Separation and Purification Technology 57 (2007) 483494

Extraction of lactic acid by phosphonium ionic liquids

Jan Martak, Stefan

Schlosser
Institute of Chemical and Environmental Engineering, Faculty of Chemical and Food Technology,
Slovak University of Technology, Radlinskeho 9, 81237 Bratislava, Slovakia

Abstract
Tetraalkylphosphonium ionic liquid (IL) with a bis 2,4,4-trimethylpentylphosphinic anion (Cyphos IL-104) is an effective extractant of lactic
acid (LA) achieving at low LA concentrations the distribution coefficients for aqueous systems above 40. L/L equilibrium data for pure Cyphos
IL-104 and its solutions in n-dodecane have been measured. With increasing acid concentration the value of the distribution coefficient of LA
decreases. Cyphos IL-104 extracts only undissociated molecules of lactic acid (LAH) via H-bonding. Increase in the concentration of IL-104
in n-dodecane results in increasing distribution coefficient of LA and the water solubility. The high water content in the solvents with Cyphos
IL-104 is connected with the formation of reverse micelles. An interesting phenomenon of the liberation of water from the solvent in extraction
of LA has been observed. It is suggested that splitting of the reverse micelles due to the formation of LAH/IL complexes occurs. With increasing
temperature the values of the distribution coefficients of LA practically does not change or only slightly increases. In contrary to this, the increase
in temperature decreased the water solubility in solvents containing IL-104. This is interpreted by the lower stability of reverse micelles at higher
temperatures. Based on experimental equilibrium data the formation of stoichiometrically defined complexes with the structure (LAH)p (IL)(H2 O)2
and (p, 1, 2) stoichiometry, where p is in the interval from 1 to 3, is suggested. The proposed model fits the equilibrium data well and indicates the
domination of the (2, 1, 2) complex at medium aqueous acid concentrations in the interval from 0.2 to 2 kmol m3 . There are two mechanisms of
water extraction into the solvents with Cyphos IL-104: the formation of reverse micelles, and the formation of hydrated complexes of LAH with
IL. A sharp decrease in the viscosity of solvents with IL-104 with increasing concentration of water or LA was observed.
2006 Elsevier B.V. All rights reserved.
Keywords: Extraction; Lactic acid; Phosphonium ionic liquids; L/L equilibrium; Modelling

1. Introduction
There is a continuing interest in application of ionic liquids
(ILs) in extractive separations documented in papers [14]. ILs
are composed of an organic cation and either an organic or an
inorganic anion. They remain liquid over a wide range of temperatures, including room temperature [5,6]. ILs are a new group
of solvents of great interest, which have been recently studied
as potential green solvents. Practically zero vapour pressure
of the most ILs and temperature stability make them attractive
solvents in many applications, e.g. in synthesis. A higher viscos-

Abbreviations: IL, ionic liquid; LA, lactic acid; LAH, undissociated

molecule of lactic acid; (p, 1, 2), stoichiometry of the complex in the solvent phase containing p molecules of LAH, 1 molecule of ionic liquid and
2 molecules of water; TOMAC, trioctylmethylammonium chloride; TIOPO,
tris(2,4,4-trimethylpentyl)phosphine oxide, which is in fact triisooctylphosphine
oxide
Corresponding author. Tel.: +421 252496743; fax: +421 252496920.
E-mail addresses: [email protected] (J. Martak),
Schlosser).
[email protected] (S.
1383-5866/$ see front matter 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.seppur.2006.09.013

ity at room temperatures could be their less favourable property

[7]. In the most papers on extraction separations with ILs are
used ILs with imidazolium cation. In metal extractions ILs are
frequently used only as a diluent for the extractant [810] as
shown also in the review [3]. Also in some papers on extraction
of organic acids ILs served as diluents [11,12]. A new group of
ILs with tetraalkylphosphonium cation has been developed [13].
A potential of their application in extraction of organic acids was
shown in papers [1,1417].
Solvent extraction of organic acids by ionic liquids was studied by Huddleston et al. [18] and in papers [1,1417], of erythromycin by Cull et al. [19], and of chlorophenols by Bekou
et al. [20]. Extraction of several amino acids studied Smirnova
et al. [12] and Wang et al. [21]. Separation of aromatic and
aliphatic hydrocarbons by imidazolium and pyridinium ILs has
been tested by Meindersma et al. [22,23]. Eight phosphonium
and imidazolium ILs, or solvents containing ILs were tested for
the extraction of several organic acids in paper [14]. Extraction
of lactic acid (LA) has been studied in papers [1417]. Matsumoto et al. [11] investigated extraction of carboxylic acids
and LA by tributylphosphate diluted in imidazolium ILs and

 Schlosser / Separation and Purication Technology 57 (2007) 483494

J. Martak, S.

484

Nomenclature
c
ca

cp,1

c0,1
cI
D
Da

H
Kp,1
m
M
p
R
S
T
V
w
z

molar concentration of the solute (lactic acid) in

undissociated form (kmol m3 )
analytical (overall) molar concentration of the
solute (LA) considering both dissociated and
undissociated forms (kmol m3 )
molar concentration of the complex
(LAH, IL) with stoichiometry (p, 1) in the solvent
(kmol m3 )
molar concentration of the free ionic liquid in the
solvent (kmol m3 )
overall molar concentration of the ionic liquid in
all forms in the solvent (kmol m3 )
distribution coefficient, defined by Eq. (9)
distribution coefficient, defined for analytical
aqueous concentration of LA in the aqueous
phase, caF
enthalpy (J mol1 )
extraction equilibrium constant of the (p, 1)
complex (m3p kmolp )
mass (kg)
molar mass (kg kmol1 )
number of solute molecules (lactic acid) in (p, 1)
acid/IL complex
gas constant (8.314 J K1 mol1 )
entropy (J K1 mol1 )
temperature (K)
volume (m3 )
mass fraction
loading of the ionic liquid defined by Eq. (7)

Greek symbols

mass balance ratio of acid defined by Eq. (10)

dynamic viscosity (mPa s)

density (kg m3 )
Indices
dry
dry solvent
F
feed (aqueous) phase
H2 O
overall water
hydrat H2 O hydration water in the (LAH, IL, H2 O)
complex
I
ionic liquid
k
solute (acid)
o
initial value
S
solvent (organic) phase
*
at equilibrium

tested also solvent toxicity for bacteria. Toxicity of ionic liquids

and organic solvents to lactic acid-producing bacteria studied
Matsumoto et al. [24].
The phosphonium IL with the 2,4,4-trimethylpentylphosphinic anion (Cyphos IL-104) is a new effective extractant
of LA [14,17] with a considerably higher value of the distribu-

tion coefficient comparing to liquid extractants described earlier.

