Resorcinarene podand with amine-functionalized side arms - Synthesis, structure, and binding properties of a neutral anion receptor

Article abstract of DOI:10.1002/ejoc.200900814

The synthesis and structure of a neutral resorcinarene host: bearing four amine-functionalized side arms is described. The anion binding properties were investigated in solution by ¹H NMR spectroscopic titration and diffusion experiments and in the gas phase by mass spectrometrlc studies. It was observed that in solution 1:2 (host/guest) complexes were formed between the resorcinarene host: and the basic fluoride and acetate anions, whereas in the gas phase 1:1 complexes with other anions (Cl^-, HCOO^-, NO₃^-, and BF₄^-) were detected additionally. Wiley-VCH Verlag GmbH & Co. KGaA.

Full text of DOI:10.1002/ejoc.200900814

FULL PAPER
DOI: 10.1002/ejoc.200900814
Resorcinarene Podand with Amine-Functionalized Side Arms – Synthesis,
Structure, and Binding Properties of a Neutral Anion Receptor
Kirsi Salorinne,[a] Dominik P. Weimann,[b] Christoph A. Schalley,[b] and Maija Nissinen*[a]
Keywords: Supramolecular chemistry / Host-guest systems / Anions / Receptors / Resorcinarene
The synthesis and structure of a neutral resorcinarene host
bearing four amine-functionalized side arms is described.
The anion binding properties were investigated in solution
by 1H NMR spectroscopic titration and diffusion experiments
and in the gas phase by mass spectrometric studies. It was
observed that in solution 1:2 (host/guest) complexes were
formed between the resorcinarene host and the basic fluoride
and acetate anions, whereas in the gas phase 1:1 complexes
with other anions (Cl–, HCOO–, NO3–, and BF4–) were de-
tected additionally.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2009)
electrostatic interactions, they are directional, which con-
tributes to the recognition and selectivity of certain anion
geometries.
Introduction
The development and synthesis of anion receptors has
gained a considerable amount of attention over the last dec-
ades.[1] The interest in anions arises from their importance
in biological, environmental and chemical processes, which
has induced a number of papers dealing with anion sensing,
transport, and extraction as well as the use of anions in
organic reactions.[2] The varying geometry and larger radii
of anions compared to the most common cations greatly
influences the binding, which again set the demands on the
design of the receptor molecule in terms of host-guest com-
plementarity and selectivity.[3] In addition, the solvent has
a more pronounced role in anion binding in terms of sol-
vation and desolvation effects, which are namely attributed
to the protic, polar, or hydrogen-bonding nature of the sol-
vent, and furthermore, the basicity of certain anions may
vary drastically depending on the solvent used.[1,3]
Our approach utilizes tetramethoxyresorcinarene[5] as the
platform, which has a well-defined macrocyclic structure
yet enough freedom to reorganize in order to best accom-
modate the guest molecule. Thus, four amine-functionalized
side arms were attached on the upper rim of the resorcinar-
ene bowl to create a neutral tetrapodal structure, in which
the resorcinarene core adopts a boat conformation. Two
binding sites are formed between the neighboring out-
stretched side arms at both ends of the resorcinarene cavity,
the amine groups offering hydrogen bonding interactions
with the anion (Scheme 1). The anion binding properties
were investigated in solution and in the gas phase. In solu-
tion, binding was observed only between the more basic F–
Calixarenes have been extensively utilized in the design
of anion receptors.[1c] The calixarene core provides a preor-
ganized three-dimensional structure, which is easily modi-
fied by incorporation of anion binding ligands in the form
of bridges or pendant side chains at the upper or lower rim
of the calixarene framework.[1c] In neutral anion receptors,
anion recognition is mainly based on hydrogen bonding in-
teractions, which are generally provided in the binding site
through amine, amide, urea, or thiourea moieties.[1c,4] Al-
though hydrogen bonding interactions are not as strong as
[a] Nanoscience Center, Department of Chemistry, University of
Jyväskylä,
P. O. Box 35, 40014 JYU, Finland
Fax: +358-14-260-4756
E-mail: maija.nissinen@jyu.fi
[b] Institut für Chemie und Biochemie, Freie Universität Berlin,
Takustrasse 3, 14195 Berlin, Germany
Scheme 1. A schematic presentation of the suggested anion binding
mode for the tetrapodal resorcinarene receptor.
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.200900814.
Eur. J. Org. Chem. 2009, 6151–6159
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6151

K. Salorinne, D. P. Weimann, C. A. Schalley, M. Nissinen
FULL PAPER
and AcO– anions and the resorcinarene host, whereas in the Structural Properties
gas phase complexes were detected with other anions (Cl–,
–
–
HCOO–, NO3 and BF4 ) as well.
The structural properties of 4 were investigated in solu-
tion by H NMR spectroscopy and in the solid state by
1
X-ray crystallography. As a result of substituting the four
hydroxy groups of tetramethoxy resorcinarene with four
amine-functionalized side arms, there no longer exists a hy-
drogen bonding network on the upper rim of the resor-
cinarene bowl to keep the molecule in a crown conforma-
Results and Discussion
Synthesis
tion, but it has to adopt a more favorable boat conforma-
1
The synthesis of tetramethoxy resorcinarene host 4, tion instead. This is seen in the H NMR spectra of 4 at
bearing four amine-functionalized side arms, was achieved room temperature as a set of averaged signals for all the
by incorporating four o-nitro-N-(2-hydroxyethyl)aniline protons belonging to the resorcinarene core due to fast
units into the tetramethoxy resorcinarene[5] platform boat-to-boat interconversion.[7] As the temperature is de-
(Scheme 2). o-Nitro-N-(2-hydroxyethyl)aniline (1) was pre- creased, the core resorcinarene resonances start to broaden
pared according to a slightly modified literature method[6] and finally give a set of doublets as –60 °C is reached corre-
by treatment of o-bromonitrobenzene with ethanolamine in sponding to the reduced C2 symmetry of the boat confor-
the presence of anhydrous copper(II) chloride in almost mation. Due to overlapping signals this was most clearly
quantitative yield. Tosylation of the hydroxy group with p- observed with the aromatic and methoxy protons of the res-
toluenesulfonyl chloride in the presence of triethylamine in orcinarene core.
