j o u r n a l of MEMBRANE SCIENCE
Journal of Membrane Science 130 (1997) 7-15
Kinetic analysis of coupled transport of thiocyanate ions through liquid membranes at different temperatures M. Kobya*, N. Topoa, N. Demircio~lu Department of Environmental Engineering, Faculty of Engineering, University of Atatiirk, 25240 Erzururn, Turkey Received 30 January 1996; received in revised form 28 November 1996; accepted 30 November 1996
Abstract Non-steady-state kinetics of coupled transport of thiocyanate ions through liquid membrane (trichloromethane), containing hexadecyl trimethyl ammonium chloride as a carrier, was examined at different temperatures. The kinetics of thiocyanate transport could be analyzed in the formalism of two, consecutive, irreversible first order reactions. The influence of temperature on the kinetic parameters (kid, k2m, Rmax, tmax, J~d~x, J~aax) have been also investigated. The membrane entrance rate, kid, and the membrane exit rates, k2m and k2a, increase with temperature. For maximum membrane entrance and exit fluxes, J~dax and J~aax, the activation energies were found from the slopes of the two linear relationships: 7.75 and 8.30 kcal/ mol, respectively. The values of the found activation energy indicate that the process is controlled by species difussion.
Keywords: Coupled transport; Thiocyanate ion transport; Transport kinetics; Liquid membranes; Wastewaters; Temperature effects
1. Introduction Liquid membranes play an important role in separation processes . Their efficiency and economic advantages designate them as the optimal solutions of some important problems in science and technology, such as precious-metal recovery , toxic product (heavy metals, organic molecules) [3-5] elimination from wastewaters, etc. Therefore, scientific research in this field continues to be very active despite some important solution-awaiting technological problems which at present prevent large-scale applications of liquid membranes in the chemical industries. Therefore, detailed investigation of kinetic *Corresponding author. 0376-7388/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PII S 0 3 7 6 - 7 3 8 8 ( 9 6 ) 0 0 3 4 8 - 1
analysis of ion transport of liquid membranes is very important in view of their selective, efficient and lowcost application in science and technology. Besides hundreds of cation-separation systems published in literature, only a few examples are known for the separation of anion. Liacono et al.  used counter-transport of chloride ion to remove CrO 2- anions from a water phase. Another example of a countertransport process, between chloride and hydroxide ion, is given by Molnar et al. . In this latter article, the transport mechanism was not entirely clear and some complexities in the chemistry of facilitated diffusion were developed for the concentration of picrate ion  and nitrate ion . The removal of toxic elements such as thiocyanate anions by liquid membranes has been of great interest
M. Kobya et aL/Journal of Membrane Science 130 (1997) 7-15
in recent years. Thiocyanate ion is an environmentally significant pollutant in industrial wastewater from coke manufacturing and iron making, petrochemical industries, and from coal gasification and liquefaction . Biological oxidation methods have been used for decomposition of biodegradable effluents, where wastewater is aerated in the presence of microorganisms that can metabolise and thus destroy many undesirable compounds. Biological treatment of thiocyanate requires long residence times and hence large tank capacities to cope with relatively high volumes of effluent streams. Furthermore, micro-organisms are extremely sensitive to such factors as pH, temperature and solid contents . Several works have been published on the removal of thiocyanate ions from industrial wastewaters using chemical oxidation, electrochemical oxidation, ion exchange, ozonization, and similar non-biological processes. The overall process claimed appears costly. The use of amines and quaternary ammonium salts in hydrometallurgical operations is well known and they provide a potentially useful group of reagents for the extraction of anionic species. The results were obtained for thiocyanate ion extraction using these quaternary ammonium salts
. In this study, the coupled transport kinetics of thiocyanate ion transport through liquid membranes has been studied at different temperatures in the range 293-308 K. The kinetics of thiocyanate transport could be analyzed in the formalism of two consecutive irreversible first order reactions. The influence of temperature on the kinetic parameters has been also investigated.
2.1. Materials The chemical reagents used in these experiments are potassium thiocyanate, KSCN (Fluka, >99%), hexadecyl trimethyl ammonium chloride (Fluka, >99.5%), trichloromethane (Merck, >99.5%), and sodium chloride, NaC1 (Merck, >99.5%). The aqueous solutions were prepared using demineralized water.
synchronous m o t o r
motor Fig. 1. Schematic representation of a liquid membrane system.
