A Comparison of the Solvation Properties of 2-Nitrophenyloctyl Ether, Nitrobenzene, and
n-Octanol as Assessed by Ion Transfer Experiments
Rubin Gulaboski,
†
Alexandra Galland,
‡
Ge´raldine Bouchard,
‡
Karolina Caban,
§
Ansgar Kretschmer,
†
Pierre-Alain Carrupt,
‡
Zbigniew Stojek,
§
Hubert H. Girault,
|
and
Fritz Scholz*
,†
Institut fu¨r Chemie und Biochemie, UniVersita¨t Greifswald, Soldmannstrasse 23,
D-17489 Greifswald, Germany, Institut de Chimie The´rapeutique, UniVersite´ de Lausanne,
CH-1015 Lausanne, Switzerland, Department of Chemistry, UniVersity of Warsaw,
Pasteura 1, PL-02-093 Warszawa, Poland, and Laboratoire d’EÄ lectrochimie Physique et Analytique,
EÄ cole Polytechnique Fe´de´rale de Lausanne, CH-1015 Lausanne, Switzerland
ReceiVed: December 2, 2003; In Final Form: February 6, 2004
The lipophilicity of the anionic forms of drugs and model compounds was assessed by their transfer across
(i) the water-2-nitrophenyloctyl ether (NPOE), (ii) the water-nitrobenzene (NB), and (iii) the water-n-
octanol interfaces by using the three-phase electrode technique. The lipophilicities, expressed in terms of
logarithm of partition coefficients, range for the studied anions from -3.46 to 0.68 (log P
A
-
,aq
QNPOE
) for NPOE,
from -3.81 to 2.62 (log P
A
-
,aq
QNB
) for NB, and from -6.20 to -3.20 (log P
A
-
,aq
Qn-oct
) for n-octanol. Although
NPOE shares with nitrobenzene the aromatic part and with n-octanol the hydrophobic carbon chain, only
very weak correlation was observed between the NPOE-based data with the n-octanol-based data, and the
same is true for the correlation of the NB-based and n-octanol-based data. However, there is a strong and
even linear correlation between the NPOE-based and the NB-based data.
1. Introduction
The processes of ion transfer across the interface of two
adjacent immiscible liquids attract significant attention due to
their wide applicability in different fields such as ion-selective
electrodes, phase-transfer catalysis, biomimetic studies of
membrane function, drug uptake, and so forth.
1-3
The standard
Gibbs energy of transfer (∆G
i,R
Qβ
) is the major parameter that
determines to what extent the ions will be distributed between
two adjacent phases. Its value provides information about the
solvation strength between the dissolved ions and the surround-
ing solvent molecules. ∆G
i,R
Qβ
is directly related to the standard
potential of ion transfer (∆φ
i,R
Qβ
) and standard partition coef-
ficient (P
i
) (see eqs 6 and 7, respectively).
The logarithm of the partition coefficient, log P, measures
the lipophilicity of compounds and is one of the major
parameters included in quantitative structure properties (QSPR)
and quantitative structure activity (QSAR) relationships, playing
a crucial role in drug design.
1,2
Moreover, log P has been used
for predicting various biological properties such as toxicity,
2
transfer through biological membranes, and affinity between
receptors and enzymes.
3
The classical techniques for the determination of log P such
as the shake-flask method or potentiometric titrations are
techniques without external control of the interfacial potential,
providing access only to the apparent values of the partition
coefficient of ions,
4
which strongly depend on the experimental
conditions, such as the phase-volume ratios and the presence,
concentration, and complexation constants of all the ions
dissolved in the biphasic system.
5,6
The introduction of four-
electrode voltammetry at the interface of two immiscible
electrolyte solutions (ITIES) led to significant progress in the
lipophilicity determinations of different organic and inorganic
ions,
7-13
providing access to the standard values of partition
coefficients of ions, which are independent of experimental
conditions. The theoretical studies applied to ITIES
14-17
have
significantly contributed to the understanding of the processes
occurring at ITIES. Moreover, various experimental modifi-
cations
18-21
have enabled to extend the available potential
windows so that the measurement of the lipophilicities even of
some heavy metal ions
22-24
could be performed. The major
restrictions of the four-electrode ITIES techniques are due to
their applicability to few organic solvents, mainly 1,2-di-
chlorethane (DCE)
13
and nitrobenzene (NB),
13
or lately to
2-nitrophenyloctyl ether (NPOE).
22,24-27
The nonpolarizability
of some important organic solvent-water interfaces such as
n-octanol-water is one of the weaknesses of this technique.