dichloromethane afforded the tosylated species 2 with 75%
Resorcinarene host 4 forms co-crystals by slow evapora-
yield, which could then be attached to the resorcinarene tion from acetonitrile (Ia), acetonitrile/ethanol (Ib), acetoni-
scaffold by nucleophilic substitution reaction using anhy- trile/TBACl (II) and pyridine (III) solutions. The crystal
drous Cs2CO3 as the base and dibenzo-18-crown-6 as the structures Ia and Ib are isomorphous with nearly identical
phase-transfer-catalyst in refluxing acetonitrile, giving res- unit cell parameters. In all of the structures, both the left-
orcinarene derivative 3 with 80% yield. The reduction of and right-handed isomers[7] of the host molecule are present
the nitro group into an amino group proved unsuccessful in the unit cell and the resorcinarene host 4 adopts a boat
with hydrazine/Raney-Ni and tin(II) chloride, but was fi- conformation, in which the resorcinarene skeleton is some-
nally accomplished after several attempts using sodium sul- what twisted (Figure 1). In the structures Ia/Ib the resor-
fide and sulfur in refluxing butanol, which afforded resor- cinarene cavity is slightly more opened than in the struc-
cinarene host 4 in 70% yield. The structures of 4 and its tures II and III with an angle of 40.3(1)° between the up-
intermediates were characterized by means of NMR spec- right aromatic rings and an angle of 145.8(1)° between the
troscopy and mass spectrometry, X-ray crystallography, and aromatic rings in the horizontal plane, whereas the respec-
elemental analysis, which all confirmed the success of the tive angles in the structures II and III are 31.6(1)° and
reactions.
166.1(1)° (average values of the two structures).
Scheme 2. Synthesis of the amine-functionalized resorcinarene host 4 with crystallographic numbering.
6152 Eur. J. Org. Chem. 2009, 6151–6159
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Resorcinarene Podand with Amine-Functionalized Side Arms
Table 1. Selected torsion angles [°] of 4.
Ia[a]/Ib[b]
II[a]
III[b]
O4/6–C41–C42–N43
O11/13–C51–C52–N53
O18/20–C61–C62–N63
O25/27–C71–C72–N73
–65.6(3)
57.9(5)
–59.6(3)
–63.6(3)
–62.9(2)
–65.0(2)
–67.0(2)
62.3(3)
65.8(3)
65.7(4)
61.8(3)
75.1(4)
[a] Left-handed isomer (O6). [b] Right-handed isomer (O4) in the
asymmetric unit.
Figure 1. ORTEP plot (at 50% probability level) and CPK model
of the crystal structure Ib of 4 (top views). The dashed lines indicate
intramolecular hydrogen bonds. Solvent molecules have been omit-
ted for clarity.
The amine-functionalized side arms fan out on the side
of the resorcinarene core with the -OCH2CH2NH- torsion
angles of each side arm close to 60° corresponding to a
gauche conformation (Table 1). In the structures Ia/Ib the
side arms lie above the plane formed by the methine bridges
(C7–C14–C21–C28) with the amine groups facing the outer
surface of the resorcinarene core (Figure 2). In structure III,
on the other hand, two opposite side arms are bent below
the plane and face the lower rim ethyl chains, whereas, in
structure II only one side arm is bent below the plane and
the amine groups of the opposite side arm point toward the
resorcinarene cavity (Figure 2). The amine groups, in the
structures Ia/Ib and III, form intra- and intermolecular hy-
drogen bonds with the phenolic oxygen atoms and/or the
amine groups of the parent or neighboring resorcinarene
unit [2.758(3)–3.358(4) Å for the intra- and 3.002(4) –
3.513(4) Å for the intermolecular N···O and N···N dis-
tances].
Figure 2. (a) Top and (b) side views of the structures Ia/Ib (dark
gray), II (black) and III (light gray) as a superposition drawing
showing the different geometries of the amine-functionalized side
arms. Hydrogen atoms and solvent molecules have been omitted
for clarity.
In both of the structures Ia/Ib and III, hydrogen bonded
pairs are formed between the left- and right-handed iso-
mers. The difference between the two structures lies in the
position of the isomers relative to one another, which also
Figure 3. The hydrogen bonded pairs formed between the left- and right-handed isomers (a) in structure Ia/Ib and (b) in structure III
excluding the solvent molecules for clarity. (c) The chains formed by the intermolecular hydrogen bonds between the neighboring resor-
cinarene molecules in structure II. Hydrogen bonds are shown by the dashed lines and only amine hydrogen atoms are shown for clarity.
Eur. J. Org. Chem. 2009, 6151–6159
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6153

K. Salorinne, D. P. Weimann, C. A. Schalley, M. Nissinen
FULL PAPER
explains the placement of the side arms above or below the
plane. In structures Ia/Ib the pairs are interconnected to
one another through the horizontal aromatic rings of the
opposite facing resorcinarene molecules, and the amine-
functionalized side arms thus settle above the plane (Fig-
ure 3, a). In structure III, on the other hand, the pairs are
composed of two opposite facing resorcinarene molecules
interlinked together by the ethyl chains, and therefore, in
order to form the intermolecular hydrogen bonds between
the neighboringmolecules, the side arms are forced to bend
below the plane (Figure 3, b).
In structure II, however, only intermolecular hydrogen
bonds are formed between the amine groups of two neigh-
boring resorcinarene molecules [average N···O distance of
3.398(4) Å] which results in a formation of chains in the
crystal packing. The acetonitrile solvent molecules in the
crystal lattice act as linkers between the parallel chains,
which form uniform layers (Figure 3, c).
1
Figure 4. H NMR titration spectra of 4 with F– in [D6]DMSO at
30 °C showing the changes in the chemical shifts of the aromatic
and amine protons upon the addition of guest as a TBA salt. (a)
Free host, (b) 0.5, (c) 1.0, (d) 1.6, (e) 2.5 and (f) 4.0 equiv. of guest
added in [D6]DMSO. (᭿ = ArH, ᭺ = NH and + = NH2).
