Complexation of Phenol and Thiophenol by Amine N-Oxides: Isothermal Titration Calorimetry and ab Initio Calculations

Article abstract of DOI:10.1002/cphc.201000482

To develop a new solvent-impregnated resin (SIR) system for removal of phenols from water, the complex formation of dimethyldodecylamine N-oxide (DMDAO), trioctylamine N-oxide (TOAO), and tris(2-ethylhexyl)amine N-oxide (TEHAO) with phenol (PhOH) and thiophenol (PhSH) is studied. To this end we use isothermal titration calorimetry (ITC) and quantum chemical modeling (on B3LYP/6-311G(d,p)-optimized geometries: B3LYP/6-311+G(d,p), B3LYP/6-311++G(2d,2p), MP2/6-311+G(d,p), and spin component scaled (SCS) MP2/6-311+G(d,p); M06-2X/6-311+G(d,p)//M06-2X/6-311G(d,p), MP2 with an extrapolation to the complete basis set limit (MP2/CBS), as well as CBS-Q). The complexes are analyzed in terms of structural (e.g., bond lengths) and electronic elements (e.g., charges). Furthermore, complexation and solvent effects (in benzene, toluene, and mesitylene) are investigated by ITC measurements, yielding binding constants K, enthalpies ΔH⁰, Gibbs fre energies ΔG⁰, and entropies ΔS⁰ of complex formation, and stoichiometry N. The ITC measurements revealed strong 1:1 complex formation between both DMDAO-PhOH and TOAO-PhOH. The binding constant (K=1.7-5.7×10⁴ M^-1) drops markedly when water-saturated toluene was used (K=5.8×10³ M^-1), and Π-Π interaction with the solvent is shown to be relevant. Quantum mechanical modeling confirms formation of stable 1:1 complexes with linear hydrogen bonds that weaken on attachment of electron-withdrawing groups to the amine N-oxide moiety. Modeling also showed that complexes with PhSH are much weaker than those with PhOH, and in fact too weak for ITC determination. CBS-Q incorrectly predicts equal or even higher binding enthalpies for PhSH than for PhOH, which invalidates it as a benchmark for other calculations. Data from the straightforward SCS-MP2 method without counterpoise correction show very good agreement with the MP2/CBS values.Extraction of (thio)phenol in industrial processes requires detailed information on the nature of the H bonds between the extractants (here a series of amine N-oxides) and (thio)phenol. This is provided by a combination of high-level quantum chemical calculations and calorimetry. The picture shows raw data (top) together with peak-integration data and fitted curve (bottom) for the calorimetric titration of phenol with dimethyldodecylamine N-oxide. Copyright

Full text of DOI:10.1002/cphc.201000482

DOI: 10.1002/cphc.201000482
Complexation of Phenol and Thiophenol by Amine N-
Oxides: Isothermal Titration Calorimetry and ab Initio
Calculations
Ruud Cuypers,[a] Sukumaran Murali,[a] Antonius T. M. Marcelis,[a] Ernst J. R. Sudhçlter,[b] and
Han Zuilhof*[a]
To develop a new solvent-impregnated resin (SIR) system for
removal of phenols from water, the complex formation of di-
methyldodecylamine N-oxide (DMDAO), trioctylamine N-oxide
(TOAO), and tris(2-ethylhexyl)amine N-oxide (TEHAO) with
phenol (PhOH) and thiophenol (PhSH) is studied. To this end
we use isothermal titration calorimetry (ITC) and quantum
chemical modeling (on B3LYP/6-311G(d,p)-optimized geome-
tries: B3LYP/6-311+G(d,p), B3LYP/6-311++G(2d,2p), MP2/6-
311+G(d,p), and spin component scaled (SCS) MP2/6-311+
G(d,p); M06-2X/6-311+G(d,p)//M06-2X/6-311G(d,p), MP2 with
an extrapolation to the complete basis set limit (MP2/CBS), as
well as CBS-Q). The complexes are analyzed in terms of struc-
tural (e.g., bond lengths) and electronic elements (e.g., charg-
es). Furthermore, complexation and solvent effects (in ben-
zene, toluene, and mesitylene) are investigated by ITC meas-
urements, yielding binding constants K, enthalpies DH0, Gibbs
fre energies DG0, and entropies DS0 of complex formation, and
stoichiometry N. The ITC measurements revealed strong 1:1
complex formation between both DMDAO–PhOH and TOAO–
PhOH. The binding constant (K=1.7–5.7ꢀ104 mÀ1) drops mark-
edly when water-saturated toluene was used (K=5.8ꢀ103 mÀ1),
and p–p interaction with the solvent is shown to be relevant.
Quantum mechanical modeling confirms formation of stable
1:1 complexes with linear hydrogen bonds that weaken on at-
tachment of electron-withdrawing groups to the amine N-
oxide moiety. Modeling also showed that complexes with
PhSH are much weaker than those with PhOH, and in fact too
weak for ITC determination. CBS-Q incorrectly predicts equal or
even higher binding enthalpies for PhSH than for PhOH, which
invalidates it as a benchmark for other calculations. Data from
the straightforward SCS-MP2 method without counterpoise
correction show very good agreement with the MP2/CBS
values.
1. Introduction
Solvent-impregnated resins[1] (SIRs, Figure 1) are used as three-
phase separation systems to isolate or concentrate dilute sol-
utes from an aqueous phase. The SIRs are porous inert resin
particles containing a stationary organic liquid that impregna-
tes its pores. When a solute in an aqueous phase is brought
into contact with an SIR, the solute will diffuse from the aque-
ous phase to the solvent phase. Ion-exchange SIRs are already
widely used[2–5] to separate heavy-metal ions from aqueous
phases in a fast and simple way, by making use of the solubili-
ty of the solute in the liquid organic phase, in exchange for
protons. Only recently have investigations on other applica-
tions been performed, for example, the recovery of apolar or-
ganics.[6,7] Bench-scale or pilot-scale recovery of polar organics
like organic acids,[8,9] amino acids,[10] and flavonoids has also
been realized.[11] To enhance extraction, complex-forming ex-
tractants can be added to the organic phase. Complex forma-
tion inside the organic solvent increases the overall equilibrium
concentration of solute in the SIR, which significantly enhances
the overall extraction capacity.
In principle, molecular recognition and complexation of the
solute inside the SIR particles, as indicated in Figure 1, can pro-
vide a new and fast way to separate even rather polar organic
solutes from aqueous streams on a kilotonne per annum scale.
For effective removal of polar organic solutes such as ethers
[a] Ir. R. Cuypers, Dr. S. Murali, Dr. A. T. M. Marcelis, Prof. Dr. H. Zuilhof
Laboratory of Organic Chemistry, Wageningen University
Dreijenplein 8, 6703 HB Wageningen (The Netherlands)
Fax: (+31)317-484914
E-mail: Han.Zuilhof@wur.nl
[b] Prof. Dr. E. J. R. Sudhçlter
Department of Chemical Engineering
Laboratory of Nano-organic chemistry
Delft University of Technology
Julianalaan 136, 2628 BL Delft (The Netherlands)
Figure 1. Principle of solute extraction by a solvent-impregnated resin (SIR)
with an active extractant in the pores.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cphc.201000482.