The recent introduction of the so-called three-phase electrode
technique
28
has overcome the limitations regarding the non-
polarizability of some organic solvent-water interfaces such
as that of n-octanol,
29,30
menthol,
31
or nitrophenyl-nonyl ether,
32
leading to significant progress in the determination of the
lipophilicity of a variety of inorganic and organic ions.
33-37
Whereas Compton and his group used three-phased electrodes
to study the electrochemistry of electroactive liquids,
32,38-40
Scholz and co-workers
28-31,33-37
developed the technique of
determining the Gibbs energies of ion transfer with the help of
solutions of electroactive compounds in droplets of organic
solvents immobilized on electrode surfaces.
* Corresponding author. Tel: 0049 3834 864 450. Fax: 0049 3834 864
451. E-mail: fscholz@mail.uni-greifswald.de
†
Universita¨t Greifswald.
‡
Universite´ de Lausanne.
§
University of Warsaw.
|
EÄ cole Polytechnique Fe´de´rale de Lausanne.
4565J. Phys. Chem. B 2004, 108, 4565-4572
10.1021/jp037670m CCC: $27.50 © 2004 American Chemical Society
Published on Web 03/12/2004
The traditionally used measure of lipophilicity as a predictor
of solute membrane partitioning is the partition coefficient in
the n-octanol-water system (log P
aq
n-oct
, since the long lipo-
philic chain of n-octanol coupled with its hydrophilic hydroxyl
group, make it an ideal mimic for the major constituents of the
biological membranes (i.e., phospholipides). However, in many
other situations, log P
aq
n-oct
cannot give a good estimate of a
drug’s absorption or permeation and other solvent systems are
needed to yield information which is complementary to log
P
aq
n-oct
data. Particularly the use of the aprotic solvent 1,2-
dichloroethane (DCE) has experimental benefits in structure-
pharmacokinetic relationships.
1
However, the high volatility and
the toxicity of DCE limit its value and call for its replacement
by a more appropriate organic solvent. Given its promising
physicochemical properties and absence of known toxicity,
o-nitrophenyl octyl ether (NPOE) has recently been introduced
in electrochemistry
22,24-27
and in medicinal chemistry.
41,42
The
intermolecular forces expressed in NPOE-water partitioning
were proven similar to the forces encoded in log P
aq
DCE
,
suggesting log P
aq
NPOE
to offer a convenient alternative to log
P
aq
DCE
.
42
In this paper the lipophilicity of 27 anionic drugs and model
compounds (see Figure 1) were studied in the NPOE-water
system and its closely related nitrobenzene (NB)-water system
using the three-phase electrode technique. The standard partition
coefficient values in these two structurally similar systems were
compared with results previously obtained in the n-octanol-
water systems.
30
2. Experimental Section
2.1. Compounds. All the studied compounds were purchased
from Fluka (Buch, Switzerland), except sulfinpyrazone (kindly
donated by Novartis Pharma, Basel, Switzerland). NPOE and
n-octanol were purchased form Acros Organics (Geel, Belgium).
All chemicals were of analytical grade and supplied by Fluka.
2.2. Electrochemical Method. A 0.05 mol L
-1
or 0.1 mol
L
-1
solution of decamethylferrocene (noted dmfc) was prepared
by dissolving the compound in water-saturated NPOE, NB, or
n-octanol. A droplet of this solution (1 µL) was attached to the
surface of a paraffin-impregnated graphite electrode that was
subsequently immersed in a NPOE, NB, or n-octanol-saturated
aqueous solution containing 0.001 to 1.0 mol L
-1
drug dissolved
in 1 mol L
-1
NaOH (0.5 mol L
-1
buffer made of sodium
hydrogen phthalate and HCl was used in the case of maleic
acid). An Ag/AgCl (saturated NaCl solution) reference electrode
was used, while a platinum wire served as an auxiliary electrode.
All experiments were performed at room temperature. Nitrogen
was purged through the solutions for 4 min prior to each
measurement and a nitrogen blanket was maintained thereafter.
Square-wave (SW) voltammograms were recorded using the
electrochemical measuring system AUTOLAB PGSTAT 30
(Eco-Chemie, Utrecht, Netherlands). Typical instrumental pa-
rameters used for the square-wave voltammetry were the
following: SW frequency f ) 10 Hz, SW amplitude E
sw
) 50
mV, scan increment dE ) 1 mV, and starting potential E
s
)
-0.5 V.
