Supramolecular sensing with phosphonate cavitands

Article abstract of DOI:10.1002/chem.200800327

Phosphonate cavitands are an emerging class of synthetic receptors for supramolecular sensing. The molecular recognition properties of the third-generation tetraphosphonate cavitands toward alcohols and water at the gas-solid interface have been evaluated

Full text of DOI:10.1002/chem.200800327

DOI: 10.1002/chem.200800327
Supramolecular Sensing with Phosphonate Cavitands
Monica Melegari,[a] Michele Suman,[a] Laura Pirondini,[a] Davide Moiani,[a]
Chiara Massera,[b] Franco Ugozzoli,[b] Elina Kalenius,[c] Pirjo Vainiotalo,[c]
Jean-Christophe Mulatier,[d] Jean-Pierre Dutasta,[d] and Enrico Dalcanale*[a]
In memory of Dmitry M. Rudkevich
Abstract: Phosphonate cavitands are
an emerging class of synthetic recep-
tors for supramolecular sensing. The
molecular recognition properties of the
third-generation tetraphosphonate cav-
itands toward alcohols and water at the
gas–solid interface have been evaluated
by means of three complementary
techniques and compared to those of
the parent mono- and diphosphonate
cavitands. The combined use of ESI-
MS and X-ray crystallography defined
precisely the host–guest association at
the interface in terms of type, number,
strength, and geometry of interactions.
Quartz crystal microbalance (QCM)
measurements then validated the pre-
dictive value of such information for
sensing applications. The importance of
energetically equivalent multiple inter-
actions on sensor selectivity and sensi-
tivity has been demonstrated by com-
paring the molecular recognition prop-
erties of tetraphosphonate cavitands
with those of their mono- and di-
phosphonate counterparts.
Keywords: cavitands
chemistry mass spectrometry
molecular recognition · sensors
· host–guest
·
·
Introduction
practical applications. For example, in chemical sensing, se-
lectivity is the key parameter in defining success or failure
of a given sensor, together with ruggedness.[1]
Molecular recognition at interfaces constitutes an important
theme in contemporary supramolecular chemistry. Address-
ing this issue is essential both for basic knowledge and for
Reliable methods to assess and predict the complexation
properties of molecular receptors at interfaces are therefore
in demand. Both in natural and synthetic receptors, the
basic tenet of most molecular recognition phenomena is
their operation in solution, in which general dispersion inter-
actions largely cancel out in moving the substrate from the
solvent to the receptor site and in which the entropic cost of
binding is partly paid by solvent release in the bulk. The
same does not hold for recognition processes involving gas-
eous species, for which general, nonspecific dispersion inter-
actions predominate in moving to the liquid/solid state.[2] In
developing highly selective gas sensors, achieving effective
molecular recognition at the gas–solid interface is therefore
a demanding task that requires a fresh approach, both in
terms of receptor design and characterization tools.[3]
[a]Dr. M. Melegari, Dr. M. Suman, Dr. L. Pirondini, Dr. D. Moiani,
Prof. E. Dalcanale
Dipartimento di Chimica Organica ed Industriale and
Unità INSTM, UdR Parma, Università degli Studi di Parma
Viale G.P. Usberti 17/a, 43100 Parma (Italy)
Fax : (+39)0521-905-472
E-mail: enrico.dalcanale@unipr.it
[b]Dr. C. Massera, Prof. F. Ugozzoli
Dipartimento di Chimica Generale ed Inorganica
Chimica Analitica, Chimica Fisica
Università degli Studi di Parma
Viale G.P. Usberti 17/a, 43100 Parma (Italy)
[c]Dr. E. Kalenius, Prof. P. Vainiotalo
Department of Chemistry, University of Joensuu
P.O. Box 111, 80101 Joensuu (Finland)
The goal of this work is to provide a clear understanding
of the parameters affecting molecular recognition of phos-
phonate cavitands at the gas–solid interface, with predicting
values on their gas-sensing properties. To achieve this goal
we set up an effective methodology for probing the molecu-
lar recognition properties of a given receptor at the gas–
solid interface. It relies on the combined use of ESI-MS,
[d]J.-C. Mulatier, Dr. J.-P. Dutasta
Laboratorie de Chimie
École Normale SupØrieure de Lyon
CNRS, 46 AllØe dꢀItalie, 69364 Lyon 07 (France)
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
5772
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 5772 – 5779

FULL PAPER
X-ray crystallography, and quartz crystal microbalance
(QCM)[1] measurements. ESI-MS and X-ray crystallography
were employed for the evaluation of the complexation prop-
erties of the receptors in the gas phase and solid state, re-
spectively. They provide precise information on the type,
number, strength, and geometry of the weak interactions re-
sponsible for the host–guest associations. The predictive
value of such information for gas–solid interactions is vali-
dated by coating QCM transducers with the receptors and
by exposing them to the analytes. By comparing the QCM
responses of different receptors, the contributions of specific
and nonspecific interactions can be dissected.
p-basic cavity, which requires an inward (i) orientation of
the P=O bridges.[5]
2) A rigid cavity that provides a permanent free volume for
the analyte around the inward facing P=O groups, pivo-
tal for effective hydrogen bonding.[6]
3) The presence of two energetically equivalent hydrogen-
bonding options available to the analyte in the case of
AB/AC derivatives.[7]
The last factor suggests that increasing the number of con-
vergent P=O groups in the cavitand should enhance alcohol
complexation at the gas-solid interface.