Mainly the synergetic effect of the anion is responsible for the
increased distribution coefficient of LA by Cyphos IL-104. The
extraction of LA with this IL is carried out by a coordination
mechanism. The presented results document a potential in the
designing of ionic liquids with specific properties via a combination of the cation and the anion exhibiting synergy effect. A
high solubility of water in the studied phosphonium ILs is connected with the formation of reverse micelles in the solvent. A
dual mechanism of water extraction to hydrophobic phosphonium ILs was identified. This involves incorporation of water
into reverse micelles, and formation of the hydrated complex
of LA with IL. The extraction of LA causes splitting of reverse
micelles with liberation of water from the solvent. In the saturated solvent only hydration water remains in the complex of
LA with IL with the dominating structure (LAH)2 (IL)(H2 O)2
[17].
Extraction of azo dyes from waste solutions of leather
industry by N-butyl,N-methyl pyrrolidinium bis(trifluoromethanesulfonyl) amide IL is reported in paper [25]. A large
effort is devoted to heavy and radioactive metals extraction with
solvents based on IL diluents as reviewed in paper [3].
Transport of amines and neutral organic substances through
liquid membranes from IL is of concern in the papers of Fortunato [26], and Branco [27]. Pertraction of organic acids through
liquid membranes was described in papers [15,16], where prolonged operation of phosphonium IL was documented. Transport of organic acids facilitated by enzymatic reactions on L/L
interfaces, using IL as a liquid membrane, has been studied by
Miyako et al. [28,29].
The aim of this work was a more detailed study of the solvent
properties of the phosphonium ionic liquid Cyphos IL-104 with
bis 2,4,4-trimethylpentylphosphinate anion whose structure is
shown in Fig. 1 in extraction of lactic acid. Pure IL-104 and its
solutions in n-dodecane have been studied.
2. Theory
Phosphonium ILs with a hydrophobic anion form complexes
with LA and other acids via a coordinating mechanism shown
in works [14,16,17]. This fact will be supported with further
data presented in this paper. Monobasic carboxylic acids usually form in reactive extraction complexes containing a variable
number of acid molecules and only one molecule of extractant, the so called (p, 1) complexes. This has been found, for
example in the extraction of carboxylic acids by trialkylamines
[30,31]. Based on the data presented in this paper, as discussed
in Section 4.2, the formation of stoichiometrically defined (p,
1, 2) complexes with the structure (LAH)p (IL)(H2 O)2 is suggested where p is from the interval from 1 to 3. A simplified
notation of the stoichiometry (p, 1) will also be used. The
reactive extraction of LA by the ionic liquid IL-104 can be
described by the following reaction of a hydrated complex
formation
pLAH + 2H2 O + IL  (LAH)p (IL)(H2 O)2

(1)

 Schlosser / Separation and Purication Technology 57 (2007) 483494

J. Martak, S.

485

Fig. 1. Structure of trihexyl(tetradecyl)phosphonium bis 2,4,4-trimethylpentylphosphinate (Cyphos IL-104).

as will be discussed in Section 4.2. The stripping of acid from

IL to an aqueous alkaline solution follows the reaction
(LAH)p (IL)(H2 O)2 + pOH  p(LA) + IL + (p + 2)H2 O
(2)
These reactions proceed on the water/solvent or solvent/stripping solution interfaces. The overbar designates
species in the solvent phase. The extraction constants of individual complexes based on reaction (1) are defined by the following
relation
Kp,1 =

cp,1
[(LAH)P (IL)(H2 O)2 ]
= p
(cF ) c0,1
[LAH]p [IL]

(3)

where c0,1 is the concentration of the free ionic liquid. Because

of an excess of water in the aqueous phase its concentration is
considered to be constant and is included in the extraction constant. The stripping reaction (2) proceeds quantitatively when
an excess of alkali is maintained, as shown at the end of Section
3.2. From this follows that the value of equilibrium constant of
the stripping reaction is approaching infinity.
For the concentration of the complex in the solvent can be
derived from Eq. (3)
cp,1 = c0,1 Kp,1 (cF )p

(4)

The concentration of the free ionic liquid can be calculated from

the total IL concentration in the solvent after subtraction of the
ionic liquid in the complexes
c0,1 = cI c0,1

i

p=1

Kp,1 (cF )p =

1+

i

cI

p
p=1 Kp,1 (cF )

(5)

The concentration of LA in the solvent phase is equal to the sum

of concentrations of complexes multiplied by the number of LA
molecules in the complex

cI ip=1 pKp,1 (cF )p

cS =
(6)

1 + ip=1 Kp,1 (cF )p
The loading of IL by LA is defined as
i
p
cS
p=1 pKp,1 (cF )
z=
=
i
cI
1 + p=1 Kp,1 (cF )p

(7)

For the correlation of experimental data three complexes with

(LAH, IL) stoichiometry (1, 1), (2, 1), and (3, 1) were considered
and the following equation was used
z=

K1,1 cF + 2K2,1 (cF )2 + 3K3,1 (cF )3

1 + K1,1 cF + K2,1 (cF )2 + K3,1 (cF )3

(8)

in estimation of the extraction constants. Eq. (7) shows that loading is independent of the IL concentration. For the distribution
coefficient follows relation
i
p1
cS
p=1 pKp,1 (cF )
D= =
cI
(9)
i
cF
1 + p=1 Kp,1 (cF )p
From this relation results that the distribution coefficient is
dependent on both the aqueous concentration of undissociated acid cF and the overall concentration of IL in the solvent
cI .
3. Experimental
3.1. Materials
Cyphos IL-104 (Cytec, CA) is an ionic liquid, which
is trihexyl-(tetradecyl)phosphonium bis 2,4,4-trimethylpentylphosphinate with a purity better than 95%. Its structure is
shown in Fig. 1. In the technical product of Cyphos IL104, tris(2,4,4-trimethylpentyl)phosphine oxide (TIOPO) is
present as an impurity in a concentration up to 5 mass%
what is around 0.11 kmol m3 . Cyphos IL-104 is hydrophobic what is documented by its low solubility in water (Table 1)
and is completely miscible with aromatic hydrocarbons, isopropanol, diethylether and partly miscible with alkanes, trioctylamine and DEHPA. Cyphos IL-101 (Cytec) is trihexyl(tetradecyl)phosphonium chloride with a purity of 96 to
97 mass%. Cyanex 272 (Cytec) is a widely used metal extractant bis-(2,4,4-trimethylpentyl)phosphinic acid with a purity of
about 85 mass% and it is a protonated form of the Cyphos IL104 anion. Cyanex 923 (Cytec) is a mixture of trialkylphosphine
oxides with C6 and C8 alkyls with a purity of 93%. The solubility
of water in pure Cyphos IL-104 and its solutions in n-dodecane
and in other compounds studied together with the density of dry
solvents at 25 C are presented in Table 1.
All solvents have been conditioned before experiments.
Cyphos IL-101 was washed with equal volume of 0.5 kmol m3