3. Principles of the Three-Phase Electrode Approach
The determination of the lipophilicity of ions by using the
three-phase electrode approach is based on the concept of
preserving the electroneutrality of both phases. The term “three-
phase electrode” denotes an electrochemical system where three
different phases, in our case a solid electrode (electron conduc-
tor), an organic liquid, and an aqueous phase are brought into
intimate contact. According to this approach, a neutral electro-
active and lipophilic compound is dissolved in the water-
immiscible organic solvent without any additional supporting
electrolyte. A droplet of that solution is attached to the surface
of the working electrode. This latter, with the attached droplet,
is then immersed into the aqueous electrolyte solution. By
applying a difference of potential between the working and the
reference electrodes in the common three-electrode arrangement,
the electrochemical processes occurring at such modified
electrodes can be studied (see Figure 2).
Due to the absence of supporting electrolyte in the organic
phase, at the beginning of the experiment, the applied potential
Figure 1. Chemical structures of the compounds that have been studied
in their anionic form. 1: Phenol; 2: 2-Nitrophenol; 3: 3-Nitrophenol;
4: 4-Nitrophenol; 5: 2,4-Dinitrophenol; 6: 2,5-Dinitrophenol; 7:
Benzoic acid; 8: 4-Bromobenzoic acid; 9: 4-Chlorobenzoic acid; 10:
3-Chlorobenzoic acid; 11: 4-Iodobenzoic acid. The names of the other
compounds are given in the figure.
4566 J. Phys. Chem. B, Vol. 108, No. 14, 2004 Gulaboski et al.
can only act at that three-phase boundary line, inducing the
electrochemical reaction of dmfc in the organic phase, simul-
taneously coupled with the anion transfer across the organic
solvent-water interface. The overall process occurring at the
three-phase electrode can be described by the following reaction
scheme:
At equilibrium, the potential difference E between the
working electrode and the reference Ag/AgCl electrodes is
defined as follows:
In a first approximation, the activities in the Nernst equation
have been replaced by concentrations. In eq 1, E
dmfc
+
(o)
|dmfc
(o)
Q
is
the standard redox potential of the couple dmfc
+
/dmfc in the
organic phase (vs Ag/AgCl), n is the number of the exchanged
electrones (n ) 1 in this case), F is the Faraday constant (96485
C mol
-1
), R is the gas constant (8.314 J mol
-1
K
-1
), and T the
temperature in Kelvin; c
dmfc
+
(o)
and c
dmfc
(o)
are, respectively, the
concentrations of dmfc
+
and dmfc in the organic phase.
∆φ
Q
°
A
-
,aq
is the standard transfer potential of the anions A
-
from
the aqueous to the organic phase, while c
A
-
(o)
and c
A
-
(aq)
are the
concentrations of the studied anion in, respectively, the organic
and aqueous phases.
The electroneutrality in the organic phase induces eq 2.
Moreover, the mass conservation law in the organic phase leads
to
where c
dmfc
(o)
/
is the initial concentration of dmfc in the organic
phase.
When the applied potential is equal to the standard redox
potential of the couple dmfc
+
/dmfc in organic phase, then
Then, combining eqs 1-4, the formal redox potential (E
f
)of
the couple dmfc
+
/dmfc in the organic phase is obtained:
Equation 5 shows that in these experimental conditions the
formal redox potential of dmfc in the organic phase (E
f
) depends
on the nature and the concentration of the transferable anions
in aqueous phase.
The three-phase junction line is a prerequisite for performing
these experiments. Since no electrolyte is deliberately added to
the organic phase, the applied potential can start the electro-
chemical reaction only at the three-phase junction where all three
phases are in intimate contact. The reason for this is that a
natural partition will introduce a small amount of salt from the
aqueous phase into the very edge region of the droplet. This
way the ohmic resistance will decrease to such an extent that
the applied potential will drive the electron transfer across the
graphite-organic liquid interface and the ion transfer across
the organic liquid-aqueous phase interface. The ohmic drop
will be negligible at that point. The reaction starts at the three-
phase line, and extends further toward the center of the
droplet
43-45
due to the electrochemical generation of ions.
The standard transfer potential (∆φ
Q
°
A
-
,aq
), the standard Gibbs
energy of transfer (∆G
Q
°
A
-
,aq
) and the standard partition coef-
ficient (log P
Q
°
A
-
,aq
) of the studied anions are deduced from the
formal potential using eqs 5, 6, and 7.
where z is the charge number having a negative sign for anions.