Solution Binding Studies
ment to form a bifluoride HF2– anion. In fact, a closer look
at the spectrum of 4 with 2 equiv. of F– anions showed a
triplet further downfield at δ = 16.2 ppm, which was de-
noted to HF2– anion.[11,12] The free HF2– anion, as its TBA
salt, has been reported[12] to show up at δ = 15.4 ppm in
[D6]DMSO, which would imply that the F– anion was actu-
Anion binding studies were performed by 1H NMR spec-
troscopic titration experiments with various anions in [D6]-
DMSO, which was chosen as the solvent due to solubility
considerations, and because the amine resonances could
only be seen in [D6]DMSO as opposed to other available
solvents (i.e. CDCl3). The anions were added as their tetra-
butylammonium (TBA) salts, since the larger cation was
thought to be less susceptible to interact with the host mole-
cule compared to the smaller tetramethylammonium
counter cation.[8]
–
ally bound to the receptor 4 as HF2 anion. This has also
been described as a “frozen” proton release or an incipient
proton transfer reaction in a recent review.[13] With acidic
-NH hydrogen atoms complete deprotonation is also pos-
sible, which results in a release of HF. This, however, was
ruled out as the amine resonances did not disappear even
after a large excess of F– anion had been added, and
furthermore, inspection of the DMSO pKa values[10] of ani-
line (30.6) as a model for the receptor and HF (15) clearly
shows that the fluoride alone is not basic enough to depro-
tonate a significant fraction of the receptor.
–
–
–
With Cl–, Br–, I–, NO3 , PF6 and BF4 anions the ad-
dition of ten equivalents of the anion caused no changes in
1
the H NMR spectra of 4, which suggested that very weak
or no binding was taking place with these anions in solu-
tion. However, upon the addition of F– anions significant
1
changes were observed in the H NMR spectrum of 4 as
both the -NH and -NH2 amine resonances shifted down-
field, which was attributed to anion binding taking place
(Figure 4). Changes in the chemical shifts of the aromatic
and the methine protons of the resorcinarene core were also
observed, which suggest alteration in the host conformation
upon binding, and also, possible CH···anion interaction[9]
between the host and the F– anion. However, there were no
changes observed in the chemical shifts belonging to the
TBA+ counter cation, which indicated that the TBA+ cation
was not interacting with the host molecule, and that the
changes observed for the resorcinarene host 4 were solely
caused by the anion alone.
The stoichiometry was confirmed by a modified Job plot
analysis,[14] which gave a maximum corresponding to a 1:2
host-guest complex formation, and the stability constants
were calculated by the EQNMR[15] computer program for
the two step binding equilibrium according to the equations
–
(1) and (2), where G denotes F– and G2 becomes F2 in the
second equilibrium with an overall binding constant of β12
= K11K12 (Table 2). The chemical shift of the aromatic pro-
ton of the resorcinarene core was followed since both the
-NH and -NH2 amine resonances would broaden and fi-
nally disappear[16] in the course of the titration experiment.
The behavior of the F– anion and its ability to act as
a strong base in [D6]DMSO[10] and deprotonate the -NH
hydrogen of the receptor molecule has been reported in the
literature by Fabbrizzi and others.[11] It has been postulated
H + G Ǟ HG
(1)
(2)
HG + G Ǟ HG2
The addition of AcO– anion also caused similar changes
that there is a two-step process that involves a formation of in the spectrum of 4, although, at higher guest concentra-
genuine hydrogen-bonded complex between the neutral tion. Host 4 appears to readily form hydrogen bonds with
host and the F– anion in the first binding equilibrium while the more electronegative F– and AcO– anions compared to
in the next step a second F– anion abstracts the HF frag- the other anions investigated, and that the basicity of the
6154
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 6151–6159

Resorcinarene Podand with Amine-Functionalized Side Arms
Table 2. Diffusion coefficients[a] (10–5 cm2 s–1) and binding con- NMR spectra and was therefore examined. It was found
stants[b] (–1) of 4 with selected anions in [D6]DMSO at 303 K.
that the diffusion coefficient of TBA+ cation in the complex
was lower than that of the “free” TBA+ cation, which indi-
cates that host 4 and TBA[HF2] ion pair forms a moder-
ately bound entity that is comparable with the binding
strength determined by 1H NMR spectroscopic titration ex-
periment (Figure 5, Table 2).
Although solid-state investigations would reveal more in-
formation about anion binding mode in host 4, unfortu-
nately, all attempts to obtain good quality crystals of the
anion complexes of 4 for crystal structure determination
have so far failed.
DDMSO
0.89Ϯ0.01 0.258Ϯ0.006
1.05Ϯ0.01
0.97Ϯ0.01 0.262Ϯ0.004
DHost
DTBA+
K
4
–
–
–
F–
–
0.572Ϯ0.002
0.414Ϯ0.001 K11 = 360
K12 = 37
4 + F–
4+ AcO–
–
–
–
K11 = 40
K12 = 11
[a] Values are average Ϯ standard deviation of three experiments.
The solutions studied were 5 m in 4 and 10 m in anion as its
TBA salt. [b] Binding constants were calculated from 1H NMR
spectroscopic titration data fitted to a 1:2 (host/guest) binding
model using EQNMR;[15] errors less than 15%. Anions were added
as their TBA salts.
Gas-Phase Binding Studies
anions in [D6]DMSO determines the binding strength,
which was the strongest with F– anion (Table 2). Moreover,
the lack of binding observed with the other anions could
additionally result from the solvation effects induced by
[D6]DMSO, as it is able to form strong hydrogen bonds
with the host molecule, and therefore, only the more
strongly binding anions will be able to overcome this com-
petition at the binding site.
In order to provide additional evidence for the binding
of anions to the neutral receptor and to gain insight into
similarities or solvation-induced differences between the sit-
uation in solution and the gas-phase, negative-mode electro-
spray-ionization mass spectrometric (ESI-MS) experiments
were performed. For this purpose, an equimolar solution of
4 and each anion (or a combination of two different anions
in the case of fluoride and acetate, respectively) was electro-
sprayed, and the formation of complexes was monitored by
high-resolution Fourier-transform ion-cyclotron-resonance
(FTICR) mass spectra.