To this purpose we prepared and tested third-generation
cavitand receptors presenting four inward facing P=O
bridges and compared them with their mono- and diphosph-
onate analogues, representing the first- and second-genera-
tion species, respectively. The corresponding tetrathio-
phosphonate cavitand TSiiii[C11H23,H,Ph]was also prepared
to test the influence of hydrogen bonding on the sensing re-
sponses.
Results and Discussion
The compounds used in the present work are the phospho-
nate cavitands shown below. They present an open, confor-
mationally rigid cavity, delimited by one (Mi), two (ABii/
ACii/ACio), or four (Tiiii) phosphonate bridging groups at
the upper rim.[4]
In previous studies we have shown that the key factors af-
fecting the sensing performances of mono- and diphospho-
nate cavitands toward alcohols are:
Synthesis:
Cavitands
Mi[C11H23,H,Ph],[4]
ABii[C11H23,H,Ph],[6] and TSiiii[C11H23,H,Ph][8] were pre-
pared following reported procedures. The Tiiii[C11H23,H,Ph]
and Tiiii[H,CH3,CH3]cavitands were prepared according to
the procedure reported in reference [9]. The preparation of
1) The simultaneous presence of hydrogen bonding with
one of the P=O groups and CH–p interactions with the
the AC–diphosphonate cavitands ACii
ACHTRE[UGN H,CH3,Ph]and
ACio[H,CH3,Ph]followed
A
a
two-step synthesis, in which
partial methylene bridging of
the starting resorcinarene was
performed first, leaving the
phosphonate bridging as last
step (see Scheme S1, in the
Supporting Information). The
removal of the alkyl “feet” at
the lower rim requires the
presence of a methyl group in
the 2-position of the starting
resorcinol to avoid polymeri-
zation in the acid-catalyzed re-
sorcinarene condensation reac-
tion with formaldehyde.[10]
Therefore all cavitands pre-
pared for crystal structure
studies have four methyl sub-
stituents at the upper rim.
Solid-state studies
ACii/ACio cavitands: Mono-
and diphosphonate cavitands
present a totally different be-
havior as receptor layers for al-
cohols in mass sensors accord-
ing to the relative orientation
of the P=O bridges. A single
Chem. Eur. J. 2008, 14, 5772 – 5779
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5773

E. Dalcanale et al.
P=O group pointing outward is sufficient to switch off the
sensor responses.[5] Ideally, this remarkable difference
should be reflected in the crystal structure of the respective
alcohol complexes. Moreover, in the case of diphosphonate
cavitands, the presence of a second inward facing P=O
group enhances the sensor responses, independently from
the relative position of the two moieties. The additional con-
verging P=O group offers a second energetically and geo-
metrically equivalent hydrogen bond to the guest, as re-
vealed
by
the
crystal
structure
of
the
ABii[C2H5,H,Ph]·MeOH complex for the AB isomer.[7]
Whether or not this behavior is also followed by the AC
isomer remains to be proven. Compact cavitands ACii-
Figure 2. Molecular structure of one of the two symmetry independent
cavitands in 2ACio[H,CH3,Ph]·EtOH·CH2Cl2 (color code: P, orange; O,
red; C, grey). The hydrogen atoms and the solvent molecules in the lat-
tice are omitted for clarity.
ACHTREUNG[H,CH3,Ph]and ACio ACHTRE[UGN H,CH3,Ph]were synthesized to clarify
their interaction mode with alcohols, removing the R sub-
stituents at the lower rim to facilitate single-crystal forma-
tion.
The crystals were obtained from a liquid diffusion of
EtOH into a solution of the cavitands in CH2Cl2. The struc-
It is interesting to note that the ethanol molecules in the
lattice (see Figure S1, Supporting Information) do not gener-
ate any hydrogen-bonding interaction with the P=O group
ture of the ACii
A
oriented outside the cavity. Also for the ACioACHTRE[UNG H,CH3,Ph]
Figure 1. The ethanol molecule is located inside the cavitand
isomer, crystallographic data are in agreement with sensors
measurements, showing a reduced affinity toward alcohols.