Eventually, it should be emphasized that the three-phase
technique with immobilized droplets of organic solvents is
attractive because even very expensive solvents can be studied
as some microliters are sufficient.
4. Results and Discussion
With the three-phase electrode approach, a pre-requirement
in the determination of the standard partition coefficient of
anions by using eq 5 is the estimation of the standard redox
potential of dmfc in the organic solvent used for the droplet
(E
dmfc
+
(o)
|dmfc
(o)
Q
). A possible method is the measurement of the
standard redox potential of dmfc in an electrolytic solution of
the studied organic solvent using the common three-electrode
configuration (i.e., common measurements in nonaqueous
solution in the presence of an internal standard). An alternative
method was previously used for the evaluation of the redox
potentials of dmfc in NB
34
and n-octanol,
29
consisting in
applying the three-phase electrode approach to measure the
formal potential of dmfc in the presence of various transferring
anions with standard transfer potentials known from literature.
The obtained y-intercept of the linear relationship between E
f
and ∆φ
Q
°
A
-
,aq
is related to E
dmfc
+
(o)
|dmfc
(o)
Q
according to eq 5. The
Figure 2. Schematic representation of the processes occurring at the three-phase electrode.
∆G
Q
°
A
-
,aq
)-zF∆φ
Q
°
A
-
,aq
(6)
log P
Q
°
A
-
,aq
)-
∆G
Q
°
A
-
,aq
2.3RT
(7)
dmfc
(o)
+ A
-
(aq)
- 1e
-
/ dmfc
+
(o)
+ A
-
(o)
(I)
E ) E
dmfc
+
(o)
/dmfc
(o)
Q
+ ∆φ
Q
°
A
-
,aq
+
RT
nF
ln
(
c
dmfc
+
(o)
c
dmfc
(o)
)
+
RT
F
ln
(
c
A
-
(o)
c
A
-
(aq)
)
(1)
c
dmfc
+
(o)
) c
A
-
(o)
(2)
c
dmfc
+
(o)
+ c
dmfc
(o)
) c
dmfc
(o)
/
(3)
c
dmfc
+
(o)
) c
dmfc
(o)
(4)
E
f
) E
dmfc
+
(o)
/dmfc
(o)
Q
+ ∆φ
Q
°
A
-
,aq
-
RT
F
ln(c
A
-
(aq)
) +
RT
F
ln
(
c
dmfc
(o)
/
2
)
(5)
Solvation Properites of 2-NPOE, NB, and n-Octanol J. Phys. Chem. B, Vol. 108, No. 14, 2004 4567
transfer processes of picrate, perchlorate, thiocyanate, and
iodide were studied at the water-NPOE interface (see
Figure 3), yielding a value of -0.140 mV (vs Ag/AgCl) for
E
dmfc
+
(o)
|dmfc
(o)
Q
.
The oxidation of dmfc (dissolved either in NPOE or in NB)
studied with the help of a three-phase electrode leads to well-
defined square-wave voltammograms (see Figure 4 A and 4B,
respectively). The values of the peak potentials of dmfc depend
on the nature and the concentrations of the anions present in
the aqueous phase, as predicted by eq 5. In all cases, the slopes
of the linear relationship between E
f
and log c
A
-
(aq)
are close to
the value of -60 mV as expected by eq 5. This feature, taken
together with the stability of the consecutive cyclic voltammo-
grams, served as a strong argument to assume reversibility of
the entire system. The physicochemical data obtained in the
NPOE-water and NB-water systems are compared with the
previously determined values in the n-octanol-water system
in Table 1.
The lipophilicity range of the studied anions was wider for
NPOE (-3.46 < log P
A
-
,aq
QNPOE
< 0.68) and NB (-3.81 < log
P
A
-
,aq
QNB
< 2.62) than for n-octanol (-6.20 < log P
A
-
,aq
QN-oct
<
-3.20), indicating that the differences of solvation energy of
anions between water and NPOE (respectively NB) were larger
than the difference between n-octanol and water. The solubilities
of water at 298 K in NPOE, NB, and n-octanol are given in
Table 2. The high water content of n-octanol suggests that water
molecules dissolved in n-octanol strongly participate in the
solute solvation, inducing a smaller differentiation of the
standard Gibbs energies of transfer of anions across the water-
n-octanol interface.
