competing products occurs depends on the formation of
kinetic products, the concentrations of the educts, the ligand-
to-metal ratio, pH, solvent effects, template molecules which di-
rect the course of the assembly, and temperature [1,20–22].
Among these, the concentrations and the ligand-to-metal ratio
are the most crucial for the self-assembly of cyclic or spherical
self-assemblies.
1
The exact number and kind of coordinated solvent molecules is difficult to
determine for this class of substances [4a] and was chosen according to elemental and
thermogravimetric analyses.

4
222
M. Tonigold, D. Volkmer / Inorganica Chimica Acta 363 (2010) 4220–4229
dissociation of simple dinuclear copper(II) paddle wheel com-
plexes, that the nanoball 3 seems to be virtually undissociated in
organic solvents (even at high concentrations of H L). The structure
2
remains intact and no transformation is observed even at very high
ligand-to-metal ratios (500:1).
For 3, thermogravimetric analysis (TGA) shows a negligible
weight loss of 5.1% in the temperature range of 30–269 °C (ex-
pected: 8.8% for complete loss of methanol). Further weight loss
of 67.4% occurs from 269 to 400 °C which is ascribed to the decom-
position of the nanoball 3. These results indicate a thermodynamic
stability against decomposition similar to related compounds [5b]
up to at least 269 °C. For 4, a weight loss of 11.6% is observed in the
temperature range of 30–260 °C (expected: 13.6% for complete loss
of DMF and water). Further weight loss of 66% occurs from 260 to
6
00 °C due to decomposition.
2.1. X-ray structure analyses
To avoid rapid loss of occluded solvent molecules, crystals of 3b
and 4 were directly covered with mineral oil after removing them
from the mother liquor. Crystals of compound 3b were found to be
completely stable if stored in the mother liquor. However, upon
exposing them to air, crystallinity is lost within minutes, leading
to an uncertain number of occluded solvent molecules, which is
common for metal-organic structures with open porosity [4a,5b].
Whereas problems of crystal aging could be minimised experimen-
tally, some unavoidable problems in refining the structure of 3b ar-
ose from the fact that the solvent molecules occluded in the crystal
lattice as well as the 2-(2-hydroxyethoxy)ethoxy substituent of the
linker are largely disordered. These disordered groups had to be re-
fined as rigid fragments using restraints with appropriate atomic
coordinates taken from literature [24]. The solvent atoms coordi-
nated to the copper atoms could be refined as DMSO, water and
methanol. Due to their high disorder we were not able to deter-
mine the exact amount and identity of the solvent molecules in-
Fig. 1. (a) Ball-and-stick representation of the X-ray single-crystal structure of 3b.
Cu, blue; O, red; C, grey. Hydrogen atoms and coordinated and guest solvent
molecules are omitted for clarity. The yellow sphere represents the internal cavity.
(
b) Schematic representation of 3b as a cuboctahedron (showing the positions of
the triangular and square apertures) comprised from (c) Cu (CO ) paddle-wheel
2
2 4
cluded in the voids of 3b, thus PLATON
/
SQUEEZE [25] was used to
(squares) linked at an 120° angle by 2. (For interpretation of the references to colour
in this figure legend, the reader is referred to the web version of this article.)
correct the data for the presence of the disordered solvent. A po-
tential solvent volume of 22 437 Å3 containing 7848 electrons
ꢁ3
was found, corresponding to an electron density of 0.350 e Å (ex-
ꢁ3
ꢁ3
pected for n-butanol: 0.489 e Å , for DMSO: 0.664 e Å calcu-
groups to the copper atoms of the neighboured layer, thus inter-
connecting the two-dimensional grids to a three-dimensional net-
work (Fig. 3). Two different kind of windows are alternatingly
present: one in which all linker molecules are facing the same
direction, and another where only the pairs of opposing ligands
are facing the same direction (and the other pair the opposite
direction), a MOF structural motif which is already known in liter-
ature [18].
Compared to the two-dimensional layers described in literature
to date, which are not covalently connected with each other, the
extensive 2-(2-hydroxyethoxy)ethoxy side chains in 4 are long en-
ough to coordinate with their hydroxy groups to the copper atoms
of the neighboured layer, thus interconnecting the two-dimen-
sional grids to a three-dimensional network (Fig. 3).
Only half of the side chains are coordinated to the copper atoms
of the next layer: the side chains of the ligands which connect the
copper atoms parallel to the a-axis bind to the next layer, whereas
the side chains of the ligands which connect the copper atoms par-
allel to the b-axis protrude into the voids of the crystal. Torsional
strain that has to be overcome seems to be the reason that not
all side chains are coordinated. The ligands which bridge the pad-
dle wheel units parallel to the a-axis have torsion angles between
the carboxylate groups and the benzene ring of 14.6° and 18.9°,
whereas the ones bridging parallel to the b-axis have torsion angles
of only 4.1° and 6.9°. As a result, the ligands bridging along the a-
axis are more bent towards the c-axis (see Fig. 2), and are therefore
in enough short distance to coordinate to the next layer (the dis-
lated from the molar mass and density at room temperature),
whose contribution was removed by the PLATON
/SQUEEZE program
[
25]. Beside these problems, the general connectivity of the rigid
part of nanoball 3b is proven, and the structure is further con-
firmed by SAXS measurements in solution (see below).
3
b has an outer dimension (measured from opposite groups at
the periphery) of 3.2 nm and consists of 12 Cu paddle wheel units
interconnected via 24 linkers (Fig. 1a). The internal cavity has a
diameter of 1.6 nm without terminal ligands, and the 12 paddle
wheel units form a cuboctahedron (Fig. 1b) composed of linked
vertices of squares (Fig. 1c).
Neighboured paddle wheel units in 3b are 1.0 nm apart from
each other (measured from the centres of the dinuclear units)
and represent the vertices of triangular and square apertures.
These openings have widths of about 0.6 and 1.2 nm, respectively
(
height of the triangular and diagonal of the square aperture, tak-
ing into account the Van-der-Waals radii) [26]. The dimensions
of the windows as well as the distances and angles within 3b are
statistically identical to those observed in structurally related
nanoballs featuring a {Cu24(bdc)24} core [5,6].
The coordination polymer 4 is formed solely from square win-
dows (Fig. 2) by bridging the paddle wheel units via the carboxyl-
ate groups into two-dimensional layers along the a- and b-axis.
These layers are piled along the c-axis, providing a one-dimen-
sional pore system. Half of the 2-(2-hydroxyethoxy)ethoxy side
chains in one two-dimensional layer coordinate with their hydroxy

M. Tonigold, D. Volkmer / Inorganica Chimica Acta 363 (2010) 4220–4229
4223
Fig. 4. Main plot: SAXS intensity (I) vs. momentum transfer (q = 0.2–3.2 nm) for a
solution of 3 in DMF (18 g/L). The symbols and the solid line correspond to the
experimental data points and the numerical fit using GNOM
/DAMMIN simulated
annealing ( = 2.941), respectively. Inset: spacefilling model of 3 and bead model
v
Fig. 2. (a) View along the c-axis of a ball-and-stick representation of the X-ray
single-crystal structure of 4. Cu, blue; O, red; C, grey. Hydrogen atoms, coordinated
and guest solvent molecules and bisethyleneglycol side chains are omitted for
clarity. (For interpretation of the references to colour in this figure legend, the
reader is referred to the web version of this article.)
from low-resolution particle shape reconstruction for 3 obtained by the GNOM/
DAMMIN fit.