Tiiii cavitands: The influence of multiple hydrogen-bond ac-
ceptor sites on alcohol complexation has been assessed by
synthesizing tetraphosphonate cavitands, with all four P=O
bridges pointing into the cavity. Cavitand Tiiii[H,CH3,CH3]
was specially synthesized for crystal structure determina-
tions. Figure 3 shows the molecular structure of the
Figure 1. Molecular structure of the ACii[H,CH3,Ph]·EtOH complex
showing the two orientations of the hydroxyl hydrogen atoms (color
code: P, orange; O, red; C, grey; H, white). The hydrogen atoms not in-
volved in hydrogen bonding are omitted for clarity.
with its methyl group forming weak CH–p interactions with
the aromatic walls of the cavity. The complex is further sta-
bilized by the hydrogen bond between the OH group and
the P=O moieties. As observed in the ABii-
ACHTREUNG[H,CH3,Ph]·MeOH case, the probability of interaction is
Figure 3. Molecular structure of the Tiiii[H,CH3,CH3]·MeOH complex
showing the four orientations of the hydroxyl hydrogen atoms (color
code: P, orange; O, red; C, grey; H, white). The hydrogen atoms not in-
volved in the weak interactions are omitted for clarity.
maximized: the solvent is disordered with 50% probability
over two equivalent orientations, thus forming alternatively
a hydrogen bond with each of the two opposite P=O groups
(O···O=P 2.795(3) Š). Therefore there is no difference in
the binding mode in the solid state between AB and AC iso-
mers, despite of the different spatial orientation of the P=O
groups, as reflected in their comparable responses as sen-
sors.[7]
Tiiii[H,CH3,CH3]·MeOH complex. The complex crystallizes
À
in the tetragonal P4/n space group, with the methanol C O
bond lying on the fourfold axis, so that the methanol hydro-
gen atoms are statistically disordered over four different ori-
entations. The alcohol interacts with the cavity through
three CH–p interactions involving the methyl group and the
aromatic rings of the cavitand (the CH···centroid distances
span from 2.845(7) to 3.060(7) Š) and forming weak hydro-
gen bonds between the hydroxyl group and the phosphonate
The crystals of cavitand ACio
grown under the same conditions of those of complex ACii-
ACHTREUNG[H,CH3,Ph]·EtOH. The phenyl ring pointing inside the
ACHTRE[UNG H,CH3,Ph](Figure 2) were
cavity prevents any solvent inclusion, so that the disordered
ethanol and dichloromethane molecules can only fill the
empty spaces in the crystal lattice.
5774
www.chemeurj.org
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 5772 – 5779

Phosphonate Cavitands
FULL PAPER
moieties at the upper rim (MeOH···O=P 3.059(6) Š; Fig-
ure S2, Supporting Information).[11] Due to its statistical dis-
order around the fourfold axis, the methanol can switch
among four different but isoenergetic triplets of CH···p in-
teractions with the host cavity and at the same time it can
also switch among four different isoenergetic MeOH···O=P
attractive interactions with the host, leading to an entropic
stabilization of the complex.
Gas-phase studies: No ethanol complexes were formed in
the gas phase with cavitands presenting one or two P=O
groups directed outward from the cavity.[12] These results
clearly indicate that the cooperativity between hydrogen
bonding and cavity inclusion is essential for complexation.
The outward facing P=O are ineffective in hydrogen bond-
ing both in the solid state and in the gas phase. Therefore
ESI-MS competition experiments were restricted to all
inward facing P=O/P=S cavitands to estimate their relative
affinity toward ethanol. All competition experiments were
carried out in 1:1 ethanol/acetonitrile solvent mixtures con-
taining equimolar amounts of two cavitands. Mean intensi-
ties of ethanol complexes formed by Mi[C11H23,H,Ph]versus
Figure 5. Dissociation
Mi[C11H23,H,Ph], ABii[C11H23,H,Ph]and Tiiii[C 11H23,H,Ph]. Normalized
intensities as a function of activation energy in eV.
(CID)
of
the
ethanol
complexes
of
ones. The former gives an estimate of the relative thermody-
namic and kinetic stability of the complexes as function of
the number of inward directed P=O groups present, which
cannot be inferred from the latter.
ABii[C11H23,H,Ph],
and
ABii[C11H23,H,Ph]versus
Tiiii[C11H23,H,Ph]are plotted in Figure 4. The results imply
that the thermodynamic stability of the complexes strongly
depends on the number of P=O groups pointing into the
cavity.
QCM
measurements:
Cavitands
Mi[C11H23,H,Ph],
ABii[C11H23,H,Ph], Tiiii[C11H23,H,Ph], TSiiii[C11H23,H,Ph],
and reference polymer polyepichlorohydrin (PECH) were
deposited by spin coating on both sides of a 10 MHz QCM
transducer[14] and exposed to vapors of C1–C4 linear alcohols
at different concentrations. The presence of long alkyl feet
at the lower rim is essential to impart permeability to the
layer.[3d] Comparison of the responses of the different coat-
ings to MeOH vapors (Figure 6) highlights the positive
effect of the number of inward facing P=O units on the
sensor performances, as predicted.
The dependence of the sensor responses on hydrogen
bonding is evidenced by comparing the behavior of
Tiiii[C11H23,H,Ph]and TSiiii[C 11H23,H,Ph](Figure 7a). The
latter behaves similarly to the unspecific polymeric coating
PECH (Figure 7b). The structural similarity of the two cavi-
tands allows a proper comparison of the sensor data, with-
Figure 4. Mean intensities of the ethanol complexes obtained in competi-
tion experiments for the pairs Mi/ABii and ABii/Tiiii.