Figure 5 shows the relationships between the standard
partition coefficients of the studied anions in the NPOE-water
and NB-water (respectively n-octanol-water) systems. We also
included the lipophilicity values taken from literature for 5 small
monoanions: ClO
4
-
(log P
A
-
,aq
QNPOE
)-3.10, log P
A
-
,aq
QNB
)
-1.36), SCN
-
(log P
A
-
,aq
QNPOE
)-4.60, log P
A
-
,aq
QNB
)-2.93, log
P
A
-
,aq
Qn-oct
)-5.00), I
-
(log P
A
-
,aq
QNPOE
)-4.91, log P
A
-
,aq
QNB
)
-3.46, log P
A
-
,aq
Qn-oct
)-5.20), NO
3
-
(log P
A
-
,aq
QNPOE
)-6.58, log
P
A
-
,aq
QNB
)-4.47, log P
A
-
,aq
Qn-oct
)-5.90), and Cl
-
(log P
A
-
,aq
QNPOE
)
-9.22, log P
A
-
,aq
QNB
)-8.03, log P
A
-
,aq
Qn-oct
)-5.80).
20-22,29
A
good correlation was obtained between log P
A
-
,aq
QNPOE
and log
P
A
-
,aq
QNB
(eq 8), with a slope close to unity and a y-intercept of
-0.47. These results are in agreement with those recently
presented by Wilke et al.
22
In these and the following equations, 95% confidence limits
are given in parentheses; n is the number of compounds, r
2
the
Figure 3. Relationship between the formal potential (i.e., the peak potential in square wave voltammetry) of dmfc in NPOE and the standard
transfer potentials of picrate, perchlorate, thiocyanate, and iodine anions across the water-NPOE interface.
Figure 4. SW voltammograms of dmfc recorded by three phase
electrode approach coupled to the transfer of the anions of some of the
studied compounds across the water-NPOE interface (A) and water-
NB interface (B). The currents in Figure 4 are normalized.
log P
A
-
,aq
QNB
) 0.99 ((0.07) log P
A
-
,aq
QNPOE
- 0.47 ((0.27)
n ) 32; r
2
) 0.85; s ) 0.73; F ) 160 (8)
4568 J. Phys. Chem. B, Vol. 108, No. 14, 2004 Gulaboski et al.
squared correlation coefficient, s the standard deviation, and F
the Fischer test.
In general, the Gibbs energy of solvation comprises three
terms:
46
the first one is the Gibbs energy of making a cavity
for accommodating the solute in the solvent, while the second
term arises from the interactions between the solvent and solute
molecules after the solute is accommodated. The third contribu-
tion is the reorganization Gibbs energy of solvent molecules
around the solute molecules.
The first theory that described the ion-solvent interactions
was given from Born
46
at the beginning of the 20th century.
The Born theory describes the ion as a rigid sphere with radius
r (equivalent to the crystallographic radius of the ion) and charge
z
i
. The solvent, which is polarized in the vicinity of an ion, is
represented by a structureless continuum of uniform dielectric
constant
r
, corresponding to its bulk value. Despite its limita-
tions (the Born theory neglects the charge delocalization effects
and the dielectric saturation, assuming that the dielectric constant
around the ion is equal to that in the bulk, which results in an
overestimation of ion-solvent interactions values), the Born
equation provides good estimates of ionic solvation energy
where ∆G
IS
R
is the Gibbs energy of solvation of the ions in the
solvent R, z and r are, respectively, the charge and the radius
of ions, N
A
is the Avogadro constant,
r
R
is the dielectric
constant of phase R, and
0
is the permittivity of vacuum.
Figure 6 shows the relationship between ∆G
A
-
,aq
QNPOE
and the
reciprocal van der Waals radii of the anions for the 27
compounds shown in Figure 1. The van der Waals radii were
calculated with the standard software MOLSV and the atomic
radii given by Gavezzotti.
47
Calculations were performed on a
Silicon Graphics Indy R4400 175 MHz using the Sybyl software
(Tripos Associates, St. Louis, MO).This relationship was linear
for ions with a van der Waals radius less than 3.3 Å (full circles
in Figure 6), suggesting that the Born solvation model is
adequate for describing the solvation of small anions (see eq
10).
This observation suggests that for small anions the electrostatic
contribution to the solvation energy is predominant, but for
bigger anions (see open circles in Figure 6) the other energetic
contributions such as the cavity formation and the repulsive/
dispersive terms can no longer be neglected.
Opposite to eq 8, a poor correlation is observed between log
P
A
-
,aq
QNPOE
and log P
A
-
,aq
Qn-oct
(see the empty squares in Figure 5).