1H NMR spectroscopic diffusion measurements were ad-
ditionally carried out with F– anion in [D6]DMSO, in order
to substantiate whether the changes in the chemical shifts
of 4 during titration experiments were rightfully attributed
to anion binding interactions taking place instead of just
deprotonation, since diffusion is less sensitive to proton
transfer than chemical shifts.[17] A 1:2 host-guest solution
Figure 6 shows the ESI-FTICR mass spectrum obtained
from a 50 µ acetonitrile solution of 4, acetate, and fluo-
ride. In addition to the expected complexes of 4 with fluo-
ride and acetate, the complexes with chloride and formate
are also formed from background anions. Chloride is al-
most omnipresent in the negative mode and the formate
anion is due to a hard-to-remove memory effect from pre-
vious experiments in the positive mode, in which formic
acid was used. It is advantageous to use AgOAc to generate
the acetate complexes, because the silver ions help to scav-
enge background chloride. Irrespective of the ionization
conditions, 1:2 complexes of the receptor and the anion
have not been observed – a finding which can be attributed
to the significantly stronger repulsion between the two
charges in the desolvated complex. Instead, 2:1 complexes
are observed, which however are likely due to “unspecific”
–
in [D6]DMSO was substantiated, and, although the HF2
anion was also identified in the spectra, the diffusion coeffi-
–
cient could not be accurately determined from the HF2
chemical shift due to the low intensity and broadness of the
resonance. However, it seems reasonable to presume that
–
the formed HF2 anion forms an ion pair with its counter
cation, TBA+, which was conveniently visible in the 1H
–
binding. An HF2 complex of receptor 4 is not observed in
the gas phase. This is not unexpected, because it would eas-
ily lose HF during the ionization/desolvation process, even
when initially formed in solution. Consequently, these find-
ings already show that there are differences between the
solution situation and that in the gas-phase.
The most remarkable difference between solution and
gas-phase binding behavior of 4 is that not only fluoride
and acetate complexes are observed in the mass spectra.
The gas-phase experiments furthermore provide evidence
for chloride, nitrate, tetrafluoroborate, and formate to form
1:1 complexes. Likewise, many other anions will certainly
bind. Consequently, a number of anions – among them
Figure 5. Normalized signal decay ln(I/I0) as a function of b values
for 4 (᭿) and TBAF (∆) in the free state and for the TBA+ cation
(᭡) in a 1:2 host-guest solution of 4 and TBAF, respectively, in
[D6]DMSO at 303 K.
–
even the weakly binding ones such as BF4 – experience
attractive interactions with the receptor, although no bind-
Eur. J. Org. Chem. 2009, 6151–6159
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6155

K. Salorinne, D. P. Weimann, C. A. Schalley, M. Nissinen
FULL PAPER
Figure 6. Negative mode ESI-FT-ICR mass spectrum of an equi-
molar solution of 4, AgOAc, and TBAF (50 µ).
ing is observed in solution. This finding can be easily traced
back to the competition with the solvent. In the solution
studies, DMSO was used, which is quite a strong hydrogen-
bond acceptor and present in large excess. Consequently,
weakly hydrogen-bonding anions cannot compete with the
solvent. In the gas phase, however, the competition does not
exist and the intrinsic attractive forces are seen that hold the
complex together. It should be briefly noted that we do not
use the mass spectroscopic data to monitor solution con-
centrations of the complexes, but simply refer to the fact
that the mere existence of a complex in the gas phase pro-
vides evidence for an attractive binding interaction. In that
sense, the spray solvent (acetonitrile or methanol in our ex-
periments) does not play a role.
Figure 7. (a) ESI-FTICR mass spectrum of HCOO–·4 after mass
selection. The small signal at m/z 1102 corresponds to stray radia-
tion. (b) IRMPD experiment with 500 ms irradiation time of a
25 W IR laser. (c) IRMPD experiment conducted with Cl–·4 under
the same conditions. (d) Analogous IRMPD experiment with F–·4.
Furthermore, electrostatic interactions are significantly
increased in the gas phase as compared to solution due to
the change in permittivity of the medium as expressed in
the dielectric constants. The vacuum dielectric constant is 1
by definition, while that of DMSO is 48. Consequently, any
electrostatic interaction between the anion and the dipoles
of the receptor should increase in strength by a factor of
approximately 48 upon the transition into the gas phase.
In a tandem mass spectrometric experiment, the complex
ions of interest can be mass-selected and subjected to gas-
phase chemical experiments. After isolation of the parent
ions, a 25 W CO2 laser (10.6 µm wavelength) was used to
fragment the ions in the FTICR cell in an infrared radiative
multiphoton dissociation (IRMPD) experiment. The corre-
sponding spectra for three selected anions, i.e. the formate,
chloride, and fluoride complexes of 4, are shown in Fig-
ure 7. These complexes show interesting differences in their
fragmentation behavior (Scheme 3):
Scheme 3. Fragmentation reactions observed in the tandem MS
spectra shown in Figure 7.
(a) The fragmentation spectrum of the formate complex
is very simple. The only dominating fragmentation channel
(c) For the fluoride complex, no loss of the complete re-
is the loss of the complete neutral receptor (Scheme 3, chan- ceptor is observed anymore. Instead, the anion is so nucleo-
nel A). philic in the absence of solvent that channel B becomes by
(b) For the chloride complex, the complete receptor loss far the major one. The fact that the fragment appears at the
is again by far the most abundant fragmentation channel, same m/z as that observed in the MS/MS spectrum of the
but in addition a quite low-intensity fragment is observed chloride complex also confirms that the halide is incorpo-
at m/z 1057. It can be attributed to a nucleophilic attack of rated in the neutral fragment in both cases. This reaction is
the almost naked chloride at the spacer connecting one of already quite remarkable in that a reaction involving coval-
the branches to the resorcinarene scaffold (Scheme 3, chan- ent bonds smoothly proceeds as the major reaction pathway
nel B).
within a noncovalent complex.[18] Noncovalent bond frag-
6156
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 6151–6159

Resorcinarene Podand with Amine-Functionalized Side Arms
and neutralized by the addition of 2  HCl solution. The organic
layer was separated and washed two times with water, dried with
MgSO4 and the solvents evaporated to dryness under vacuum.