 Schlosser / Separation and Purication Technology 57 (2007) 483494

J. Martak, S.

486

Table 1
Solubility of water in pure Cyphos IL-104 and its solutions in n-dodecane and in other compounds studied together with the density and viscosity of dry solvents at
25 C
Solvent

cIo,dry
(kmol m3 )

Molar mass
(g mol1 )

I,dry (kg m3 )

I,dry (mPa s)

Solubility in
water (g m3 )

ws , H2 O 102
(mass%)

cS,H2 O
(kmol m3 )

Pure Cyphos IL-104

1.102

773.3

885.67

707,2

9.1

14.4

7.28

49.4a

Cyphos IL-104 in
n-dodecane

0.724

837.81

0.324

788.98

Pure Cyphos IL-101

Pure Cyanex 272

1.661
2.716

519.3
290.4

889.36b
917.17b

Pure Cyphos IL-105

Pure Cyanex 923

1.581
2.369

549.9
344.6

896.54b
877.56b

16 at pH 2.6c ;
38 at pH 3.7c
72.7d
<10c

a
b
c
d

9.24

4.36

3.55

1.58

14.4
1.28

7.21
0.653

3.31
8.00

1.69
3.94

Water content in the solvent 0.58 kmol m3 .

Estimated from the density of the water saturated compound considering additivity of volumes.
The manufacturer data.
Unconditioned ionic liquid.

NaOH and Cyanex 272 with equal volume of 1 kmol m3

H2 SO4 . All other solvents were conditioned with 1 kmol m3
H2 SO4 and then with 0.5 kmol m3 NaOH. After conditioning
with H2 SO4 and/or NaOH all solvents were repeatedly washed
with distilled water till a pH of water was attained.
Lactic acid extra pure was purchased from Merck (DE) as a
90 mass% aqueous solution. NaOH and H2 SO4 with analytical
grade purity was obtained from Mikrochem (SK). n-Dodecane
has been supplied by Fluka (CH) had a purity above 98 mass %.
3.2. Methods
L/L equilibrium has been estimated by contacting 2 to 24 cm3
of an aqueous solution of LA with 12.5 cm3 of solvent using
4, 10 or 25 cm3 vials according to the final volume of the
water/solvent system. Aqueous solutions of LA have been prepared from a stock solution. This was prepared from concentrated LA, which was diluted with deionised water (5 times)
and boiled under the total reflux of distillate for more than 5 h
to split the acid dimer [32,33]. A rotational shaking water bath
with a frequency of 180 min1 was used. Its temperature was
kept at a selected value with an accuracy of 0.1 C. In the case
of solvents containing Cyphos IL-104 a thorough shaking of the
water/solvent systems induced dispergation of phases and hence
the formation of an emulsion. This can be avoided when the solvent is contacted gently with the aqueous phase and agitation is
done by rotational shaking. The intensity of agitation had to be
below the limit at which dispergation occurred. These conditions
were satisfactory for the achievement of equilibrium when the
contacting time was prolonged to more than 10 h. The emulsion
was difficult to separate by settling, but it could be broken by
centrifugation. After separation of phases, equilibrium pH of the
aqueous phase was measured and the content of lactic acid was
analysed. Equilibrium pH in most experiments was well below
pKa = 3.86 and all LA was in an undissociated form. The material balance in the L/L equilibrium was checked using the mass

balance ratio defined as

=

c F VF + c S VS
cFo VFo

(10)

where VFo and VF are the initial and the equilibrium volumes of
the aqueous phase. The value of the mass balance ratio should
be ideally equal to 1. In the presented experiments the difference
from this value was less than 0.05, but in the most measurements it was better than 0.03.
The volumes of both aqueous and organic phases are varying
due to the extraction of acid from the aqueous to the organic
phase and liberation of water in the opposite direction and
because molar volumes of these solutes are not small. Similar
opposite transfers occur also in the stripping and have to be considered when calculating the content of acid in the organic phase
from the estimated stripping solution concentration. Moreover,
the volume variations of both phases in the equilibrium must be
considered when checking the material balance according to Eq.
(10).
When the density of the organic phase in equilibrium (S ) is
available then VS can be estimated from the equilibrium concentration data. The mass of the organic phase is given by
equation
mS = mS,dry + mS,H2 O + mS,k
= mSo (1 wSo,H2 O ) + mS wS,H2 O +

c S m S Mk
S

(11)

After rearrangement of this equation a relation for the solvent

mass is obtained. Dividing it with the density of the equilibrium
solvent phase the following relation holds for its volume
VS =

VSo So (1 wSo,H2 O )
S (1 wS,H2 O ) cS Mk

(12)

 Schlosser / Separation and Purication Technology 57 (2007) 483494

J. Martak, S.

For the volume of the aqueous phase in equilibrium can be written

VF = VFo (cFo VFo cF VF )

Mk
k

+ (cSo,H2 O VSo cS,H2 O VS )

M H2 O
H2 O

(13)

where the second term is the volume of acid extracted from the
aqueous phase and the third term represents the volume of water
liberated from the solvent phase. After rearrangement of Eq.
(13) the following relation for VF
VFo + (cSo,H2 O VSo cS,H2 O VS )(MH2 O /H2 O )
cFo VFo (Mk /k )
VF =
1 cF (Mk /k )

487

The mass fraction of water in the solvents was determined

by Karl Fischer titration with the Karl Fischer single component
reagent 5 (Merck, DE), by the automatic titrator DL 53 (MettlerToledo, CH) with a Karl Fischer adapter.
The density of the organic phases was determined by the
vibrational density meter DMA-5000 (Anton Paar, A) with an
accuracy of 0.01 kg m3 according to the manufacturer. The
viscosity of the solvents was measured by the rotating cone and
plate method with Rheotest 2.1 (MLW, DE) with a thermostated
plate. The preliminary identification of reverse micelles was
performed with a light scattering goniometer BI-200SM utilizing an 136-channel BI-2030 AT correlator (Brookhaven Instruments Inc., USA) and an Ar ion laser with a 514.5 nm radiation
(Lexel 95).