2.2. Small-angle X-ray scattering (SAXS)
SAXS measurements, which provide the shape and internal
structure of particles in solutions [27], were conducted to confirm
that the structural features of 3 persist in solution. The logarithmic
plot of SAXS intensity (I) versus momentum transfer (q) for 3 is
shown in Fig. 4 together with the numerical fit to the SAXS data
evaluating the shape in solution. The representative reconstruction
of the low resolution three-dimensional particle shape was pro-
vided by simulated annealing using the GNOM/DAMMIN software pack-
ages [28,29] and matches the shape and dimensions of 3 well (see
Fig. 4).
g
The radius of gyration (R ) obtained from the SAXS data for 3 is
1
.1 nm, in excellent agreement with the expected shape: due to the
crystal structure analysis, the distances of the interior and exterior
copper ions from the centre of the nanoballs 3 are 0.8 and 1.1 nm,
respectively. The maximal distance from the centre to the periph-
ery is 1.9 nm.
Principally, bead models derived from experimental SAXS data
show an effective spatial resolution of approximately 0.5 nm for
molecules with radii of gyration below 1.5 nm [30].
Using the X-ray single-crystal structure data of 3b, we addition-
ally calculated its hypothetical scattering intensities with the pro-
gram CRYSOL [31], which matches the observed experimental
intensities well (Fig. S2, Supplementary Material).
2.3. Thermodynamic stability
Fig. 3. (a) View along the b-axis of a ball-and-stick representation of the X-ray
single-crystal structure of 4. Cu, blue; O, red; C, grey. Hydrogen atoms and
coordinated and guest solvent molecules are omitted for clarity. (For interpretation
of the references to colour in this figure legend, the reader is referred to the web
version of this article.)
To compare the thermodynamic stability of the nanoball 3 with
simple dinuclear copper(II) paddle wheel complexes, dilution
experiments were performed, measuring the absorbances for dif-
ferent initial concentrations of the complexes. The ionic strength
3
was fixed at 0.5 M using KNO . Lower initial concentrations (Cutot)
lead to an increased dissociation, which results in decreased molar
absorbances (Fig. 5), since the typical band I of copper(II) paddle
wheel complexes [32] (maximum absorption at 710 and 730 nm
tance between the oxygen atom in the 5-position of bdc to the
closest copper atom in the next layer is 6.7 Å for the bent ligands,
whereas it is 8.0 Å for the ones with uncoordinated side chains).
2 2
in H O and DMF:H O 1:1, respectively) disappears to yield isolated
and fully solvated copper(II) ions (maximum absorption at 800 and

4
224
M. Tonigold, D. Volkmer / Inorganica Chimica Acta 363 (2010) 4220–4229
To facilitate the dissociation of the complexes as described in
the original literature [11a] for copper(II) acetate in water,
.140 mol/L acetic acid or 4-hydroxybenzoic acid, respectively,
were added prior to the measurements. The obtained course of mo-
lar absorbances (ꢀ
, calculated for Cutot, the total concentration of
0
e
copper atoms in solution) versus total copper concentrations Cutot
was evaluated at 770 (Fig. 5) and 680 nm according to Eq. (1) (der-
ivation as described in Supplementary Material) [11a],
eCuðO2CRÞ2 ½CuðO
2
CRÞ ꢂ þ
e
CRÞþ
Cutot
½CuðO
2
CRÞ ꢂ þ eCu2þ ½Cu2þ
þ
ꢂ
2
CuðO
2
ꢀ
e
¼
ð1Þ
2 2
simplifying the dinuclear complexes as [Cu(O CR) ], giving
CRÞ ꢂ þ ½Cu2þꢂ
þ
ð2Þ
Cutot ¼ ½CuðO
2
CRÞ ꢂ ¼ ½CuðO
2
CRÞ ꢂ þ ½CuðO
2
2
0
2
2 2 2 2 4
The formulation Cu(O CR) (and not Cu (O CR) ) was chosen
according to literature, since the experimental data can thus be
fit to a much simpler model (containing only two dissociation steps
and three compounds, see Supplementary Material for more de-
tails). The fact that this simpler model is able to completely de-
scribe the experimental course of absorbances justifies the
avoidance of a more complicated model. Furthermore, only the
extinction coefficients for the dincluear complexes and Cu2+ are di-
rectly available experimentally, whereas the extinction coefficients
of the intermediate species have to be determined during the data
fitting. Therefore, the extinction coefficients for the additional
intermediate species and the respective equilibria constants
Fig. 5. Apparent molar absorbances at 770 nm for copper(II) acetate (d), copper(II)
4
Cutot
-hydroxybenzoate (N) and 3 (j) in H
2
O:DMF 1:1 vs. the total copper concentration
.
+
Cu2+ + 2 (O CR)
–
Cu(O CR)
Cu(O CR) + (O CR)–
2
2
2
2
2
1
:1
2:1
(
which would be introduced using a more complex model for the
R = CH or C H OH
•
•
3 6 5
dissociation of [Cu (O CR) ]) would not be determined as accu-
2
2
4
+
Cu(O CR)
Cu(O CR) + (O CR)–
K1
K2
rately as the constants from the model described in Eq. (1) via
spectrophotometric measurements. As described [11a], the molar
absorbance eCu2þ in Eq. (1) is equal to the molar absorbance coeffi-
cient of copper(II), whereas the molar absorption of the copper(II)
2
2
+
2
2
Cu(O CR)
Cu2+ + (O CR)–
2
2
Scheme 2. Two-step dissociation equilibria for simple copper(II) carboxylates.
carboxylates (
eCuðO2CRÞ2 ) was obtained by adding sodium acetate or
8
2 2
15 nm in H O and DMF:H O 1:1, respectively). At very low con-
sodium 4-hydroxybenzoate to shift the equilibrium towards
centrations, the molar absorbances approach the values for cop-
per(II) ions in solution, whereas at high concentrations, they
approximate the values of the dinuclear complexes or the nanoball,
respectively.
Already these courses of the molar absorbances versus the ini-
tial concentrations (Cutot) of the complexes reveal an obviously dif-
ferent dissociation behaviour for the nanoball 3 compared to the
simple dinuclear paddle wheel complexes. The nanoball shows a
much steeper increase in absorbance (equivalent to a sharp free-
to-bound transition over a narrow concentration range), and fur-
thermore a sigmoidal (S-shaped) course of absorbances in contrary
to the dinuclear complexes.
Cu(O
and K
2
CR)
2
at high concentrations.
e
CRÞþ
as well as finally K
1
CuðO
2
2
were obtained by fitting the theoretical course of the molar
absorbances to the experimental data at 770 and 680 nm (see Sup-
plementary Material for details). Since 3 as well as copper(II) 4-
hydroxybenzoate show poor solubility in pure water, we examined
the complex dissociation constants K
hydroxybenzoate in H O:DMF 1:1, and for copper(II) acetate both
in water and in H O:DMF 1:1 (Table 1).
Both dissociation constants for copper(II) acetate are approxi-
mately halved when switching from H O to H O:DMF 1:1, which
displays a negligible increase in stability when compared to the
more drastic changes when water is completely replaced by an or-
ganic solvent [12,13]. Compared to copper(II) acetate, copper(II) 4-
hydroxybenzoate shows similar stability against dissociation (with
1 2
and K for copper(II) 4-
2
2
2
2
2.4. Thermodynamic stability of simple paddle wheel complexes in
aqueous solvents
a slightly decreased first dissociation constant K
1
and a slightly in-
creased second dissociation constant K ).