This correlation is slightly improved when phenol, 2-nitro-
phenol, 3-nitrophenol, and 4-nitrophenol are excluded, confirm-
TABLE 1: Standard Gibbs Energies of Transfer and Partition Coefficients of the Studied Anions
anion of
∆G
A
-
,aq
QNB
/
kJ mol
-1
∆G
A
-
,aq
QNPOE
/
kJ mol
-1
∆G
A
-
,aq
Qn-oct
/
kJ mol
-1 a
log P
A
-
,aq
QNB
log P
A
-
,aq
QNPOE
log P
A
-
,aq
QN-oct
a
r/Å
b
1 phenol 20.45 19.50 23.13 -3.62 -3.46 -4.10 2.78
2 2-nitrophenol 14.60 18.00 22.56 -2.59 -3.19 -4.00 2.98
3 3-nitrophenol 20.00 18.75 23.70 -3.54 -3.32 -4.20 2.98
4 4-nitrophenol 21.48 17.20 24.82 -3.81 -3.05 -4.40 2.98
5 2,4-dinitrophenol 8.70 11.62 24.25 -1.54 -2.06 -4.30 3.15
6 2,5-dinitrophenol 14.00 14.85 25.40 -2.48 -2.63 -4.50 3.15
7 benzoic acid 21.00 18.95 32.72 -3.72 -3.36 -5.80 2.97
8 4-bromobenzoic acid 12.00 17.32 28.20 -2.13 -3.07 -5.00 3.15
9 4-chlorobenzoic acid 12.25 15.23 31.59 -2.17 -2.70 -5.60 3.09
10 3-chlorobenzoic acid 15.25 16.65 27.08 -2.70 -2.95 -4.80 3.09
11 4-iodobenzoic acid 14.00 12.00 28.77 -2.48 -2.13 -5.10 3.21
12 naphthoic acid 15.50 17.80 31.05 -2.74 -3.15 -5.50 3.77
13 ketoprofen 19.33 18.05 28.20 -3.42 -3.20 -5.00 3.84
14 suprofen 15.80 16.47 32.15 -2.80 -2.92 -5.70 3.79
15 naproxen 11.50 12.86 28.77 -2.04 -2.28 -5.10 3.70
16 pirprofen 5.65 9.55 28.75 -1.00 -1.69 -5.10 3.76
17 flurbiprofen 10.50 12.35 24.80 -1.86 -2.19 -4.40 3.75
18 ibuprofen 17.40 15.34 28.77 -3.08 -2.72 -5.10 3.59
19 carprofen -14.80 -3.85 18.05 2.62 0.68 -3.20 3.82
20 indomethacin 11.00 15.05 26.50 -1.95 -2.67 -4.70 4.06
21 phenylbutazone 3.70 12.85 24.25 -0.65 -2.28 -4.30 4.13
22 sulfinpyrazone 7.10 12.85 24.25 -1.26 -2.28 -4.30 4.40
23 warfarine 9.80 12.30 23.13 -1.74 -2.18 -4.10 4.05
24 phenobarbital 26.75 24.10 31.02 -4.74 -4.27 -5.50 3.64
25 phenytoine 17.70 19.65 30.45 -3.14 -3.48 -5.40 3.78
26 maleic acid 20.30 19.50 31.02 -3.60 -3.46 -5.50 2.75
27 picric acid -3.00 3.65 n.m.
c
0.53 -0.65 n.m.
c
3.28
a
Taken from reference 30.
b
van der Waals radius of the ion.
c
Not measured
TABLE 2: Physicochemical Properties of NPOE, NB, DCE, and n-Octanol, at 298 K
48
properties NPOE NB n-octanol
molar mass [g mol
-1
] 251.33 123.11 130.2
density [g cm
-3
] 1.041 1.198 0.825
molar volume [cm
3
mol
-1
] 241.4 102.7 160.8
solubility of the solvent in water [mol L
-1
] 2.0 × 10
-6
1.5 × 10
-2
4.1 × 10
-3
solubility of water in the solvent [mol L
-1
] 4.6 × 10
-2
0.2 2.4
relative permittivity 24.2 34.8 10.3
∆G
IS
R
)-
z
2
e
2
N
A
8π
0
r
(
1 -
1
r
R
)
(9)
∆G
A
-
,aq
QNPOE
) 146.4((15.62)
(
1
r
)
- 31.34((5.72)
n ) 18; r
2
) 0.85; s ) 3.87; F ) 89 (10)
Solvation Properites of 2-NPOE, NB, and n-Octanol J. Phys. Chem. B, Vol. 108, No. 14, 2004 4569