2.2. Kinetic procedure Coupled-transport experiments were conducted using a thermostated (TECHNE mark, model TE8D) apparatus, shown in Fig. 1. The donor ('d,' 150 ml) and acceptor ('a,' 150 ml) water phases were stirred at variable speed by a mechanical stirrer (IKA mark, model RE 166). The membrane phase ('m,' 200 ml) (density >1) under the water phases was stirred magnetically (HEIDOLPH mark, model MR 3003 SD). Stirring speeds for the donor, membrane and acceptor phases were 150, 200 and 150rpm, respectively. The initial composition of the phases: the donor phase was an aqueous KSCN solution (initial thiocyanate ion concentration, Cdo=4.31 X 10 -4 M; pH 10.50) while the acceptor phase was an aqueous 5 M sodium chloride solution. The organic membrane phase was made up by dissolving carrier hexadecyl trimethyl ammonium chloride in the trichloromethane (Ccarrier~-10-3 M). The duration of a kinetic run was 360 min. The surfaces were as follows: Sd. . . . pha~c/ membrane~-0.921 c m 2,
Sacceptor/membran e phase ~
0.743 cm 2. Samples (0.5 ml) were taken from both water phases (acceptor and donor phases) at regular time intervals and the thiocyanate ion concentration was analyzed by spectrophotometric method . Each experimental result reported is the arithmetic mean for
M. Kobya et al./Journal of Membrane Science 130 (1997) 7-15
two independent samples and volume change by spectrophotometric method (error <1%).
written as: [QSCN] [C1-] K = [QCl] [SCN-]
3. Resultsand discussion In coupled transport, the carrier agent couples the flow of two or more species, e.g. thiocyanate ion and chloride. In this case, the carder must contain, of course, chloride-ionizable groups. Thanks to this coupling, thiocyanate ion can be transported against its concentration gradient provided the concentration gradient of chlorides is sufficiently large. The mechanism of coupled transport of thiocyanate ion is given in Fig. 2. QC1 represents the chloride-ionizable carrier. At the interface d/m, thiocyanate ion reacts with one chloride-ionizable carrier molecule QC1, liberating one chloride ions. Then the complex, QSCN, diffuses through the membrane. At the interface m/a, the carrier molecules are protonated and the thiocyanate ions are liberated into the receiving phase. Finally, the neutral carder diffuses back across the membrane. Thus, the thiocyanate ions move from left to right and electrical neutrality is maintained by the movement of chloride ions in the opposite direction. A simple theoretical approach can be used to obtain flux equation for coupled-transport systems ; we must consider the equilibrium existing at the two interfaces (membrane-water phases): QC1 4- SCN- ~, QSCN 4- C1-
The equilibrium constant of this reaction can be
It should be noted that only [QSCN] and [QCl] are measurable in the organic phase (phase m) and [el-] and [SEN-], in the aqueous phases (d and a). Taking into account Fick's law at steady state, the flux of the thiocyanate complex across the liquid membrane is given by JQSCN -- DQSC.___~N([QSCN]d _ [QSCN]a) l
where DQSCN is the mean diffusion coefficient of the complex in the membrane of thickness I. The coupling effects can be demonstrated by taking into account the mass-balance expression for the carrier molecule: [QSCN] + [QC1] =C
where C is the total concentration of the pure carrier QC1. By combining Eqs. (2)-(4), we obtain the final flux equation:
-[C1-] 1 + K[SCN-]/a
1 1 -~
[C1-] i K[SCN-]/a
(5) It can be seen that active transport will exist as long as we have . [C1-] -~ < (" [C1-]d ~ K[QC1][SCN-I}a \K[QC1][SCN-]] a
[SCN-]d [Cl-]d [SCN-]a > [C1-]-----~
The maximum separation is attained when the flux stops, i.e. when [SCN-]d --[C1-]d [SCN-]a [C1-]a w~et-mem~ane
Fig. 2. Apparatus used for coupled thiocyanate ion transport through liquid membranes.
