A comprehensive study of the complexation of alkali metal cations by lower rim calix[4]arene amide derivatives

Article abstract of DOI:10.1039/c7cp03920d

The complexation of alkali metal cations by lower rim N,N-dihexylacetamide (L1) and newly synthesized N-hexyl-N-methylacetamide (L2) calix[4]arene tertiary-amide derivatives was thoroughly studied at 25 °C in acetonitrile (MeCN), benzonitrile (PhCN), and methanol (MeOH) by means of direct and competitive microcalorimetric titrations, and UV and ¹H NMR spectroscopies. In addition, by measuring the ligands' solubilities, the solution (transfer) Gibbs energies of the ligands and their alkali metal complexes were obtained. The inclusion of solvent molecules in the free and complexed calixarene hydrophobic cavities was also investigated. Computational (classical molecular dynamics) investigations of the studied systems were also carried out. The obtained results were compared with those previously obtained by studying the complexation ability of an N-hexylacetamidecalix[4]arene secondary-amide derivative (L3). The stability constants of 1:1 complexes were determined in all solvents used (the values obtained by different methods being in excellent agreement), as were the corresponding complexation enthalpies and entropies. Almost all of the examined reactions were enthalpically controlled. The most striking exceptions were reactions of Li⁺ with both ligands in methanol, for which the entropic contribution to the reaction Gibbs energy was substantial due the entropically favourable desolvation of the smallest lithium cation. The thermodynamic stabilities of the complexes were quite solvent dependent (the stability decreased in the solvent order: MeCN > PhCN ? MeOH), which could be accounted for by considering the differences in the solvation of the ligand and free and complexed alkali metal cations in the solvents used. Comparison of the stability constants of the ligand L1 and L2 complexes clearly revealed that the higher electron-donating ability of the hexyl with respect to the methyl group is of considerable importance in determining the equilibria of the complexation reactions. Additionally, the quite strong influence of intramolecular hydrogen bond formation in compound L3 (not present in ligands L1 and L2) and that of the inclusion of solvent molecules in the calixarene hydrophobic cone were shown to be of great importance in determining the thermodynamic stability of the calixarene-cation complexes. The experimental results were fully supported by those obtained by MD simulations.

Full text of DOI:10.1039/c7cp03920d

View Article Online
View Journal
PCCP
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: G. Horvat, L.
Frkanec, N. Cindro and V. Tomisic, Phys. Chem. Chem. Phys., 2017, DOI: 10.1039/C7CP03920D.
Volume 18 Number
1
7
January 2016 Pages 1–636
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
PCCP
Physical Chemistry Chemical Physics
www.rsc.org/pccp
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1463-9076
PERSPECTIVE
Darya Radziuk and Helmuth Möhwald
Ultrasonically treated liquid interfaces for progress in cleaning and
separation processes
rsc.li/pccp

Page 1 of 28
Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C7CP03920D
A comprehensive study of alkali metal cations complexation by
lower rim calix[4]arene amide derivatives
a
b
a
a
Gordan Horvat,* Leo Frkanec, Nikola Cindro and Vladislav Tomišić*
a
Division of Physical Chemistry, Department of Chemistry, Faculty of Science, University of Zagreb,
Horvatovac 102a, 10000 Zagreb, Croatia
b
Laboratory for Supramolecular Chemistry, Division of Organic Chemistry and Biochemistry, Ruđer
Bošković Institute, Bijenička cesta 54, 10000 Zagreb, Croatia
Abstract
Complexation of alkali metal cations by lower rim N,Nꢀdihexylacetamide (L1) and newly synthesized Nꢀ
hexylꢀNꢀmethylacetamide (L2) calix[4]arene tertiaryꢀamide derivatives was thoroughly studied at 25 °C in
acetonitrile (MeCN), benzonitrile (PhCN), and methanol (MeOH) by means of direct and competitive
1
microcalorimetric titrations, as well as UV and H NMR spectroscopies. In addition, by measuring ligands
solubilities, solution (transfer) Gibbs energies of the ligands and their alkali metal complexes were obtained.
The inclusion of solvent molecules in the free and complexed calixarene hydrophobic cavity was also
investigated. Computational (classical molecular dynamics) investigations of the studied systems were carried
out as well. The obtained results were compared with those previously obtained by studying complexation
abilities of Nꢀhexylacetamidecalix[4]arene secondaryꢀamide derivative (L3). The stability constants of 1:1
complexes were determined in all solvents used (the values obtained by different methods being in excellent
agreement), as were the corresponding complexation enthalpies and entropies. Almost all of the examined
+
reactions were enthalpically controlled. The most striking exceptions were reactions of Li with both ligands
in methanol, for which entropic contribution to the reaction Gibbs energy was substantial due the entropically
favourable desolvation of the smallest lithium cation. Thermodynamic stabilities of the complexes were quite
solvent dependent (stability decreased in the solvent order: MeCN > PhCN >> MeOH), which could be
accounted for by considering the differences in the solvation of the ligand as well as free and complexed
alkali metal cations in the solvents used. Comparison of the stability constants of ligands L1 and L2
complexes clearly revealed that higher electronꢀdonating ability of the hexyl with respect to methyl group is
of considerable importance in determining the equilibria of the complexation reactions. Additionally, the
quite strong influence of intramolecular hydrogen bonds formation in compound L3 (not present in ligands
L1 and L2) as well that of inclusion of solvent molecules into the calixarene hydrophobic cone were shown to
be of great importance in determining the calixareneꢀcation complex thermodynamic stability. The
experimental results were fully supported by those obtained by MD simulations.
*
Authors to whom correspondence should be addressed (Eꢀmail: ghorvat@chem.pmf.hr, vtomisic@chem.pmf.hr)
1

Physical Chemistry Chemical Physics
Page 2 of 28
View Article Online
DOI: 10.1039/C7CP03920D
1
. Introduction
Calixarenes are macrocyclic ligands which, when substituted at the upper and/or lower rim, can
1–5
efficiently and in some cases selectively bind cations, anions, and neutral molecules. Due to the
specific spatial arrangement of phenolic subunits in these compounds, a wellꢀdefined binding site
6–10
can be achieved by introducing substituents with appropriate functional groups.
The lower rim
derivatives comprising carbonyl functional groups, i.e. ketones, esters, or amides, have been
extensively studied and were shown to form quite stable complexes with alkaliꢀ, alkalineꢀearth, and
7
,8,10–12
transitionꢀmetal cations in various solvents.
In such compounds a cationꢀbinding site is
formed by nucleophilic ether and carbonyl oxygen atoms. Particularly interesting derivatives are
those which bear secondaryꢀ and tertiaryꢀamide substituents at the lower rim due to their high, but
1
3–27
mutually quite different affinities towards alkali metal cations.
The difference in the stabilities of
the corresponding complexes is mostly a consequence of the fact that secondaryꢀamide derivatives
are able to form intramolecular N–HꢀꢀO=C hydrogen bonds, which significantly decrease their
14,18,27–29
abilities to bind cations as compared to tertiaryꢀamide derivatives.
The calixarene cationꢀbinding ability is in general rather strongly influenced by the
compatibility of the sizes of cation and ligandꢀbinding site. In addition, solvation of reactants and
products of the complexation reaction plays a very important role in the binding process. In this
respect, the inclusion of solvent molecules in the calixarene hydrophobic cone can be of great
4
,16,23,24,30–32
importance.
This specific interaction with solvent molecules is facilitated by cation
binding which leads to an appropriate ligand’s cone preorganization. It has been shown that such an
inclusion of solvent molecule can significantly increase the cationꢀcalixarene complex
1
6,23,24,33
thermodynamic stability.
As already mentioned, calixarene amides have been known as very efficient binders of alkali
metal cations. However, the factors governing this efficiency, e.g. the nature of substituents at the
amide groups and solvation effect, have not been fully explored. Recently we have reported on
detailed thermodynamic, structural, and computational studies of the complexation reactions of a
secondaryꢀamide calix[4]arene derivative L3 (Fig. 1) with alkali metal cations in several
1
6,23
solvents.
In this paper we present the investigations of complexation affinities of two tertiaryꢀ
amide calixarene derivatives L1 and L2 (Fig. 1) towards alkali metal cations. In order to explore the
solvent effect on the equilibria of complexation reactions, three solvents were used, namely
acetonitrile, benzonitrile, and methanol. The solvents were chosen considering differences in their
cation and ligands solvation abilities as well as affinities for hydrogen bonding and for inclusion in
the calixarene basket. Stability constants of the complexes, as well as complexation enthalpies and
entropies, were determined by means of direct and competitive microcalorimetric titrations (the
latter were necessary because of the high complex stabilities). In some cases complexation reactions
were also quantitatively studied spectrophotometrically. A detailed structural insight into the alkali
2