Under the same conditions, cavitand TSiiii[C11H23,H,Ph]
did not form complexes with alcohols, due to the reduced
hydrogen-bond acceptor ability of the P=S group compared
to the P=O analogue.[13] The relative kinetic stability of all
three ethanol complexes studied in competition experiments
has been assessed by means of CID (collision induced disso-
ciation) MS/MS experiments. Ethanol complexes of all cavi-
tands dissociated producing the protonated cavitand
[M+H]+. The results shown in Figure 5 clearly indicate that
also the kinetic stability of the ethanol complexes is directly
related to the number of inward facing P=O groups present
in the cavitand. ESI-MS experiments underline the impor-
tance of multiple hydrogen-bonding options available to the
guest for effective complexation in the gas phase.
Figure 6. Selectivity
ABii[C11H23,H,Ph], Tiiii[C11H23,H,Ph], and polymer PECH to methanol
at 1500 ppm.
patterns
of
cavitands
Mi[C11H23,H,Ph],
It is important to stress that the ESI-MS results are not
only supportive, but also complementary to the solid-state
Chem. Eur. J. 2008, 14, 5772 – 5779
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5775

E. Dalcanale et al.
tands, not for the analytes. The influence of dispersion inter-
actions on the QCM responses for long-chain alcohols jeop-
ardizes the identification of a clear trend in alcohol detec-
tion.
The water case: A closer match between prediction and ex-
periment has been achieved in the case of water, an analyte
which does not experience dispersion interactions.
The QCM data (Figure 8) show that TSiiii[C11H23,H,Ph]
and Mi[C11H23,H,Ph]coatings are almost insensitive to
water and that the response of Tiiii[C11H23,H,Ph]is twice
Figure 8. Selectivity
Mi[C11H23,H,Ph], ABii[C11H23,H,Ph], and Tiiii[C11H23,H,Ph]to water
(230 ppm).
patterns
of
cavitands
TSiiii[C11H23,H,Ph],
Figure 7. a) Responses of Tiiii[C11H23,H,Ph]and TSiiii[C 11H23,H,Ph]to
C1–C4 linear alcohols (1500 ppm each); b) responses of cavitands
Tiiii[C11H23,H,Ph], TSiiii[C11H23,H,Ph], and polymer PECH to C1–C5 al-
cohols (25 ppm each).
out any bias due to different solid-state packing and coating
morphology. Substitution of the four P=O groups with the
P=S analogues strongly reduces the responses across the
entire alcohol series, in line with ESI-MS data. The in-
creased weight of dispersion interactions, associated with al-
cohol chain length, determines the general enhancement of
the responses exhibited by both cavitands. This effect is par-
ticularly pronounced at high alcohol concentrations at which
the layer is saturated (1500 ppm, Figure 7a). Nevertheless,
the progressive dilution of specific responses is less evident
at low analyte concentrations (25 ppm, Figure 7b), at which
that of ABii[C11H23,H,Ph]. The same trend is present in the
gas phase: TSiiii[C11H23,H,Ph]and Mi[C 11H23,H,Ph]do not
form any complex with water, while ABii[C11H23,H,Ph]
forms
a
protonated [M+H+H2O]+ complex and
Tiiii[C11H23,H,Ph]a protonated [ M+H+2H2O]+ complex.
The composition of these complexes was verified by a CID
experiment (Figure 9).
The crystal structure of the Tiiii[H,CH3,CH3]·2H2O com-
plex, grown from trifluoroethanol, supports the complex sto-
ichiometry indicated by the gas-phase experiments[15] and re-
veals the preferred host–guest hydrogen-bonding pattern
(Figure 10, see also Figure S4, Supporting Information). The
two water molecules are involved in a zig-zag chain of hy-
drogen bonds with the two distal P=O groups, instead of
binding each of them to a vicinal P=O couple. Therefore, in
the absence of nonspecific dispersion interactions, the bind-
ing mode stoichiometry is fully reflected in the correspond-
ing mass-sensor responses (Figure 8). Such level of defini-
tion of host–guest interactions at the gas–solid interface is
unprecedented, allowing reliable predictivity in sensor re-
sponses for the analyte/receptor pair.
the energetically favorable cavity inclusions dominate.[6]
A
significant enhancement of the responses was only observed
for n-butanol and n-pentanol, indicating that nonspecific
extra cavity physisorption contribute significantly to the
overall QCM response. In all other cases the response is
comparable, suggesting that the inclusion mode is the same
for short-chain alcohols.
The threshold value for alcohol detection has been deter-
mined for the most sensitive Tiiii[C11H23,H,Ph]in the case of
ethanol in the 5–100 ppm regime (see Figure S3, Supporting
Information). With a reproducible response of 8Æ0.6 Hz at
5 ppm and a noise level of Æ1 Hz for 10 MHz transducers,
the lower limit of ethanol detection can be estimated in
2 ppm.