2
To compare the thermodynamic stability of the nanoball 3 with
simple dinuclear copper(II) paddle wheel complexes, we first eval-
uated the dissociation constants of copper(II) acetate and cop-
per(II) 4-hydroxybenzoate (which is more comparable to the
2.5. Thermodynamic stability of nanoballs in aqueous solvents
To investigate the thermodynamic stability of the nanoball,
aromatic dicarboxylic acid 2) via spectrophotometry according to
the procedure described in literature.2
solutions of 3 in H O:DMF 1:1 with different initial concentrations
2
were prepared. To facilitate the dissociation of the complexes, two
In water, copper(II) acetate behaves essentially as a 1:1 electro-
lyte at concentrations greater than about 0.1 M, whereas at lower
concentrations it dissociates as a 2:1 electrolyte, according to the
two equilibria shown in Scheme 2.
Table 1
Stepwise complex dissociation constants (mol Lꢁ1) and overall dissociation constants
(pK = ꢁlog K
1
ꢀK
2
) for the simple dinuclear complexes copper(II) acetate and copper(II)
4
-hydroxybenzoate.
2
R
Solvent
K
K
pK
The dissociation constants
1 2
K and K of copper (II) acetate in water
1
2
þ
ꢁ
CuðOAcÞ2 ꢀ CuðOAcÞ þ OAc
K1,
CH
CH
HOC H
6 4
3
H
2
H
2
H
2
O
0.11(3)
0.052(14)
0.023(6)
0.016(4)
0.0085(22)
0.017(4)
2.8(5)
3.4(5)
3.4(5)
þ
2þ
ꢁ
CuðOAcÞ ꢀ Cu þ OAc
K2
3
O:DMF 1:1
O:DMF 1:1
were earlier determined to be ꢁlog K
1
= 0.9–1.0 and ꢁlog K
2
= 1.6–2.4 (temperature
range 15–35 °C; ionic strength range 0–3.0) [11].

M. Tonigold, D. Volkmer / Inorganica Chimica Acta 363 (2010) 4220–4229
4225
equivalents of acetic acid per copper(II) atom in the solution were
added. The absorbances slowly decreased after the addition of ace-
tic acid and reached their final value after 2–3 months (as moni-
tored by spectrophotometry). No effect on the finally obtained
absorbances (and thus the equilibrium concentrations) was ob-
served for different concentrations of acetic acid per copper(II)
atom (until up to four equivalents), ensuring that acetic acid is only
acting as a catalyst for the dissociation. After 6 months, the molar
absorbances for the different initial concentrations of the nanoball
3
were measured. The recorded molar absorbances per copper
atom versus the initial concentrations of 3 were evaluated at 770
Fig. 5) and 680 nm, showing an obviously different course of
(
absorbances compared to simple copper(II) carboxylates.
Below Cutot = 1.7 mmol/L, the molar absorbances which we ob-
tained show only a slight increase with the concentration and
nearly equal the molar absorbances of copper(II). This indicates
that below 1.7 mmol (which can therefore be seen as a lower crit-
ical self-assembly concentration) only copper(II) and presumably
small fragments of the nanoball are present in solution. Above this
concentration, a distinct increase in molar absorbance is observed,
which converges at high concentrations to the molar absorbance of
the nanoball. This behaviour resembles supramolecular micelle
formation and the behaviour at the critical micelle concentration
Fig. 6. Modified Hill plot for the Nanoball 3 in DMF:H
origin with a slope of one and is given as a guide for the eye (displaying the
behaviour of a system without cooperativity). The x-axis was scaled relative to Cu50
the concentration at 50% site occupancy. (For interpretation of the references to
colour in this figure legend, the reader is referred to the web version of this article.)
2
O 1:1. The red line passes the
,
(
cmc) [33] and is in agreement with the theoretical course of con-
centrations for supramolecular assemblies around the critical self-
assembly concentration [22]. In analogy to the dissociation equilib-
ria of the simple dinuclear copper(II) paddle wheel complexes, the
experimental data fit a dissociation model in which the nanoball
fragmentates into simple species (‘‘CuL”), which is further on in
equilibrium with copper(II) ions and free deprotonated ligand 2
in solution (Scheme 3). Although the structure of the nanoball is
quite complicated, and principally there could be many intermedi-
ate products involved, the simple theoretical model is suitable to
fit the experimental data, showing that the concentrations of pos-
sible intermediates are too low to significantly contribute to the fi-
nal absorbances.
ꢁ32.5
mol11/L11 can be
ciation steps). Thus, the value of K = 10
1
thought to be theoretically composed of 11 distinct steps, in each
of which one dinuclear unit is released from the nanoball (Eq. (5)).
Cu12
Cu11
L
12 ꢀ Cu11
11 ꢀ Cu10
L
11 þ CuL
10 þ CuL
K
K
1;1
1;2
L
L
Cu10L10 ꢀ Cu L þ CuL
9
9
K1;3
ð5Þ
.
. .
Cu
2
L
2
ꢀ 2CuL
K
1;11
The obtained course of molar absorbances versus concentra-
tions was evaluated at 770 and 680 nm according to Eq. (3)
Since each of these steps involve the disconnection of at least
one carboxylic group from a copper paddle wheel complex of the
(
regarding the apparent molar absorbances per copper atom),
nanoball, each release step of CuL can be generally compared to
ꢁ
the release of (O
plex. Thus K
comparable to each of the distinct dissociation steps (K1,1, K1,2
2
CR) from a simple, dinuclear paddle wheel com-
¼ 1
2 ꢀ
eCu12L12 ½Cu12
L
12ꢂ þ
e
CuL½CuLꢂ þ eCu2þ ½Cu2þꢂ
ð3Þ
1
from the dissociation of these simple complexes is
ꢀ
e
Cutot
,
and so on) of the nanoball. Simply assuming that all dissociation
constants of these distinct steps are equal provides us an average
thus simplifying the nanoball as [Cu12L12], giving
12ꢂ þ ½CuLꢂ þ ½Cu2þꢂ
value for each distinct dissociation step of
K
1,1 = K1,2 = K1,3
Cutot ¼ 12 ꢀ ½Cu12
L
ð4Þ
pffiffiffiffiffi
1
1
=. . .= K1,11
=
K
1
= 0.0011 mol/L, which is 20–50 times less than
Without the addition of acid, the nanoball 3 is kinetically inert
in solution, thus its molar absorption Cu12L12 was simply obtained
from solutions of 3 in H O:DMF 1:1. CuL as well as finally
pK = 32.5 ± 0.41 and pK = 2.74 ± 0.25 were obtained by fitting
K
1
for the simple dinuclear paddle wheel complexes (0.051 and
e
0.023 mol/L for copper(II) acetate and 4-hydroxybenzoate, respec-
tively). A similar increase in stability is observed when regarding
the K value (0.0018 mol/L for the nanoball, compared to 0.0085
2
2
e
1
2
the theoretical course of the molar absorbances to the experimen-
tal data at 770 and 680 nm (see Supplementary Material). The
overall dissociation constant (dissociation of the nanoball into
and 0.017 mol/L for copper(II) acetate and 4-hydroxybenzoate,
respectively).
To illustrate the increased stability of the nanoball compared to
the comprising dinuclear complexes, we can exemplarily calculate
2
+
2ꢁ
Cu and ligand L ) is pK = p(K
1
ꢀ12K
2
) = 65.4.
2
ꢁ
ꢁ1
A direct comparison of the overall dissociation constant of the
that at a total concentration Cutot = L = 0.055 mol L
of cop-
ꢁ65.4
23 ꢁ23
nanoball (K = 10
wheel complexes (K = 10
the different units of these stability constants. However, a compar-
ison is possible for the stepwise dissociation constants K and K
when K for the nanoball is divided into appropriate equal disso-
mol
L
) with simple dinuclear paddle
per(II) ions present in solution (the sum of all copper atoms pres-
ent in any species), 95% of copper(II) ions can be found in the
nanoball, whereas under corresponding conditions only 51% or
39% of totally present copper(II) ions can be found in binuclear
copper(II) acetate or 4-hydroxybenzoate, respectively, showing
qualitatively the increased stability of the nanoball over simple
binuclear complexes. Moreover, 50% of the copper(II) ions still re-
ꢁ3.4
2
ꢁ2
mol L for both) is difficult due to
1
2
(
1
ꢁ1
Cu L
2 12
12 CuL
K1
K2
main in the nanoball at a concentration of Cutot = 0.0046 mol L ,
1
Cu2+ + L2-
whereas only 8% or 5% of totally present copper(II) ions can be
found in binuclear copper(II) acetate or 4-hydroxybenzoate,
respectively.