The variation of thiocyanate ion concentration with time was directly measured in both donor (Cd in M) and acceptor (Ca in M) aqueous phases, while the corresponding time dependence in the membrane phase (Cm in M) was calculated from the material
M. Kobya et al./Journal of Membrane Science 130 (1997) 7-15 1.00
. 1 O0
Fig. 3. Time dependence of reduced concentration of thiocyanate ions, Rd (N), Rm (+), Ra (m) phases in coupled transport through liquid membranes (T--298+0.1 K) Theoretical curves are calculated from Eqs. (15), (16) and (17), respectivley.
Fig. 5. Time variation of reduced concentration of thiocyanate ions in the membrane phase (Rm) during coupled transport through liquid membranes at different temperatures: 293 K (Fq), 298 K (11), 303 K (+) and 308 K (O). Theoretical curves are calculated from Eq. (16).
,. . . .
-r . . . .
,. . . .
= 0 o 0 ,. . . .
-r . . . .
, , , , i , , , l l l , , , l l , l , l , , , , l l , , i f , , , l l
Fig. 4. Time variation of reduced concentration of thiocyanate ions in the donor phase (Rd) during coupled transport through liquid membranes at different temperatures: 293 K (Fq), 298 K (m), 303 K (+) and 308 K (O). Theoretical curves are calculated from Eq. (15).
Fig. 6. Time variation of reduced concentration of thiocyanate ions in the acceptor phase (Ra) during coupled transport through liquid membranes at different temperatures: 293 K ([3), 298 K (11), 303 K (+) and 308 K (O). Theoretical curves are calculated from Eq. (17).
balance. The results are represented in Figs. 3-6, where dimensionless, reduced concentrations are used for practical reasons:
material-balance equation is reduced to the simple expression:
Rd ---- Cd Cdo'
Rm = Cm , Ra - - Ca Ca•
where Coo is the initial concentration of thiocyanate ion in the donor phase at t=0. With this notation, the
Ra+Rm+Ra = 1
In all cases, Ro decreases mono-exponentially with time, R a follows a monotonically increasing sigmoid-type curve, while the time evolution of Rm presents a maximum. These observations suggest
M. Kobya et al./Journal of Membrane Science 130 (1997) 7-15
that, in the present case, thiocyanate transport obeys the kinetic laws of two, consecutive, irreversible first order reactions according to the kinetic scheme:
d ~k~m ~ ak2
where d, m and a stand for thiocyanate ions in the donor, membrane and acceptor phases, respectively. The irreversibility is suggested by the fact that in favorable cases, Rd and R m become zero while R a reaches the limiting value of 1 at the end of the transport experiments (520 min, also see below). In other words, the transport from the donor phase to the acceptor phase is virtually complete without any back leakage due to kinetic reversibility. This fact could be further confirmed by the chloride material balance. As a matter of fact, the final chloride ion concentration value in the donor phase corresponded exactly to the amount of transported copper ions. Since these latter ions are complexed by two carrier molecules [ 14], the complete transport of a given number of ion-grams of thiocyanate was accompanied by twice greater a number of counter-transported ion-grams of chloride. The kinetic irreversibility of the whole transport process is imposed by the important chloride gradient across the membrane, since we have [C1-]ao/[C1-]do=5 × 105 at t=0. The resulting chloride flux is coupled to the thiocyanate ion flux through the ionizable cartier molecules. It was shown Eq. (7) that, for steady state situation, coupled transport is maintained. It is clear from the experimental conditions that this relationship is verified in the present case, even at the transport yield of 99.99%. Obviously, we can write
[SEN ]do -- [SCN-]a ~-C N--~-a
([El-]do "[- [SUN-la b > ~,~ + [SCN_]aj
with [SCN-]do=4.31 x 10 - 4 M , [C1-]do=l x 10 -5 M, [C1-]ao=5M and [SCN-]a=0.9999[SCN]do, from which we obtain 10x 10-5>8.827 x 10 -5, i.e. the transport of thiocyanate ion from the donor phase still continues. It can be expressed in Eq. (8) that the thiocyanate ion flux stops. In this case, this will occur when we have 99.99%