ChemPhysChem 2010, 11, 3465 – 3473
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3465

H. Zuilhof et al.
and phenols in water-cleaning processes, the additional com-
plexation step of the solute in the organic solvent phase turns
out to be crucial for large-scale application, due to the relative-
ly high solubility of the solute in water. A rational way of opti-
mizing an extractant for these applications is by using molecu-
lar design. For this a thorough understanding of the complexa-
tion at the molecular level is required. Molecular modeling can
facilitate this understanding by providing useful data about
complex formation that can be compared with those of experi-
ments. Here we investigate a set of amine N-oxide extractants
for extraction of phenol from an aqueous environment, to es-
tablish a basis for future industrial design. For the extraction of
phenols, strong hydrogen bonds are desired for SIR-based pro-
cesses. It was shown that complex-forming compounds con-
taining a phosphoryl moiety are suitable extractants.[7,12] How-
ever, Torgov et al.[13–15] indicated that amine N-oxides would
likely be even better for phenol recovery, which prompted this
study with amine N-oxides as potential candidates that can
form hydrogen-bonded complexes. For example, an investiga-
tion[15] of the removal of low concentrations of phenol (10–
100 mgLÀ1) from sodium sulfate solutions (0.25–0.4m) with tri-
octylamine N-oxide (TOAO) dissolved in toluene (0.25–0.4m)
yielded distribution coefficients KD ([phenol]org/[phenol]water) of
1100 to 3300.
addition, we report for the first time intrinsic complexation
constants, which were obtained in water-free and water-satu-
rated organic solvents, in order to quantify the expected inhib-
iting effect of an aqueous environment, as this is relevant for
industrial purposes.
Measurements by ITC[21–25] can be used to directly obtain the
equilibrium constant K, stoichiometry N, and enthalpy of com-
plex formation DH0 in complexation reactions, and the
changes in entropy DS0 and free energy DG0 can be calculated
from those values. Although ITC measurements are usually car-
ried out in aqueous solutions,[26,27] its use is also unproblematic
with cosolvents like methanol or DMSO,[28] or with organic sol-
vents like chloroform,[29,30] toluene,[7,29,31,32] and others.[33] Ben-
zene, toluene, and mesitylene were used as apolar model sol-
vents. For toluene, a relatively small influence on hydrogen-
bonding interactions has been reported.[29,34,35]
For the development of an SIR extraction setup for the re-
moval of phenols from aqueous systems, these hydrogen-
bonding interactions are of vital importance. The present
paper discusses these factors, yielding an optimal candidate by
using ITC measurements in combination with quantum chemi-
cal theoretical methods. A range of different methods (B3LYP/
6-311+G(d,p)//B3LYP/6-311G(d,p),
M06-2X/6-311+G(d,p)//
M06-2X/6-311G(d,p), MP2/6-311+G(d,p)//B3LYP/6-311G(d,p),
Alkylated amine N-oxides combine low solubility in water
with a high negative charge on the oxygen atom, which is fa-
vorable for a hydrogen-bond acceptor. The most important hy-
drogen-bond characteristics to be explored in the current
study are the hydrogen-bond strength and the geometry of
the resulting complexes. From
MP2/6-311++G(2d,2p)//B3LYP/6-311G(d,p), spin component
scaled (SCS) MP2 using the same basis sets, with and without
counterpoise (cp) corrections, and a complete basis set extrap-
olation (MP2/CBS) method) were applied to the model com-
pounds depicted in Table 1. Since analysis of a small data set
the literature, neutral OÀH···O
Table 1. Model amine N-oxide compounds used in the quantum chemical calculations and geometric parame-
hydrogen bonds are expected to
ters describing the hydrogen-bond geometry.[a]
have
a
strength
of
4–
a
15 kcalmolÀ1[16,17] and cover
Compound
Molecular formula
Structure
range of hydrogen-bond lengths
trimethylamine N-oxide
dimethylethylamine N-oxide
triethylamine N-oxide
tri(trifluoromethyl)amine N-oxide
pyridine N-oxide
(CH3)3N+ÀOÀ
(CH3)2(C2H5)N+ÀOÀ
(C2H5)3N+ÀOÀ
(CF3)3N+ÀOÀ
and OÀH···O angles,[18,19] as these
parameters depend strongly on
the molecular geometry of the
C5H5N+ÀOÀ
system at hand. Hydrogen
bonds between charged donor
[a] Definitions analogous for thiophenol (PhSH).
or acceptor atoms are usually
significantly stronger, with typi-
cal DH0 values of about À15 to 40 kcalmolÀ1.[16,17] For amine N-
oxide hydrogen bonds a binding strength in the range of 5–
14 kcalmolÀ1 is expected. Recent computational studies show
that accurate geometric information is required to make mean-
ingful predictions.[20]
of hydrogen-bonded complexes suggested that the M06 set of
density functionals might perform somewhat better in describ-
ing hydrogen-bonding interactions,[36] M06-2X/6-311+G(d,p)//
M06-2X/6-311G(d,p)[37,38] calculations were also performed. Fi-
nally, the composite CBS-Q method, a complete basis set
method of Petersson and co-workers for obtaining very accu-
rate energies,[39] was used to obtain an absolute benchmark for
the energy values. The different methods were chosen to di-
minish the chance of finding highly method-dependent results.
Recently, SCS-MP2 methods, which use the default scaling fac-
tors of 6/5 for antiparallel spins and 1/3 for parallel spins,
showed a dramatic increase in accuracy over normal MP2 ener-
gies and provided a more balanced description of many sys-
tems,[40–43] and sometimes even reached QCISD or QCISD(T)
quality of data.[44] Since the application of SCS-MP2 on the sta-
Herein, the interaction between the phenols and the extrac-
tants is experimentally investigated in detail by isothermal ti-
tration calorimetry (ITC) binding experiments. Although some
information about hydrogen bonding of phenols to amine N-
oxides is already available,[6,13–15] it was obtained indirectly
from IR measurements and from liquid–liquid extraction ex-
periments. The current work presents the measured binding
affinities, full thermodynamic description (stoichiometry, K,
DH0, DG0, DS0) and a preliminary investigation of solvent ef-
fects, complemented by theoretical investigations in vacuo. In
3466
www.chemphyschem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 2010, 11, 3465 – 3473

Complexation of Phenol and Thiophenol by Amine N-Oxides
bility of H-bonding is relatively
Table 2. Thermodynamic parameters derived from the ITC experiments at 308C.
unexplored, its performance is
described in detail here, and its
results from the latter method
are compared to those of bench-
mark MP2/CBS method.
Complex
N
[–]
K
DH0
DG0
TDS0
[mÀ1
]
[kcalmolÀ1
]
[kcalmolÀ1
]
[kcalmolÀ1
]
DMDAO–PhOH[a]
DMDAO–PhOH[b]
TOAO–PhOH[a]
TOAO–PhOH[b]
1.04Æ0.07
0.98Æ0.06
1.14Æ0.12
1.19Æ0.15
(2.8Æ0.5)ꢀ104
(5.8Æ2.3)ꢀ103
(2.3Æ1.1)ꢀ104
(5.8Æ2.6)ꢀ103
À7.3Æ0.4
À7.1Æ1.2
À4.7Æ1.4
À2.9Æ1.0
À6.2Æ0.1
À5.4Æ0.3
À6.0Æ0.3
À5.1Æ0.4
À1.1Æ0.4
À2.4Æ1.4
1.3Æ1.2
2.2Æ0.4
[a] In dry toluene. [b] In water-saturated toluene.