CuL
Scheme 3. Two-step dissociation equilibria for the nanoball 3.

4
226
M. Tonigold, D. Volkmer / Inorganica Chimica Acta 363 (2010) 4220–4229
Positive cooperativity [14] is the underlying reason for the in-
However, despite the highly increased overall dissociation con-
stant, the metal-to-ligand ratio plays a crucial role: when (by add-
ing additional diprotonated ligand) the ratio in solution
mismatches the stochiometry of the nanoball, slow transformation
into a three-dimensional network is observed.
In a wider context, these results are also capable of explaining
the thermodynamic stability of the related class of metal-organic
frameworks (MOFs) [35]. Also here, the stability of these three-
dimensional supramolecular architectures should exceed the sta-
bility of the underlying simple complexes from the secondary
building units. In this light, the results give a promising outlook
for the future design of stable and catalytically active MOFs, possi-
bly remaining framework topography and preventing metal leach-
ing even when the simple underlying complexes are less stable.
creased stability of the supramolecular nanoball 3 against disso-
ciation compared to simple dinuclear paddle wheel complexes.
The sigmoidal course, which is already visible from the course
of absorbances (Fig. 5), is also represented in the binding iso-
3
therms of 3 (Supplementary Material), and a modified Hill plot
reveals a value of n
D
ꢃ 2 and furthermore shows all indications
of positive cooperativity for the supramolecular nanoball (Fig. 6).
In contrary, the simple dinuclear copper(II) paddle wheel com-
plexes show no signs of cooperativity (see Supplementary Material
for an extended discussion).
A further simple explanation of the enhanced dissociation sta-
bility of the nanoball 3 in comparison with simple copper(II) pad-
dle wheel complexes is given by the chelate effect (the underlying
driving force for positive cooperativity in this kind of supramolec-
ular assembly). The chelate effect in general is predominantly due
to a gain in translational entropy (whereas the contribution of ring,
conformational or steric strain energies or preorganisation as in
macrocycles or cryptants are usually comparably small) [34]. Since
the base of the chelate effect is predominantly a result of changes
in translational entropy, chelation principally causes a similar de-
crease of overall dissociation constants: A general decrease of pK
by a value of 1.5–4 is observed per exchange of two monodentated
with one comparable bidentate ligand [34]. Applying this general
observation to the case of the nanoball, we have one molecule
which is comprised of 12 dinuclear copper(II) paddle wheel sub-
units (bridged by 24 bifunctional ligand molecules), which has to
be compared with 12 simple dinuclear copper(II) paddle wheel
complexes (possessing 48 monofunctional carboxylic ligands in
total). When comparing the nanoball with the comprising sub-
units, we therefore replace 48 monofunctional ligands with 24
bifunctional ligands, leading to an expected increase of the overall
dissociation constant ꢁlog K by a value between 36 and 96 (equal
to 24 times the effect for exchanging two monodentated with one
comparable bidentate ligand). The overall dissociation constants of
the simple dinuclear paddle wheel complexes is pK = 3.4 for both
4
. Experimental section
4.1. Materials and methods
Melting points (uncorrected): Büchi Melting Point B-540 appa-
ratus. Infrared spectra: Bruker FTIR IFS 113v spectrometer; KBr pel-
lets. UV–Vis spectra: J&M TIDAS. NMR spectroscopy: Bruker DRX
4
00; data given as ppm; spectra referenced to the residual solvent
peak. CI mass spectra: Finnigan MAT, SSQ 7000 mass spectrometer,
single-stage-quadrupol system, intensities relative to base peak.
TGA: Mettler Toledo TGA/SDTA 851, heating rate: 10 °C/min, nitro-
gen flow: 50 mL/min. SAXS: Bruker AXS Nano-STAR equipped with
a HI-STAR area detector. Potentiometric titrations: Schott Instru-
ments ProLab 2000 equipped with an Ag/AgCl glass electrode. Tet-
0
rakis(
l
2
-p-Hydroxybenzoato-O,O )-bis(dimethylsulfoxide-O)-
dicopper dimethylsulfoxide solvate 5 was synthesized according to
literature [36].
DMF was distilled over CaH in vacuum and stored over molec-
2
ular sieve 4 Å prior to use. All other solvents and reagents were of
reagent grades and used as received.
copper(II) acetate and copper(II) 4-hydroxybenzoate in DMF:H
2
O
4
.2. Dimethyl-5-(2-(2-hydroxyethoxy)ethoxy)
1
:1, respectively. Therefore, an overall dissociation constant for
benzene-1,3-dicarboxylic acid (1)
the nanoball between 40 and 100 would be expected. Indeed, the
overall dissociation constant of the nanoball (pK = 65.4) lies in
the expected range for an increased stability due to the classical
chelate effect.
Under argon, a suspension of 10 g (47.6 mmol) dimethyl-5-hy-
droxy-1,3-benzenedicarboxylic acid, 13.4 g (80.9 mmol) potassium
iodide, 7.1 g (57.1 mmol) potassium carbonate and 17.8 g
(
142.8 mmol) 2-(2-chloroethoxy)ethanol in 500 mL acetonitril
was refluxed for 24 h. The solvent was evaporated after filtration
and 1000 mL 1 M HCl were added to the residue. The aqueous
phase was extracted ethyl acetate, and the combined organic
phases were washed with 100 mL brine and dried over sodium sul-
fate. The solvent was evaporated, and the crude product was sub-
jected to chromatography (silica, ethyl acetate:cyclohexane 2:1) to
give 1 (11 g, 79%) as a white powder mp 59.0–61.3 °C. Anal. Calc.
3
. Conclusions
We have presented the synthesis of a new, highly soluble hy-
droxy-functionalised nanoball which is based on the fundamental
copper(II) nanoball coordination unit previously reported by Za-
worotko and co-workers [5c]. Nanoballs are thermodynamically
more stable than simple dinuclear copper(II) paddle wheel com-
plexes such as copper acetate or benzoate due to the chelate effect
ꢁ1
C14H18O : C, 56.4; H, 6.1. Found: C, 56.3; H, 6.1%. mmax(KBr)/cm
7
3300, 2958, 2925, 2883, 1726, 1595, 1461, 1435, 1347, 1310,
1244, 1133, 1051, 997, 874, 756, 719, 668, 616, 590, 549 and
458; d (400 MHz; DMSO-d ) 8.25 (1 H, t, J 1.3, Aryl), 7.76 (2 H,
(
or, in other words, positive chelate cooperativity). The increased
stability compared to simple dinuclear copper(II) complexes repre-
sents a promising perspective for using water-soluble nanoballs for
technological applications or in antifouling coatings in the future.
Furthermore, it must be emphasised that the dissociation is only
observed in aqueous solvents, whereas in organic solvents no indi-
cations for dissociation can be detected.