Figs. 3 and 6 show that thiocyanate ion accumulates in the membrane phase during the transport process. As a consequence, the thiocyanate ion concentration gradient varies permanently and, therefore, non-steadystate kinetics will govern the whole transport process. Such kinetic behavior may be observed whenever the amounts of carrier and that of thiocyanate ion are comparable [14-16]. Using reduced concentrations, the following rate equations may be written for the above-proposed kinetic scheme (Eq. (10)): dRd -dt dRm
k l R d =- Jd
-- klRd -
= k 2 R m ~ Ja
Integration of this system of first order differential equations (Eqs. (13) and (14)) leads to relationships giving the time variation of the reduced thiocyanate ion concentrations in the donor, membrane and acceptor phases (non-steady-state kinetic regime) and one obtains when kl # k2 : Rd = exp(-klt)
kl Rm -- k2 - k----~[exp(-klt) - exp(-k2t)]
and Ra = 1
k2 - k~ [k2 exp(-klt) - k, exp(-k2t)] (17)
where kl (kid) and k 2 (k2m , k2a) are the apparent membrane entrance and exit rate constants, respectively. It is apparent that Rd decreases mono-exponentially with time, Ra follows a monotonically increasing sigmoid-type curve, while the time variation of Rm presents a maximum. The maximum value of R m (when dRm/dt=0), =
is reached at time, taking the logarithm of Eq. (18) and rearranging Eq. (19)obtained below : kl tmax= (kl-~1k~) 1n (~2)
M. Kobya et al./Journal of Membrane Science 130 (1997) 7-15
Ra vs. t yields an increasing sigmoid curve. It has an inflection point (when d2Ra/dt2=0) (20)
Ria~ = 1 _ (k~) -kz/(k'-k~) (1 + kk-!21) occurring at (for detail see ):
We see that, at t = tmax = tinfl, the system is in steady state since the concentration of thiocyanate ion in the membrane (Rm) does not vary with time (Eq. (26)), because the penetration (Jd) and exit (Ja) fluxes are equal but are of opposite sign (for details, see Eq. (14)): (28)
_j~nax = ..~jTax
tinn = tmax
First order time differentiation of Eqs. (15)-(17) leads to the final forms of flux equations: dRd dt dRm
-- - - [exp(-klt) - exp(-kzt)] dt k2 - kl
dR~a -- klk~2 [exp(-klt) - exp(-k2t)] dt k2 - kl
The complexity of these equations prevents simple comparison of kinetics observed for different membrane materials. Therefore, it is useful to examine and compare maximum release rates which can be attained in a given experimental condition. Substituting t in Eqs. (22)-(24) by its maximum value (Eqs. (19) and (21)) one obtains (for detail, see ):
dt Imax= - k l
dRm dt [max = 0
and dt [max = kz
The actual numerical analysis was carried out by non-linear curve fitting using a BASIC iteration programme. The first rate constant, kl, was obtained from Eq. (15) using the donor-phase data (kid), while the membrane exit-rate constant, kz, may be obtained either directly from the acceptor-phase kinetic data (kza) using Eq. (17) or indirectly from the membranephase data calculated on the basis of Eq. (16) (kzm). In both cases, the kid value obtained from Eq. (15) was used in the calculations. The obtained kinetic parameters, kld, k2m, kza, Rmax, tmax, J~dax and J~a~x are given in Table 1. The variation of reduced thiocyanate ion concentration with time in donor, membrane and acceptor phases was shown in Figs. 4, 5 and 6, respectively. Coupled transport of thiocyanate ion experiments was executed at temperatures 293, 298, 303 and 308 K. Rd deceases with time increasing the temperature; at the other temperatures except for 293 K, it was realised that a large amount of thiocyanate ion has penetrated into the membrane phase at ~ 160 min. During the transport experiments at different temperatures, it can be seen that the diffusion of thiocyanate ion into the membrane phase, kl, takes place at higher speed than that of its release from the membrane phase to the acceptor phase, k2, (kl>k2). The final value of Rd value after 360 min of reaction time is expressed with R~n. Then, permeability degree (Rp = 1 - R d n") values
Table 1 Non-steady-state kinetic parameters for coupled transport of thiocyanate ions through liquid membranes at different temperatures (k~d, k2m,
~a, n ~ x, '=x, ~ x an~ ~ ) T (K)
kid X 102 (min- 1)
k2mX 103 (min- 1)
k2a x 103 (min- 1)