Page 3 of 28
Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C7CP03920D
metal cations complexation by ligands L1 and L2, as well as into the solventꢀmolecule inclusion in
1
the hydrophobic calixarene cavity, was achieved by H NMR measurements and molecular
dynamics simulations. The obtained thermodynamic results are thoroughly discussed and compared
to those corresponding to compound L3 regarding structural characteristics of the ligands. The
solvent effect, i.e. that of reactants and product solvation, on the complexation reactions has been
particularly addressed.
t-Bu
L1 R = R = C H
13
1
2
6
L2 R = CH3 R = C H
1
2
6
13
13
4
L3 R = H
R = C H
2 6
O
1
O
N
R1
R2
Fig. 1 Structures of compounds L1 to L3.
2
. Experimental
2
.1 Synthesis
Compounds L1 and L3 and the acid chloride of calix[4]arene tetracetic acid (the cone conformer)
21
were prepared according to the procedures described elsewhere.
5
,11,17,23ꢀTetraꢀtertꢀbutylꢀ25,26,27,28ꢀtetrakis(NꢀhexylꢀNꢀmethylꢀcarbamoylmethoxy)calix[4]arene
(L2)
Tetraacid (1.00 g, 1.13 mmol) and thionyl chloride (5.0 mL) were refluxed for one hour, evaporated
and then coevaporated with dry toluene. Resulting acylꢀchloride was dissolved in 50 mL of dry
o
DCM and cooled under argon to 0 C. With vigorous stirring mixture of pyridine (1.00 mL, 12.9
mmol) and Nꢀhexylmethylamine (0.86 mL, 5.7 mmol) dissolved in 10 mL of dry DCM was added
dropwise. Mixture was stirred for 24 hours, diluted with DCM, washed 5 times with miliQ water and
evaporated without drying. Crude compound was dissolved in 20 mL of boiling HPLC grade
acetonitrile and left to crystalize in refrigerator, yielding 570 mg (40 %) of pure compound.
Analytically pure sample was obtained by recrystallization (again from acetonitrile).
3

Physical Chemistry Chemical Physics
Page 4 of 28
View Article Online
DOI: 10.1039/C7CP03920D
1
13
+
Following signals were found in the H NMR and C NMR spectra of NaL2 complex in
CDCl3:
1
H NMR (CDCl3) δ/ppm 7.14 (s, 8H), 4.63ꢀ4.38 (m, 12H), 3.46ꢀ3.27 (m, 8H), 3.12 (bs, 4H), 3.01ꢀ
.86 (m, 12H), 1.58 (bs, 8H), 1.40ꢀ1.24 (m, 24H), 1.17 (s, 36H), 0.94ꢀ0.83 (m, 12H);
2
1
3
C NMR (CDCl3) δ/ppm 168.67 (s), 168.55 (s), 151.36 (s), 151.24 (s), 151.08 (s), 150.86 (s),
1
48.08 (s), 135.03 (s), 125.78 (d), 117.11 (s), 74.09 (t), 73.97 (t), 48.55 (t), 48.35 (t), 48.16 (t), 34.28
(
t), 33.86 (q), 33.24 (q), 31.90 (t), 31.60 (t), 31.39 (q), 30.19 (t), 28.19 (t), 27.37 (t), 26.99 (t), 26.48
t);
(
The structure of the L2 ligand was confirmed by HRMS (MALDIꢀTOF) spectrometry:
+
m/z [M + H ]− calculated for ((C20
H
31
O
N)
) − 1269,9497, found 1269,9474.
2
4
2
.2 Materials for physicochemical measurements
The solvents, acetonitrile (MeCN, Merck, Uvasol and J. T. Baker, HPLC grade), benzonitrile (Sigma
Aldrich, Chromasolv, 99.9 %), and methanol (J. T. Baker, HPLC grade) were used without further
purification. The salts used for the investigation of L complexation were LiClO
4
(Sigma Aldrich,
9
9.99 %), NaClO (Sigma Aldrich, 98+ %), KClO (Merck, p.a.), KSCN (Sigma Aldrich, ≥ 99 %),
4
4
4
RbCl (Sigma Aldrich, 99 %), RbBPh (Sigma Aldrich, 95 %), CsCl (Sigma Aldrich, 99.5 %).
2
.3 Solubility determinations
Saturated solutions of L1 and L2 in acetonitrile and methanol were prepared by adding an excess
amount of the solid substance to the solvent. The obtained mixtures were left in a thermostat at 25
°C for several days in order to equilibrate. The concentrations of saturated solutions were determined
at 25.0 °C spectrophotometrically by means of a Varian Cary 5 spectrophotometer equipped with a
thermostatting device. Calibration curves were obtained by measuring the absorbances of L1 and L2
solutions of known concentrations.
2
.4 Spectrophotometry
UV titrations of L1 and L2 with alkali metal salt solutions in acetonitrile and methanol were
performed at (25.0 ± 0.1) °C by means of a Varian Cary 5 spectrophotometer. The metal salt
solutions were added to a solutions of ligands’ placed in the quartz cell (l = 1 cm or l = 10 cm). After
each addition UV spectrum was recorded with a sampling interval of 1 nm and integration time of
0
.2 s. The stability constant of the complexation of cesium cation with L1 in acetonitrile was
determined by spectrophotometric titrations (Table 1, Fig. 7). Due to the rather small affinity of L1
4

Page 5 of 28
Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C7CP03920D
+
towards Cs cation, the largest concentration of cesium salt solution that was prepared was
+
insufficient to ensure the reliable determination of CsL1 complex stability constant by a titration in
which the titrant was the metal salt solution. For that reason, the titration was performed in a way
that the concentration of cesium cation was varied by the removal of a certain volume of the reaction
+
mixture that contained L1 and Cs cation which was replaced by the same volume of L1 solution of
the exact analytical concentration as the one in the reaction cell. By this method it was possible to
+
reach an adequate range of analytical concentration of Cs and, in return, determine a reliable value
+
+
of the CsL1 stability constant. The value of stability constant of RbL2 complex was determined
using cuvettes with a 10 cm optical path that enabled the measurements of reaction mixture spectra
in an adequate range of analytical concentrations of L2 for a reliable stability constant
determination. All measurements were done in triplicate. The obtained data were analyzed by
34
multivariate nonꢀlinear regression analysis using Hyperquad program.
1
13
2
.5 H NMR and C NMR studies
1
o
H NMR titrations were carried out at 25 C by means of a Bruker Avance 600 MHz with a solvent
signal used as standard for titrations in CD CN and CD OD or with a TMS signal as standard in
3
3
CDCl
3
solutions. In titrations of L1 and L2 with the alkali metal cations in deuterated acetonitrile,
the solutions salts were added to the solutions of L1 and L2. Spectra were recorded at 32 pulses.
1
13
+
H and C NMR spectra of NaL2 complex (Fig. S1, ESI) were recorded on Bruker Ascend
00 MHz at rt using TMS as a reference in proton spectra and middle signal of CDCl (77.16 ppm)
in carbon spectra, chemical shifts are reported in ppm.
4
3
+
+
+
Solutions of LiL2 , NaL2 and KL2 complexes in CDCl were prepared by dissolving
3
ligand in deuterated chloroform followed by the addition of appropriate metal salt solid. Sonification
of the solutions was used to ensure complete complexation.
2
.6 Mass spectrometry
HRMS spectrum of L2 was obtained on a Bruker Microflex MALDI / TOF instrument.
2
.7 Calorimetry
Microcalorimetric measurements were conducted with an isothermal titration calorimeter Microcal
VPꢀITC at 25.0 °C. The calorimeter was calibrated electrically, and its reliability was additionally
checked by carrying out the complexation of barium(II) by 18ꢀcrownꢀ6 in aqueous medium at 25 °C.
–
1
The results obtained (log K = 3.75, ꢁ H = –31.7 kJ mol ) were in excellent agreement with the
r
5

Physical Chemistry Chemical Physics
Page 6 of 28
View Article Online
DOI: 10.1039/C7CP03920D
–1 35
r
literature values (log K = 3.73, ꢁ H = –31.5 kJ mol ). Thermograms were processed using the
Microcal OriginPro 7.0 program.
In the calorimetric studies of alkali metal cations complexation by L, the enthalpy changes
were recorded upon stepwise additions of acetonitrile, benzonitrile or methanol solution of metal salt
3
+
+
0
into solution of ligand (V = 1.4182 cm ). Titrations of the M–L1 and M–L2 complexes with
acetonitrile in benzonitrile were carried out by the addition of MeCN solution in PhCN to the
solution of complexes in the same solvent. The heats measured in the titration experiments were
corrected for heats of titrant dilution obtained by blank experiments. The dependence of successive
enthalpy changes on the titrant volume was processed by nonꢀlinear leastꢀsquares fitting procedure
using OriginPro 7.5 program. All measurements were repeated three or more times.
2
.8 Molecular dynamics simulations
3
6–42
The molecular dynamics simulations were carried out by means of the GROMACS
package
(version 5.1.4). Intramolecular and nonbonded intermolecular interactions were modelled by the
43
OPLSꢀAA (Optimized Parameters for Liquid SimulationsꢀAll Atoms) force field. Partial charges
assigned to ring carbons bound to CH groups that link the monomers were assumed to be zero. The
2
initial structures of ligands L1 and L2, as well as their alkali metal complexes, were derived from
23
the crystal structure of sodium complex of similar secondary amide calixarene derivative. The
calixarene ligands, their alkali metal complexes and corresponding acetonitrile adducts were
solvated in cubical boxes containing between 2100 and 3700 molecules of acetonitrile, benzonitrile
or methanol, and with periodic boundary conditions. The solute concentration in such a box was
–3
about 0.01 mol dm . The solvent boxes were equilibrated prior to inclusion of ligands and their
complexes with the box density after equilibration in all cases being close to the experimental one
–
within 2 %. During the simulations of the systems comprising calixarene and metal cations, Cl ion
was included to neutralize the box. The chloride counterion was kept fixed at the box periphery
whereas the complex was initially positioned at the box center. In all simulations an energy
minimization procedure was performed followed by a molecular dynamics simulation in NpT
44
conditions for 50.5 ns, where the first 0.5 ns were not used in data analysis. The Verlet algorithm
with a time step of 1 fs was employed. The cutoff radius for nonbonded van der Waals and shortꢀ
range Coulomb interactions was 16 Å. Longꢀrange Coulomb interactions were treated by the Ewald
45
method as implemented in the PME (Particle Mesh Ewald) procedure. The simulation temperature
4
6,47
was kept at 298 K with the NoséꢀHoover
algorithm using a time constant of 1 ps. The pressure
4
8
was kept at 1 bar by MartynaꢀTuckermanꢀTobiasꢀKlein algorithm and the time constant of 1 ps.
49
Figures of calixarene molecular structures were created using VMD software.
6