H
6
1,4
d, J1,4 1.3, Aryl), 4.21 (2 H, dd, J1,3 5.4, J1,3 3.7, CH ), 3.91 (6 H, s,
2
OCH ), 3.87 (2 H, dd, J 5.4, J1,3 3.7, CH ), 3.75 (2 H, dd, J 5.4,
3
1,3
2
1,3
J1,3 3.7, CH ), 3.66 (2 H, dd, J 5.4, J1,3 3.7, CH ) and 2.33 (1 H,br
2
1,3
2
s, OH); d (100 MHz; DMSO-d ) 166.1, 158.8, 131.7, 123.2, 120.0,
C
6
ꢁ
72.7, 69.5, 68.0, 61.8 and 52.5; m/z (CI) 565 (2M-CH O , 100%),
99 (M+H , 18%)and 267 (M-CH
3
+
ꢁ
3
O , 39).
2
3
The slope at the origin of this modified Hill plot, n
D
, is a measure for the
> 1. For
cooperativity of the system, generally indicating positive cooperativity for n
D
4.3. 5-(2-(2-Hydroxyethoxy)ethoxy)benzene-1,3-dicarboxylic acid (2)
large cyclic oligomers (such as our nanoball), the modified Hill plot starts parallel to
the reference line for low concentrations, and deviates over the concentration range
1
0 g
(33.6 mmol)
dimethyl-5-(2-(2-hydroxyethoxy)eth-
D
where the macrocycle is the most stable species (with a slope n > 1 at the origin) in
the case of strong positive cooperativity.
oxy)benzene-1,3-dicarboxylic acid 1 and 11.3 g (201 mmol) potas-

M. Tonigold, D. Volkmer / Inorganica Chimica Acta 363 (2010) 4220–4229
4227
sium hydroxide were dissolved in 200 mL ethanol: water (1:1) and
refluxed for 6 h. The solvent was evaporated, the residue was dis-
solved in 50 mL water and adjusted with 1 M HCl to pH 1–2. The
aqueous phase was extracted ethyl acetate, and the combined or-
ganic phases were washed with 100 mL brine and dried over so-
dium sulfate. The solvent was evaporated to give 2 (8.8 g, 97%) as
cules is not possible due to their strong disorder. Due to weak scat-
tering of the crystals of 3b, reflexes above 15° were omitted
H
from the structure solution and all side chain atoms were refined
isotropically, and certain atom distances were specified as fixed
due to problems caused by the diffuse electron density of the dis-
ordered solvent, guest molecules and 2-(2-hydroxyethoxy)ethoxy
groups. Details for crystal data collection and final refinement
parameters of 3b and 4 are summarized in Tables 2 and 3. See Sup-
plementary Material for details.
a white powder mp 173.5–175.2 °C. Anal. Calc. for C12
H
14
O
7
: C,
ꢁ1
5
2
1
3.3; H, 5.2. Found: C, 56.4; H, 5.4%. mmax(KBr)/cm 3186, 2937,
620, 1725, 1698, 1612, 1457, 1438, 1394, 1286, 1191, 1128,
063, 981, 933, 886, 861, 798, 734, 633, 588, 543, 508 and 450;
d
H
(400 MHz; DMSO-d
Aryl), 7.65 (2 H, d, J1,4 1.3, Aryl), 4.65 (1 H, br s, OH), 4.21 (2 H, t,
1,3 4.4, CH ), 3.77 (2 H, t, J1,3 4.4, CH ) and 3.45-3.55 (4 H, m,
CH ); d (100 MHz; DMSO-d
6
) 13.3 (2 H, s, COOH), 8.08 (1 H, t, J1,4 1.3,
4
.8. Determination of equilibrium constants
J
2
2
Spectra were recorded on a J&M TIDAS equipped with 1 cm
2
C
6
) 166.9, 159.1, 133.1, 122.8, 119.5,
ꢁ
quartz cuvettes at 25.0 (2) °C in the range of 500–1100 nm, the io-
nic strength was fixed at 0.5 M using KNO . Studies were carried
out using a precision piston microburet (Eppendorf) to add the
copper(II) carboxylates to 2 mL of solvent (H O:DMF 1:1). Between
8 and 25 solutions were prepared with a [CuL]tot concentration
ranging between 0.0004 and 0.05 mol/L. For 3, a fixed ratio of
HOAc]/[CuL]tot = 2 was used to facilitate the dissociation of the
7
2.9, 69.2, 68.3 and 60.7; m/z (CI) 253 (M-OH , 100%).
3
4
4
.4. Tetracosakis(l -5-(2-(2-hydroxyethoxy)ethoxy)benzene-1,3-
2
dicarboxylato)-tetracosa-copper(II) methanol solvate (3)
1
0
.9 mL (7.4 mmol) 2,6-dimethylpyridine were added to a solu-
tion of 0.9 g (3.7 mmol) Cu(NO O and 1.0 g (3.7 mmol) 2 in
ꢀ3H
00 mL methanol within 1 min under vigorous stirring. The imme-
[
3
)
2
2
complexes, whereas copper(II) acetate and copper(II) 4-hydroxy-
1
diately formed blue suspension was stirred for 30 min, filtered and
washed four times with 100 mL methanol, dried over night in air
and afterwards in vacuum at room temperature to give 3 (1.2 g,
Table 2
Crystallographic data for 3b.
9
5%) as a blue powder. Anal. Calc. for [Cu24(C12
12 7 3
H O )(CH OH)24]:
ꢁ
1
Formula
306 192 6
C H372Cu24O S
8839.65
green, block
C, 42.9; H, 4.4. Found: C, 42.8; H, 4.5%. max(KBr)/cm
m
3382,
Formula weight
Crystal colour, habit
Crystal dimensions (mm)
Crystal system
Space group
2
7
926, 2860, 1636, 1589, 1457, 1380, 1267, 1126, 1056, 884, 774,
31 and 492.
0.31 ꢄ 0.31 ꢄ 0.27
cubic
Pa 3ꢀ (#205)
4
4.5. Tetracosakis(l -5-(2-(2-hydroxyethoxy)ethoxy)benzene-1,3-
a (Å)
38.119(4)
55387(11)
4
dicarboxylato)-dodecaaqua-hexakis(dimethylsulfoxid)-hexamethanol
tetracosa-copper(II) (3b)
3
V (Å )
Z
D
calc (g cm 3
ꢁ
)
1.060
F(0 0 0)
(mmꢁ1)
Number of observations (all reflections)
18144
0.993
38537
0.1004
0.2910
0.745
0.290 and ꢁ0.213
Green single crystals of 3b were recrystallised by vapour diffu-
sion of n-butanol into a solution of 3 in DMSO, mounted in inert oil
and transferred to the cold gas stream of the diffractometer.
l
a
R
R
[I > 2
(all reflections)
r
(I)]
b
4.6.
5
l -5-(2-(2-Hydroxyethoxy)ethoxy)benzene-1,3-dicarboxylato)-
Goodness-of-fit (GOF)
D
qmax and
D
q
min (e Åꢁ3)
hemiaqua-tetracosa-copper(II)N,N-dimethylformamide solvate (4)
P
2
2
2
Þꢂ1=2.
GOF = ½ wðF ꢁ F Þ =ðN
o
ꢁ N
P
v
o
c
P
R ¼ jjF
P
P
a
1
2 mg (33
lmol) of 3 and (32.4 mg, 120
l
mol) of 2 were dis-
o
j ꢁ jF
c
jj= jF
o
j.
b
wR = ½ ðwðF ꢁ F Þ Þ= wðF Þ ꢂ1=2.
2
o
2
c
2
2
2
o
solved in 3 mL of H
2
O:DMF (1:1) and left undisturbed at room tem-
perature for 8 weeks to obtain 4 (10.9 mg, 90%) as blue crystals
which were transferred into inert oil and transferred to the cold
Table 3
Crystallographic data for 4.
gas stream of the diffractometer. Anal. Calc. for [Cu
2 12 7 2
(C12H O ) (-
H
2
1
2
O)]ꢀ(C
.4%. max(KBr)/cm 3422, 2932, 2863, 1634, 1590, 1458, 1418,
378, 1322, 1271, 1124, 1066, 892, 775, 927 and 492.