R. . . .
293 298 303 308
1.594-0.03 2.954-0.02 4.254-0.01 5.14±0.01
4.364-0.01 5.134-0.02 5.614-0.03 7.054-0.02
3.984-0.03 5.09-4-0.03 5.56±0.06 6.93+0.03
0.61 0.69 0.74 0.73
dJ~ax (min- 1)
~a ~ (min 1)
112.09 71.80 54.85 44.78
2.67 3.55 4.12 5.14
2.51 3.53 4.01 5.07
M. Kobya et al./Journal of Membrane Science 130 (1997) 7-15 6.20
Table 2 Distribution of thiocyanate ion in donor, membrane and acceptor phases in coupled transport (t=360 min)
293 298 303 308
0.03 0.01 0.02 0.02
0.27 0.17 0.15 0.09
0.70 0.82 0.83 0.89
0.98 0.99 0.98 0.98
can be obtained. As long as the Rp value approaches 1, the k2m and k2a values closely approach each other. The case Rp=l shows that the numerical analysis which applied to experimental values, which has very little standard errors. This shows that in the non-steady-state kinetic regime, thiocyanate ions may accumulate little in the membrane in which thiocyanate ion concentration increases in a rapidly transporting process. As can be seen from Table 2, greater values for Rp (0.98) was obtained. This is verified by a good correlation between experimental and theoretical data. While the variation of thiocyanate ion in membrane phase with time reaches maximum for 112 rain, 293 K; 72 min, 298 K; 55 min, 303 K; and 45 min, 308 K, the separation yield can be characterized completely by Ra values. After 360rain reaction time, Ra is 0.70 at 293 K and 0.93 at 308 K. This phenomenon shows that temperature is effective on separation yield. Furthermore, it is shown in Figs. 5 and 6 that Ra approaches 1 and R m approaches zero, both increasing with temperature. For the coupled transport of thiocyanate ion, it is considered that the difference between the membrane entrance rate, kl, and membrane exit rate, k2, is high. Due to this difference, R~ ax increases and tmax decreases with increasing temperature. Thus, thiocyanate ions are rapidly transported into the membrane and less cumulation appears in the membrane. For diffusion-controlled processes, E~ values are quite low, while for chemical-controlled processes, Ea values are much higher because of the usually stronger influence of the temperature on the rate constants. Thus, the values of Ea obtained for a given process can serve as an indicator whether
5 . 0 0
t I l l l l l l l l l l l
iW i l l l , l =
( 1/T). 1O~(K-') Fig. 7. Arrhenius plot of thiocyanate ion transport in J~aax ( i ) .
diffusion or chemical reaction is the rate-controlling step. For diffusion-controlled processes, the value of the apparent activation energy is below 5 kcal/mol, whereas those for the chemical reactions have been reported to be above 10 kcal/mol . It was concluded that temperature is effective on the kinetic constants which was obtained for the coupled transport of thiocyanate ions. Since the coupled transport of thiocyanate ions was formed with the consecutive irreversible reactions, calculating the activation energy of the transport process with respect to membrane entrance and exit rates is not proper. Thus, maximum membrane entrance f l u x (J~ndax) and membrane exit permeate flux (J~aax) values including kl and k2 (Eq. (29)) were used for transport processes. Activation energy (Ea) values were obtained from maximum membrane exit flux (aJ~ax) versus 1/T plots given in Fig. 7. ln(J) = ln(a) - E - f fa ( 1 )
Therefore, values of the activation energies for the membrane entrance and exit steps are obtained from the slopes of the linear relationships: Ea,d=7.75 kcal/ mol and Ea,r=8.30 kcal/mol. Those calculated activation energies indicate that temperature has an influence on the transportation rate constants of thiocyanate ion. It may be assumed that the temperature effect is mainly exerted on the maximum membrane entrance and exit fluxes of the coupled-
M. Kobya et al./Journal of Membrane Science 130 (1997) 7-15
transport reactions taking place in the reaction zones of ionic interaction, adjacent to the interfaces. The values of the activation energy obtained for the two transport stages revealed that the first and second steps are most probably controlled by diffusion of both species.
a d do m
In this study, the kinetics of thiocyanate ion transport has been analyzed in the formalism of two, consecutive, irreversible first order reactions. The membrane entrance rate, kid and membrane exit rates, k2m and k2a, increased with temperature. With the increasing temperature, Rmax increases, tmax decreases, and thiocyanate ions are rapidly transported into the membrane and less cumulation appears in the membrane. For maximum membrane entrance and exit fluxes, J~dax and Jamax, the activation energies were calculated as 7.75 and 8.30 kcal/mol, respectively. The values of the apparent activation energy indicate that the process is diffusionally controlled.