2. Results and
Discussion
2.1. Isothermal Titration Calorimetry
study. The negative entropy factor TDS0 indicates an overall in-
crease in order on binding and enthalpy-driven formation of
hydrogen-bonded complexes.[22]
An example of a calorimetric titration of phenol with dimethyl-
dodecylamine N-oxide (DMDAO) in dry toluene at 308C is
given in Figure 2. The isotherm shows a clear sigmoidal curve,
indicating strong binding. By integrating the binding isotherm
In the calorimetric titration of phenol with DMDAO in water-
saturated toluene, the binding enthalpy (DH0 =À7.1 kcalmolÀ1
)
is marginally less negative than in dry toluene (DH0 =
À7.3 kcalmolÀ1), but the binding affinity is markedly lower (K=
5.8ꢀ103 mÀ1 vs. K=2.8ꢀ104 mÀ1 in dry toluene). This influence
of water on the complex-forming behavior is likely caused by
entropic complex-ordering effects of water molecules, as can
be seen from the different change in entropy. These water
molecules (at 308C [H2O] in water-saturated toluene is
28mm;[7] [amine N-oxide] is typically 10–20 mm) will preferably
be present near the most polar positions in the molecules,
that is, near the N+ÀOÀ moiety in amine N-oxides and near
the OH and SH groups in phenol and thiophenol, respectively.
Multiple water molecules[45] will be replaced on phenol bind-
ing, and this water, which is freed from every amine N-oxide
and phenol molecule, will form larger water clusters in the
apolar solvent and on average become more structured, even
if it remains weakly attached to the amine N-oxide–phenol
complex as a hydrogen-bonded large cluster (Figure 3C). This
will likely cause a more negative entropic effect. Accordingly,
on phenol binding the TDS0 value in the presence of water is
more negative than that in its absence, and the DH0 value is
nearly the same with or without the presence of water.
In the calorimetric titration of phenol with trioctylamine N-
oxide (TOAO, Figure 3B) in dry toluene again 1:1 complexation
is observed, with a binding affinity slightly lower than for
DMDAO (K=2.3ꢀ104 mÀ1 vs. 2.8ꢀ104 mÀ1 for DMDAO). In addi-
tion, DH0 is markedly less negative (DH0 =À4.7 kcalmolÀ1 vs.
À7.3 kcalmolÀ1 for DMDAO). This effect could be explained by
a slight dimerization of the TOAO molecules, much as is the
case for phosphine oxide molecules.[46] By means of B3LYP-
based calculations, it was shown that indeed dimers of amine
N-oxides can be formed, in both head-to-head and head-to-tail
configurations (head-to-head dimers tend to be more stable
than head-to-tail dimers; see Supporting Information, Figure S-
1). The presence of three alkyl tails per molecule yields favora-
ble van der Waals interactions between two TOAO molecules
when they are near one another. A micelle-like dimer of two
TOAO molecules can be formed, and breaking up of the appar-
ent dimer and subsequent binding of a phenol molecule to
both TOAO molecules will lead to an overall lower binding en-
thalpy than was found for the DMDAO–phenol complex, in
Figure 2. Calorimetric titration of phenol with DMDAO in dry toluene at
308C. Top panel: raw data; bottom panel: peak integration data and fitted
curve.
(top panel), the binding curve is obtained (bottom panel). Fit-
ting yields the stoichiometry N, binding enthalpy DH0, and
binding affinity K (see Table 2). The binding of DMDAO with
phenol is exothermic, with DH0 =À7.3 kcalmolÀ1 and K=2.8ꢀ
104 mÀ1. The molar ratio of DMDAO/phenol of 1.04 at the
equivalence point indicates formation of a 1:1 hydrogen-
bonded complex without other effects that interfere with
direct binding of phenol to DMDAO (Figure 3A). Similar 1:1
stoichiometry was observed for the other complexes under
ChemPhysChem 2010, 11, 3465 – 3473
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org
3467

H. Zuilhof et al.
the mutual interaction of the
amine N-oxides could not be
quantified separately by ITC due
to the very low heat effect. ITC
measurements of the interac-
tions between the extractants
and thiophenol also gave only
very weak heat effects, which
make it impossible to accurately
quantify any of the thermody-
namic parameters.
The ITC titration of phenol
with tris-(2-ethylhexyl)amine N-
oxide (TEHAO) in dry toluene
(see Supporting Information, Fig-
ure S-2) shows a binding iso-
therm which is markedly differ-
ent from that observed for the
other amine N-oxide–phenol in-
teractions. It shows a binding
behavior characterized by a low
exothermic heat effect (DH0 =
0.5–1 kcalmolÀ1), a low binding
affinity (absence of a sigmoidal
curve; Kꢀ300–600mÀ1), and a
Figure 3. Representation of the proposed complex formation in dry toluene (A and B), and water-saturated tolu-
ene (C and D).
which these mutual van der Waals interactions will be weaker
because only one long alkyl tail is present. Moreover, we find
an entropic effect that is positive (TDS0 =+1.3 kcalmolÀ1), indi-
cating an overall higher degree of freedom after phenol bind-
ing, and can be partly attributed to increased disorder in the
alkyl chains.
large scatter, which imply that basically no phenol binding
takes place with TEHAO, likely due to the steric hindrance
posed by the three ethyl chains at the 2-position of the hexyl
tails.
2.2. Modeling
The TOAO–phenol complex in water-saturated toluene (Fig-
ure 3D) shows a markedly lower binding affinity (K=5.8ꢀ
103 mÀ1 vs. 2.3ꢀ104 mÀ1 in water-free toluene) and less negative
DH0 (DH0 =À2.9 kcalmolÀ1 vs À4.7 kcalmolÀ1 in water-free tol-
uene). This effect is much larger
As can be seen from the structures of the complexes (Figure 4
and Supporting Information, Figure S-3), optimized by molecu-
lar modeling, the hydrogen bonds of phenol or thiophenol to
than for the DMDAO–phenol
complex. Due to the apparent
dimer formation of TOAO and
the presence of water near the
N+ÀOÀ moiety, replacement of
water by phenol is more difficult.
In TOAO–phenol complexation
in water-saturated toluene, TDS0
is positive (+2.2 kcalmolÀ1), that
is, more positive than for water-
free toluene, and indicated a less
ordered situation on phenol
Figure 4. Optimized B3LYP/6-311G(d,p) structures of some selected amine N-oxide–phenol complexes a) DMEAO–
phenol, b) DMEAO–thiophenol, and c) PNO–phenol.
binding. In line with Figure 3B dimer formation of TOAO yields
rather structured H2O clusters. If the dimer is broken by addi-
tion of phenol, then TOAO–phenol–water complexes are
formed, in which the water is bound with more conformational
freedom, which implies DS greater than 0. Hydrophobic inter-
actions could further play a stabilizing role here.[47]
the N+ÀOÀ moiety are all formed approximately along the OÀ
H/SÀH bond, and thus nearly linear hydrogen-bond geometries
are formed. In the dimethylethylamine N-oxide (DMEAO)–
phenol complex (a model complex for the experimentally used
dimethyldodecylamine N-oxide complexes, see complex C1 in
Figure 4a) the hydrogen-bond length is short (d=1.62 ꢂ, see
Supporting Information, Table S-1) and the bond angle is
almost linear (a=1698). In addition, the bond lengthenings of
the hydrogen-bond donor and acceptor are quite substantial
The aggregation number of water will depend on the con-
centration, the chemical environment, and the temperature.