3 7
H
NO): C, 43.0; H, 4.4; N, 1.9. Found: C, 41.8; H, 4.9; N,
ꢁ1
m
Formula
C
27
H
33Cu
754.64
green, plate
2
NO16
Formula weight
Crystal colour, habit
Crystal dimensions (mm)
Crystal system
Space group
a (Å)
0.08 ꢄ 0.39 ꢄ 0.46
4.7. Crystal structure determination of 3b and 4
orthorhombic
1 1
P2 2 2 (#18)
Intensity data for 3b and 4 were collected at 153 K on a STOE-
18.253(4)
18.660(4)
11.928(2)
4062.6(14)
4
b (Å)
IPDS image-plate diffractometer (graphite monochromator) using
Mo K radiation (k = 0.7107 Å). The data were corrected for Lorentz
c (Å)
a
3
V (Å )
and polarisation effects, and multi-scan absorption correction [37]
was applied for 4. Programs used: data collection: Bruker XSCANS
Z
D
calc (g cm 3
ꢁ
)
1.234
F(0 0 0)
(mmꢁ1)
Number of observations (all reflections)
1552
1.106
32129
0.1227
0.3017
1.295
2.015 and ꢁ1.316
[
38], data reduction: Bruker SHELXTL [39], structure solution: SHEL-
l
XS-97 [40], structure refinement: SHELXL-97 [41], molecular graph-
ics: Diamond V3.1. [42], Mercury V1.5 [43]. Structures were
solved using direct methods and refined by full-matrix least
squares on |F| . All disordered atoms and atoms with non-positive
U
placement parameters. Hydrogen atoms were placed in geometri-
cally calculated positions. For 3b, an exact determination of the
total amount and identity of the non-coordinated solvent mole-
a
R
R
[I > 2
(all reflections)
r
(I)]
b
2
Goodness-of-fit (GOF)
D
qmax and
D
q
min (e Åꢁ3)
eq were calculated with isotropic, all others with anisotropic dis-
P
2
2
2
Þꢂ1=2.
GOF = ½ wðF ꢁ F Þ =ðN
o
ꢁ N
P
v
o
c
P
R ¼ jjF
P
P
a
o
j ꢁ jF
c
jj= jF
o
j.
b
wR = ½ ðwðF ꢁ F Þ Þ= wðF Þ ꢂ1=2.
2
o
2
c
2
2
2
o

4
228
M. Tonigold, D. Volkmer / Inorganica Chimica Acta 363 (2010) 4220–4229
benzoate were measured with
a
constant concentration of
Venturi, K. Yase, New J. Chem. 22 (1998) 959–972;
c) Monographs in Supramolecular Chemistry: J.F. Stoddard (Ed.), Royal Society
of Chemistry, vols. 1-6, Cambridge, UK, 1989, 1991, 1994–1996;
d) F.A. Cotton, P. Lei, C. Lin, C.A. Murillo, X. Wang, S.Y. Yu, Z.X. Zhang, J. Am.
Chem. Soc. 126 (2004) 1518–1525.
4] (a) M. Tominaga, K. Suzuki, T. Murase, M. Fujita, J. Am. Chem. Soc. 127 (2005)
(
0
.140 mol/L acetic acid or 4-hydroxybenzoic acid, respectively.
For 3, the spectra were measured after 6 months to ensure com-
plete equilibration. Equilibrium constants were calculated in a
similar way as already described for copper acetate [11a] using
Gauss–Newton non-linear least square fitting (See Supplementary
Material).
(
[
1
(
1950–11951;
b) M. Tominaga, K. Suzuki, M. Kawano, T. Kusakawa, T. Ozeki, S. Sakamoto, K.
Yamaguchi, M. Fujita, Angew. Chem. 116 (2004) 5739–5743;
c) N. Takeda, K. Unemoto, K. Yamaguchi, M. Fujita, Nature 398 (1999) 794–
96;
(d) B. Olenyuk, J.A. Whiteford, A. Fechtenkötter, P.J. Stang, Nature 398 (1999)
96–799;
e) Z. Ni, A. Yassar, T. Antoun, O.M. Yaghi, J. Am. Chem. Soc. 127 (2005) 12752–
2753;
(
7
4.9. Small-angle X-ray scattering (SAXS)
7
(
1
Compound 3 was dissolved in DMF at a concentration of 18 g/L.
The solutions were filtered using 0.2
tering experiments were done with a Bruker AXS Nano-STAR
equipped with a HI-STAR area detector. The light source was a con-
ventional X-ray tube operating at 35 mA and 40 kV (Cu
l
m pore-size filters. The scat-
(f) P. Guinan, M. McCann, H. Ryan, Polyhedron 11 (1992) 205–210;
(g) A.C. Sudik, A.R. Millward, N.W. Ockwig, A.P. Côté, J. Kim, O.M. Yaghi, J. Am.
Chem. Soc. 127 (2005) 7110–7118;
(
(
h) F.A. Cotton, C. Lin, C.A. Murillo, Inorg. Chem. 40 (2001) 6413–6417;
i) F.A. Cotton, L.M. Daniels, C. Lin, C.A. Murillo, Chem. Commun. (1999) 841–
K
a
= 0.154 nm). The blank-scattering contributions of the solvent,
842;
j) S. Leininger, B. Olenyuk, P.J. Stang, Chem. Rev. 100 (2000) 853–908.
5] (a) Y. Ke, D.J. Collins, H.-C. Zhou, Inorg. Chem. 44 (2005) 4154–4156;
b) B. Moulton, J. Lu, A. Mondal, M.J. Zaworotko, Chem. Commun. 9 (2001) 863–
864;
c) M. Eddaoudi, J. Kim, J.B. Wachter, H.K. Chae, M. O’Keeffe, O.M. Yaghi, J. Am.
Chem. Soc. 123 (2001) 4368–4369;
d) H. Abourahma, A.W. Coleman, B. Moulton, B. Rather, P. Shahgaldian, M.J.
Zaworotko, Chem. Commun. 22 (2001) 2380–2381;
(
sample holder, and incoherent scattering were subtracted prior
to any calculations. The scattering data were transformed into
pair-distance distribution functions (PDDF) via the program GNOM
[
(
(
[
28]. X-ray solution scattering from the atomic coordinates of 3b
was evaluated with the program CRYSOL [31]. The program DAMMIN
(
[
(
29] was used for ab initio shape determination from the PDDF
by simulated annealing using a single phase dummy atom model).
(
(
e) G.J. McManus, Z. Wang, M.J. Zaworotko, Cryst. Growth Des. 4 (2004) 11–13;
f) H. Furukawa, J. Kim, K.E. Plass, O.M. Yaghi, J. Am. Chem. Soc. 128 (2006)
8
398–8399.
[6] M. Tonigold, J. Hitzbleck, S. Bahnmüller, G. Langstein, D. Volkmer, Dalton Trans.
2009) 1363–1371.
Acknowledgement
(
[
7] Dodecanuclear Pd(II)-based nanoballs (Pd(II)12L24) complexes are remarkably
stable and the half-lives are much longer than those for comparable
monodentate Pd(II)-pyridine complexes by a factor of 10 s. Once formed,
the 36-component molecular spheres behave like covalent compounds: S.
Sato, Y. Ishido, M. Fujita, J. Am. Chem. Soc. 131 (2009) 6064-6065.
The authors are grateful to the Landesgraduiertenförderung Ba-
den-Württemberg for financial support.