(14)

was obtained. The density of pure LA, k , estimated from

concentration dependences of density at 15, 25 and 35 C, is
1300.63, 1285.80, and 1272.02 kg m3 , respectively. For pure
Cyphos IL-104 the percentage of the volume change of the
organic phase varies with increasing acid concentration from 1
to about 0.9 and then to +7 vol.%, as shown in Fig. 2a. For the
solvent with 0.72 kmol m3 of Cyphos IL-104 in n-dodecane the
volume change is relatively smaller (Fig. 2b). Volume changes
may contribute to the overall error of the experimental data when
they are not taken into consideration.
3.2.1. Analysis
The concentration of lactic acid in aqueous solutions was
analysed by capillary isotachophoresis using an analyser EA
100 (Villa, SK). The 10 mol m3 histidine chloride buffer with
pH 6 containing 0.1% methylhydroxyethylcellulose was used
as a leading electrolyte and 5 mol m3 of histidyl morpholineethanesulphonate as a termination electrolyte. The concentration of acid in the organic phase was estimated after stripping
its sample with a solution of NaOH and the aqueous solution
was analysed for the content of LA. It was found that 1.5-fold
molar excess of sodium hydroxide is enough for the quantitative
stripping of LA.

4. Results and discussion

4.1. Solubility of water in ILs
The solubility of water in phosphonium ILs Cyphos IL-104
and Cyphos IL-101 with the same cation is surprisingly high
and achieves 14.4 mass% at 25 C for both, as shown in Table 1.
Cyphos IL-104 is a hydrophobic IL, what can be documented by
its very low solubility in water at 25 C, which is about 10 g m3 .
Cyphos IL-104 and its solutions in n-dodecane when saturated
with water are transparent and not turbid. A high solubility of
water in them suggests the formation of reverse micelles. From
Fig. 3 follows that in the solvents containing IL-104 and IL101 saturated with water there are about 48 mol of water per
1 mol of ionic liquid. This is in a good agreement with the
results reported by Cardoso et al. [34], who investigated the
amino acid solubilization in reversed micelles with solutions of
trioctylmethylammonium chloride (TOMAC) with hexanol in
n-heptane and found the same ratio. They have found reverse
micelles with a hydrodynamic radius in the interval from 4 to
6 nm. TOMAC, as well as other quaternary ammonium extractants, e.g. Aliquat 336, are in fact also ionic liquids used in
solvent extraction for a long time, even in industrial applications. The formation of reverse micelles is supported also by

Fig. 2. Normalized volume changes of the solvent phase in extraction of LA by pure ionic liquid Cyphos IL-104 (a) and by the solvent with 0.72 kmol m3 of Cyphos
IL-104 in n-dodecane. Volume change: totaloverall; LAdue to acid extraction; H2 Odue to water liberation from the solvent.

488

Schlosser / Separation and Purication Technology 57 (2007) 483494

J. Martak, S.

Fig. 4. Dependence of the molar ratio of water to Cyphos IL-104 on the concentration of Cyphos IL-104 in n-dodecane and in pure IL (1.10 kmol m3 ) at
various temperatures.

Fig. 3. Dependence of the molar ratio of water to IL or other extractants on

temperature in the solvents saturated with water. Cyphos IL-104 was diluted in
n-dodecane.

the sensitivity of IL-104 and IL-101 and their solutions to form

W/O emulsions when they are in contact with aqueous solutions
via dispergation. The prerequisite for a surface activity of the
tested phosphonium ILs is clear when looking on the IL structure in Fig. 1. The long hydrophobic tetradecyl tail is connected
to phosphorus forming a tetraalkylphosphonium cation with a
polar head. The presence of reverse micelles has been identified
in Cyphos IL-104 solutions saturated with water in preliminary
measurements performed by a laser light scattering. Formation
of reverse micelles in ILs has been reported in papers [3538].
A tendency of the solvents with Cyphos IL-104 to form emulsion when dispersed with an aqueous solution can complicate
their application in classical L/L extraction. To solve this problem membrane based solvent extraction or pertraction both with
immobilized L/L interface(s) can be used [39,40].
Although, the mass fraction of water in both Cyphos IL-101
and IL-104 are the same (Table 1), the molar ratio of water to IL
is of about one-third smaller for IL-101 (Fig. 3). This difference
indicates that also the anion contributes to the formation of
reverse micelles and influences the water content in the solvent.
The acidic form of IL-104 anion, what is Cyanex 272, has a relatively low solubility of water (Table 1). Contrary to Cyanex 272
in IL-104, the use of the dicyanamide anion instead of chloride
fairly decreased the water solubility in Cyphos IL-105, as shown
in Table 1 and Fig. 3. The presence of two oxygen atoms in the
phosphinic group of the Cyphos IL-104 anion, which are able to
form H-bonds, can explain these differences in water solubility.
The molar ratio of water to IL in the solvents with Cyphos IL104 increases with increasing IL concentration and achieves the
value of 7.6 in pure IL (cS,Io dry = 1.10 kmol m3 ) at 298.15 K
(Table 1; Fig. 4). With increasing temperature the water content in the solvents with Cyphos IL-104 and IL-101 moderately
decreases (Fig. 3). This is a surprising fact, because water sol-

ubility in organics usually increases with temperature. For this

fact a lower stability of reverse micelles at higher temperatures
can be responsible. For other tested compounds, shown in Fig. 3,
a significant temperature dependence was not found. For both
temperatures of 295.15 and 308.15 K about 1 mol of water per
1 mol of Cyphos IL-105, and about 2 mol of water per 1 mol of
Cyanex 923 have been observed. At such low water contents the
formation of reverse micelles is not likely and rather hydration
via H-bonds is supposed.
The viscosity of Cyphos IL-104 is high and greatly decreases
with dilution by n-dodecane as shown in Table 1. With increasing concentration of water in the solvent also the value of the
dynamic viscosity of pure Cyphos IL-104 and its solutions in
n-dodecane decreases, as shown in Figs. 5 and 6. The full line in
Fig. 5 is correlated according to the relation similar to that used
in paper [41]
 c

S,H2 O
= I,dry exp
(15)
a

Fig. 5. Dependence of the dynamic viscosity of the pure Cyphos IL-104 on

water concentration at the share rate of 300 s1 and temperature 298.15 K. The
line is correlated according to Eq. (15).