Recrystallization from hot chloroform/hexane afforded 5.6 g (75%)
of orange powdery solid; m.p. 122–124 °C. 1H NMR (CDCl3): δ =
mentation cannot efficiently compete indicating how
strongly bound the fluoride indeed is. But a second aspect
is also important to note: As a minor fragment, a loss of
HF is observed at m/z 1191. In view of the proton affinities
of fluoride (1529 kJ/mol)[19] and anilide (1502 kJ/mol)[20] as
a model compound for the receptor for which the gas-phase
thermochemical data is known, fluoride is capable of de-
protonating the receptor in the gas phase. This aspect nicely
3
3
8.13 (dd, J = 1.5, J = 7.0 Hz, 1 H, ArH), 7.98 (br. s, 1 H, NH),
7.73 (m, 2 H, ArTs), 7.41 (m, 1 H, ArH), 7.25 (m, 2 H, ArTs), 6.75
(dd, 3J = 0.9 and 3J = 7.8 Hz, 1 H, ArH), 6.69 (m, 1 H, ArH), 4.27
3
3
(t, J = 5.5 Hz, 2 H, CH2), 3.63 (q, J = 5.7 Hz, 2 H, CH2), 2.40
(s, 3 H, TsCH3) ppm. 13C NMR (CDCl3, 126 MHz): δ = 145.2,
closes the cycle back to the above discussed solution-phase
–
binding of the HF2 anion. These fragmentation spectra 144.4, 136.2, 132.6, 132.5, 129.8, 127.8, 127.0, 116.1, 113.2, 67.3,
41.5, 21.6 ppm. MS (ESI-TOF): m/z = 359.03 [M + Na]+.
clearly reflect the properties of the different anions, for
C15H16N2O5S (336.37): calcd. C 53.56, H 4.79, N 8.33; found C
example the much higher nucleophilicity and basicity of
53.46, H 4.66, N 8.17.
fluoride.
Tetramethoxy-tetrakis{[2-(2-nitrophenyl)amino]ethoxy}resorcinarene
(3): A mixture of tetramethoxyresorcinarene[5] (1.0 g, 1.5 mmol),
Cs2CO3 (4.1 g, 12.6 mmol) and dibenzo-18-crown-6 (0.41 g,
1.1 mmol) was suspended in dry acetonitrile (60 mL) under nitro-
gen and refluxed for 15 min before the dropwise addition of 2
(2.2 g, 6.5 mmol) in acetonitrile (40 mL) with vigorous stirring. The
Conclusions
In conclusion, the synthesis and structural properties of
a neutral resorcinarene host 4, bearing four amine-function-
alized side arms, is reported. The complexation properties resulting yellow suspension was refluxed overnight. After cooling
–
of 4 toward various anions (halides, AcO–, HCOO–, NO3 ,
to room temperature the reaction mixture was filtered by suction
–
–
and solvent was evaporated under vacuum. The residue was dis-
solved in dichloromethane and washed with water. The organic
layer was separated, dried with MgSO4 and the solvents evaporated
to dryness under vacuum. Recrystallization from methanol/chloro-
form afforded 1.6 g (80%) of orange crystalline solid; m.p. 195–
PF6 , and BF4 ) were investigated both in solution and in
the gas phase. In solution, resorcinarene host 4 readily
formed 1:2 host-guest complexes with fluoride and acetate
anions, and the basicity of these anions seemed to be the
driving force in the formation of hydrogen-bonding interac-
tions with receptor 4. With fluoride, in particular, the bind-
ing was identified as a two step process, which involved the
1
3
198 °C. H NMR (CDCl3): δ = 8.23 (br. t, J = 5.3 Hz, 4 H, NH),
8.17 (dd, 3J = 1.6, 3J = 7.0 Hz, 4 H, ArH), 7.42 (m, 4 H, ArH),
3
6.88 (d, J = 8.0 Hz, 4 H, ArH), 6.68 (s, 4 H, ArHreso), 6.65 (m, 4
–
H, ArH), 6.25 (s, 4 H, ArHreso), 4.43 (t, 3J = 7.6 Hz, 4 H, CH),
formation of a bifluoride, HF2 anion in the second stage
of the binding. Fluoride also represents a special case in the 4.07 (m, 4 H, CH2CH2), 4.83 (m, 4 H, CH2CH2) 3.51–3.46 (over-
lapping s and m, 20 H, OCH3 and CH2CH2), 1.86 (m, 8 H,
CH2CH3), 0.89 (t, 3J = 7.3 Hz, 12 H, CH3) ppm. 13C NMR
(CDCl3, 126 MHz): δ = 155.8, 154.8, 145.3, 136.1, 132.3, 127.0,
126.9, 126.7, 126.3, 115.5, 113.7, 97.9, 67.5, 55.8, 42.6, 37.0, 27.8,
gas phase. Its high basicity leads to HF losses, while no HA
losses have been observed for any of the other anions. In
addition, fluoride is also the most nucleophilic anion under
study. It is interesting to see how the high hydrogen-bond
energies lead to nucleophilic substitution reactions rather
than the cleavage of the noncovalent bonds. The strong ef-
fects of the surrounding environment become again clearly
visible.
12.6 ppm. MS (ESI-TOF): m/z
=
1335.54 [M
+
Na]+.
C72H80N8O16·0.5CHCl3 (1373.17): calcd. C 63.40, H 5.91, N 8.16;
found C 63.39, H 5.71, N 8.03.