ꢁ
5
Appendix A. Supplementary material
[8] (a) J. Kang, J. Rebek, Nature 385 (1997) 50–52;
(
(
(
b) E.S. Barrett, J.L. Irwin, A.J. Edwards, M.S. Sherburn, J. Am. Chem. Soc. 126
2004) 16747–16749;
c) J. Sherman, Chem. Commun. (2003) 1617–1623;
CCDC 766959 and 766958 contains the supplementary crystal-
lographic data for 3b and 4. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.ica.2010.06.022.
(d) K. Rissanen, Angew. Chem. Int. Ed. 44 (2005) 3652–3654;
e) F. Hof, S.L. Craig, C. Nuckolls, J. Rebek Jr., Angew. Chem. Int. Ed. 41 (2002)
488–1508;
f) M.A. Mateos-Timoneda, M. Crego-Calama, D.N. Reinhoudt, Chem. Soc. Rev.
33 (2004) 363–372;
(
1
(
(
(
(
(
(
(
g) L.R. MacGillivray, J.L. Atwood, Angew. Chem. Int. Ed. 38 (1999) 1018–1033;
h) L.R. MacGillivray, J.L. Atwood, Nature 389 (1997) 469–472;
i) A. Shivanyuk, J. Rebek, Chem. Commun. (2001) 2424–2425;
j) L.J. Prins, J. Huskens, F. De Jong, P. Timmerman, D.N. Reinhoudt, Nature 398
1999) 498–502;
References
k) L.J. Prins, F. De Jong, P. Timmerman, D.N. Reinhoudt, Nature 408 (2000)
[
1] (a) A. Ziv, A. Grego, S. Kopilevich, L. Zeiri, P. Miro, C. Bo, A. Müller, I.A.
Weinstock, J. Am. Chem. Soc. 131 (2009) 6380–6382;
181–184;
(
(
l) X. Liu, R. Warmuth, J. Am. Chem. Soc. 128 (2006) 14120–14127;
m) D.J. Cram, M.E. Tanner, R. Thomas, Angew. Chem. 103 (1991) 1048–1051.
(
b) A. Müller, S. Sarkar, S.Q. Shah, H. Bögge, M. Schmidtmann, P. Kögerler, B.
Hauptfleisch, A.X. Trautwein, V.V. Schünemann, Angew. Chem., Int. Ed. Engl. 38
[
9] (a) A. Marquis-Rigault, A. Dupont-Gervais, P.N.W. Baxter, A. Van Dorsselaer, J.-
M. Lehn, Inorg. Chem. 35 (1996) 2307–2310;
(
(
1999) 3238–3241;
c) A. Müller, S.Q.N. Shah, H. Bögge, M. Schmidtmann, Nature 397 (1999) 48–
(
b) C. Piguet, J.C. Bünzli, G. Bernardinelli, C.G. Borchert, P. Froidevaux, J. Chem.
Soc., Dalton Trans. (1995) 83–97;
c) N. Fatin-Rouge, S. Blanc, A. Pfeil, A. Rigault, A.-M. Albrecht-Gary, J.-M. Lehn,
Helv. Chim. Acta 84 (2001) 1694–1711;
d) N. Fatin-Rouge, S. Blanc, E. Leize, A. Van Dorsselaer, P. Baret, J.-L. Pierre, A.-
M. Albrecht-Gary, Inorg. Chem. 39 (2000) 5771–5778;
e) M. Fujita, F. Ibukuro, K. Yamaguchi, K. Ogura, J. Am. Chem. Soc. 117 (1995)
175–4176;
f) B. Hasenkopf, J.-M. Lehn, N. Boumediene, E. Leize, A. Van Dorsselaer, Angew.
Chem. Int. Ed. 37 (1998) 3265–3268.
10] (a) J. Chen, S. Körner, S.L. Craig, D.M. Rudkevich, J. Rebek Jr., Nature 415 (2002)
85–386;
5
(
1
0;
d) D.J. Cram, M.E. Tanner, R. Thomas, Angew. Chem., Int. Ed. Engl. 30 (1991)
024–1027;
(
(
(
(
e) X. Liu, G. Chu, R.A. Moss, R.R. Sauers, R. Warmuth, Angew. Chem., Int. Ed. 44
2005) 1994–1997;
f) M. Ziegler, J.L. Brumaghim, K.N. Raymond, Angew. Chem., Int. Ed. 39 (2000)
(
(
4
(
4
119–4121;
(
(
g) J. Kang, J. Rebek, Nature 385 (1996) 50–52;
h) R. Warmuth, J.-L. Kerdelhué, S. Sánchez Carrera, K.J. Langenwalter, N.
Brown, Angew. Chem., Int. Ed. 41 (2002) 96–99;
i) M. Yoshizawa, Y. Takeyama, T. Kusukawa, M. Fujita, Angew. Chem., Int. Ed.
1 (2002) 1347–1349;
j) M. Yoshizawa, T. Kususkawa, M. Fujita, S. Sakamoto, K. Yamaguchi, J. Am.
Chem. Soc. 123 (2001) 10454–10459;
[
(
4
(
3
(
(
b) E.S. Barrett, T.J. Dale, J. Rebek Jr., J. Am. Chem. Soc. 129 (2007) 8818–8824;
c) T. Gottschalk, B. Jaun, F. Diederich, Angew. Chem. Int. Ed. 46 (2007) 260–
264;
(
(
(
k) M. Yoshizawa, Y. Takeyama, T. Okano, M. Fujita, J. Am. Chem. Soc. 125
2003) 3243–3247;
l) M. Yoshizawa, S. Miyagi, M. Kawano, K. Ishiguro, M. Fujita, J. Am. Chem. Soc.
(
(
(
d) T. Amaya, J. Rebek Jr., J. Am. Chem. Soc. 126 (2004) 6216–6217;
e) L.C. Palmer, J. Rebek Jr., Org. Biomol. Chem. 2 (2004) 3051–3059;
f) L.J. Prins, E.E. Neuteboom, V. Paraschiv, M. Crego-Calama, P. Timmerman,
1
26 (2004) 9172–9173;
m) M.L. Merlau, M.P. Mejia, S.T. Nguyen, J.T. Hupp, Angew. Chem., Int. Ed. Engl.
0 (2001) 4239–4242;
n) M. Yoshizawa, M. Tamura, M. Fujita, Science 312 (2006) 251–254.
2] (a) C.L.D. Gibb, B.C. Gibb, J. Am. Chem. Soc. 126 (2004) 11408–11409;
b) S.T. Mough, J.C. Goeltz, K.T. Holman, Angew. Chem. Int. Ed. 43 (2004) 5631–
635.
3] (a) D.B. Amabilino, J.F. Stoddart, Chem. Rev. 95 (1995) 2725–2828;
b) D.B. Amabilino, M. Asakawa, P.R. Ashton, R. Ballardini, V. Balzani, M.
Belohradsky, A. Credi, M. Higuchi, F.M. Raymo, T. Shimizu, J.F. Stoddart, M.
D.N. Reinhoudt, J. Org. Chem. 67 (2002) 4808–4820.
(
4
(
[
11] (a) Y. Doucet, G.W. Marion, Compt. Rend. 240 (1955) 1616–1618;
(
(
(
b) Y. Doucet, R. Cogniac, Compt. Rend. 240 (1955) 968–972;
c) S.C. Sircar, S. Aditya, B. Prasad, J. Ind. Chem. Soc. 30 (1953) 633–636;
d) J.W. Bunting, K.M. Thong, Can. J. Chem. 48 (1970) 1654–1656.
[
[
(
5
[
[
12] R.L. Martin, A. Whitley, J. Chem. Soc. (1958) 1394–1402.