Page 7 of 28
Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C7CP03920D
3
. Results and discussion
3
.1 Synthesis
Compound L2 was synthesized according to the previously described method, starting from the
corresponding tetraester which was hydrolized to tetraacid. In reaction with thionyl chloride
tetrachloride was obtained which was reacted with excess of Nꢀmethylhexylamine to yield the target
compound (Scheme 1).
1
. SOCl2
NaOH
H O/EtOH
2. MeNHHex
4
4
2
4
O
O
DCM
O
O
N
O
O
OH
O
95 %
40 %
Scheme 1 Synthesis of compound L2.
3
.2 Structural studies of L1 and L2 in solution
1
Molecular structures of ligands L1 and L2 in solution were studied by means of H NMR
1
spectroscopy and MD simulations. In the H NMR spectrum of L1 in CDCl
3
(Fig. 2, Table S1, ESI)
a single set of ligand proton signals (tertꢀbutyl, ArꢀH, and ArꢀOꢀCH proton singlets and two
2
doublets for equatorial and axial protons) is present. Such a signal pattern is characteristic for C
4
1
cone conformation, or represents dynamic C
2
–C
2
cone exchange that is fast on the H NMR time
5
scale. The spectra of L1 in CD CN and CD OD are similar to that in CDCl in terms of protonꢀ
3
3
3
signal multiplicity. However, in acetonitrile the signal of ArꢀH protons is significantly downfield
shifted compared to chloroform and methanol (Fig. 2, Table S1, ESI). This is usually considered as
an indication of the inclusion of acetonitrile molecule into the hydrophobic calix[4]arene
1
8,24,50
1
hydrophobic cavity
which is fast on the H NMR time scale.
7

Physical Chemistry Chemical Physics
Page 8 of 28
View Article Online
DOI: 10.1039/C7CP03920D
b
t-Bu
c
CD OD
a2
d
3
a2
b
a1
H
b
4
H
c
O
a2
c
a2
a1
d
CD CN
3
O
N
d
d
b
c
a2
a1
d
CDCl3
7
6
5
4
3
δ
/ ppm
1
o
Fig. 2 H NMR spectra of L1 in deuterated chloroform, acetonitrile, and methanol at 25 C.
Ligand L2 is asymmetrically substituted tertiaryꢀamide calix[4]arene derivative in which a
rotational isomerism is possible regarding the orientation of methyl and hexyl substituents with
respect to the amide oxygen atom, the corresponding reorientation activation enthalpy being ≈70 kJ
–1 51
1
mol . As a consequence, the rotation is relatively slow at room temperature at the H NMR
timescale, leading to an appearance of separate proton signals for each of the rotamers. There are 6
different rotamers of L2 possible in C
in flattened cone conformation of C
4
cone conformation and 9 rotamers when calixarene basket is
1
2
symmetry. In H NMR spectra of L2 in CD
3
Cl there are
between 9 and 11 signals present for each of the ligand's proton type (Fig. 3), where the most notable
differences in relative chemical shifts were found for aromatic and tertꢀbutyl protons. Similar
phenomenon was observed in the spectra of L2 in deuterated acetonitrile, methanol, and dimethyl
o
sulfoxide (Fig. 3). Upon temperature increase of dimethyl sulfoxide L2 solution from 25 C to 100
o
C, multiple signals collapsed into a single one (Fig. 4). The above findings can be explained by
taking into account a possibility of the existence of an equilibrium between L2 rotamers, each
having different set of proton signals. Again a chemical shift of ArꢀH protons in CD
considerably higher than in CDCl and CD OD, indicating a possible inclusion of acetonitrile
molecule into the hydrophobic L2 basket. Due to the complexity of L2 spectrum in CD CN, and the
3
CN is
3
3
3
fact that ArꢀH proton shifts lie somewhere between values of those corresponding to the free L1 and
L1MeCN species, it was not possible to elucidate with certainty whether L2MeCN adduct is present
in the acetonitrile solution of L2.
8

Page 9 of 28
Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C7CP03920D
t-Bu
b
a1
H
b
CD OD
3
4
H
O
a2
c
b
O
d
a2
a2
CD CN
3
N
d
a1
a2
c
b
CDCl3
7
6
5
4
3
δ
/ ppm
1
o
Fig. 3 H NMR spectra of L2 in deuterated chloroform, acetonitrile, and methanol at 25 C.
t-Bu
b
o
a2
d
a
1
c
1
00 C
b
a1
H
4
O
H
a2
c
o
O
5
0 C
N
d
o
2
5 C
7
6
5
4
3
δ
/ ppm
1
Fig. 4 Temperature dependence of L2 H NMR spectrum in deuterated dimethyl sulfoxide.
The structures of calixarene ligands L1 and L2 were also investigated by molecular
dynamics simulations. Due to the existence of several rotamers of L2, two isomers with most
distinct structural features were simulated, Z-L2 in which carbonyl oxygen atom lie on the same side
of amide bond as hexyl group in all four calixarene subunits, and E-L2 with the opposite orientation
of the named groups at the calixarene lower rim (Figs. S2 and S3, ESI). During simulations in
acetonitrile, an inclusion of solvent molecules in the calixarene basket was observed for both ligands
(Figs. S2 and S3, ESI). This process had an influence on the shape of the calixarene cone which
changed from the flattened to the regular one upon MeCN molecule inclusion. On the other hand, the
9

Physical Chemistry Chemical Physics
Page 10 of 28
View Article Online
DOI: 10.1039/C7CP03920D
inclusion of benzonitrile molecule in the free ligands was not observed during MD simulations in
that solvent.
3
.3 Studies of alkali metal cations complexation in acetonitrile
Stability constants of L1 and L2 complexes with alkali metal cations were determined by
means of microcalorimetric and spectrophotometric titrations (Table 1). Due to the large affinities of
+
+
+
the studied macrocycles for Li , Na , and K cations, the determination of the corresponding
equilibrium constants was not possible by direct calorimetric titrations. Therefore, these constants,
along with the corresponding reaction enthalpies and entropies, were determined by means of
displacement microcalorimetric titrations. The values of complexation thermodynamic quantities
were obtained by the combination of the results of competition titrations and those obtained by direct
titrations of ligands L1 and L2 with rubidium cation (Fig. 5 and S5, ESI). The stability constants of
complexes of both ligands with potassium cation were determined from the titration of solution of
rubidium complex with the solution of potassium cation (Fig. 6 and S20, ESI). The equilibrium
constants for the formation of sodium complexes were determined from the results of the
displacement titration of potassium complexes with sodium cation solution (Figs. S6 and S21, ESI).
+
+
+
The affinities of both ligands for Na were greater than for Li , so the stability constants of LiL1
+
and LiL2 complexes were determined by the competition microcalorimetric titration of lithium
complex solution with sodium cation solution (Figs. S7 and S22, ESI). Due to the difference in the
+
+
stabilities of LiL and KL complexes, a more adequate procedure to determine the equilibrium
constants of complexation of L1 and L2 with lithium cation would be by means of displacement
+
titrations of the complexed potassium cation with Li . However, the similar values of reaction
enthalpies for the complexation of the studied ligands with these cations cause a rather small heat
effect due to the cation substitution, which then lead to unreliable experimental data.
1
0