Tetrakis{[2-(2-aminophenyl)amino]ethoxy}-tetramethoxyresorcinar-
ene (4): A mixture of 3 (1.6 g, 1.2 mmol), 60% sodium sulfide
(5.9 g, 75.6 mmol) and sulfur (1.3 g, 40.5 mmol) were suspended in
1-butanol (150 mL) and refluxed overnight. After cooling to room
temperature, water (100 mL) was added in the reaction mixture
with vigorous stirring. The organic layer was separated, washed
repeatedly with water and finally treated with hexane. The precipi-
tate formed was separated by suction filtration, washed with meth-
anol and dried in vacuo. Yield 1.0 g (70%) of beige powdery solid;
Experimental Section
General: 1H and 13C NMR spectra were recorded with a Bruker
Avance DRX 500 spectrometer and chemical shifts were calibrated
to the residual proton and carbon resonance of the solvent. Rou-
tine ESI mass spectra were measured with Micromass LCT ESI-
TOF instrument. Elemental analyses were determined with Vario
EL III instrument. Melting points were measured in open capillar-
ies with a Stuart Scientific SMP3 melting point apparatus and are
uncorrected. Tetramethoxy resorcinarene[5] and 1[6] were prepared
according to literature procedures. All other reagents used were
commercial unless otherwise noted. Dichloromethane and acetoni-
trile were distilled from CaCl2 and stored over 3 Å molecular sieves
under N2 atmosphere.
1
m.p. 186–188 °C. H NMR (CDCl3): δ = 6.78 (m, 4 H, ArH) 6.73
(s, 4 H, ArHreso), 6.72–6.64 (m, 12 H, ArH), 6.22 (s, 4 H, ArHreso),
3
4.47 (t, J = 7.6 Hz, 4 H, CH), 4.09 (m, 4 H, CH2CH2), 3.87 (m,
4 H, CH2CH2), 3.48 (s, 12 H, OCH3), 3.38 (m, 4 H, CH2CH2),
3.29 (m, 4 H, CH2CH2), 1.88 (m, 8 H, CH2CH3), 0.92 (t, 3J =
7.2 Hz, 12 H, CH3) ppm. 13C NMR (CDCl3, 126 MHz): δ = 155.7,
154.8, 136.7, 135.6, 126.2, 126.1, 126.0, 119.9, 119.4, 116.0, 112.9,
96.9, 67.2, 56.0, 44.0, 36.9, 27.9, 12.7 ppm. MS (ESI-TOF): m/z =
1193.69 [M]+. C72H88N8O8 (1193.54): calcd. C 72.46, H 7.43, N
9.39; found C 72.13, H 7.47, N 9.00.
o-Nitro-N-(2-tolylsulfonyloxyethyl)aniline (2): A mixture of 1 (4.1 g,
22.5 mmol) and p-toluenesulfonyl chloride (5.0 g, 26.2 mmol) was
dissolved in dry dichloromethane (50 mL) under nitrogen with stir-
ring. Triethylamine (3.6 mL, 26.0 mmol) in dichloromethane was
added dropwise and the reaction mixture was stirred at room tem-
perature for two days. Water was added to the reaction mixture
NMR Spectroscopic Methods: 1H NMR titrations were done by
subsequently adding increasing aliquots of the F– and AcO– anions
as their TBA salts in [D6]DMSO in the solution of the host in [D6]-
DMSO and recording the spectra at 303 K with Bruker Avance
Eur. J. Org. Chem. 2009, 6151–6159
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6157

K. Salorinne, D. P. Weimann, C. A. Schalley, M. Nissinen
FULL PAPER
DRX 500 spectrometer. The obtained titration data was analyzed
by the computer program EQNMR.[15] For the other investigated
1.231 Mgm–3; 2θ range = 3.19–67.29°; 17646 reflections collected;
11837 independent reflections (Rint = 0.237); 912 parameters; R1 =
0.059, wR2 = 0.130 for observed data [IϾ2σ(I)]; R1 = 0.092, wR2
= 0.150 for all data; GOF = 1.062. Disordered solvent acetonitrile
over two positions with site occupancy of 0.6:0.4. The bonding
distances are restrained to be equal (SADI). DFIX used for amine
hydrogen atoms H53, H70B, H80A and H80B.
anions (Cl–, Br–, I–, NO3 , BF4 and PF6 ) a 10 fold excess of the
salt (added as TBA salts) confirmed there was no interaction be-
tween the host and the guest species. Job plot samples were pre-
pared with 0.3, 0.5, 1, 2, and 3 equiv. of the salt while keeping the
sum of host and guest concentration equal.
–
–
–
Ib: Colorless block crystals (0.10ϫ0.10ϫ0.05 mm3) were grown by
slow evaporation from acetonitrile/ethanol solution of 4. Crystal
data: C72H88N8O8·CH3CN·CH3CH2OH; Mr = 1280.63; triclinic;
NMR diffusion measurements were performed by using Bruker Av-
ance 400 MHz spectrometer equipped with a Great 1/10 pulsed gra-
dient unit and a direct probe at 303 K. A LED[21] pulse sequence
was used for the diffusion measurements with a sine-shape pulsed
gradient duration δ of 2.5 ms incremented from 0 to 27.0 Gcm–1 in
twelve steps. The pulsed gradient separation ∆ was 100 ms and the
eddy current delay was 5 ms. The reported diffusion coefficients are
the average Ϯ standard deviation of at least three different mea-
surements. The samples were prepared 5 m in 4 and 10 m in the
anion as its TBA salt.
¯
space group P1; a = 15.2189(5), b = 15.8530(5), c = 16.5455(5) Å;
α = 74.357(3), β = 83.113(2), γ = 64.094(2)°; V = 3457.7(2) Å3; Z =
2; ρcalcd. = 1.230 Mgm–3; 2θ range = 3.19–67.02°; 14508 reflections
collected; 11223 independent reflections (Rint = 0.136); 912 param-
eters; R1 = 0.083, wR2 = 0.199 for observed data [IϾ2σ(I)]; R1 =
0.142, wR2 = 0.240 for all data; GOF = 1.033; Disordered solvent
acetonitrile over two positions with site occupancy of 0.2:0.8.