Conclusion
So far, the proposed methodology has shown the ability
to predict the trend in the sensor responses for the cavi-
The third-generation tetraphosphonate cavitand receptors
have been prepared and the corresponding mass sensors
5776
www.chemeurj.org
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 5772 – 5779

Phosphonate Cavitands
FULL PAPER
ing and CH–p interactions, made possible by the insertion
of phosphonate bridges on top of a preorganized cavity. Nei-
ther interaction, taken alone, is sufficient for efficient com-
plexation. QCM measurements showed that the third gener-
ation Tiiii cavitands outperforms the previous two (Mi,
ABii/ACii) in terms of sensor responses. The results demon-
strated that the entropic stabilization of host–guest com-
plexes through energetically equivalent multiple interactions
is a viable route to enhance sensitivity and selectivity in
mass sensors, while retaining full reversibility.
In conclusion, the combined use of ESI-MS and crystallo-
graphic analyses allows us to anticipate the molecular recog-
nition properties of cavitands at the gas–solid interface,
when the dominant interactions in the two phases coincide.
In the absence of dispersion interactions, this approach
allows us to predict the trend in sensor responses of a given
analyte–cavitand pair. The overall methodology is of wide
applicability to many receptor–analyte pairs, allowing an
easy and reliable evaluation of their molecular recognition
potential at interfaces.
Experimental Section
Figure 9. a) CID spectrum (Ecom =1.0 eV) of pre-isolated m/z 1391.7.
Measured from solution containing ABii[C11H23,H,Ph]and 2.5% water in
ACN; b) CID spectrum (Ecom =1.0 eV) of pre-isolated m/z 1629.6. Mea-
sured from solution containing Tiiii[C11H23,H,Ph]and 10% water in
ACN.
Cavitands synthesis
Tiiii[C11H23,H,Ph]: Dichlorophenylphosphate (1.23 mL, 7.96 mmol) and
N-methylpyrrolidine (213 mL, 2.04 mmol) were added, under nitrogen, to
a solution of undecyl-“footed” resorcinarene (2.0 g, 1.81 mmol) in anhy-
drous toluene (100 mL). The mixture was stirred at 1128C for 5 h. After
evaporation of the solvent, the crude product was purified by column
chromatography (SiO2, CH2Cl2/EtOH 9:1) to give the title compound as
white solid (1.44 g, 50%). 1H NMR (CDCl3, 300 MHz): d=8.08–7.98 (m,
8H; POArHo), 7.60–7.53 (m, 12H; POArHm +POArHp), 7.29 (s, 4H;
ArHdown), 6.97 (s, 4H; ArHup), 4.78 (t, 3J=7.6 Hz, 4H; RCHAr2), 2.37
(m, 8H; CHCH2R), 1.43–1.25 (m, 72H; -CH2-), 0.86 ppm (t, 3J=6.9 Hz,
12H; -CH3); 31P NMR (CDCl3, 162 MHz): d=5.83 ppm (s, POAr); ESI-
MS: m/z: 1595.4 [M+H]+, 1618.3 [M+Na]+.
Tiiii[H,CH3,CH3]: Dichloromethylphosphate (1.46 mL, 16.17 mmol) and
N-methylpyrrolidine (432 mL, 4.15 mmol) were added, under nitrogen, to
a suspension of resorcinarene 1 (2.0 g, 3.67 mmol) in anhydrous dichloro-
benzene (100 mL). The mixture was stirred at 1408C for 24 h. After evap-
oration of the solvent, the crude product was purified by column chroma-
tography (SiO2, CH2Cl2/MeOH 8:2) to give the title compound as white
solid (0.81 g, 28%). 1H NMR (CDCl3/MeOD, 300 MHz): d=6.99 (s, 4H;
ArHdown), 4.19 (d, 2J=13 Hz, 4H; CH2eqAr2), 3.36 (d, 2J=13 Hz, 4H;
CH2ax.Ar2), 1.98 (s, 12H; ArCH3), 1.81 ppm (d,
JH-P =18 Hz, 12H;
POCH3); 31P NMR (CDCl3/MeOD, 162 MHz): d=22.03 ppm (s,
POCH3); ESI-MS: m/z (%): 785.3 (30) [M+H]+, 823.3 (100) [M+K]+.
AC methylene dibridged resorcinarene: Dibromomethane (0.14 mL,
1.93 mmol) and K2CO3 (512 mg, 3.70 mmol) were added to a solution of
resorcinarene 1 (SI) (526 g, 0.966 mmol) in anhydrous DMA (80 mL).
The mixture was stirred at 808C for 3 h. After evaporation of the solvent,
the crude product was purified by column chromatography (SiO2,
hexane/acetone 7.5:2.5) to give the AC isomer as white solid (38 mg,
7%). 1H NMR (CDCl3, 300 MHz): d=7.08 (s, 4H; ArHdown), 6.33 (brs,
4H; ArOH), 5.89 (d, J=7 Hz, 2H; O-CH2out-O), 4.45 (d, J=13 Hz, 2H;
CH2ax.Ar2), 4.35 (d, 2J=7 Hz, 2H; O-CH2in-O), 4.05 (d, 2J=13 Hz, 2H;
CH2ax.Ar2), 3.46 (d, 2J=13 Hz, 2H; CH2eqAr2), 3.25 (d, 2J=13 Hz, 2H;
CH2eqAr2), 2.01 (s, 12H; ArCH3); ESI-MS: m/z (%): 607.3 (100) [M+K]+.