Page 11 of 28
Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C7CP03920D
0
.0
5
5
5
5
5
4
4
4
8
6
4
2
0
8
6
4
a)
-0.1
b)
-0.2
-0.3
-0.4
-0.5
0.0
0
20
40
60
80
100
0.4
0.8
1.2
1.6
2.0
2.4
+
t / min
n(Rb ) / n(L2)
–
5
–3
3
Fig. 5 a) Microcalorimetric titration of L2 (c = 9.95 × 10 mol dm , V = 1.4182 cm ) with RbNO
3
–3
–3
(
c = 1.07 × 10 mol dm ) in acetonitrile; t = 25 °C; b) Dependence of successive enthalpy change
+
on n(Rb )/n(L2) ratio. ■ experimental; — calculated.
0
.0
60
55
50
45
40
35
a)
b)
-0.2
-0.4
-0.6
-0.8
-1.0
0
20
40
60
80
100
120
0.0
0.4
0.8
1.2
1.6
2.0
+
+
t / min
n(K ) / n(RbL1 )
+
–4
–3
3
Fig. 6 a) Microcalorimetric titration of RbL1 (c = 4.95 × 10 mol dm , V = 1.4182 cm ) with
–
3
–3
KClO (c = 5.23 × 10 mol dm ) in acetonitrile; t = 25 °C; b) Dependence of successive enthalpy
4
+
+
change on n(K )/n(RbL1 ) ratio. ■ experimental; — calculated.
The complexation reaction enthalpies were also determined by direct titrations. As can be
seen from the data listed in Table 1 and Table S4, ESI, the values obtained by direct and
displacement titrations are in very good agreement. This fact can serve as a confirmation of the
reliability of the thermodynamic results obtained in this work.
All complexation reactions are enthalpically controlled, whereby the absolute entropic
contribution to the standard reaction Gibbs energy is approximately an order of magnitude smaller in
comparison to the enthalpic contribution for all complexes, except those with lithium cation. The
latter can be explained by considering that due to the strong solvation of this cation in acetonitrile, its
1
8
desolvation entropy is more favorable than for the other cations. It should be noted that the stability
1
1

Physical Chemistry Chemical Physics
Page 12 of 28
View Article Online
DOI: 10.1039/C7CP03920D
constants of L1 and L2 complexes with alkali metal cations in acetonitrile are about 5 orders of
23
magnitude larger than those corresponding to the secondary amide derivative L3. As already
mentioned, this difference in ligand affinities is mainly due to the enthalpically unfavorable
disruption of intramolecular hydrogen bonds network in L3 in the course of cation coordination by
carbonyl oxygen atoms. Such an effect is clearly reflected in the values of reaction enthalpies, which
–1
–1
are between 28 kJ mol and 35 kJ mol more negative for the cation complexation by ligands L1
+
+
and L2 compared to that of L3. Compounds L1 and L2 have similar affinities towards Li and Na
+
+
cations, although reaction enthalpies for the formation of LiL2 and NaL2 complexes are about 5 kJ
–1
+
+
mol larger than for the formation of LiL1 and NaL1 . These differences in reaction enthalpies are
somewhat compensated with more favorable entropic contributions in the cases of alkali metal
cation complexation reactions with L2 compared to L1.
The differences in affinities of L1 and L2 for potassium and rubidium cations are more
+
+
pronounced than in the case of Li and Na , i.e. L1 binds former cations significantly more strongly
+
+
–1
than L2. The reaction enthalpies of K and Rb complexation with L1 are about 5 kJ mol lower
than the ones corresponding to L2, which in turn impacts the overall stability of the complexes
formed. Such an effect is most probably the result of the different electronꢀdonating character of
methyl and hexyl substituents bound to the amide groups, i.e. hexyl group is better electron donor
4
than methyl one.
Table 1 Thermodynamic parameters for complexation of alkali metal cations with L1 and L2 in
o
acetonitrile at 25 C determined by microcalorimetric and spectrophotometric titrations.
°
°
°
(
ꢁ G ±SE
)
(
ꢁ H ±SE
)
(
ꢁ S ± SE
)
r
r
r
log K ± SE
Cation
kJ mol−
1
kJ mol−1
J mol K
−1 −1
L1
+
Li
11.27 ± 0.06
12.15 ± 0.04
9.19 ± 0.02
6.39 ± 0.02a
3.69 ± 0.01
–64.3 ± 0.4
–69.4 ± 0.2
–52.5 ± 0.1
–36.5 ± 0.1
–21.06 ± 0.05
L2
–63.3 ± 0.1
–70.0 ± 0.1
–48.2 ± 0.1
–31.99 ± 0.08
–53.7 ± 0.3
–70.3 ± 0.3
–54.1 ± 0.3
–35.3 ± 0.3
36.4 ± 3.2
–3.1 ± 1.0
–5.5 ± 0.9
3.8 ± 0.8
+
Na
+
K
+
Rb
Cs
+
+
Li
11.09 ± 0.01
12.27 ± 0.01
8.46 ± 0.01
5.61 ± 0.01 a
–48.8 ± 0.2
–64.6 ± 0.2
–48.6 ± 0.2
–29.9 ± 0.2
48.4 ± 0.8
18.0 ± 0.8
–1.1 ± 0.8
6.9 ± 0.7
+
Na
+
K
+
Rb
5.690 ± 0.007
a
determined spectrophotometrically. SE = standard error of the mean (N = 3–5).
1
2

Page 13 of 28
Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C7CP03920D
Spectrophotometric titrations of ligands L1 and L2 in acetonitrile were also conducted. In all
of the titrations the addition of salt solution resulted in the decrease of absorbance in the larger part
of the ligand UV spectrum. In the cases of titrations with lithium, sodium, potassium and rubidium
cations, there was no change of the solution spectrum after the equimolar cation/ligand ratio (Figs.
S8–S11, S23–S25, ESI). That indicated strong complexation and formation of complexes with 1:1
stoichiometry, in accordance with the result of microcalorimetric measurements. Stability constants
+
+
of CsL1 and RbL2 (Table 1) were calculated by processing spectrophotometric data collected in a
way described in the Experimental section (Fig. 7 and S26, ESI).
2
.0
1.8
.7
A
1.8
a)
b)
1
1
1
1
1
0
0
0
0
0
.6
.4
.2
.0
.8
.6
.4
.2
.0
A
279nm
1
.6
.5
.4
1.3
.2
1
1
1
1.1
0.0
2
50
260
270
280
290
300
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
+
λ
/ nm
n(Cs ) / n(L1)
–4
–3
3
Fig. 7 a) Spectrophotometric titration of L1 (c = 3.81 × 10 mol dm , V
0
= 2.5 cm ) with CsNO
3
in
o
+
■
acetonitrile. l = 1 cm, t = 25.0 C. b) Dependence of absorbance at 279 nm on n(Cs ) / n(L1) ratio.
experimental; ― calculated.
1
In order to assess the structure of alkali metal cation complexes in acetonitrile, H NMR
titrations were carried out. When L1 was titrated with alkali metal cation solutions the observed
1
changes in H NMR spectra were similar in all cases (Figs. S12–S15, ESI). Namely, the ion
1
exchange was slow on the H NMR time scale leading to the appearance of two distinct sets of
signals, one that can be attributed to the free L1 and the other assigned to the complexed form of the
calixarene ligand. From the obtained results it cannot be ascertained whether or not the formed
complexes contained specifically bound acetonitrile molecule in their hydrophobic cavities, as was
23
the case for lithium, sodium, and potassium complexes of L3 in acetonitrile. Titration of L2 with
alkali metal cation solutions in CD CN led to the unification of several proton signals of ArꢀH and
3
tertꢀbutyl protons of free L2 (Figs. S27–S30, ESI) into one signal for each proton, separate from the
1
original signals, which was an indicator of a slow cation exchange on the H NMR timescale. Newly
emerged signals were assigned to the L2ꢀcation complex for all cations studied. The reduction of a
number of upperꢀrim proton signals can be rationalized by the simetrization of calixarene basket
1
3

Physical Chemistry Chemical Physics
Page 14 of 28
View Article Online
DOI: 10.1039/C7CP03920D
4
conformation, a conformational transition already experimentally and
upon cation binding into a C
23
computationally observed for calix[4]arene amide derivatives. A similar effect can be seen in the
1
+
+
+
H NMR spectra of L2 complexes with Li , Na and K cations in deuterated chloroform (Fig. 8).
t-Bu
b
a1
+
KL2
H
4
O
H
+
a2
NaL2
c
O
N
d
+
LiL2
L2
7.2
7.1
7.0
6.9
6.8
/ ppm
6.7
6.6
δ
1
+
+
+
Fig. 8 H NMR spectra of ArꢀH protons of L2 , LiL2 , NaL2 and KL2 in deuterated chloroform at
o
2
5 C.
Molecular dynamics simulations of alkali metal cation complexes of L1, E–L2 and Z–L2 in
acetonitrile rendered additional details about the molecular structures (Figs. 9, S16, and S17, ESI),
binding site geometries and cation coordination, as well as the information on the process of solvent
molecule inclusion in the calixarene hydrophobic cavity. During these simulations, alkali metal
cations were coordinated by all four ether oxygen atoms and by variable number of carbonyl oxygen
atoms (Tables S2–S3, S5–S8, ESI). The ligandꢀcation interaction energies in the complexes of L1
and rotamers of L2 were more favorable for smaller cations, whereby lithium cation was the most
strongly bound to the calixarene ligands and cesium cation had the weakest interactions with these
compounds. The coordination number of the cations depended on size as well. While lithium cation
was coordinated by two carbonyl oxygen atoms on average, cesium was coordinated by nearly all
four, owing to its large ionic radius. The distributions of distances between carbonylꢀoxygen atoms
and metal cations in the complexes are given in Figs. S18, S19, and S31–34, ESI. The inclusion of
acetonitrile molecules in the hydrophobic cavities of complexes of L1 and L2 with alkali metal
cations was more pronounced than in the cases of free ligands. Specifically bound MeCN molecule
was present during most of the simulation time for all complexes, whereby 3 to 6 different molecules
were exchanged during a single MD simulation. Upon the inclusion of the solvent molecule, the
shape of calixarene cone became more regular. That was reflected by a decrease of the absolute
1
4