II: Colorless block crystals (0.30ϫ0.25ϫ0.10 mm3) were grown by
slow evaporation from acetonitrile/TBACl solution of 4. Crystal
data: C72H88N8O8·CH3CN; Mr = 1234.56; triclinic; space group
Gas-Phase Binding Studies: All gas-phase experiments described
herein were conducted with an Ionspec QFT-7 FT-ICR mass spec-
trometer (Varian Inc., Lake Forest, CA), equipped with a 7 T su-
perconducting magnet and a Micromass Z-Spray electrospray ion-
ization (ESI) source (Waters Co., Saint-Quentin, France). The sam-
ples were introduced into the source as 50 µ solutions of 4 and
the corresponding anion in acetonitrile (or methanol) at flow rates
of 1–2 µL/min. A constant spray and highest intensities were
achieved with a capillary voltage of 3800 V at a source temperature
of 40 °C. The parameters for sample cone and extractor cone volt-
age were optimized for maximum intensitites of the desired com-
plexes. Multiple scans (10–20) were recorded and averaged for each
spectrum in order to improve the signal-to-noise ratio. After ac-
cumulation and transfer into the instrument’s FTICR analyzer cell,
the ions were detected by a standard excitation and detection se-
quence.
¯
P1; a = 14.9523(3), b = 14.9859(4), c = 17.0271(4) Å; α =
109.464(2), β = 100.650(2), γ = 104.694(2)°; V = 3324.4(1) Å3; Z =
2; ρcalcd. = 1.233 Mgm–3; 2θ range = 2.88–63.45°; 16387 reflections
collected; 10787 independent reflections (Rint = 0.045); 910 param-
eters; R1 = 0.049, wR2 = 0.119 for observed data [IϾ2σ(I)]; R1 =
0.071, wR2 = 0.133 for all data; GOF = 1.017; disordered ethyl
chain C35-C36 and methoxy group O11–C38 over two positions
with the site occupancies of 0.6:0.4 and 0.7:0.3, respectively. Disor-
dered solvent acetonitrile over two positions with site occupancy
of 0.7:0.3 and the bonding distances are restrained to be equal
(SADI).
III: Colorless block crystals (0.15ϫ0.15ϫ0.10 mm3) were grown
by slow evaporation from pyridine solution of 4. Crystal data:
C72H88N8O8·3C5H5N; Mr = 1430.80; monoclinic; space group P21/
c; a = 13.6942(3), b = 22.2544(4), c = 25.7929(5) Å; β = 96.985(1)°;
V = 7802.2(3) Å3; Z = 4; ρcalcd. = 1.218 Mgm–3; 2θ range = 2.63–
67.09°; 23691 reflections collected; 13318 independent reflections
(Rint = 0.061); 1025 parameters; R1 = 0.066, wR2 = 0.154 for ob-
served data [IϾ2σ(I)]; R1 = 0.114, wR2 = 0.185 for all data; GOF
= 1.026; Disordered solvent pyridine over two positions with site
occupancy of 0.8:0.2. Restraints SAME and EADP used for the
disordered part.
For the fragmentation experiments, the ions of interest were mass
selected in the ICR cell and irradiated with a 25 W CO2 laser in the
IR region [infrared multiphoton dissociation (IRMPD), 10.6 µm
wavelength] to induce fragmentation.
Crystal Structure Determination: Data were recorded by using a
Nonius Kappa CCD diffractometer having an Apex II detector
with graphite-monochromatized Cu-Kα [λ(Cu-Kα) = 1.54178 Å] ra-
diation at temperature of 173.0 K. The data were processed with
Denzo-SMN v0.97.638.[22] The structures were solved by direct
methods (SHELXS-97[23]) and refinements based on F2 were made
by full-matrix least-squares techniques (SHELXL-97[24]). The hy-
drogen atoms were calculated to their idealized positions with iso-
tropic temperature factors (1.2 or 1.5 times the C temperature fac-
tor) and refined as riding atoms, except for the amine hydrogen
atoms, which were located from the difference Fourier map. Ab-
sorption correction[25] was made to all structures, but used only in
the structure Ib since it worsened the R-value with all the other
structures.
Supporting Information (see also the footnote on the first page of
1
this article): H and 13C NMR spectra of 2, 3, and 4; crystal struc-
ture of 3; Job plot and NMR titration data; gas-phase binding
data.
Acknowledgments
K. S. wishes to thank the Academy of Finland for funding (project
116503), Mr. Reijo Kauppinen for his help with the NMR measure-
ments and B. Sc. Katri Moilanen for her help with the synthesis.
CCDC-705756 (for Ia), -705757 (for Ib), -735535 (for II) and
-735536 (for III) contain the supplementary crystallographic data
for this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
[1] For a review see: a) P. Gale, Acc. Chem. Res. 2006, 39, 465–
475; b) P. A. Gale, R. Quesada, Coord. Chem. Rev. 2006, 250,
3219–3244; c) S. E. Matthews, P. D. Beer, Supramol. Chem.
2005, 17, 411–435; d) P. D. Beer, P. A. Gale, Angew. Chem. Int.
Ed. 2001, 40, 486–516; e) K. Bowman-James, Supramolecular
Chemistry of Anions (Eds.: K. Bowman-James, E. García-Es-
paña), Wiley-VCH, New York, NY, 1997.
Ia: Colorless block crystals (0.30ϫ0.15ϫ0.05 mm3) were grown
by slow evaporation from acetonitrile solution of 4. Crystal data:
¯
C72H88N8O8·2CH3CN; Mr = 1275.61; triclinic; space group P1; a
= 15.3148(5), b = 15.8823(5), c = 16.3810(5) Å; α = 74.421(1), β =
[2] Recent examples: a) P. A. Gale, Chem. Commun. 2008, 4525–
4540; b) Y. H. Qiao, H. Lin, H. K. Lin, J. Incl. Phenom. Macro-
82.453(1), γ = 63.744(1)°; V = 3441.7(2) Å3; Z = 2; ρcalcd.
=
6158
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 6151–6159

Resorcinarene Podand with Amine-Functionalized Side Arms
cyclic Chem. 2007, 59, 211–215; c) A. P. Davis, Coord. Chem.