Unfortunately, the interaction of water with amine N-oxides or
3468
www.chemphyschem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 2010, 11, 3465 – 3473

Complexation of Phenol and Thiophenol by Amine N-Oxides
Table 3. Enthalpy of complexation DH [kcalmolÀ1] of the amine N-oxide complexes with phenol and thiophenol (B3LYP/6-311G(d,p)-optimized geometries)
in vacuo.
No.
Complex
DH
B3LYP
MP2
no cp
SCS-MP2
MP2/CBS
M06-2X[e]
no cp
CBS-Q
small[b]
no cp cp
large[c]
small[b]
large[c]
no cp[d]
cp
no cp
cp
no cp
cp
C1
C2
C3
C4
C5
C6
(C2H5)(CH3)2N+ÀOÀ–HOPh
(C2H5)(CH3)2N+ÀOÀ–HSPh
C5H5N+ÀOÀ–HOPh
À14.7
À8.2
À9.8
À5.1
À6.4
À3.3
À18.9
À12.7
À12.1
À9.3
À15.1
À9.1
À9.0
À5.8
À7.5
À4.7
À19.0
À12.6
À12.4
À9.0
À16.3
À10.2
À10.2
À6.8
À17.4
À11.1
À11.2
À8.1
À13.8
À7.6
À8.2
À4.8
À6.6
À3.9
À17.3
À10.8
À11.4
À7.7
À14.6
À8.4
À9.1
À5.6
À7.4
À4.4
À18.4
À12.4
À11.8
À8.8
À19.3
À13.4
À13.5
À9.6
À13.7
À13.6
À10.7
À12.1
À9.5
C5H5N+ÀOÀ–HSPh
(CF3)3N+ÀOÀÀHOPh
À12.0
À8.8
À11.8
À8.2
À8.5
À10.9
À7.7
À10.8
À7.2
À11.4
À8.0
À10.5
À7.6
(CF3)3N+ÀOÀ–HSPh
À5.4
À11.8
[a] B3LYP/6-311+G(d,p). [b] 6-311+G(d,p). [c] 6-311++G(2d,2p). [d] cp=a posteriori counterpoise correction. [e] M06-2X/6-311+G(d,p)//M06-2X/6-
311G(d,p).
(Dr1 =0.022, Dr2 =0.043 ꢂ) and the dihedral angle of phenol is
close to zero (f=À1.28), and these parameters are all indica-
tive of strong hydrogen bonds.[7]
the dominant electrostatic interaction, some covalent character
is also present, as can be deduced from the atom–atom over-
lap-weighted NAO bond orders (see Supporting Information,
Table S-3). For example, the bond order between the DMEAO
oxygen and the phenol hydrogen atoms amounts to 0.13 in
the optimized geometry. The OÀH bond order is 0.57, that is,
0.09 lower than in uncomplexed phenol. These data point to
strong complex formation, even stronger than was seen before
in phosphine oxide–phenol hydrogen bonds, in which a maxi-
mum hydrogen-bond order of 0.08 and a decrease in OÀH
bond order of only 0.06 were found.[7,48]
The enthalpies of complexation for some selected amine N-
oxide complexes are summarized in Table 3 (for DH of all in-
vestigated amine N-oxides, see Table S-2 in the Supporting In-
formation). The enthalpy of complex formation DH (in vacuo)
of complex C1 is found to be À18.4 kcalmolÀ1 at the elaborate
MP2/CBS level, which is considered to be accurate. The cp-cor-
rected MP2/6-311++G(2d,2p) method provides a reasonable
estimate (DH=À16.3 kcalmolÀ1), whereas B3LYP (in vacuo)
shows only qualitative agreement (DH=À14.7 kcalmolÀ1).
The cp-uncorrected MP2 values are all more negative than
the MP2/CBS values due to the basis set superposition error
(BSSE). The SCS-MP2 values are all less negative than the MP2/
CBS values, but typically agree with the MP2/CBS value to
within 1 kcalmolÀ1, an observation also made for phosphine
oxide–phenol and phosphate–phenol complexes and their thio
analogues.[48] With counterpoise correction, the agreement of
both methods with the MP2/CBS method becomes worse. In-
terestingly, all cp-corrected MP2 values deviate more from
MP2/CBS values than the cp-uncorrected MP2 values, and this
suggests that counterpoise correction overcorrects the BSSE.
The binding affinities and geometric and electronic parameters
of the complexes of phenol and thiophenol with the two other
alkylamine N-oxides [trimethylamine N-oxide (TMAO) and trie-
thylamine N-oxide (TEAO)], that were investigated for the
direct comparison to experiment using amine N-oxides with
longer alkyl tails were found to be almost equal to those
found for DMEAO complexes, and are therefore only given in
the Supporting Information (Tables S-1 and S-2). The binding
enthalpies DH are found to be a little more negative (all within
ca. 1 kcalmolÀ1) for the more highly substituted complexes,
which have a slightly higher electron-donor capacity.
Thiophenol complexes differ in several aspects from phenol
complexes, as is also seen in the ITC measurements. Binding
enthalpies of thiophenol complexes are much less negative
than those of phenol complexes. For thiophenol complexes,
DH is on average about 6 kcalmolÀ1 (MP2/CBS) less negative
than of phenol complexes; hydrogen-bond lengths d are
found to be longer for thiophenol complexes than for phenol
complexes (d=1.8 vs. d=1.6 ꢂ for phenol complexes; see Sup-
porting Information, Table S-2), and donor and acceptor bond
length increases Dr1 and Dr2 are both markedly smaller than
for phenol complexes. Also, the dihedral angles f of the thio-
phenols with respect to the benzene ring were found to devi-
ate from the optimal angle of 08, whereas f of the phenols
was always found to be close to 08. This can be understood
from the energy barriers for rotation around the CÀO/CÀS
bond. For phenol, the energy barriers EOÀH at 90 and 08 with re-
spect to the plane of the benzene ring are calculated to be 3.5
and 3.0 kcalmolÀ1 (B3LYP/6-311+G(d,p) and MP2/6-311+
G(d,p), respectively). For thiophenol, the corresponding calcu-
lated values are 0.7 and 0.4 kcalmolÀ1, respectively, that is, ro-
tation in a thiophenol complex is nearly free. Therefore, large
dihedral angles are more often observed if this is otherwise at-
tractive in the complex, which apparently is the case here. The
charge on the thiophenol hydrogen atom of +0.2 is signifi-
cantly lower than of the phenol hydrogen atom (+0.5). For thi-
ophenol, the bond order of the hydrogen bond is lower than
for the phenol complexes (0.10 for DMEAO–thiophenol vs.
0.13 for DMEAO–phenol, see below). Clearly, all these data
agree with the much lower binding energies found for the thi-
ophenol complexes.