Page 15 of 28
Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C7CP03920D
difference between the average values of the distances between opposite aromatic carbon atoms
bound to tertꢀbutyl groups and a referent distance for a pure cone shape (Tables S2–S4, S5–S8,
2
3
ESI). Also, the calixarene cone became more rigid when the hydrophobic cavity was filled with the
solvent molecule, which was reflected by a decrease in the standard deviation of the distances
between opposite aromatic carbon atoms bound to tertꢀbutyl groups (Tables S2–S3, S5–S8, ESI).
The interaction energy between specifically bound acetonitrile molecule and calixarene ligand was
–
1
almost the same for all complexes of L1 and L2, being around –50 kJ mol .
b)
a)
c)
d)
f)
e)
1
5

Physical Chemistry Chemical Physics
Page 16 of 28
View Article Online
DOI: 10.1039/C7CP03920D
g)
h)
j)
i)
+
+
+
+
Fig. 9 Structures of a) LiZ–L2MeCN , b) LiE–L2MeCN , c) NaZ–L2MeCN , d) NaE–L2MeCN ,
+
+
+
+
+
e) KZ–L2MeCN , f) KE–L2MeCN , g) RbZ–L2MeCN , h) RbE–L2MeCN , i) CsZ–L2MeCN
+
o
and j) CsE–L2MeCN obtained by MD simulations in acetonitrile at 25 C. Hydrogen atoms are
omitted for clarity.
3
.4 Studies of alkali metal cations complexation in benzonitrile
The complexation reactions of ligands L1 and L2 with alkali metal cations were also studied in
+
+
+
benzonitrile. We have already investigated the complexation of compound L3 with Li , Na , and K
cations in benzonitrile, and have observed the coordination of metal cation by PhCN molecule
+
16
included in the ligand hydrophobic cone in the case of LiL3 complex. In addition, we performed
calorimetric titrations of calixareneꢀmetal complexes with acetonitrile in benzonitrile by which the
process of the inclusion of acetonitrile molecule in the calixarene cone was thermodynamically
1
6
characterized.
Like in acetonitrile, calixarenes L1 and L2 were shown to strongly bind most of the alkali
metal cations in benzonitrile. The values of stability constants of the corresponding complexes,
along with the reaction enthalpies and entropies, were determined by means of direct
+
+
+
+
+
microcalorimetric titrations (CsL1 , RbL1 , CsL2 , RbL2 , KL2 , Figs. S37, S38, and S56–S58,
1
6

Page 17 of 28
Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C7CP03920D
ESI) or by competition titration experiments (lithium, sodium, and potassium complexes of L1 and
L2, Fig. 10, S39, S40, S59, S60, ESI, and Table 2). The values of the reaction enthalpies for the
–1
reactions of L1 with lithium, sodium, and potassium cations are lower about 5 kJ mol in
comparison to those of the complexation reactions of L2 with the same cations. This effect,
accompanied by the more favorable complexation entropies for the reactions of L1, results in the
+
+
+
more stable complexes of LiL1 , NaL1 , and KL1 species with respect to the analogous complexes
of ligand L2.
It should be noted that contrary to acetonitrile (where sodium complexes with L1 and L2 are
+
the most stable), in benzonitrile both macrocycles show highest affinities towards Li cation. The
same was previously observed in the case of the derivative L3 and accounted for by the favorable
coordination of the smallest lithium cation by a nitrile nitrogen atom of the benzonitrile molecule
16
included in the calixarene hydrophobic cone. The same arguments could be applied in the cases of
L1 and L2 complexes (see also below).
0
.0
50
45
40
35
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
-0.7
a)
b)
20
30
40
50
60
70
80
90
100
0.25
0.50
0.75
1.00
1.25
1.50
+
+
t / min
n(K ) / n(RbL1 )
+
–4 –3
3
Fig. 10 a) Microcalorimetric titration of RbL1 (c = 4.73 × 10 mol dm , V = 1.4182 cm ) with
–
3
–3
KSCN (c = 3.33 × 10 mol dm ) in benzonitrile; t = 25 °C; b) Dependence of successive enthalpy
+
+
change on n(K ) / n(RbL1 ) ratio. ■ experimental; — calculated.
1
7

Physical Chemistry Chemical Physics
Page 18 of 28
View Article Online
DOI: 10.1039/C7CP03920D
Table 2 Thermodynamic parameters for complexation of alkali metal cations with L1 and L2 in
o
benzonitrile at 25 C determined by direct and competitive microcalorimetric titrations.
°
°
°
(
ꢁ G ±SE
)
(
ꢁ H ±SE
)
(
ꢁ S ± SE
)
r
r
r
log K ±SE
Cation
kJ mol−
1
kJ mol−1
J mol K
−1 −1
L1
+
Li
11.86 ± 0.05
11.53 ± 0.02
8.53 ± 0.02
6.13 ± 0.02
3.64 ± 0.02
–67.7 ± 0.3
–65.8 ± 0.1
–48.7 ± 0.1
–35.0 ± 0.1
–20.8 ± 0.1
L2
–61.0 ± 0.6
–59.4 ± 0.3
–40.8 ± 0.3
–31.00 ± 0.05
–18.3 ± 0.4
–51.6 ± 0.4
–58.0 ± 0.2
–39.9 ± 0.2
–25.9 ± 0.2
–12.6 ± 0.5
54 ± 2
+
Na
26.0 ± 0.6
29.4 ± 0.6
30.4 ± 0.3
27 ± 2
+
K
+
Rb
Cs
+
+
Li
10.68 ± 0.09
10.40 ± 0.05
7.15 ± 0.05
5.43 ± 0.01
3.20 ± 0.07
–46 ± 1
–53.82 ± 0.7
–35.5 ± 0.7
–23.5 ± 0.2
–7.5 ± 0.1
51 ± 5
18.8 ± 3
18 ± 3
25.3 ± 0.7
36 ± 2
+
Na
+
K
+
Rb
Cs
+
SE = standard error of the mean (N = 3–5).
3
.5 Inclusion of acetonitrile molecule into the calixarene–cation complexes in benzonitrile
In order to study the inclusion of acetonitrile molecule in the hydrophobic cone of L1 and L2
complexes, the benzonitrile solutions of these complexes were microcalorimetrically titrated by the
solution of MeCN in PhCN. It was reasonably (according to the complex stabilities) assumed that
under the conditions used nearly all of the calixarene molecules were in the complexed form. As an
+
example, titration of NaL1 complex is shown in Fig. 11 (all other titrations are presented in Figs.
S41–S43, ESI). The results of these titrations are presented in Table 3.
3
3
2
2
1
1
5
0
5
0
5
0
5
0
5
-
0.3
b)
a)
-0.6
-
-
-
-
0.9
1.2
1.5
1.8
-
-10
-15
-2.1
20
40
60
80
100
120
140
160
0
10
20
30
40
50
60
70
80
+
t / min
n(MeCN) / n(NaL1 )
+
–4
–3
3
Fig. 11 a) Microcalorimetric titration of NaL1 (c = 2.90 × 10 mol dm , V = 1.4182 cm ) with
–3
MeCN (c = 0.100 mol dm ) in benzonitrile; t = 25 °C; b) Dependence of successive enthalpy change
+
on n(MeCN) / n(NaL1 ) ratio. ■ experimental; ― calculated.
1
8