 Schlosser / Separation and Purication Technology 57 (2007) 483494

J. Martak, S.

Fig. 6. Temperature dependences of the dynamic viscosity of the solvent with

0.72 kmol m3 of Cyphos IL-104 in n-dodecane for two water concentrations
in the solvent at the share rate of 300 s1 . Lines are correlated according to Eq.
(16) with parameters shown in Table 2.

where the value of I,dry is presented in Table 1 and

a = 4.8 kmol1 m3 . Water influence on the viscosity of IL-104
solutions is in agreement with data published for imidazolium
ILs in papers [4144]. On the other hand, this course is surprising when compared with the reverse micelles systems, e.g.
those formed with AOT where the viscosity is slightly increasing with increasing volume fraction of water [4547]. Seddon
et al. [41] attributed the decrease in viscosity to the reduction
of electrostatic interactions due to inclusion of water and lowering the cohesive force between IL molecules. This is supported
by the data on viscosity of the solvent with 0.72 kmol m3
of Cyphos IL-104 in n-dodecane saturated at 35 C with LA
(cF = 1.29 kmol m3 ), which does not change much despite the
molar ratio of water to IL dropped from 5.7 to 2.6 (Fig. 6). LA
forming a complex with IL-104 saturates its coordinating centres and prevents aggregation of IL molecules and in this way the
viscosity of the solvent decreases in a similar way than water. A
positive effect of increased temperature towards decreasing the
viscosity of the solvent is shown in Fig. 6 where full lines are
correlated according to the VTF equation, which is an empirical
modification of the Arrhenius equation, and was also used for
ILs in papers [43,48]:
= AT 0.5 exp

k
(T To )

(16)

A, k, and To are adjustable parameters and their values estimated

for data given in Fig. 6 are shown in Table 2.
Table 2
Parameters of Eq. (16) for temperature dependences of dynamic viscosity of
Cyphos IL-104 solutions shown in Fig. 6
cS,H2 O
(kmol m3 )

cS,LA
(kmol m3 )

A 102
(mPa s K0.5 )

k (K1 )

To (K)

0.58
4.34
1.01

0
0
1.10

116
0.216
2.06

15.0
1887
469

281
20.1
198

489

Fig. 7. The concentration dependence of the distribution coefficients of lactic

acid in solvent/water systems. Solvents pure Cyphos IL-104 (1.10 kmol m3 ), its
solutions in n-dodecane and Hostarex A327 (mixture of C8 C10 trialkylamines,
TAA) with 0.8 kmol m3 of isodecanol (modifier) in a fraction of C10 C13 nalkanes. All measurements were done at 298.15 K. The lines are correlated
according to the presented model.

4.2. Extraction of lactic acid

It has been found that Cyphos IL-104 is an effective extractant of LA. Extraction properties of the tested solvents will be
characterized by the loading and distribution coefficients defined
by Eqs. (7) and (9), respectively. The values of the distribution
coefficient are above 2 for the LA aqueous concentrations below
1 kmol m3 , as shown in Fig. 7. The values of the distribution
coefficient of LA in the solvents with Cyphos IL-104 are increasing with decreasing aqueous acid concentration. In pure Cyphos
IL-104 and at low acid concentrations they may exceed a value
of 40. This type of concentration dependence is advantageous
in the extraction or pertraction, where low acid concentration
in the raffinate is welcomed. From Eq. (9) follows that for the
given cF the distribution coefficient is directly proportional to
the IL concentration. This is shown also in Fig. 7 where the
data for solvents with various IL concentrations are presented.
A very moderate increase (if any) of the distribution coefficient
with increase in the temperature from 288.15 to 308.15 K has
been observed for 0.72 kmol m3 Cyphos IL-104 in n-dodecane,
as shown in Fig. 8. This is notable only for LA concentrations
below 0.1 kmol m3 . In extraction of acids by solvents with tertiary amine the value of the distribution coefficient decreases
with increasing temperature, as shown for butyric acid in paper
[31], where a 20% decrease in the distribution coefficient was
observed, when the temperature increased from 20 to 35 C. The
possibility to work at higher temperatures without negative effect
on the distribution coefficient is advantageous and will enable
to work at typical fermentation temperatures, e.g. at 32 C in
the fungal production of LA [49]. To work at higher temperature is advantageous also in terms of the decreasing solvent
viscosity what should have a positive effect towards increasing
mass-transfer and reaction rate. By increasing the temperature
from 25 to 35 C the solvent viscosity decreases from 49.4 to
33 mPa s or from 33.2 to 26.2 mPa s for the solvent saturated
with water (Fig. 6).

490

Schlosser / Separation and Purication Technology 57 (2007) 483494

J. Martak, S.

Fig. 8. The concentration dependence of the distribution coefficients of lactic

acid at various temperatures for the solvent with 0.72 kmol m3 of Cyphos IL104 in n-dodecane. The line is correlated according to the presented model.

Fig. 10. The concentration dependence of the loading of Cyphos IL-104 in

solvent/water systems for pure IL-104 (1.10 kmol m3 ) and its solutions in
n-dodecane was tested at 298.15 K. The line is correlated according to the suggested model.

Water insoluble trialkylamines are widely studied for the

extraction of LA. The distribution coefficients of LA for the
solvent containing 0.4 kmol m3 Hostarex A327, which is a mixture of C8 C10 trialkylamines, hardly exceed the value of 1 (TAA
in Fig. 7). Especially at lower acid concentrations, this value is
fairly lower than that for solvents with the lowest concentration
of Cyphos IL-104 of 0.32 kmol m3 .
The distribution coefficients Da based on analytical concentration (overall) of LA in the aqueous phase ca , considering both
undissociated and dissociated forms of LA, for solvents with
Cyphos IL-104 strongly decreased with increasing pH, and at pH
7 they approached zero (Fig. 9). From this follows that Cyphos
IL-104 extracts LA only in its undissociated form (LAH) via
H-bond coordination. As estimated from pKa = 3.86 (25 C) at
pH < 2, LA is present in the aqueous solution only in an undissociated form. This supports the hypothesis of reactive extraction
of LAH by the coordination mechanism.