Figure 10. Molecular
structure
of
the
2
2
Tiiii[H,CH3,CH3]·2CF3CH2OH·2H2O complex showing the hydrogen-
bond interactions (color code: P, orange; O, red; C, grey; H, white; F,
light blue). The hydrogen atoms are omitted for clarity.
ACiiACTHER[NGU H,CH3,Ph] and ACioACHTRE[UNG H,CH3,Ph]: Dichlorophenylphosphate (20 mL,
tested in their ability to detect alcohols and water. The alco-
hols complexation both in the solid state and in the gas
phase is driven by the cooperative effect of hydrogen bond-
0.142 mmol) and N-methylpyrrolidine (8 mL, 0.076 mmol) were added,
under nitrogen, to a solution of AC methylene dibridged resorcinarene
(38 mg, 0.067 mmol) in anhydrous toluene (33 mL). The mixture was
Chem. Eur. J. 2008, 14, 5772 – 5779
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5777

E. Dalcanale et al.
stirred at 1128C for 5 h. After evaporation of the solvent, the crude prod-
uct was purified by column chromatography (SiO2, hexane/THF 5:5) to
give the ACii isomer (7 mg, 13%) and the ACio isomer (10 mg, 18%) as
a white solid.
tand2 ratio of 1:1 in the presence of ethanol. Each experiment was car-
ried out on five different samples and each sample was measured five
times. The overall variance was calculated from the standard deviation of
sampling and the standard deviation of the measurement (s2tot =s12 +s22). In
collision induced dissociation (CID) experiments, collisionally cooled
precursor ions were isolated by the CHEF procedure.[22] Isolated ions
were thermalized during 3.0 s delay, translationally activated by an on-
resonance radio frequency (RF) pulse, and allowed to collide with pulsed
argon background gas. Each spectrum was a collection of 32 scans.
ACii: 1H NMR (CDCl3, 300 MHz): d=8.13 (m, 4H; POArHo), 7.66–7.56
(m, 6H; POArHm + POArHp), 7.04 (s, 4H; ArHdown), 5.81 (d, 2J=7 Hz,
2
2
2H; O-CH2out-O), 4.71 (d, J=7 Hz, 2H; O-CH2in-O), 4.51 (d, J=13 Hz,
2H; CH2ax.Ar2), 4.45 (dd, 2J=13 Hz, JH-P =3 Hz, 2H; CH2ax.Ar2), 3.46 (d,
2J=13 Hz, 2H; CH2eqAr2), 3.29 (d, 2J=13 Hz, 2H; CH2eqAr2), 2.15 ppm
(s, 12H; ArCH3); 31P NMR (CDCl3, 162 MHz): d=6.61 ppm (s, POAr);
ESI-MS: m/z (%): 835.1 (100) [M+Na]+.
Sensors measurements: Sensing measurements were performed using a
10 MHz AT-cut quartz. Cavitands films were deposited on gold electrode
areas on both sides of the quartz transducers by spin-coating deposition
technique. Each microbalance was coated with the same amount of cavi-
tand, which was verified by measuring the frequency shift of Df=20Æ
0.5 kHz on the final coated QCM. The measurement system (Gaslab
20.1; IFAK, Magdeburg) was equipped with a flow chamber, containing
four coated quartz crystals, a reference quartz crystal, and a thermocou-
ple. The temperature of the chamber was thermostated at 20Æ0.18C.
The QCM chamber was connected with two mass flow controllers
(Brooks 5850S): one allowed control of the flow rate of alcohol mixture
between 2 and 50 mLminÀ1 and the other controlled the flow rate of
pure nitrogen from 150 to 200 mLminÀ1. The starting stream of N2 (200Æ
1
ACio: H NMR (CDCl3, 300 MHz): d=8.12 (m, 4H; POArHo), 7.69–7.60
(m, 3H; POArHm + POArHp), 7.32 (m, 1H; POArH), 7.13 (s, 2H;
ArHdown), 7.12 (s, 2H; ArHdown), 6.82 (m, 2H; POArH), 6.72 (m, 2H;
POArH), 5.70 (d, 2J=7 Hz, 2H; O-CH2out-O), 4.72 (d, 2J=13 Hz, 1H;
2
2
CH2ax.Ar2), 4.51 (d, J=13 Hz, 2H; CH2ax.Ar2), 4.42 (dd, J=13 Hz, JH-P
=
3 Hz, 1H; CH2ax.Ar2), 3.91 (d, 2J=7 Hz, 2H; O-CH2in-O), 3.58 (d, 2J=
13 Hz, 1H; CH2eqAr2), 3.51 (d, 2J=13 Hz, 1H; CH2eqAr2), 3.32 (d, 2J=
13 Hz, 2H; CH2eqAr2), 2.08 (s, 6H; ArCH3), 1.62 ppm (s, 6H; ArCH3).
31P NMR (CDCl3, 162 MHz): d=8.19 (s, 1P, POAr), 4.32 ppm (s, 1P;
POAr); ESI-MS: m/z (%): 835.1 (100) [M+Na]+.