Page 19 of 28
Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C7CP03920D
Table 3 Thermodynamic parameters for the inclusion of acetonitrile in the complexes of L1 and L2
o
with alkali metal cations in benzonitrile at 25 C determined by microcalorimetric titrations.
°
°
°
(
ꢁ G ±SE
)
(
ꢁ H ±SE
)
(
ꢁ S ± SE
)
r
r
r
log K ±SE
Complex
kJ mol−
1
kJ mol−1
J mol K
−1 −1
+
ML1
+
LiL1
1.08 ± 0.01
2.266 ± 0.006
1.955 ± 0.004
2.07 ± 0.01
–6.14 ± 0.05
–12.94 ± 0.04
–11.16 ± 0.02
–11.80 ± 0.06
–25.4 ± 0.2
–43.8 ± 0.3
–47.1 ± 0.3
–35.2 ± 0.3
–64.7 ± 0.4
–103 ± 1
–121 ± 1
–78 ± 1
+
NaL1+
KL1
+
RbL1
+
ML2
+
LiL2
1.015 ± 0.008
2.359 ± 0.007
2.09 ± 0.01
–5.80 ± 0.05
–13.47 ± 0.04
–11.90 ± 0.06
–11.30 ± 0.03
–25.1 ± 0.3
–37.2 ± 0.2
–34.8 ± 0.5
–31.2 ± 0.1
–65 ± 1
–79.6 ± 0.7
–77 ± 2
+
NaL2+
KL2
+
RbL2
1.981 ± 0.006
–66.7 ± 0.4
SE = standard error of the mean (N = 3–5).
+
+
Complexes LiL1 and LiL2 have almost the same affinity towards acetonitrile in
benzonitrile, and this parity also holds for the corresponding values of reaction enthalpies and
entropies. More pronounced differences in the affinities for MeCN molecule were found in the cases
of sodium and potassium complexes. Although reaction enthalpies for the inclusion of acetonitrile in
+
+
the cone of NaL1 and KL1 are more favorable than that for their L2 analogs, more negative
+
+
reaction entropies in the cases of NaL1 and KL1 complexes eventually result in the lower stability
+
+
+
of NaL1MeCN and KL1MeCN ternary species compared to that of NaL2MeCN and
+
+
+
KL2MeCN . It is interesting to note that stability constants of LiL1MeCN and LiL2MeCN species
are an order of magnitude lower that those corresponding to acetonitrile adducts of L1 and L2
complexes with other alkali metal cations. This can be, at least partially, explained by taking into
account already mentioned inclusion of benzonitrile molecule in the calixarene cone, previously
+
16
observed in the case of LiL3 complex. As an evidence of the benzonitrile inclusion in the lithium
complexes can serve a comparison of reaction enthalpies and entropies of acetonitrile inclusion
reactions (Table 3). The reaction enthalpies for the MeCN inclusion in the lithium complexes are
less favorable than for the other complexes, which can be rationalized by considering the possibility
+
+
of substitution of benzonitrile molecule included in the LiL1 or LiL2 cone by acetonitrile
molecule. Such a substitution is most probably enthalpically less favorable than the inclusion of
acetonitrile molecule in an unoccupied cone. Similar reasoning can be applied in the analysis in
terms of entropy which is more favorable for the lithium L1 and L2 complexes compared to the
others. That could be explained by taking into account that during the substitution process the loss of
1
9

Physical Chemistry Chemical Physics
Page 20 of 28
View Article Online
DOI: 10.1039/C7CP03920D
acetonitrile translation entropy is compensated by the release of benzonitrile molecule bound to the
calixarene cone.
To additionally confirm the importance of specific interaction with PhCN molecule, we
o
calculated the ꢁꢁ
t
G values, i.e. the standard transfer Gibbs energies of sodium, potassium, and
rubidium complexes relative to the standard transfer Gibbs energies of the corresponding lithium
complexes, Eqs. (1 and 2):
o
+
+
o
+
o
+
∆
+
∆ G (MLMeCN (MeCN)→ M–L (PhCN)) = (∆ G (M + L, PhCN) – ∆ G (M + L, MeCN) +
t
r
r
o
+
o
+
o
+
o
+
∆
t
G (M , MeCN→PhCN)) – (∆
r
G (Li + L, PhCN) – ∆
r
G (Li + L, MeCN) + ∆
t
G (Li , MeCN→PhCN))
(1)
(2)
o
+
+
o
+
+
∆
+
∆
t
G (MLMeCN (MeCN)→ MLMeCN (PhCN)) = ∆∆
t
G (MLMeCN (MeCN)→ M–L (PhCN)) +
o
+
o
+
∆
r
G (M –L + MeCN, PhCN) – ∆
r
G (Li –L + MeCN, PhCN)
+
where M denotes Na, K, or Rb, M–L corresponds to acetonitrileꢀfree complexes with or without
o
+
o
+
benzonitrile molecule included in the ligand cone whereas ∆
r
G (M + L, solvent) and ∆
r
G (Li + L,
+
solvent) stand for the corresponding reaction Gibbs energies in the given solvent. The ꢁ
t
G°(M ,
+
+
+
+
MeCN→PhCN) values for Li , Na , K and Rb were calculated by combining the Gibbs energies of
52
53
54
transfer of cations from water to acetonitrile or benzonitrile based on Ph
4
AsPh
4
B convention
+
+
+
(ꢁ G°(M , MeCN→PhCN) = ꢁ G°(M , H G°(M , H
t t 2 t 2
O→PhCN) – ꢁ O→MeCN)).
In that way, the standard transfer Gibbs energies of ligands L1 and L2 cancel out of the equation,
and all conclusions are drawn based on comparison with the transfer of lithiumꢀcalixarene
o
+
complexes. From the data listed in Table 4 it is evident that the ∆∆
t
G (MLMeCN (MeCN)→ M–
+
–1
L (PhCN)) values for all complexes are about 6 kJ mol more positive than those corresponding to
+
+
the MLMeCN (MeCN)→ MLMeCN (PhCN) transfer, relative to the transfer of lithium complex.
The reason for these consistent differences can be assigned to the process of benzonitrile molecule
inclusion which is most pronounced in the case of lithium complex and less expressed for the other
alkali metal cation complexes of L1 and L2 in benzonitrile. In the above analysis it was assumed
that all alkali metal cations complexes in MeCN are preferably in the form of acetonitrile adduct
(Figs. 9, S16 and S17, ESI), as suggested by the results the MD simulations.
2
0

Page 21 of 28
Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C7CP03920D
Table 4 Standard transfer Gibbs energies of sodium, potassium, and rubidium complexes relative to
the standard transfer Gibbs energies of the corresponding lithium complexes (Eqs. 1 and 2).
°
°
(
ꢁꢁ G ±SE
)
(
ꢁꢁ G ±SE
)
t
t
Complex
kJ mol−
1
kJ mol−1
+
+
MLMeCN (MeCN)
→
MLMeCN (MeCN)
+
+
M–L (PhCN)
→ MLMeCN (PhCN)
L1
+
NaL1+
7.0
6.2
1.9
0.2
1.2
–3.8
KL1
+
RbL1
L2
+
NaL2+
8.3
4.1
–4.3
0.6
–2.0
–9.8
KL2
+
RbL2
In order to assess the molecular structures of alkali metal cation complexes of L1 and L2
ligands in benzonitrile, along with their affinities towards the inclusion of benzonitrile molecule,
MD simulations were performed.
During the simulations of the alkali metal cation complexes of
L1, E–L2 and Z–L2 in benzonitrile the inclusion of solvent molecules was observed. In these
adducts the included benzonitrile molecules were oriented in two different ways, one in which the
+
nitrile group was pointing towards the bulk (structures denoted as MLPhCN where M stands for the
alkali metal cation and L is either L1, E–L2 or Z–L2 ligand), and the other where the nitrile group
+
was inside the calixarene cone and oriented towards the alkali metal cation (denoted as MLPhCN' ).
The presence of the latter adduct was more pronounced in the simulations of lithium and sodium
complexes, but was also found in the simulations including other cations (Tables S9–S12 and S14–
S21, ESI). The most favorable interaction between included benzonitrile molecule with a metal
+
+
cation was that found for LiLPhCN' complexes followed by those of MLPhCN' type with sodium,
potassium, and rubidium (Tables S9–S12 and S14–S21, ESI). Upon the coordination of cation by the
nitrile group, in most cases an average number of coordinating carbonyl groups decreased. This was
+
+
most evident for the NaL1 and NaL2 complexes where a reduction of the number of coordinating
–1
carbonyl groups resulted in an increase of the cationꢀligand interaction energies by up to 40 kJ mol
with respect to the same interaction in acetonitrile. As was the case for the acetonitrile adducts with
alkali metal cation complexes of L1, E–L2, and Z–L2, the inclusion of benzonitrile molecule in the
hydrophobic cavity resulted with a more rigid basket of more regular cone shape. Molecular
+
dynamics simulations of MLMeCN complexes in benzonitrile were also carried out in order to
assess the structures and energetics of these species in this solvent. The compounds had similar
interaction energies between complexed cations, ligands, and included acetonitrile molecules as the
corresponding adducts in acetonitrile. As expected, the most pronounced difference was found in the
solvation energies of the complexes in MeCN and PhCN.
2
1

Physical Chemistry Chemical Physics
Page 22 of 28
View Article Online
DOI: 10.1039/C7CP03920D
3
.6 Studies of alkali metal cation complexation in methanol
Complexation reactions of L1 and L2 were studied in methanol as well. This solvent was used
because of its abilities of forming rather strong hydrogen bonds, both as proton donors and as proton
acceptors. Complexation of alkali metal cations by L1 and L2 was followed by direct and
competition microcalorimetric titrations (Table 5, Figs. S91–S94, S98–S101, ESI).
Table 5 Thermodynamic parameters for complexation of alkali metal cations with L1 and L2 in
o
methanol at 25 C determined by microcalorimetric titrations.
°
°
°
(
ꢁ G ± SE
)
(
ꢁ H ±SE
)
(
ꢁ S ± SE
)
r
r
r
log K ±SE
Cation
kJ mol−
1
kJ mol−1
J mol K
−1 −1
L1
+
Li
4.06 ± 0.03
7.78 ± 0.06
5.84 ± 0.05
3.77 ± 0.01
–23.2 ± 0.2
–44.5 ± 0.4
–33.4 ± 0.3
–21.50 ± 0.07
L2
–20.72 ± 0.02
–41.9 ± 0.1
–30.45 ± 0.07
–18.36 ± 0.08
–5.8 ± 0.1
–45.9 ± 0.4
–40.1 ± 0.4
–21 ± 1
58 ± 1
–5 ± 2
–23 ± 2
1 ± 4
+
Na
+
K
+
Rb
+
Li
3.630 ± 0.004
7.33 ± 0.03
5.34 ± 0.01
3.22 ± 0.01
–7.5 ± 0.2
–45.6 ± 0.2
–40.1 ± 0.2
–22.1 ± 0.3
44.3 ± 0.5
–13 ± 1
–32.3 ± 0.6
–13 ± 1
+
Na
+
K
+
Rb
SE = standard error of the mean (N = 3–5).
To get more detailed thermodynamic information about the solvent effect on the reactions
studied, an effort was made to determine the values of standard Gibbs energies of solution for L1
and L2 in acetonitrile and methanol. For that reason, the solubilities of the ligand in MeCN and
MeOH were measured (Table 6). Standard solution Gibbs energies for methanol and acetonitrile as
solvents were calculated from the solubility data using equation:
∆
solG° = –RTln K° = –RT ln(
γ
L
s/c°) ≈ –RT ln(s/c°)
(3)
–3
where s denotes solubility, c° = 1 mol dm is standard concentration, and stands for the activity
γ
L
coefficient of the ligand which is assumed to be close to unity.
2
2