The concentration dependence of the loading of IL, defined

by Eq. (7), shows that it is independent of the IL-104 concentration as shown also by the experimental data in Fig. 10. This
is generally typical for (p, 1) complexes containing a variable
number of acid molecules and only one molecule of IL acting
as a reactive extractant. A second organic phase is formed in the
solvent with 0.32 kmol m3 of Cyphos IL-104 in n-dodecane at
the loading above 1.3, probably due to the limited solubility of
the complex in n-dodecane.
The loading of Cyphos IL-104 by LA approaches the
value of 2.4 in the presented experiments (Figs. 10 and 11).
This suggests also formation of a complex with a higher
number of acid molecules than 2. Thus, for the modelling of
equilibrium, presented in Section 2, complexes with (LAH, IL)
stoichiometry (1, 1), (2, 1), and (3, 1) were supposed in Eq. (8).
The equilibrium data measured at 298.15 K were correlated
according to the model with a good fit (Figs. 10 and 11). The

Fig. 9. pH dependence of the distribution coefficient for various equilibrium LA

concentrations. Solvent with 0.72 kmol m3 IL-104 in n-dodecane was used at
298.15 K.

Fig. 11. The concentration dependence of the loading of Cyphos IL-104 in

solvent/water systems for the solvent with 0.72 kmol m3 of Cyphos IL-104
in n-dodecane at various temperatures. The line is correlated according to the
suggested model for data at 298.15 K.

 Schlosser / Separation and Purication Technology 57 (2007) 483494

J. Martak, S.
Table 3
Extraction constants, estimated according to Eq. (7), and entropies of the complexes formation at 25 C
Complex
(LAH, IL) (p, 1)

Kp,1 (m3p kmolp )

Error
(m3p kmolp )

S
( K1 mol1 )

(1, 1)
(2, 1)
(3, 1)

40.9
123
38.4

16.5
47.9
14.9

30.9
40.0
30.35

resulting values of extraction constants are presented in Table 3.

Using the extraction constants the concentration dependences
of the molar ratios of individual complexes to the total IL
content (Fig. 12a) and the concentration dependences of the
molar ratios of LA in individual complexes to the total LA
content in the solvent were estimated (Fig. 12b). These results
indicate that the dominant (LAH, IL) complex is the (2, 1)
complex in the interval of aqueous acid concentrations from
0.2 to 2.0 kmol m3 . At higher concentrations the complex (3,
1) is also important. The complex (1, 1) is being formed mainly
at aqueous LA concentrations below 0.2 kmol m3 .
The loading of IL-104 does not depend on the concentration
of IL what proves the assumption of formation of solely (p, 1)

Fig. 12. The concentration dependence of the molar ratio of individual LAH-IL
complexes and free IL (0, 1) to the total IL-104 concentration (a) and the molar
ratio of LA in individual LAH-IL complexes to the total LA concentration in
the solvent (b) in solvent/water systems at 298.15 K. The lines were calculated
according to the presented model using the extraction constants in Table 3.

491

complexes (Fig. 10). The value of loading is also independent

of temperature in the interval from 288.15 to 308.15 K (Fig. 11).
The curve calculated from extraction constants in Table 3 shows
a reasonable fit to experimental data in Fig. 10 with the value
of the correlation coefficient of 0.982. The same curve fits the
data for IL solution in Fig. 11 with the correlation coefficient of
0.965. Both the enthalpy and the entropy of the complexation
reaction were estimated according to the following equation
ln Kp,1 =

H
S
+
RT
R

(17)

Since it was shown that the extraction of LA by Cyphos IL104 is practically independent of temperature and the value of
Kp,1 is constant, the first term in Eq. (17) is zero. The calculated
values of entropy are shown in Table 3.
An interesting phenomenon observed was the liberation of
water from the solvent due to reactive extraction of LA by IL104 (Fig. 13). With increasing loading of IL in extraction of
LA to the solvent the water concentration in the solvent steeply
decreases and accordingly the molar ratio of water to IL (Fig. 13).
The molar ratios of water to IL-104 at zero loading differ for
pure Cyphos IL-104 and its solutions in n-dodecane, but with
increasing loading this ratio becomes the same and for loadings
above about 1.4 tends to the value of 2 mol of water per 1 mol
of IL (Fig. 13a). This shows, that also in the complex with the
highest acid loading for p = 3 the number of hydration water
molecules is 2 and the (LA, IL, H2 O) stoichiometry is (3, 1,
2). When considering two molecules of hydration water in all
complexes the stoichiometry is (p, 1, 2) and from the model
presented in Section 2 follows a relation for estimation of the
molar ratio of hydration water to IL in the solvent

2 ip=1 Kp,1 (cF )p
cS,hydrat H2 O
=
(18)

cI
1 + ip=1 Kp,1 (cF )p
The line estimated for hydration water from this equation
and shown in Fig. 13a is in a good agreement with experimental data at loadings higher than about 1.4. This indicates that
at higher loadings no reverse micelles are present in the solvent. The difference in the water ratio value given by this line
and the experimental water ratio is related to water encapsulated in reverse micelles. Temperature has only a small effect on
the value of the molar ratio of water to IL, especially at higher
loadings (Fig. 13b), comparing to the pure solvent without LA
where the temperature effect is present (Figs.13b and 4). The
molar ratio of water to LA in the solvent achieved a plateau at
the value slightly below 1 for the aqueous LA concentrations
above 0.8 kmol m3 for all tested temperatures and IL concentrations (Fig. 14). This plateau corresponds to the loadings above
1.9 according to Figs. 10 and 11.
Liberation of water is interpreted by a hypothesis that
the reverse micelles containing water are broken due to the
formation of the LAH/IL complex. Removal of IL molecules
from the outer layer of the reverse micelle destabilizes them
and results in the splitting of reverse micelles and liberation
of water. A similar behaviour was observed also for Cyphos
IL-101 as it is shown in paper [17], where reverse micelles were

492

Schlosser / Separation and Purication Technology 57 (2007) 483494

J. Martak, S.

Fig. 13. Dependence of the molar ratio of water to Cyphos IL-104 in the organic phase on IL loading by LA in extraction of LA by pure Cyphos IL-104 (1.10 kmol m3 )
and its solutions in n-dodecane at 298.15 K (a) and temperature dependence of this ratio for 0.72 kmol m3 Cyphos IL-104 (b). The lines represent the calculated
values for hydration water bound in all (p, 1, 2) complexes according to Eq. (18).