The B3LYP-derived natural population analysis (NPA) charges
on the oxygen atoms of the amine N-oxides and the hydrogen
atom of the phenol OH group were found to be À0.7 and
+0.5, respectively, and these values did not change significant-
ly from one method (B3LYP, M06-2X, etc.) to the others. These
charges indicate a strong electrostatic tendency to form hydro-
gen bonds (Supporting Information, Table S-1).[7,17,48] Besides
ChemPhysChem 2010, 11, 3465 – 3473
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org
3469

H. Zuilhof et al.
The theoretical findings show that a strong 1:1 hydrogen-
bonded complex is formed between amine N-oxides and
phenol, in line with the ITC results (high K, highly negative DH0
and DG0 values, and N=1) obtained in toluene as solvent. To
study the influence of the solvent in slightly more detail, we
also performed preliminary investigations of the solvent effects
on complex formation, using mesitylene, toluene, and benzene
as model solvents for the apolar phase in SIRs (Table 4).
tion of the resulting CBS-Q geometries with MP2/6-311++
G(2d,2p) yielded energies within 1 kcalmolÀ1 of what was ob-
tained on the B3LYP-optimized geometries. Therefore, option 2
(incorrect energy calculation on correct geometries) is most
likely. Since no systematic errors have been reported for com-
putations of oxygen-containing hydrogen bonds using models
like CBS-Q, this indicates that the CBS-Q values for sulfur com-
pounds contain a systematic error, as was also found previous-
ly for hydrogen bonds between phenol/thiophenol and anoth-
er class of compounds.[48] A closer investigation of the energies
obtained by CBS-Q revealed that the first steps (up to the
MP4SDQ contributions) were in accordance with expected
binding energies, while the erroneous output of the CBS-Q
method is in fact produced in the last modeling step (CBS ex-
trapolation within the CBS-Q method), which may be related
to the size of the basis sets from which the extrapolation to in-
finity is performed. As a result, and to our surprise, CBS-Q is
not useful as a benchmark method in this case.
Table 4. Thermodynamic parameters [kcalmolÀ1
]
for DMDAO–phenol
Benzene
complexes in various solvents, obtained by ITC.[a]
Parameter
Mesitylene
Toluene
N [–]
0.85Æ0.03
(5.7Æ1.8)ꢀ104
À5.6Æ0.5
À1.0Æ0.6
À6.6Æ0.2
1.04Æ0.07
0.87Æ0.02
(1.7Æ0.2)ꢀ104
À7.8Æ0.3
2.0Æ0.3
K [mÀ1
DH0
]
(2.8Æ0.5)ꢀ104
À7.3Æ0.4
1.1Æ0.4
TDS0
DG0
À6.2Æ0.1
À5.9Æ0.1
[a] All solvents were dried before use.
The effects of electron-withdrawing groups (EWG) on the
complex-forming ability of amine N-oxides were investigated
in order to possibly fine-tune the complexation behavior in
future applications. Both pyridine N-oxide (PNO; complexes C3
and C4) and tri(trifluoromethyl)amine N-oxide (TTFMAO; com-
plexes C5 and C6) show lower phenol binding enthalpies than
the alkylamine N-oxides for all levels of theory (Table 3). In ac-
cordance with this, the hydrogen-bond lengths are found to
be longer than for the alkylamine N-oxide complexes (Support-
ing Information, Table S-1). This can be explained by a reduc-
tion of the negative charge on the amine N-oxide oxygen
atoms in both PNO and TTFMAO (À0.6 vs. À0.7 in alkylamine
N-oxide), and by the much lower H···O bond order (0.05 for
TTFMAO and 0.07 for PNO vs. 0.13 for the alkylamine N-oxide–
phenol complexes; for more extensive data sets, see Support-
ing Information, Tables S-1 to S-3).
The ITC experiments for the DMDAO–phenol complex yield
a value of DH0 =À7.8 kcalmolÀ1 in benzene, and slightly less
negative values in toluene and mesitylene (À7.3 and
À5.6 kcalmolÀ1, respectively; Table 4). This trend is opposite to
that followed by the measured K values: in benzene, toluene,
and mesitylene K values of 1.7ꢀ104, 2.8ꢀ104, and 5.7ꢀ104 mÀ1
were observed, respectively. This disparity with increasing
methyl substitution of the aromatic solvent points to signifi-
cant entropic effects, in line with the above discussion. While a
full discussion based on additional experiments and computa-
tions involving solvents fall outside the scope of this study and
will be published elsewhere, a factor of importance is likely the
p–p interaction between phenol and the solvent.
An interesting linear correlation can be observed when the
binding enthalpy ÀDH (in vacuo, MP2/CBS) of the ten investi-
gated complexes (TMAO, EDMAO, TEAO, PNO, and TTFMAO
with both phenol and thiophenol) is plotted against the NÀO
bond lengthening Dr1 (Figure 5a). The linear correlation pa-
rameters are indicated in the plot and show good correlation
for both the alkyl-substituted and the EWG-substituted amine
N-oxides. A much higher slope was found for alkylamine N-
oxides than for EWG-amine N-oxide molecules, and it indicates
a much more rigid NÀO bond for the former amine N-oxides.
A plot of the binding enthalpy ÀDH of these ten complexes
against the OÀH/SÀH bond lengthening in the phenol or thio-
phenol Dr2 (Figure 5b) is also linear, and confirms that, while
complexes are much weaker for thiophenol than for phenol,
SÀH bond lengthening is more easily accomplished than OÀH
bond lengthening.
The CBS-Q results, initially added for the sake of having
highly accurate values for comparison of the other methods
used here, are found to be in disagreement with the results
obtained by these other calculation methods. In the CBS-Q cal-
culations some of the amine N-oxide–thiophenol complexes
show nearly equal or more negative binding enthalpies than
the corresponding amine N-oxide–phenol complexes (cf.
TEAO–phenol and TEAO–thiophenol with DH values of À15.7
and À16.5 kcalmolÀ1, respectively; see also Supporting Infor-
mation, Table S-2). However, experimental data from this and
previous work[7] clearly show that phenol uniformly yields
stronger hydrogen bonds than thiophenol (binding constant K
is higher for phenol by typically 1–2 orders of magnitude).