Crystal structures: The crystal structure of complexes ACii-
2 mLminÀ1
)
was then replaced by
a
N2 +alcohol mixture (200Æ
A
2ACioACHTRENU[G H,CH3,Ph]·EtOH·CH2Cl2,
2 mLminÀ1); the N2/alcohol ratio was imposed by the desired final alco-
hol concentration considering that the total amount of the stream had to
be 200Æ2 mLminÀ1. After reaching of the flat characteristic plateau
(equilibrium of the partition coefficient) the chamber was flushed with
pure N2 to restore the starting conditions. During the whole process the
coated quartz crystal frequency was measured as a function of the time
every 1 s. All measurements were repeated at least four times, with varia-
tions in response of less than 3%. Alcohols used in low ppm measure-
ments were supplied by SAPIO S.r.l. in gas cylinders with a certified con-
centration of in ppm. The graduated cylinders were prepared following
the standard gravimetric procedure of the normative ISO 6142. In the
case of high ppm measurements (1500 ppm) the organic vapors were gen-
erated by bubbling a stream of nitrogen carrier gas through the volatile
liquids to produce a continuous steam saturated with vapor, the concen-
tration of which depended on the vapor pressure of the liquid (values
were obtained by experimental data interpolation[23]). This stream was di-
luted with nitrogen by the second mass flow controller to obtain the de-
sired analyte concentration.
Tiiii[H,CH3,CH3]·MeOH and Tiiii[H,CH3,CH3]·2CF3CH2OH·2H2O were
determined by single-crystal X-ray diffraction methods. Crystallographic
and experimental details for the structures are summarized in Tables S1
and S2 in the Supporting Information. Intensity data and cell parameters
were recorded at room temperature on a Bruker AXS Smart 1000 single-
crystal diffractometer, equipped with a CCD area detector with graphite
monochromated MoKa radiation. The structures were solved by direct
methods by using the SIR97 program[16] and refined on Fo2 by full-matrix
least-squares procedures, with the SHELXL-97 program.[17] Both pro-
grams were used in the WinGX suite.[18] The data reductions were per-
formed by using the SAINT[19] and SADABS[20] programs. All the non-
hydrogen atoms were refined with anisotropic atomic displacements, with
the exclusion of the disordered ethanol guest in ACii
of two carbon atoms of one phenyl ring and of the disordered ethanol
and dichloromethane solvent in 2ACio[H,CH3,Ph]·EtOH·CH2Cl2, of the
ACHTRE[UNG H,CH3,Ph]·EtOH,
AHCTREUNG
methanol guest in Tiiii[H,CH3,CH3]·MeOH, and of the trifluoroethanol
and water molecules in Tiiii[H,CH3,CH3]·2CF3CH2OH·2H2O. The hydro-
À
gen atoms were included in the refinement at idealized geometries (C H
0.95 Š) and refined “riding” on the corresponding parent atoms. The
weighting scheme used in the last cycle of refinement was w=1/[s2Fo2 +
(0.1535P)2], w=1/[s2Fo2 +(0.1155P)2], w=1/[s2F2o +(0.1305P)2], and w=
Acknowledgements
1/[s2F2o +(0.2181P)2]with
P=(F2o +2Fc2)/3 for ACii[H,CH3,Ph]·EtOH,
2ACio[H,CH3,Ph]·EtOH·CH2Cl2, Tiiii[H,CH3,CH3]·MeOH, and Tiiii[H,
CH3,CH3]·2CF3CH2OH·2H2O, respectively.
This work was supported by MUR through Progetto Galileo (French-
Italian exchange program to E.D. and J.P.D.) and by EU through NoE
MAGMANet (3-NMP 515767-2). The instrumental facilities at the
Centro Interfacoltà di Misure G. Casnati of the University of Parma
were used. The financial support from the Magnus Ehrnrooth foundation
and the MaBio project by the European Social Fund and the State pro-
vincial office of Eastern Finland, the Department of Education and Cul-
ture (E.K.) is also acknowledged.
CCDC-678546 (ACii[H,CH3,Ph]·EtOH), 678547 (2ACio[H,CH3,Ph]·
EtOH·CH2Cl2), 655612 (Tiiii[H,CH3,CH3]·MeOH), and 655313 (Tiiii[H,
CH3,CH3]·2CF3CH2OH·2H2O) contain the supplementary crystallo-
graphic 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. Geometric calculations were performed with the
PARST97 program.[21]
ESI-MS studies: Mass spectrometry experiments were performed with
the BioApex 47e Fourier transform ion cyclotron resonance mass spec-
trometer equipped with an InfinityTM cell, a passively shielded 4.7 T
160 mm bore superconducting magnet, and an external ApolloTM electro-
spray ionization source. The sample was introduced to a 708 off-axis
sprayer (stainless steel metal capillary) through a syringe infusion pump
at a flow rate of 1.5 mLminÀ1. Room-temperature nitrogen (N2) was used
as a nebulizing and counter-current drying gas. The measurements and
data handling were accomplished with Bruker XMASS software, version
6.0.2. More precise description of instrument and the parameters used
have been published.[12] Cavitand concentration in samples was 2.0–
4.0 mm. The samples for water complexation contained cavitand (2 mm),
1–20% (v/v) H2O, 0.5% (v/v) acetic acid and ACN as a solvent. Compe-
tition experiments with cavitands were performed with a cavitand1/cavi-
[1]J. W. Grate, G. C. Frye, in Sensors Update, Vol. 2 (Eds: H. Baltes, W.