Fig. 14. The concentration dependences of the molar ratio of water to lactic acid in the organic phase on the equilibrium concentration of lactic acid in the aqueous
phase in extraction of LA by pure Cyphos IL-104 (1.10 kmol m3 ) and its solutions in n-dodecane at 298.15 K (a) and temperature dependence of this ratio for the
solvent with 0.72 kmol m3 of Cyphos IL-104 (b). The lines represent the calculated values for hydration water bound in all (p, 1, 2) complexes according to the
presented model and Eq. (18).

also considered. This shows that there are two mechanisms of

water extraction into the solvent with Cyphos IL-104. These
are: the formation of reverse micelles, and the inclusion of
water into the hydrated complex of LA with IL containing two
molecules of water in the complex molecule.
All facts discussed above support the hypothesis that the
extraction of LA by Cyphos IL-104 is stoichiometricaly defined,
indicating the reactive mechanism of acid extraction by IL-104.
For comparison, in extraction of LA by phosphonium chloride
(Cyphos IL-101) an ion-exchange mechanism occurs or considerably contributes to extraction as suggested in paper [17].
The suggestions for on-line extraction of LA from the fermentation broth for fungal fermentation is discussed in paper
[49] and generally in paper [39]. The pH of fermentation broths
are mostly higher than pKa of lactic acid and should be decreased
by the addition of a mineral acid before the extraction step.
Cyphos IL-104 has an affinity also to mineral acids. The distribution coefficients of HCl, H2 SO4 and lactic acids are similar,

as presented in paper [15]. A competitive extraction of mineral

acids may decrease the extraction of LA, similarly as it was
observed in the extraction of phenol with octylmethylimmidazolium hexafluorophosphate from sulphate solutions reported in
paper [14] and for extraction of methylpyrazinecarboxylic acid
with trioctylamine described in papers [39,50].

5. Conclusions
Phosphonium IL Cyphos IL-104 is an effective extractant of
lactic acid achieving at low LA concentrations the distribution
coefficients above 40. In contrast to this, for the solvent with
0.4 kmol m3 trialkylamine, what is a classical solvent for LA,
the distribution coefficient is about 1. With increasing acid
concentration the value of the distribution coefficient decreases.
Cyphos IL-104 extracts only undissociated LA via H-bond
coordination.

 Schlosser / Separation and Purication Technology 57 (2007) 483494

J. Martak, S.

The high water content in solvents with phosphonium ionic

liquid Cyphos IL-104, which is quite hydrophobic, suggests the
formation of reverse micelles. An interesting phenomenon of
the liberation of water from the solvent to the aqueous phase in
extraction of lactic acid has been observed. This suggests splitting of the reverse micelles due to the formation of LAH/IL complexes and destabilisation of reverse micelles. With increasing
temperature the value of the distribution coefficient practically
does not change or only slightly increases. In contrary to this,
the increase in temperature decreases the water solubility in solvents containing IL-104, what is not typical for other organic
compounds. This is interpreted by the lower stability of reverse
micelles at higher temperatures.
A sharp decrease in viscosity of solvents containing IL-104
with increasing concentration of water was observed. Viscosity
of the solvent with IL-104 in n-dodecane saturated with LA does
not change much despite the molar ratio of water to IL dropped
from 5.7 to 2.6. Both water and LA saturate the coordinating
centres in IL molecules and prevent their structuring or aggregation. In this way the viscosity of the solvent phase decreases.
The fact that loading is independent of the IL-104 concentration and the existence of the limiting molar ratios of LAH/IL and
H2 O/IL in the saturated solvent phase with the value 2.4 and 2,
respectively, enabled to formulate the reactive mechanism of LA
extraction. The formation of stoichiometrically defined (p, 1, 2)
complexes with the structure (LAH)p (IL)(H2 O)2 is suggested
where p is from the interval ranging from 1 to 3. The proposed
model fits the equilibrium data well and indicates the domination
of the (2, 1, 2) complex at the medium acid concentrations in the
interval from 0.2 to 2 kmol m3 . LAH/IL/H2 O complexes with
stoichiometry (1, 1, 2) and (3, 1, 2) dominate at the aqueous LA
concentrations below 0.2 and above 2 kmol m3 , respectively.
There are two mechanisms of water extraction into the solvent
with Cyphos IL-104: the formation of reverse micelles, and the
inclusion of water into hydrated complexes of LA with IL having
two molecules of water in the complex.
Acknowledgements
This work was supported by the Slovak Grant Agency under
the Grant No. VEGA No. 1/2468/05. We thank Al J. Robertson and Cytec for kind providing samples of ionic liquids and
discussions. We are grateful to Dr. D. Lath from the Polymer
Institute of the Slovak Academy of Sciences in Bratislava for
preliminary laser light scattering measurements.
References
[1] J. McFarlane, W.B. Ridenour, H. Luo, R.D. Hunt, D.W. DePaoli, R.X. Ren,
Sep. Sci. Technol. 40 (2005) 1245.
[2] J.F. Liu, J.A. Jonsson, G.B. Jiang, Trends Anal. Chem. 24 (2005) 20.
[3] H. Zhao, S.Q. Xia, P.S. Ma, J. Chem. Technol. Biotechnol. 80 (2005) 1089.
[4] S. Pandey, Anal. Chim. Acta 556 (2006) 38.
[5] R.D. Rogers, K.R. Seddon, Ionic liquids, in: Industrial Applications to
Green Chemistry, ACS Symposium Series, vol. 818, 2002.
[6] R.D. Rogers, K.R. Seddon, Ionic liquids as green solvents, in: Progress and
Prospects, ACS Symposium Series, vol. 856, 2003.
[7] P. Wasserscheid, W. Keim, Angew. Chem. Int. Ed. 39 (2000) 3773.