13] R.W. Larsen, G.J. McManus, J.J. Perry IV, E. Rivera-Otero, M.J. Zaworotko, Inorg.
Chem. 46 (2007) 5904–5910.
(
[
14] C.A. Hunter, H.L. Anderson, Angew. Chem. Int. Ed. 48 (2009) 7488–7499.

M. Tonigold, D. Volkmer / Inorganica Chimica Acta 363 (2010) 4220–4229
4229
[
[
15] B. Zhang, R. Breslow, J. Am. Chem. Soc. 115 (1993) 9353–9354.
16] P. Ballester, A.I. Oliva, A. Costa, P.M. Deyà, A. Frontera, R.M. Gomila, C.A.
Hunter, J. Am. Chem. Soc. 128 (2006) 5560–5569.
17] A. Marquis-Rigault, A. Dupont-Gervais, P.N.W. Baxter, A.V. Dorsselaer, J.-M.
Lehn, Inorg. Chem. 35 (1996) 2307–2310.
[30] J. Lipfert, S. Doniach, Annu. Rev. Biophys. Biomol. Struct. 36 (2007) 307–327.
[31] D.I. Svergun, C. Barberato, M.H.J. Koch, J. Appl. Cryst. 28 (1995) 768–773.
[32] (a) L. Dubicki, Aust. J. Chem. 25 (1972) 1141–1149;
(b) G.W. Reimann, G.F. Kokoszka, G. Gordon, Inorg. Chem. 4 (1965) 1082–
1084;
[
[
18] (a) H. Abouhrama, G.J. Bodwell, J. Lu, B. Moulton, I.R. Pottie, R.B. Walsh, M.J.
Zaworotko, Cryst. Growth Des. 3 (2003) 513–519;
(c) L. Dubicki, R.L. Martin, Inorg. Chem. 5 (1966) 2203–2209;
(d) D.S. McClure, J. Chem. Phys. 39 (1963) 2850–2855.
(
1
b) D.-X. Xue, Y.-Y. Lin, X.-N. Cheng, X.-M. Chen, Cryst. Growth Des. 7 (2007)
332–1336.
19] (a) J.J. Perry, G.J. McManus, J.M. Zawarotko, Chem. Commun. 22 (2004) 2534–
535;
b) M. Eddaoudi, J. Kim, D. Vodak, A. Sudik, J. Wachter, M. O’Keeffe, O.M. Yaghi,
Proc. Natl. Acad. Sci. USA 99 (2002) 4900–4904;
c) B. Moulton, H. Abourahma, M.W. Bradner, J. Lu, G.J. McManus, M.J.
[33] (a) C. Tanford, J. Phys. Chem. 78 (1974) 2469–2479 (and references cited
herein);
[
[b] Y. Moroi, Micelles: Theoretical and Applied Aspects, Springer-Verlag,
Berlin, 1992, pp. 56-63 (and references cited herein)
2
(
[34] [a] A.E. Martell, in: W. Schneider, G. Anderegg, R. Gut (Eds.), Essays in
Coordination Chemistry, Berkhauser Verlag, Basel, 1964, pp. 52–64;
[b] A.W. Adamson, J. Am. Chem. Soc. 76 (1954) 1578–1579;
[c] G. Schwarzenbach, Helv. Chim. Acta 35 (1952) 2344–2359;
[d] K.B. Yatsimirskii, Theor. Exp. Chem. 16 (1980) 34–40;
[e] R.D. Hancock, A.E. Martell, Comments Inorg. Chem. 6 (1988) 237–284.
[35] Recent reviews highlighting different aspects of MOF applications: [a] M.
Eddaoudi, D.B. Moler, H. Li, B. Chen, T.M. Reineke, M. O’Keeffe, O.M. Yaghi, Acc.
Chem. Res. 34 (2001) 319–330;
(
Zaworotko, Chem. Commun. 12 (2003) 1342–1343.
20] G.F. Swiegers, J. Malefetse, Chem. Rev. 100 (2000) 3483–3537.
21] G. Ercolani, J. Phys. Chem. B 102 (1998) 5699–5703.
22] (a) H.L. Anderson, Inorg. Chem. 33 (1994) 972–981;
[
[
[
(
b) X. Chi, A.J. Guerin, R.A. Haycock, C.A. Hunter, L.D. Sarson, J. Chem. Soc.
Chem. Commun. (1995) 2563–2565;
c) D.S. Lawrence, T. Jiang, M. Levett, Chem. Rev. 95 (1995) 2229–2260.
23] (a) W. Ostwald, Z. Phys. Chem. 22 (1897) 289–330;
b) W. Ostwald, The Fundamental Principles of Chemistry, Longmans, Green
(
[b] G. Férey, C. Mellot-Draznieks, C. Serre, F. Millange, Acc. Chem. Res. 38
(2005) 217–225;
[c] U. Müller, M. Schubert, F. Teich, H. Puetter, K. Schierle-Arndt, J. Pastré, J.
Mater. Chem. 16 (2006) 626–636.
[
[
(
and Co., New York, NY, 1909; see also A.F. Holleman, Lehrbuch der
Anorganischen Chemie, de Gruyter, Berlin, New York, NY, 1925.
24] DMSO: N. Judas, Acta Cryst. Sect. E 61 (2005) 2217–2219;
[36] I. Bkouche-Waksman, C. Bois, G.A. Popovitch, P. L’Haridon, Bull. Soc. Chim. Fr. 1
(1980) 69–75.
CH
Chim. Fr. (1980) 69–71;
O:G.M. Brown, R. Chidambaram, Acta Cryst. Sect. B 29 (1973) 2393–2403;
ArOCH CH OH:K. Goubitz, E.J. Sonneveld, V.V. Chernyshev, A.V. Yatsenko, S.G.
3
OH:I. Bkouche-Waksman, C. Bois, G.A. Popovitch, P.L. Haridon, Bull. Soc.
[37] R.H. Blessing, Acta Cryst. A51 (1995) 33–38.
[38] Bruker, XSCANS, Version 2.2, Bruker Analytical X-ray Systems Inc., Madison,
WI, 1996.
[39] SHELXTL V6.12 2000 Bruker Analytical X-Ray Systems, Madison, WI.
[40] G.M. Sheldrick, Acta Cryst. A46 (1990) 467–473.
[41] G.M. Sheldrick, SHELXL-97, Program for X-ray Crystal Structure Refinement,
University of Göttingen, Göttingen, Germany, 1997.
[42] K. Brandenburg, H. Putz GbR, Diamond – Crystal and Molecular Structure
Visualization, Crystal Impact, Postfach 1251, D-53002 Bonn.
[43] C.F. Macrae, P.R. Edgington, P. McCabe, E. Pidcock, G.P. Shields, R. Taylor, M.
Towler, J. van de Streek, Mercury: visualization and analysis of crystal
structures, J. Appl. Cryst. 39 (2006) 453–457.
H
2
2
2
Zhukov, C.A. Reiss, H. Schenk, Z. Kristallogr. 214 (1999) 469–474.
25] A.L. Spek, J. Appl. Cryst. 36 (2003) 7–13.
26] The van der Waals radius of C (0.17 nm) was used to calculate the width of the
apertures: A. Bondi, J. Phys. Chem. 68 (1964) 441–451.
27] O. Glatter, J. Appl. Crystallogr. 12 (1979) 166–175.
28] (a) D.I. Svergun, J. Appl. Cryst. 24 (1991) 485–492;
[
[
[
[
(
b) D.I. Svergun, J. Appl. Cryst. 25 (1991) 495–503.
[
29] D.I. Svergun, Biophys. J. 76 (1999) 2879–2886.