These experimental observations are in line with generally ob-
served differences in binding enthalpies for O···HÀO and
O···HÀS hydrogen bonds.[49] This erroneous prediction by CBS-
Q can have two possible causes: 1) use of correct energy calcu-
lations on incorrect geometries, or 2) use of incorrect energy
calculations on correct geometries. Close inspection of the
MP2(FC)/6-31G(d’) geometries resulting from the CBS calcula-
tions showed only minor differences to the geometries found
in the B3LYP optimization, and single-point energy recalcula-
The enthalpy of complexation ÀDH of the ten complexes in
vacuo is plotted against the hydrogen-bond length d(H···O) in
Figure 6. The hydrogen-bond enthalpy is calculated to be ap-
proximately proportional to the inverse fifth power of the hy-
drogen-bond length: DH~d(H···O)À5.04. In recent papers[50–52]
a
calculated hydrogen-bond enthalpy proportional to the inverse
fifth power of the hydrogen-bond length was also found, in
3470
www.chemphyschem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 2010, 11, 3465 – 3473

Complexation of Phenol and Thiophenol by Amine N-Oxides
quate regeneration, can be re-
duced by amine N-oxides carry-
ing
electron-withdrawing
groups, which may be beneficial
for large-scale SIR extraction. The
complexes between amine N-
oxides and thiophenol are calcu-
lated to be much weaker, and
experimentally the heat effect of
the complexation was shown to
be too weak to be observed by
ITC. The weaker hydrogen bonds
found for thiophenols are largely
due to lower charges on the hy-
drogen-bond donor atom, which
Figure 5. Dependence of the binding enthalpy ÀDH on NÀO bond lengthening Dr1 for alkyl-substituted and
EWG-substituted amine N-oxides (a) and on OÀH/SÀH bond lengthening Dr2 for complexes with phenol and with
thiophenol (b). Depicted data were obtained with MP2(CBS)//B3LYP/6-311G(d,p) for all ten amine N-oxide–(thio)-
2
&
&
phenol complexes. a) : Alkylamine N-oxide–(thio)phenol (DH=1140Dr1À6.35; R =0.92); : EWG-amine N-oxide–
2
2
*
*
(thio)phenol DH=458Dr1 +1.87; R =0.86). b) : Amine N-oxide–phenol (DH=251Dr2 +7.65; R =0.95); : Amine
N-oxide-thiophenol (DH=166Dr2 +7.14; R2 =0.85).
yield smaller electrostatic interactions. CBS-Q modeling consis-
tently predicts an equally strong or even a stronger hydrogen
bond for thiophenols than for phenols, in contrast with both
experimental and the other modeling results. As a result, CBS-
Q cannot be used as a benchmark for complexes of sulfur
compounds. The results from relatively cheap SCS-MP2 and
M06-2X calculations without counterpoise correction agree
well with those of the MP2/CBS method, and are in all cases
clearly better than counterpoise-corrected calculations.
The choice of an optimal extractant is crucial for future SIR
applications. The combination of ITC and modeling experi-
ments shows that liquid amine N-oxides would be very good
candidates for phenol extraction from aqueous media.
Figure 6. Dependence of ÀDH on hydrogen-bond length d for amine N-
oxide–(thio)phenol-amine hydrogen-bonded complexes.
Experimental Section
line with a previously made claim[52] that such data should
generally fit an inverse power relation ÀDHꢀf(rÀa) with a>3.
Recently, we discussed the phenol binding capacity of a
number of phosphine oxides and phosphates, and concluded
that for phenol recovery from an aqueous environment the
liquid phosphine oxide blend Cyanex 923 would be an opti-
mum extractant for SIR applications.[7,48] This phosphine oxide
blend combined a high binding affinity for phenols with indus-
trially relevant properties such as low melting point (eliminates
the need for a solvent) and high boiling point (beneficial for
regeneration of the SIR). The present study shows that amine
N-oxides are a second highly interesting class of electronically
and sterically tunable materials for extracting phenols on a
large scale in SIR-based processes.
General: The following materials were used as received without
further purification: toluene (Riedel de Haan, puriss.), trioctylamine
(Aldrich, 98%), hydrogen peroxide (Chemlab, 35% w/w), acetic
acid (Fisher, p.a.), phenol (Merck, p.a.), thiophenol (Janssen Chimica,
97%), dimethyldodecylamine N-oxide (Fluka, 98%), tris-2-ethylhex-
ylamine (Fluka, 97%), ethyl acetate (Baker, >99.5%), methanol
(Fisher, p.a.), mesitylene (Aldrich, p.a.), cyclohexane (Acros, >99%),
benzene (Fluka, ꢁ 99.5%). Molecular sieves (3 ꢂ, Aldrich) were
used as received.
Synthesis of Trioctylamine N-Oxide (TOAO): Trioctylamine N-oxide
was synthesized by N-oxidation of trioctylamine according to a
modification of the methods of Davis and Hetzer[53] and Ejaz.[54] Hy-
drogen peroxide (35%, 8.46 g, 87.0 mmol) was added to a solution
of trioctylamine (8.2 g, 23.1 mmol) in acetic acid (50 mL). The mix-
ture was stirred at 258C for approximately 120 h. Ice-cold water
(100 mL) was then added, followed by the addition of chilled
NaOH solution until near neutralization of the acetic acid. The
white solid that separated after addition of the NaOH solution was
collected by filtration on a Bꢃchner funnel, washed repeatedly with
ice-cold water and ice-cold hexane, and then recrystallized from
hexane. Yield: 7.54 g (20.4 mmol, 88%). The absence of traces of
trioctylamine in the final product was checked by thin-layer chro-
matography (Al2O3) with EtOAc:MeOH (19:1). TOAO was stored in a
vacuo desiccator over NaOH. 1H NMR (400 MHz, CDCl3): d=0.84–
0.89 (t, 9H), 1.20–1.35 (m, 30H), 1.70–1.80 (m, 6H), 3.04–3.09 ppm
(m, 6H); 13C NMR (100.6 MHz, CDCl3): d=65.1, 31.7, 29.2, 29.0, 26.6,
22.9, 22.6, 14.0 ppm.
3. Conclusions
Isothermal titration calorimetry combined with ab initio molec-
ular modeling indicates enthalpy-driven formation of a strong
hydrogen-bonded 1:1 complexes between phenol and amine
N-oxides, which are highly attractive for solvent-impregnated
resin (SIR)-based large-scale separations. These complexes have
a high binding affinity (K=(1.7–5.7)ꢀ104 mÀ1) in dry benzene,
toluene, and mesitylene, which is reduced significantly in
water-saturated solvents (K=5.8ꢀ103 mÀ1 in water-saturated
toluene). The binding affinity, if found to be too strong for ade-
ChemPhysChem 2010, 11, 3465 – 3473
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org
3471

H. Zuilhof et al.
Synthesis of Tris(2-ethylhexyl)amine N-Oxide (TEHAO): A similar
procedure was used as for the synthesis of TOAO, but column
chromatography (neutral silica column, EtOAc:MeOH 39:1) was
used rather than recrystallization. Yield: >80%; liquid. The absence
of traces of tris-(2-ethylhexyl)amine in the final product was
checked by thin-layer chromatography (Al2O3) with EtOAc:MeOH
Although analysis of a small data set of hydrogen-bonded com-
plexes suggests that other density functionals might perform
somewhat better in describing hydrogen-bonding interactions,[36]
no detailed analysis has been presented in the literature that in-
cludes a series of amine N-oxides, and therefore the M06 function-
al was also used along with the widely used B3LYP functional to
compare the results.
1
(19:1). H NMR (400 MHz, CDCl3): d=0.80–0.95 (m, 18H), 1.15–1.40
(m, 24H), 1.40–1.60 (m, 3H), 2.49 ppm (d, 6H); 13C NMR
(100.6 MHz, CDCl3): d=52.9, 38.5, 31.3, 28.9, 24,1, 23.3, 14.1,
10.2 ppm.
Finally, since electrostatic interactions are crucial in determining
the strength of hydrogen bonds,[17,65–68] natural population analysis
(NPA)[69,70] was performed as part of the calculations. As this analy-
sis exhibits a relatively small method and basis-set dependence, it
has been proven to be a reliable charge indicator for related theo-
retical studies of hydrogen bonds. In addition, atom–atom overlap-
weighted natural atomic orbital (NAO) bond orders were obtained.