Gçpel, J. Hesse), Wiley-VCH, Weinheim (Germany), 1996, pp. 10–
20.
[2]J. Janata, in Principles of Chemical Sensors, Plenum, New York
(USA), 1989.
[3]a) F. L. Dickert, H. Stathopulos, M. Reif, Adv. Mater. 1996, 8, 525–
529; b) D. M. Rudkevich, Angew. Chem. 2004, 116, 568–581;
Angew. Chem. Int. Ed. 2004, 43, 558–571; c) S. J. Dalgarno, P. K.
Thallapally, L. J. Barbour, J. A. Atwood, Chem. Soc. Rev. 2007, 36,
236–245; d) L. Pirondini, E. Dalcanale Chem. Soc. Rev. 2007, 36,
695–706; e) A. G. Grechin, H.-J. Buschmann, E. Schollmeyer
Angew. Chem. 2007, 119, 6619–6621; Angew. Chem. Int. Ed. 2007,
46, 6499–6501; Angew. Chem. Int. Ed. 2007, 46, 6499–6501.
5778
www.chemeurj.org
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 5772 – 5779

Phosphonate Cavitands
FULL PAPER
[4]For the nomenclature adopted for phosphonate cavitands see: R. Pi-
nalli, M. Suman, E. Dalcanale, Eur. J. Org. Chem. 2004, 451–462.
[5]R. Pinalli, F. F. Nachtigall, F. Uguzzoli, E. Dalcanale, Angew. Chem.
1999, 111, 2530–2533; Angew. Chem. Int. Ed. 1999, 38, 2377–2380.
[6]R. Paolesse, C. Di Natale, S. Nardis, A. Magagnano, A. DꢀAmico, R.
Pinalli, E. Dalcanale, Chem. Eur. J. 2003, 9, 5388–5395.
[7]M. Suman, M. Freddi, C. Massera, F. Uguzzoli, E. Dalcanale, J. Am.
Chem. Soc. 2003, 125, 12068–12069.
[8]B. Bibal, J. P. Declercq, J. P. Dutasta, B. Tinant, A.-G. Valade, Tetra-
hedron 2003, 59, 5849–5854.
[9]P. Delangle, J.-C. Mulatier, B. Tinant, J.-P. Declercq, J.-P. Dutasta,
Eur. J. Org. Chem. 2001, 3695–3704.
[15]The MMFF-B3LYP/6–31G* computed structure in the gas phase is
fully consistent with the crystal structure both in terms of complex
stoichiometry and water binding mode (D. V. Dearden, E. Dalca-
nale, unpublished results).
[16]A. Altomare, M. C. Burla, M. Camalli, G. L. Cascarano, C. Giaco-
vazzo, A. Guagliardi, A. G. Moliterni, G. Polidori, R. Spagna, J.
Appl. Crystallogr. 1999, 32, 115–119.
[17]G. M. Sheldrick, SHELXL97, Program for Crystal Structure Refine-
ment, University of Gçttingen, Gçttingen (Germany), 1997.
[18]WinGX: L. J. Farrugia, J. Appl. Crystallogr. 1999, 32, 837–838.
[19]SAINT, Software Users Guide, 6.0, Bruker Analytical X-ray Sys-
tems, 1999.
[10]H. Konishi, Y. Iwasaki, T. Okano, J. Kiji, Chem. Lett. 1989, 1815–
1816.
[11]G. R. Desiraju, T. Steiner, in The Weak Hydrogen Bond, Oxford
University Press, Oxford (UK), 1999, pp. 12–15.
[12]E. Ventola, P. Vainiotalo, M. Suman, E. Dalcanale, J. Am. Soc. Mass
Spectrom. 2006, 17, 213–221.
[13]D. Reyntjens-Van Damme, T. Zeegers-Huyskens, J. Phys. Chem.
1980, 84, 282–285.
[20]G. M. Sheldrick, SADABS Area-Detector Absorption Correction,
2.03, University of Gçttingen, Gçttingen (Germany), 1999.
[21]M. Nardelli, J. Appl. Crystallogr. 1996, 29, 296–300.
[22]L. J. De Koning, N. M. Nibbering, S. L. Van Orden, F. H. Laukien,
Int. J. Mass Spectrom. Ion Processes 1997, 165/166, 209–219.
[23]R. H. Perry, D. W. Green, Perry Chemical Engineersꢀ Handbook, 7th
ed., McGraw-Hill, New York, 1997.
[14]The amount of material coated was controlled by measuring a fre-
quency shift of Df=20Æ0.05 kHz on deposition of each layer.
Received: February 21, 2008
Published online: May 19, 2008
Chem. Eur. J. 2008, 14, 5772 – 5779
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5779