5. List of symbols C Ccarrier
k2a k2m R
thiocyanate ion concentration (M) carrier concentration (M) activation energy (kcal/mol) membrane entrance rate constant (min -1) membrane exit rate constant (min -1) membrane entrance rate constant (min -t) membrane exit rate constant (min -1) membrane exit rate constant (min -t) reduced thiocyanate ion concentration
(--) Rma~ Rp
Rd at the end of reaction time (--) maximum reduced thiocyanate ion concentration of membrane phase (--) permeability degree (--) interface surface of phases (cm 2) time (min) maximum time (min) temperature (K)
acceptor phase donor phase initial thiocyanate ion concentration membrane phase
 R.D. Noble and J.D. Way, Liquid membrane technology, in R.D. Noble and J.D. Way (Eds.), Liquid Membranes: Theory and Applications, ACS Symp. Ser. No. 347, American Chemical Society, Washington, DC, 1987, p. 1.  L.L. Tavlarides, J.H. Bae and C.K. Lee, Solvent extraction, membranes, and ion exchange in hydrometallurgical dilute metals separation, Sep. Sci. Technol., 22 (1987) 581.  K. Kitagawa, Y. Nishikawa, J.W. Frankenfeld and N.N. Li, Wastewater treatment by liquid membrane process, Environ. Sci. Technol., l l (1977) 602.  K. Saito, K. Uezu, T. Hod, S. Furusaki, T. Sugo and J. Okamoto, Performance analysis of a fixed bed charged with capillary fiber form chelating resin for recovery of uranium from seawater, AIChE J., 34 (1988) 411.  D.T. Friesen, W.C. Babcock, D.J. Brose and A.R. Chambers, Recovery of citric acid from fermentation beer using supported liquid membranes, J. Membrane Sci., 56 (1991) 127.  O. Loiacono, E. Drioli and R. Molinari, Metal ion separation and concentration with supported liquid membranes, J. Membrane Sci., 28 (1986) 123.  W.J. Molnar, C.P. Wang, D.E Evans and E.L. Cusler, Liquid membranes for concentrating anions using a hydroxide flux, J. Membrane Sci., 4 (1978) 129.  M. Sugiura and T. Yamaguchi, Coupled transport of picrate anion through liquid membranes supported by a microporous polymer film, J. Colloid Interface Sci., 96 (1983) 454.  A.M. Neplenbroek, D. Bargeman and C.A. Smolders, Nitrate removal using supported liquid membranes: transport mechanism, J. Membrane Sci., 67 (1992) 107.  R.G. Luthy and S.G. Bruce Jr., Kinetics of reaction of cyanide and reduced sulfur species in aqueous solution, Environmental Sci. Technol., 13 (1979) 1481.  N.K. Thonchk, S.D. Jones, B.G. Reuben and P. Mahi, Extraction of thiocyanate ions from coal gasifications effluents, Chem. Eng. Res. Des., 66 (1988) 503.  B. Reuben, S. Jones and N. Kaur, Solvent extraction of thiocyanate ions with quarternary ammonium salts, Chemistry and Industry, 7 (1985) 14.  APHA, AWWA, WPCF, Standard methods for the examination of water and wastewater, 16th edn, Washington, DC, 1985.  M. Kobya, Transport kinetics of thiocyanate ions from aqueous environment with liquid membranes, Ph.D. thesis, Atattirk University, Erzurum, Turkey, 1996.
M. Kobya et al./Journal of Membrane Science 130 (1997) 7-15  M. Szpakowska and O.B. Nagy, Membrane material effect on copper coupled transport through liquid membranes, J. Membrane Sci., 64 (1991) 129.  T.M. Fyles, V.A. Malik-Diemer and D.M. Whitfield, Membrane transport systems II. Transport of alkali metal ions against their concentration gradients, Can. J. Chem., 59 (1981) 1734.
 A.A. Frost and R.G. Pearson, Kinetics and Mechanism, John Wiley, New York, NY, 1953, p. 166.  Z. Lazarova and L. Boyadzhiev, Kinetics aspects of copper(H) transport across liquid membrane containing LIX-860 as a carrier, J. Membrane Sci., 78 (1993) 239.