Isothermal Titration Calorimetry (ITC): ITC was performed by using
a MicroCal VP-ITC microcalorimeter, and the setup was described
elsewhere.[55] The results were analyzed with Origin 7SR2V7.0383
(B383) software (Origin Lab Corporation). The measuring cell
(1.46 mL) was flushed and filled with a freshly prepared and de-
gassed solution of phenol or thiophenol in mesitylene, toluene, or
benzene. The 280 mL automatic syringe was flushed and filled with
a freshly prepared and degassed solution of the complexing agent
(amine N-oxide) in the same solvent at a concentration of roughly
10 times the molarity of the cell solution. The number of injections
and the injected volume per injection were optimized depending
on the concentration of the solutions and the magnitude of the
heat effect per second. Between subsequent injections a sufficient
amount of time (usually 180 to 300 seconds) was chosen to allow
for the cell feedback to return to the baseline. The stirring speed
for all ITC measurements was 502 rpm (default value). A reference
measurement was performed by using the same procedure with
only the solvent, not the solute, in the measuring cell. By subtract-
ing the data of the reference measurement from those of the
actual measurement the net results were obtained, and they were
fitted by using the one-set-of-sites model implemented in the
Origin software.
The model compounds (Table 1) resemble the compounds used in
ITC [dimethyldodecylamine N-oxide (DMDAO), trioctylamine N-
oxide (TOAO) and tris-(2-ethylhexyl)amine N-oxide (TEHAO)]. How-
ever, the long alkyl tails were replaced by methyl or ethyl groups
in order to shorten the computational time, as the primary interac-
tion between the complex fragments was not affected by doing
so. To get some insight into the influence of electron-withdrawing
groups on the complex formation of amine N-oxides with phenol
and thiophenol, the complexation behavior was also investigated
for tris(trifluoromethyl)amine N-oxide and pyridine N-oxide. The rel-
evant geometrical parameters for hydrogen bonds in the investi-
gated complexes are defined in Table 1.
Acknowledgements
The authors thank the Dutch Technology Foundation STW for fi-
nancial support (project no. 06347). Andrꢀ ten Bçhmer is kindly
acknowledged for technical assistance and Prof. Andrꢀ de Haan
(Technical University Eindhoven) for fruitful discussions.
Molecular Modeling: Computational data of the compounds de-
picted in Table 1 were obtained with Gaussian 03, RevisionC.2,[56]
and Gaussian 09, Revision A.1.[57] The geometries of the monomers
and complexes were optimized at the B3LYP/6-311G(d,p) and
M062X/6-311G(d,p) levels of theory, which generally describe the
geometries of these systems well.[18,20,37,38,58] All B3LYP/6-311G(d,p)-
optimized stationary points were confirmed as true minima
through vibrational frequency calculations. The B3LYP/6-311+
G(d,p) method and the MP2 and SCS-MP2 methods both with the
small [6-311+G(d,p)] basis set were chosen to ensure that the con-
clusions are not influenced by significant method-dependent phe-
nomena. With such basis sets both B3LYP and MP2 typically pro-
vide accurate (within 2 kcalmolÀ1) interaction energies for hydro-
gen bonding, while MP2 calculations with larger basis sets are typi-
cally even more accurate.[58–61] An extrapolation to the complete
basis set (CBS) energy limit was obtained by using the keyword
“cbsextrapolate”.[62] The option “NMin=10” was specified to ac-
count for a minimum of 10 pair natural orbitals for extrapolation
to the CBS limit. Counterpoise (cp) correction[63,64] was applied for
all MP2 and SCS-MP2 calculations to correct for over-stabilization
due to the supermolecule approach. Given the large size of an ex-
trapolation to the complete basis set, no cp correction was applied
to the MP2/CBS energies. The B3LYP-derived geometries were
taken as input structures for the CBS-Q method, which is known to
provide very accurate energy data. The CBS-Q method consists of
several different steps, described elsewhere.[39] All CBS-Q structures
were confirmed to be true minima in the vibrational frequency
analysis that is part of the composite method. The CBS-Q energy
at 0 K was used for the presented results. Total energies are given
in Table S-4 of the Supporting Information.
Keywords: H bonding · isothermal titration calorimetry ·
phenol · quantum chemistry · solvent-impregnated resin
[1] J. L. Cortina, A. Warshawsky in Ion Exchange Solvent Extraction, Vol. 13
(Eds.: J. A. Marinsky, Y. Marcus), Marcel Dekker, New York, 1997, p. 195.
[2] S. S. El-Dessouky, E. H. Borai, J. Radioanal. Nucl. Chem. 2006, 268, 247.
[3] B. B. Gupta, Z. B. Ismail, Comp. Interf. 2006, 13, 487.
[4] J. S. Liu, H. Chen, Z. L. Guo, Y. C. Hu, J. Appl. Polym. Sci. 2006, 100, 253.
[5] K. Ohnaka, A. Yuchi, Chem. Lett. 2005, 34, 868.
[6] B. Burghoff, R. Cuypers, M. van Ettinger, E. J. R. Sudhçlter, H. Zuilhof,
A. B. de Haan, Sep. Purif. Technol. 2009, 67, 117.
[7] R. Cuypers, B. Burghoff, A. T. M. Marcelis, E. J. R. Sudhçlter, A. B. de Haan,
H. Zuilhof, J. Phys. Chem. A 2008, 112, 11714.
[8] R.-S. Juang, H.-L. Chang, Ind. Eng. Chem. Res. 1995, 34, 1294.
[9] M. Traving, H.-J. Bart, Chem. Eng. Technol. 2002, 25, 997.
[10] A. Kostova, H. J. Bart, Chem. Ing. Tech. 2004, 76, 1743.
[11] H. Kitazaki, M. Ishimaru, K. Inone, K. Yoshida, S. Nakamura in Proceed-
ings of the International Solvent Extraction Conference, ISEC’96 ed.,
Melbourne, Australia, 1996, p. 1667.
[12] B. Burghoff, E. L. V. Goetheer, A. B. de Haan, React. Funct. Polym. 2008,
68, 1314.
[13] M. K. Drozdova, I. V. Nikolaeva, V. G. Torgov, Zh. Fiz. Khim. 1997, 71, 573.
[14] Y. I. Korenman, S. I. Niftaliev, V. G. Torgov, M. K. Drozdova, Russ. J. Appl.
Chem. 1993, 66, 141.
[15] V. G. Torgov, Y. I. Korenman, M. K. Drozdova, V. V. Rebristaya, R. P. Lisit-
skaya, Russ. J. Appl. Chem. 1993, 66, 182.
[16] G. R. Desiraju, T. Steiner, The Weak Hydrogen Bond, 1st ed., Oxford Uni-
versity Press, New York, 1999.
3472
www.chemphyschem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 2010, 11, 3465 – 3473

Complexation of Phenol and Thiophenol by Amine N-Oxides
[17] G. A. Jeffrey, An Introduction to Hydrogen Bonding, Oxford University
Press, New York, 1997.
[54] M. Ejaz, Anal. Chim. Acta 1974, 71, 383.
[55] M. Chellani, Am. Biotechnol. Lab. 1999, 17, 14.