Page 23 of 28
Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C7CP03920D
Table 6 Solubilities and derived standard solution Gibbs energies of L1 and L2 in acetonitrile and
methanol at 25 °C.
L1
L2
3
–3
–1
3
–3
–1
1
0 × s / mol dm
ꢁsolG° / kJ mol
10 × s / mol dm
ꢁsolG° / kJ mol
Acetonitrile
Methanol
0.77
17.7
10.2
0.47
40.4
19.0
8.0
1
6.1
The obtained values are reported in Table 6 and were used to calculate transfer Gibbs energies of
+
+
+
+
+
LiL , NaL , KL , RbL and CsL complexes from MeCN to MeOH (Schemes 2 and S2–S3) by the
following equation:
+
+
t t t
∆ G°(ML , MeCN→MeOH) = ∆ G°(M , MeCN→MeOH) + ∆ G°(L, (MeCN→MeOH) +
+
∆
r
G°(MeOH) – ∆
r
G°(MeCN)
(4)
+
+
+
The ꢁ G°(M , MeOH→MeCN) values for Na and K were calculated by combining the Gibbs
t
52
52
energies of transfer of cations from water to methanol or acetonitrile based on Ph
4
AsPh
4
B
54
+
+
+
convention (ꢁ G°(M , MeCN→MeOH) = ꢁ G°(M , H G°(M , H
t t 2 t 2
O→MeOH) – ꢁ O→MeCN)).
From the results presented in Table 5, a decrease of L1 and L2 affinities towards alkali metal
cations is evident when compared to the stability of the corresponding complexes in acetonitrile and
benzonitrile (Table 1 andTable 2). The origin of the effect from the thermodynamic point of view
lies primarily in the favorable transfer Gibbs energies of L1 and L2 from acetonitrile to methanol
(Schemes 2 and S1, S2, ESI) and in unfavorable transfer Gibbs energies of the ligandꢀcation
complexes.
The impact of the differences in alkali metal cation solvation in the studied solvents (i.e. the
corresponding transfer Gibbs energies) on the overall decrease in the affinities is not of great
importance, except for lithium cation. In the latter case, the favorable cation transfer to methanol
increases the standard Gibbs energy of complexation in methanol with respect to the one in
acetonitrile as much as the sum of those of the free ligands and their complexes. Methanol is a
protonꢀdonating solvent which can form relatively strong hydrogenꢀbonding interactions with the
carbonylꢀoxygen atoms at the calixarene lower rim substituents. These interactions are the most
likely cause of the favorable transfer of ligands L1 and L2 from acetonitrile to methanol. The
increase in the Gibbs energies of solvation from acetonitrile to methanol of all calixareneꢀalkali
metal cation complexes studied can be ascribed to the now missing interaction between carbonyl
oxygen atoms and methanol molecules due to the cation coordination by the ligands carbonyl
groups. In addition, an important contribution to the solvation of calixarene complexes in acetonitrile
and methanol arises from the specific solvation, namely the inclusion of the solvent molecule in the
2
3

Physical Chemistry Chemical Physics
Page 24 of 28
View Article Online
DOI: 10.1039/C7CP03920D
complexed calixarene basket. Such an interaction is more favorable in the case of acetonitrile
23
23
inclusion, thus having more beneficial effect on the overall complex stability. When analyzing
the contributions to standard Gibbs energy of complexation in acetonitrile and methanol, it can be
seen that a decrease of the stability of L1 and L2 complexes is primarily influenced by the enthalpic
contribution which is less favorable in methanol (Table 1, Table 2, and Table 5). Most of the
complexes are stabilized by favorable reaction enthalpy except for the lithium ones where the
favorable entropic contribution to the reaction Gibbs energy is quite significant. This finding can be
related to the relatively high entropy of the lithium cation desolvation in methanol.
ꢁ
r
G° = –64.3 kJ mol–1
LiL1 (MeCN)
ꢁ G° = 12.6 kJ mol
+
+
a)
Li (MeCN)
+
L1(MeCN)
–1
ꢁ
t
G° = –21 kJ mol–1
ꢁ
t
G° = –7.5 kJ mol–1
t
+
+
Li (MeOH)
+
L1(MeOH)
LiL1 (MeOH)
ꢁ
G° = –23.2 kJ mol–1
r
ꢁ
r
G° = –69.4 kJ mol–1
+
+
Na (MeCN)
+
L1(MeCN)
NaL1 (MeCN)
b)
G° = –5.4 kJ mol–1 ꢁ G° = –7.5 kJ mol
−1
−1
ꢁ
t
t
ꢁ G° = 12.0 kJ mol
t
+
+
Na (MeOH)
+
L1(MeOH)
NaL1 (MeOH)
–1
r
ꢁ G° = –44.5 kJ mol
+
+
Scheme 2. Thermodynamic cycles for complexation of a) Li and b) Na with L1 in acetonitrile and
methanol expressed in terms of Gibbs energies.
Further insight into the structural aspects of the studied alkali metal cation complexes in
methanol was obtained by molecular dynamics simulations of these compounds. The same
conclusions can be drawn regarding the cation coordination and cationꢀligand interaction energies as
those presented above in the discussion of the results of MD simulations in acetonitrile. The most
+
dominant form of all complexes in methanol was an MLMeOH adduct which was present over 95
%
of simulation time during every simulation (Tables S22–S26, ESI). The interaction energies
2
4

Page 25 of 28
Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C7CP03920D
–
1
between the ligand and the included methanol molecule were around –50 kJ mol in all of the
observed adducts. Upon the inclusion of the solvent molecule, the calixarene basket became more
regular and rigid, but this structural change did not influence the coordination of the cation, which
remained the same as it was in the solventꢀfree complexes.
4
. Conclusions
In this paper we present a thorough and systematic study of complexation of alkali metal cations by
tertiary amide calix[4]arene derivatives L1 and L2 in acetonitrile, benzonitrile, and methanol.
Thermodynamic quantities related to the studied reactions as well as MeOH→MeCN transfer Gibbs
energies of reactants and products were determined.
The cationꢀbinding affinities of L1 and L2 were found to be much higher than that of
previously studied secondaryꢀamideꢀ based compound L3, which could be mostly ascribed to the
presence of intramolecular hydrogen bonds formed between amide groups in the latter ligand (these
bonds need to be disrupted in order to achieve ion binding). Compound L1 comprising two hexyl
groups bound to the amide nitrogen atom was proven to be better cation binder than L2 with one
methyl instead of hexyl, due to the better electronꢀdonating ability of hexyl compared to methyl
group, and hence higher basicity of the coordinating carbonyl oxygen atoms.
Most of the studied reactions were enthalpy driven. In some cases the favorable entropic
contribution to the reaction Gibbs energy was quite significant, which was most likely related to the
(de)solvation of the reactants and products of complexation reactions. Thermodynamic stabilities of
the complexes generally decreased in the solvent order: MeCN > PhCN >> MeOH. Apart from the
differences in the solvation of cations, ligands, and complexes, inclusion of solvent molecules in the
calixarene hydrophobic cavity could also be responsible for this finding. The occurrence of this
1
process was indicated by the results of H NMR and MD studies, and was quantitatively
characterized by means of isothermal titration calorimetry. In line with this, the highest affinities of
+
L1 and L2 for Li in benzonitrile (in other solvents stability peak corresponds to sodium cation)
could be accounted for by considering the possibility of inclusion of solvent molecule in the
calixarene cone, and coordination of this smallest cation by PhCN nitrile group.
5
. Acknowledgments
This work has been fully supported by Croatian Science Foundation under the project IPꢀ2014ꢀ09ꢀ
309 (SupraCAR). Computational resources provided by the Croatian National Grid Infrastructure
www.croꢀngi.hr) at Zagreb University Computing Centre (Srce) were used for this publication.
7
(
2
5