493

[8] A.E. Visser, R.P. Swatloski, W.M. Reichert, S.T. Griffin, R.D. Rogers, Ind.
Eng. Chem. Res. 39 (2000) 3596.
[9] A.E. Visser, J.D. Holbrey, R.D. Rogers, Chem. Commun. (2001) 2484.
[10] K. Nakashima, F. Kubota, T. Maruyama, M. Goto, Ind. Eng. Chem. Res.
44 (2005) 4368.
[11] M. Matsumoto, K. Mochiduki, K. Fukunishi, K. Kondo, Sep. Purif. Technol. 40 (2004) 97.
[12] S.V. Smirnova, I.I. Torocheshnikova, A.A. Formanovsky, I.V. Pletnev, Anal.
Bioanal. Chem. 378 (2004) 1369.
[13] C.J. Bradaric, A. Downard, C. Kennedy, A.J. Robertson, Y.H. Zhou, Green
Chem. 5 (2003) 143.
Schlosser, Screening of ionic liquids for application in solvent
[14] J. Martak, S.

extraction and pertraction, in: J. Markos, V. Stefuca

(Eds.), Proceedings of
31th Conference of SSCHI (CD ROM with full papers), Tatranske Matliare
(SK), May 2428, 2004, 12 p.
Schlosser, Ionic liquids in pertraction and extraction
[15] J. Martak, S.
of organic acids, in: B. Lenarcik, R. Wodzki (Eds.), Proceedings of
19th Symposium Ars Separatoria, Zloty Potok (PL), 2004, p. 106,
http://www.ars separatoria.chem.uni.torun.pl/MATERIALY4.htm.
Schlosser, Extraction and pertraction of lactic acid by phos[16] J. Martak, S.
phonium ionic liquid, in: Proceeding of PERMEA (CD ROM with full
papers), 1822 September, 2005, 11 p.
Schlosser, Chem. Pap. 60 (2006) 395.
[17] J. Martak, S.
[18] J.G. Huddleston, H.D. Willauer, R.P. Swatloski, A.E. Visser, R.D. Rogers,
Chem. Commun. (1998) 1765.
[19] S.G. Cull, J.D. Holbrey, V. Vargas-Mora, K.R. Seddon, G.J. Lye, Biotechnol. Bioeng. 69 (2000) 227.
[20] E. Bekou, D.D. Dionysiou, R.Y. Qian, G.D. Botsaris, Extraction of
chlorophenols from water using room temperature ionic liquids, in: R.D.
Rogers, K.R. Seddon (Eds.), Ionic Liquids As Green Solvents: Progress
and Prospects, ACS Symposium Series, vol. 856, 2003, p. 544.
[21] J.J. Wang, Y.C. Pei, Y. Zhao, Z.G. Hu, Green Chem. 7 (2005) 196.
[22] W.G. Meindersma, Anita.J.G. Podt, Andre.B.de. Haan, Fuel Process. Technol. 87 (2005) 59.
[23] G.W. Meindersma, A. Podt, M.G. Meseguer, A.B. de. Haan, Ionic liquids
as alternatives to organic solvents in liquid-liquid extraction of aromatics,
in: R.D. Rogers, K.R. Seddon (Eds.), Ionic Liquids III B: Fundamentals,
Progress, Challenges and Opportunities: Transformations and Processes,
ACS Symposium Series, vol. 902, 2005, p. 57.
[24] M. Matsumoto, K. Mochiduki, K. Kondo, J. Biosci. Bioeng. 98 (2004) 344.
[25] R. Vijayaraghavan, N. Vedaraman, M. Surianarayanan, D.R. MacFarlane,
Talanta 69 (2006) 1059.
[26] R. Fortunato, L. Branco, C.A.M. Afonso, M.A. Reis, J.G. Crespo, Membr.
News (EMS) (2002) 35.
[27] L.C. Branco, J.G. Crespo, C.A.M. Afonso, Angew. Chem. Int. Ed. 41 (2002)
2771.
[28] E. Miyako, T. Maruyama, N. Kamiya, M. Goto, Chem. Commun. (2003)
29262927.
[29] E. Miyako, T. Maruyama, N. Kamiya, M. Goto, Biotechnol. Lett. 25 (2003)
805.
Schlosser, Chem. Pap. 54 (2000) 413.
[30] J. Martak, S.
Schlosser, J. Martak, J. Chem. Eng. Data 46 (2001) 735.
[31] E. Sabolova, S.
[32] J.A. Tamada, A.S. Kertes, C.J. King, Ind. Eng. Chem. Res. 29 (1990)
1319.
[33] J. Prochazka, A. Heyberger, V. Bzek, M. Kousova, E. Volaufova, Ind. Eng.
Chem. Res. 33 (1994) 1565.
[34] M.M. Cardoso, M.J. Barradas, K.H. Kroner, J.G. Crespo, J. Chem. Technol.
Biotechnol. 74 (1999) 801.
[35] M.M. Cardoso, R.M.C. Viegas, J.P.S.G. Crespo, J. Membr. Sci. 156 (1999)
303.
[36] L. Gaillon, J. Sirieix-Plenet, P. Letellier, J. Solution Chem. 33 (2004) 1333.
[37] J. Sirieix-Plenet, L. Gaillon, P. Letellier, Talanta 63 (2004) 979986.
[38] G.M. Sando, K. Dahl, J.C. Owrutsky, Chem. Phys. Lett. 418 (2006) 402.
Schlosser, R. Kertesz, J. Martak, Sep. Purif. Technol. 41 (2005) 237.
[39] S.
Schlosser, Desalination 199 (2006) 518.
[40] J. Martak, S.
[41] K.R. Seddon, A. Stark, M.J. Torres, Pure Appl. Chem. 72 (2000) 2275.
[42] B.D. Fitchett, T.N. Knepp, J.C. Conboy, J. Electrochem. Soc. 151 (2004)
E219.

494

Schlosser / Separation and Purication Technology 57 (2007) 483494

J. Martak, S.

[43] J. Jacquemin, P. Husson, A.A.H. Padua, V. Majer, Green Chem. 8 (2006)

172.
[44] S. Pandey, K.A. Fletcher, S.N. Baker, G.A. Baker, Analyst 129 (2004) 569.
[45] R.A. Day, B.H. Robinson, J.H.R. Clarke, J.V. Doherty, J. Chem. Soc., Faraday Trans. I 79 (1979) 132.
[46] S.P. Moulik, B.K. Paul, Adv. Colloid Interface Sci. 78 (1998) 99.

[47] T. Kinugasa, A. Kondo, S. Nishimura, Y. Miyauchi, Y. Nishii, K. Watanabe,

H. Takeuchi, Colloids Surf. A, Physicochem. Eng. Aspects 204 (2002) 193.
[48] O.O. Okoturo, T.J. VanderNoot, J. Electroanal. Chem. 568 (2004) 167.
Schlosser, E. Sabolova, L. Kristofkova, M. Rosenberg, Pro[49] J. Martak, S.
cess Biochem. 38 (2003) 1573.
Schlosser, Chem. Pap. 56 (2002) 418.
[50] L. Kubisova, J. Martak, S.