[18] J. P. M. Lommerse, S. L. Price, R. Taylor, J. Comput. Chem. 1997, 18, 757.
[19] S. K. Panigrahi, G. R. Desiraju, Proteins Struct. Funct. Bioinf. 2007, 67, 128.
[20] G. B. W. L. Ligthart, D. Guo, A. L. Spek, H. Kooijman, H. Zuilhof, R. P. Sij-
besma, J. Org. Chem. 2008, 73, 111.
[56] Gaussian 03 (Revision C.2), M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven,
K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone,
B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsu-
ji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Na-
kajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P.
Hratchian, J. B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Strat-
mann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y.
Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakr-
zewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G.
Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P.
Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y.
Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wallingford,
CT, 2004.
[57] Gaussian 09 (Revision A.1), M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
gar, J. Tomasi, M. Cossi, N. J. Rega Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. E. Gomperts, O. Stratmann,
A. J. Yazyev, R. Austin, C. Cammi, J. W. Pomelli, R. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
slowski, D. J. Fox, A.1 ed., Gaussian, Inc., Wallingford, CT, 2009.
[58] S. Raub, C. M. Marian, J. Comput. Chem. 2007, 28, 1503.
[21] J. E. Ladbury, Biotechniques 2004, 37, 885.
[22] J. E. Ladbury, M. L. Doyle, Biocalorimetry 2, Wiley, New York, 2004.
[23] S. Leavitt, E. Freire, Curr. Opin. Struct. Biol. 2001, 11, 560.
[24] P. C. Weber, F. R. Salemme, Curr. Opin. Struct. Biol. 2003, 13, 115.
[25] T. Wiseman, S. Williston, J. F. Brandts, L.-N. Lin, Anal. Biochem. 1989, 179,
131.
[26] T. J. M. de Bruin, A. T. M. Marcelis, H. Zuilhof, L. M. Rodenburg, H. A. G.
Niederlander, A. Koudijs, P. E. M. Overdevest, A. Van Der Padt, E. J. R.
Sudholter, Chirality 2000, 12, 627.
[27] T. J. M. de Bruin, A. T. M. Marcelis, H. Zuilhof, E. J. R. Sudhçlter, Langmuir
2000, 16, 8270.
[28] H. J. Wiggers, J. Cheleski, A. Zottis, G. Oliva, A. D. Andricopulo, C. A.
Montanari, Anal. Biochem. 2007, 370, 107.
[29] A. Arnaud, L. Bouteiller, Langmuir 2004, 20, 6858.
[30] P. Ballester, M. Capo, A. Costa, P. M. Deya, R. Gomila, A. Decken, G. De-
slongchamps, J. Org. Chem. 2002, 67, 8832.
[31] D. Merino-Garcia, S. I. Andersen, Pet. Sci. Technol. 2003, 21, 507.
[32] D. Merino-Garcia, S. I. Andersen, Langmuir 2004, 20, 4559.
[33] J. L. Sessler, D. E. Gross, W. S. Cho, V. M. Lynch, F. P. Schmidtchen, G. W.
Bates, M. E. Light, P. A. Gale, J. Am. Chem. Soc. 2006, 128, 12281.
[34] T. Mizutani, H. Takagi, Y. Ueno, T. Horiguchi, K. Yamamura, H. Ogoshi, J.
Phys. Org. Chem. 1998, 11, 737.
[35] M. Xiang, M. Jiang, Y. Zhang, C. Wu, L. Feng, Macromolecules 1997, 30,
2313.
[36] L. F. Molnar, X. He, B. Wang, K. M. Merz, J. Chem. Phys. 2009, 131.
[37] Y. Zhao, D. G. Truhlar, Acc. Chem. Res. 2008, 41, 157.
[38] Y. Zhao, D. G. Truhlar, Theor. Chem. Acc. 2008, 120, 215.
[39] J. W. Ochterski, G. A. Petersson, J. A. Montgomery, J. Chem. Phys. 1996,
104, 2598.
[59] C. J. Cramer, Essentials Of Computational Chemistry: Theories And
Models, 2nd ed., Wiley, New York, 2004.
ˇ
ˇ
ˇ
[40] J. Antony, S. Grimme, J. Phys. Chem. A 2007, 111, 4862.
[41] R. A. Bachorz, F. A. Bischoff, S. Hofener, W. Klopper, P. Ottiger, R. Leist,
J. A. Frey, S. Leutwyler, Phys. Chem. Chem. Phys. 2008, 10, 2758.
[42] S. Grimme, J. Chem. Phys. 2003, 118, 9095.
[60] P. Jurecka, J. Sponer, J. Cerny, P. Hobza, Phys. Chem. Chem. Phys. 2006,
8, 1985.
[61] M. Kone, B. Illien, J. Graton, C. Laurence, J. Phys. Chem. A 2005, 109,
11907.
[43] T. Schwabe, S. Grimme, Acc. Chem. Res. 2008, 41, 569.
[44] M. Piacenza, S. Grimme, ChemPhysChem 2005, 6, 1554.
[45] V. Kocherbitov, V. Veryazov, O. Soderman, J. Mol. Struct. (THEOCHEM)
2007, 808, 111.
[62] M. R. Nyden, G. A. Petersson, J. Chem. Phys. 1981, 75, 1843.
[63] S. F. Boys, F. Bernardi, Mol. Phys. 1970, 19, 553.
[64] F. B. van Duijneveldt, J. G. C. M. van Duijneveldt-van de Rijdt, J. H. van
Lenthe, Chem. Rev. 1994, 94, 1873.
[46] T. Malaspina, L. T. Costa, E. E. Fileti, Int. J. Quantum Chem. 2009, 109,
250.
[47] A detailed analysis of these further factors falls outside the scope of
this paper.
[65] G. Alagona, Int. J. Quantum Chem. 1987, 32, 227.
[66] S. Cybulski, S. Scheiner, J. Phys. Chem. 1990, 94, 6106.
[67] D. Hadzˇi, Theoretical Treatments of Hydrogen Bonding, Wiley, New York,
1997.
[48] R. Cuypers, E. J. R. Sudhçlter, H. Zuilhof, ChemPhysChem 2010, 11, 2230.
[49] F. Wennmohs, V. Staemmler, M. Schindler, J. Chem. Phys. 2003, 119,
3208.
[68] H. Umeyama, K. Morokuma, J. Am. Chem. Soc. 1977, 99, 1316.
[69] E. D. Glendening, A. E. Reed, J. E. Carpenter, F. Weinhold., NBO Version
3.1.
[50] J. Chen, M. A. McAllister, J. K. Lee, K. N. Houk, J. Org. Chem. 1998, 63,
4611.
[70] A. E. Reed, R. B. Weinstock, F. Weinhold, J. Chem. Phys. 1985, 83, 735.
[51] E. Espinosa, E. Molins, C. Lecomte, Chem. Phys. Lett. 1998, 285, 170.
[52] M. Rozenberg, A. Loewenschuss, Y. Marcus, Phys. Chem. Chem. Phys.
2000, 2, 2699.
Received: June 16, 2010
Revised: August 26, 2010
[53] M. M. Davis, H. B. Hetzer, J. Am. Chem. Soc. 1954, 76, 4247.
Published online on October 22, 2010
ChemPhysChem 2010, 11, 3465 – 3473
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org
3473