Physical Chemistry Chemical Physics
Page 26 of 28
View Article Online
DOI: 10.1039/C7CP03920D
References
1
.
.
Z. Asfari, V. Böhmer, J. Harrowfield and J. Vicens, Eds., Calixarenes 2001, Kluwer
Academic Publishers, Dordrecht, 2001.
2
C. D. Gutsche, Calixarenes: An Introduction, The Royal Society of Chemistry, Cambridge,
2
nd edn., 2008.
3
.
.
J. Vicens and J. Harrowfield, Eds., Calixarenes in the Nanoworld, Springer, Dordrecht, 2007.
A. F. Danil de Namor, R. M. Cleverley, and M. L. ZapataꢀOrmachea, Chem. Rev., 1998, 98,
4
2
495–2526.
5.
6.
7.
8.
V. Böhmer, Angew. Chemie Int. Ed. English, 1995, 34, 713–745.
A. Casnati, Chem. Commun., 2013, 49, 6827–6830.
W. Sliwa and T. Girek, J. Incl. Phenom. Macrocycl. Chem., 2010, 66, 15–41.
B. Mokhtari, K. Pourabdollah, and N. Dalali, J. Incl. Phenom. Macrocycl. Chem., 2010, 69,
1
–55.
9
1
1
1
1
1
1
.
R. V Rodik, V. I. Boyko, and V. I. Kalchenko, Curr. Med. Chem., 2009, 16, 1630–1655.
0. B. S. Creaven, D. F. Donlon, and J. McGinley, Coord. Chem. Rev., 2009, 253, 893–962.
1. D. T. Schühle, J. a. Peters, and J. Schatz, Coord. Chem. Rev., 2011, 255, 2727–2745.
2. R. Ludwig and N. T. K. Dzung, Sensors, 2002, 2, 397–416.
3. N. Bregović, N. Cindro, L. Frkanec, and V. Tomišić, Supramol. Chem., 2016, 28, 608–615.
4. I. Sviben, N. Galić, V. Tomišić, and L. Frkanec, New J. Chem., 2015, 39, 6099–6107.
5. A. F. Danil de Namor, T. T. MatsufujiꢀYasuda, K. ZegarraꢀFernandez, O. A. Webb, and A. El
Gamouz, Croat. Chem. Acta, 2013, 86, 1–19.
1
1
6. G. Horvat, V. Stilinović, B. Kaitner, L. Frkanec, and V. Tomišić, Inorg. Chem., 2013, 52,
1
2702–12.
7. J. Požar, G. Horvat, M. Čalogović, N. Galić, L. Frkanec, and V. Tomišić, Croat. Chem. Acta,
012, 85, 541–552.
2
18. J. Požar, T. Preočanin, L. Frkanec, and V. Tomišić, J. Solution Chem., 2010, 39, 835–848.
19. A. Hamdi, R. Abidi, and J. Vicens, J. Incl. Phenom. Macrocycl. Chem., 2007, 60, 193–196.
20. F. ArnaudꢀNeu, S. Barboso, F. Berny, A. Casnati, N. Muzet, A. Pinalli, R. Ungaro, M.ꢀJ.
SchwingꢀWeill, and G. Wipff, J. Chem. Soc. Perkin Trans. 2, 1999, 1727–1738.
2
1. F. ArnaudꢀNeu, G. Barrett, S. Fanni, D. Marrs, W. McGregor, M. A. McKervey, M.ꢀJ.
SchwingꢀWeill, V. Vetrogon, and S. Wechsler, J. Chem. Soc. Perkin Trans. 2, 1995, 3, 453–
4
61.
2
2
2. A. Arduini, E. Ghidini, A. Pochini, R. Ungaro, G. D. Andreetti, G. Calestani, and F. Ugozzoli,
J. Incl. Phenom., 1988, 6, 119–134.
3. G. Horvat, V. Stilinović, T. Hrenar, B. Kaitner, L. Frkanec, and V. Tomišić, Inorg. Chem.,
2
012, 51, 6264–6278.
2
6

Page 27 of 28
Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C7CP03920D
2
2
2
4. A. F. D. de Namor, N. A. de Sueros, M. A. McKervey, G. Barrett, F. A. Neu, and M. J.
SchwingꢀWeill, J. Chem. Soc. Chem. Commun., 1991, 1546–1548.
5. V. Tomišić, N. Galić, B. Bertoša, L. Frkanec, V. Simeon, and M. Žinić, J. Incl. Phenom.
Macrocycl. Chem., 2005, 53, 263–268.
6. N. Galić, N. Burić, R. Tomaš, L. Frkanec, and V. Tomišić, Supramol. Chem., 2011, 23, 389–
3
97.
2
7. T. Benković, V. Tomišić, L. Frkanec, and N. Galić, Croat. Chem. Acta, 2012, 85, 469–477.
8. E. Nomura, M. Takagaki, C. Nakaoka, M. Uchida, and H. Taniguchi, J. Org. Chem., 1999,
2
6
4, 3151–3156.
2
9. L. Frkanec, A. Višnjevac, B. KojićꢀProdić, and M. Žinić, Chem. Eur. J., 2000, 6, 442–453.
0. A. F. Danil De Namor, W. B. AparicioꢀAragon, M. T. Goitia, and A. R. Casal, Supramol.
Chem., 2004, 16, 423–433.
3
3
3
3
1. A. F. Danil de Namor, M. L. ZapataꢀOrmachea, and R. G. Hutcherson, J. Phys. Chem. B,
1
999, 103, 366–371.
2. M. T. Bakić, D. Jadreško, T. Hrenar, G. Horvat, J. Požar, N. Galić, V. Sokol, R. Tomaš, S.
Alihodžić, M. Žinić, L. Frkanec, and V. Tomišić, RSC Adv., 2015, 5, 23900–23914.
3. A. S. de Araujo, O. E. Piro, E. E. Castellano, and A. F. Danil de Namor, J. Phys. Chem. A,
2
008, 112, 11885–11894.
3
3
3
4. P. Gans, A. Sabatini, and A. Vacca, Talanta, 1996, 43, 1739–53.
5. J. Tellinghuisen, J. Phys. Chem. B, 2007, 111, 11531–11537.
6. H. J. C. Berendsen, D. van der Spoel, and R. van Drunen, Comput. Phys. Commun., 1995, 91,
4
3–56.
3
7. E. Lindahl, B. Hess, and D. van der Spoel, J. Mol. Model., 2001, 7, 306–317.
8. D. Van Der Spoel, E. Lindahl, B. Hess, G. Groenhof, A. E. Mark, and H. J. C. Berendsen, J.
Comput. Chem., 2005, 26, 1701–1718.
3
3
4
4
4
4
4
9. B. Hess, C. Kutzner, D. van der Spoel, and E. Lindahl, J. Chem. Theory Comput., 2008, 4,
4
35–447.
0. S. Pronk, S. Páll, R. Schulz, P. Larsson, P. Bjelkmar, R. Apostolov, M. R. Shirts, J. C. Smith,
P. M. Kasson, D. van der Spoel, B. Hess, and E. Lindahl, Bioinformatics, 2013, 29, 845–854.
1. S. Páll, M. J. Abraham, C. Kutzner, B. Hess, and E. Lindahl, in Lecture Notes in Computer
Science, eds. S. Markidis and E. Laure, Springer, Cham, 2015, pp. 3–27.
2. M. J. Abraham, T. Murtola, R. Schulz, S. P?ll, J. C. Smith, B. Hess, and E. Lindahl,
SoftwareX, 2015, 1–2, 19–25.
3. W. L. Jorgensen, D. S. Maxwell, and J. TiradoꢀRives, J. Am. Chem. Soc., 1996, 118, 11225–
1
1236.
4. W. C. Swope, H. C. Andersen, P. H. Berens, and K. R. Wilson, J. Chem. Phys., 1982, 76,
2
7

Physical Chemistry Chemical Physics
Page 28 of 28
View Article Online
DOI: 10.1039/C7CP03920D
6
37–649.
4
4
4
4
5. T. Darden, D. York, and L. Pedersen, J. Chem. Phys., 1993, 98, 10089–10092.
6. S. Nosé, Mol. Phys., 1984, 52, 255–268.
7. W. Hoover, Phys. Rev. A, 1985, 31, 1695–1697.
8. G. J. Martyna, M. E. Tuckerman, D. J. Tobias, and M. L. Klein, Mol. Phys., 1996, 87, 1117–
157.
1
4
5
9. W. Humphrey, A. Dalke, and K. Schulten, J. Mol. Graph., 1996, 14, 33–38.
0. A. F. Danil de Namor, S. Chahine, E. E. Castellano, and O. E. Piro, J. Phys. Chem. A, 2005,
1
09, 6743–6751.
1. K. B. Wiberg, P. R. Rablen, D. J. Rush, and T. A. Keith, J. Am. Chem. Soc., 1995, 117, 4261–
270.
5
4
5
2. Y. Marcus, Ion Properties, M. Dekker, New York, 1997.
5
3. A. F. D. de Namor and H. B. de Ponce, J. Chem. Soc. Faraday Trans. 1, 1988, 84, 1671–
1
678.
5
4. J. Burgess, Metal Ions in Solution, Ellis Horwood, Chichester, 1978.
2
8