Application of pseudo-emulsion based hollow fiber strip dispersion (PEHFSD) for recovery of Cr(III) from alkaline solutions Francisco Jose Alguacil*, Manuel Alonso, Felix Antonio Lopez, Aurora Lopez-Delgado Centro Nacional de Investigaciones Metalurgicas (Agencia CSIC), Avda. Gregorio del Amo 8, 28040 Madrid, Spain. *E-mail corresponding author: fjalgua@cenim.csic.es Abstract The permeation of chromium (III) using PEHFSD technology and the quaternary ammonium salt TOMACl (trioctyl methylammonium chloride) dissolved in n-decane as mobile carrier has been investigated. Factors affecting chromium permeability, such as hydrodynamic conditions, carrier concentration in the organic phase, metal and NaOH concentrations in the feed phase, etc., have been analyzed. The value of the overall permeation coefficient obtained under standard experimental conditions was 4.0x10-4 cm s-1. Mass transfer modelling was performed and the aqueous and membrane resistances were estimated as 1066 and 1.7-17x103 s cm-1, respectively. Keywords: Chromium (III); TOMACl; Pseudo-emulsion based hollow fiber strip dispersion (PEHFSD) 1. Introduction Chromium (III) compounds are used in a variety of industries, specially as tanning agents in the leather industry, thus, leading to the appearance of the element in the corresponding industrial effluents. Though Cr(III) compounds are not as toxic as Cr(VI) compounds [1], the possibility of Cr(III) oxidation to Cr(VI) is a potential hazard which makes the removal and recovery of Cr(III) from effluents a primary target, because it reduces risk of pollution of environment and at the same time allowing to reuse the recovered chromium (III). Methods to remove or recover chromium (III) for effluents had been recently reviewed in the literature [2]. Among separations technology, liquid membranes processes combine extraction and stripping into one single step, and having non-equilibrium mass transfer characteristics, their separations performance are not restringed by the conditions of equilibrium which characterized other processes. Including in liquid membranes, emulsion liquid 2 membranes and supported liquid membranes had known different grade of interest, though i.e., in the case of supported liquid membranes its apparent instability has been the major disadvantage for scaling-up of the processes. Thus, a technology is required to provide all the advantages of supported liquid membranes but without stability problems, and hollow fiber modules have been used under various process configurations with their advantages: minimization of the required membrane area and high mass transfer due to a large area per unit volume, while incorporating the advantages of highly efficient emulsion liquid membranes . Named by various designations, i.e. pseudo-emulsion based hollow fiber strip dispersion (PEHFSD), recent references about the viability of this new technology can be found elsewhere [3-7]. In the present work, the transport of Cr(III) from alkaline aqueous solutions using PEHFSD with trioctyl methylammonium chloride (TOMACl) as mobile carrier is studied under various experimental conditions. A permeation model is also reported. 2. Experimental 2.1. Materials TOMACl (Fluka) was used as carrier for transport experiments in this investigation; n-decanol (Merck) and n-decane (Fluka) were used as modifier and diluent, respectively. Stock chromium (III) solution was prepared by dissolving the required amount of chromium (III) nitrate (Fluka) in deionized water. All other chemicals used in the present study were of AR grade. The organic solution was prepared by dissolving the appropriate volume of TOMACl in n-decanol and n-decane to obtain organic phases of different concentrations. The hollow fiber device used for the investigation was obtained from Hoechst Celanese (now Membrana): Liqui-Cel 8x28 cm PCG-259 contactor and 5PCS-1002 Liqui-Cel LLE. The main characteristics are given in Table 1. 2.2. PEHFSD experiments The hollow fiber strip dispersion process comprises a unique membrane module for extraction and stripping, one stirred tank for feed phase homogenization and one stirred tank for preparation of a pseudo-emulsion containing TOMACl+n-decanol+n-decane and H2SO4. Both phases are pumping to the module by two gear pumps capable of 3 variable flows. The organic phase wet the microporous wall of the fiber because of its hydrophobic nature. The interface was maintaining at the pore by applying a higher pressure to the feed phase than to the pseudo-emulsion phase. The differential pressure was always kept below the breakthrough pressure. In this experimental work, the pressure of the feed phase was 0.2 bar higher than in the pseudo-emulsion phase. The view of PEHFSD using a single contactor in recirculation mode is shown in Fig. 1. The operation was carried out by passing alkaline feed containing Cr(III) through the tube side and pseudo-emulsion through the shell side in countercurrent mode. The characteristics of the pseudo-emulsion should be such that it should have clear and fast phase separation when mixing is stopped. The recovery of chromium from pseudoemulsion can be accomplished, pseudo-emulsion breaks down after the mixing of the organic and receiving phases stopped, and strip and organic phases separate (near 1 minute) automatically. The volume of pseudo-emulsion phase used in the experimentation was 800 cm3 (400 cm3 of the organic solution + 400 cm3 of the H2SO4 solution), whereas 3000 cm3 of feed solution of the desired Cr(III) concentration was prepared by taking a suitable aliquot from the stock solution. At predetermined time, small aliquots of the aqueous streams were taken and analyzed for chromium concentration by AAS (Perkin Elmer 1100B spectrophotometer). Permeabilities values were calculated from equation (11), see below. 