Rev. 2006, 250, 2939–2951; d) M. D. Lankshear, P. D. Beer, Co-
ord. Chem. Rev. 2006, 250, 3142–3160.
ney, H. D. P. Ali, G. M. Hussey, J. Org. Chem. 2005, 70, 10875–
10878; f) D. E. Gómez, L. Fabbrizzi, M. Licchelli, E. Monzani,
Org. Biomol. Chem. 2005, 3, 1495–1500; g) L. S. Evans, P. A.
Gale, M. E. Light, R. Quesada, Chem. Commun. 2006, 965–
967; h) C. Liu, X. Qian, J. Wang, Z. Li, Tetrahedron Lett. 2008,
49, 1087–1090; i) C. Caltagirone, G. W. Bates, P. A. Gale, M. E.
Light, Chem. Commun. 2008, 61–63.
[3] a) K. Bowman-James, Acc. Chem. Res. 2005, 38, 671–678; b)
S. O. Kang, M. A. Hossain, K. Bowman-James, Coord. Chem.
Rev. 2006, 250, 3038–3052.
[4] M. M. G. Antonisse, D. N. Reinhoudt, Chem. Commun. 1998,
443–448.
[5] M. J. McIldowie, M. Mocerino, B. W. Skelton, A. H. White,
Org. Lett. 2000, 2, 3869–3871.
[6] a) J. M. McManus, R. M. Herbst, J. Org. Chem. 1959, 24,
1042–1044; b) L. Westerheide, F. K. Müller, R. Than, B. Krebs,
Inorg. Chem. 2001, 40, 1951–1961.
[12] S. O. Kang, D. Powell, V. W. Day, K. Bowman-James, Angew.
Chem. Int. Ed. 2006, 45, 1921–1925.
[13] T. Steiner, Angew. Chem. Int. Ed. 2002, 41, 48–76.
[14] K. Hirose, J. Incl. Phenom. Macrocyclic Chem. 2001, 39, 193–
209.
[15] M. J. Hynes, J. Chem. Soc., Dalton Trans. 1993, 311–312.
[16] The broadening of the amine resonances was associated with
the increasing amount of the intrinsic water in the system in
the course of the titration experiment.
[7] M. Luostarinen, K. Salorinne, H. Lähteenmäki, H. Man-
sikkamäki, C. A. Schalley, M. Nissinen, K. Rissanen, J. Incl.
Phenom. Macrocyclic Chem. 2007, 58, 71–80.
[8] The aromatic resorcinarene cavity is known to bind ammonium
cations, see for example: a) H. Mansikkamäki, C. A. Schalley,
M. Nissinen, K. Rissanen, New J. Chem. 2005, 29, 116–127; b)
J. L. Atwood, A. Szumna, J. Supramol. Chem. 2003, 2, 479–
482; c) A. R. Mustafina, S. V. Fedorenko, N. A. Makarova,
E. K. Kazakova, Z. G. Bazhanova, V. E. Kataev, A. I. Konova-
lov, J. Incl. Phenom. Macrocyclic Chem. 2001, 40, 73–76.
[9] S. S. Zhu, H. Staats, K. Brandhorst, J. Grunenberg, F. Gruppi,
E. Dalcanale, A. Lützen, K. Rissanen, C. A. Schalley, Angew.
Chem. Int. Ed. 2008, 47, 788–792.
[17] a) O. Mayzel, Y. Cohen, J. Chem. Soc., Chem. Commun. 1994,
1901–1902; b) S. Simova, S. Berger, J. Inclusion Phenom.
Macrocyclic Chem. 2005, 53, 163–170.
[18] For an intriguing example of covalent bond formation in a
noncovalent complex, also see: R. R. Julian, J. A. May, B. M.
Stoltz, J. L. Beauchamp, Angew. Chem. Int. Ed. 2003, 42, 1012–
1015.
[19] C. Blondel, C. Delsart, F. Goldfarb, J. Phys. B 2001, 34, L281–
L288.
[20] J. E. Bartmess, J. A. Scott, R. T. McIver Jr., J. Am. Chem. Soc.
1979, 101, 6046–6056.
[21] J. S. Gibbs, C. S. Johnson, J. Magn. Reson. 1991, 93, 395–402.
[22] Z. Otwinowski, W. Minor, in: Molecular Crystallography, Part
A, vol. 276 (Eds: C. W. Carter, R. M. Sweet), Academic Press,
New York, 1997, pp. 307–326.
[23] a) G. M. Sheldrick, Acta Crystallogr., Sect. A 2008, 64, 112–
122; b) G. M. Sheldrick, Acta Crystallogr., Sect. A 1990, 46,
467–473.
[24] G. M. Sheldrick, A program for crystal structure refinement,
University of Göttingen, Göttingen (Germany), 1997.
[10] F. G. Bordwell, Acc. Chem. Res. 1988, 21, 456–463. For com-
parison some -NH group acidities in DMSO: aniline pKa
=
30.6, urea pKa = 26.9, diphenylamine pKa = 24.9, thiourea pKa
= 21.1. The acidities (pKa values) of HF, HCl, HBr, AcOH and
H2O in DMSO are 15, 1.8, 0.9, 12.3 and 32, respectively.
[11] a) T. Gunnlaugsson, P. E. Kruger, P. Jensen, F. M. Pfeffer,
G. M. Hussey, Tetrahedron Lett. 2003, 44, 8909–8913: ; b) S.
Camiolo, P. A. Gale, M. B. Hursthouse, M. E. Light, Org. Bi-
omol. Chem. 2003, 1, 741–744; c) . M. Boiocchi, L. Del Boca,
D. E. Gómez, L. Fabbrizzi, M. Licchelli, E. Monzani, J. Am.
Chem. Soc. 2004, 126, 16507–16514; d) V. Amendola, M. [25] R. H. Blessing, Acta Crystallogr., Sect. A 1995, 51, 33–38.
Boiocchi, L. Fabbrizzi, A. Palchetti, Chem. Eur. J. 2005, 11,
5648–5660; e) T. Gunnlaugsson, P. E. Kruger, P. Jensen, J. Tier-
Received: July 20, 2009
Published Online: October 27, 2009
Eur. J. Org. Chem. 2009, 6151–6159
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6159