3. Results and discussion For extraction of Cr(III) through a PEHFSD containing TOMACl as a mobile carrier, the concentration profile across the hollow fiber membrane pores in the hollow fiber module is schematically shown in Fig. 2. The Cr(III) ions in alkaline media represented as Cr(OH)4- exchanged via an anion-exchange reaction with organic extractant TOMA+Cl- dissolved in n-decane [2]. The extraction reaction is expressed as: CrOH4aq TOMA Cl org TOMA CrOH4org Cl aq (1) The organic complex TOMA+Cr(OH)4- then diffuses through the pore of the membrane toward the membrane-pseudo-emulsion interface, where Cr(III) is stripped in 4 the pseudo-emulsion phase after coming into contact with H2SO4. This reaction is fast and instantaneous. This stripping reaction is expressed as: 2TOMA CrOH4org 4H 2SO 4aq TOMA 2 SO 24org Cr2 SO 4 3aq 8H 2 O (2) Thus, regenerated organic phases contained both trioctyl methylammonium chloride and trioctyl methylammonium sulphate. It was shown [2], that the last reagent efficiently extracts tetrahydroxo chromate (III) anions from alkaline solutions. In the present work, and for easiness and as a first approach, eq. (1) was only considered in further calculations. In any case, the regenerated carrier diffuses back to the feed through the pore of the hollow fiber, after which the process is repeated. The extraction constant can be described by the equation: TOMA CrOH Cl CrOH TOMA Cl 4 org K ext 4 aq aq (3) org The value of log Kext for the above was found to be 3.316±0.024 [8]. The distribution coefficient could be presented as: log DCr log K ext log TOMA Cl org log Cl aq (4) The model for the transport of Cr(III) in PEHFSD system operating in the recycling mode consists of equations describing: 1) the change of Cr(III) concentration in the feed and reception solutions when circulating through the hollow fiber module, and 2) the change in Cr(III) concentration in the feed and pseudo-emulsion tanks, where the aqueous solutions are continuously recirculated, based on the complete mixing supposition. Considering linear concentration gradient and absence of back-mixing, these equations are: 5 i) module mass balance for the feed solution: M M CrF CrF A M PCr CrM u F F Cr R t z VM in (5) ii) tank mass balance for the feed solution: dCr F Q F CrMF,z L CrMF,Z0 dt VF T (6) iii) module mass balance for the reception solution: M M Cr R Cr R A M PCr Cr M u R F Cr R t z VM out (7) iv) tank mass balance for the reception solution: Q dCr R M M R Cr R ,z 0 Cr R ,z L dt VR M (8) where PCr is the overall permeability coefficient, [Cr] is the chromium (III) concentration, L is the fiber length, Q is the flow rate, u is the linear velocity and V is the volume. The superscripts M and T refer to the membrane module and tank, respectively. When a sulphuric acid solution is used as the stripping agent, an instantaneous reaction is assumed to occur at the fiber outside, thus [Cr]RM= 0 and [Cr]RT= 0. Then, the solution to equations [5]-[8] leds to: Cr F, t 0 VF ln CrF 2 PCr L t Q F 1 exp u r F i (9) 6 and experimental results can be fitted to a first-order kinetic law: Cr TF, t 0 VF ln Cr F B t (10) where B is a coefficient dependent on the geometry of the fibers and the module, linear velocity of the solutions, and the overall permeability of the system. This overall permeability coefficient can be estimated from the experimental value of the slope B, as: PCr u F ri 2L B ln 1 QF (11) for the system operated in the recycling mode. The design of hollow fiber modules for the separation-concentration of chromium using the overall permeability coefficient centers on three mass transfer resistances which are due to: i) the liquid flowing through the hollow fiber tube, ii) the chromium-species diffusion across the liquid membrane immobilized on the microporous wall of the fiber, and iii) the aqueous interface created on the outside of the fiber. The reciprocal of the overall permeability coefficient is: r r 1 1 1 1 i i PCr k i rlm PM ro k o (12) where ri, ro and rlm are the inner, outer radii of the hollow fiber and the hollow fiber logmean radius, respectively; ki and ko are the interfacial coefficients corresponding to the inner and outer aqueous boundary layers. PM is the membrane permeability, which is related to the distribution coefficient of chromium (DCr) with TOMACl by: 7 PM D Cr k m k m K ext TOMA Cl Cl 1 aq org (13) where DCr is defined by: D Cr Crorg Craq (14) where [Cr]org and [Cr]aq are the total analytical concentrations of chromium in the organic and aqueous feed phases. Inserting equation (13) in equation (12) gives: Cl aq r 1 1 i PCr k i rlm k m K ext TOMA Cl org ri 1 ro k o (15) When the reaction is instantaneous on the reception side, the contribution of the outer aqueous interface is removed from equation (15) and the overall permeability coefficient is given by: Cl aq r 1 1 i PCr k i rlm k m K ext TOMA Cl (16) org 3.1. Influence of the feed flow in the feed and pseudo-emulsion phases on chromium transport A series of experiments was conducted to establish adequate hydrodynamic conditions. The permeability coefficient was studied as a function of the flow in the feed solution, keeping the flow of the pseudo-emulsion phase constant at 100 cm3 min-1. Results obtained are shown in Table 1. PCrx104 cm s-1 increased from 0.9 to 4.0 for an increase in feed flow from 50 to 200 cm3 min-1 and decreased with further increase in feed flow. The increase of the permeability coefficient with feed flow was caused by a decrease in the thickness of the aqueous feed boundary layer, whereas the decrease of 8 PCr value could be lower residence time at higher flow rates, which provides insufficient time to complex Cr(III) with TOMACl. The dependence of the permeability coefficient on pseudo-emulsion flow was also investigated. In this case, PCr was not found to be dependent on this variable (50-150 cm3 min-1). 3.2. Influence of pseudo-emulsion phase composition on permeability of Cr(III) 3.2.1 Influence of sulphuric acid concentration and third phase formation Sulphuric acid solutions are good strippants for Cr(III)-loaded in TOMACl+ndecanol+n-decane organic phases, thus the effect of H2SO4 concentration in the stripping phase on permeability of Cr(III) has been studied in the range 0.5-2 M H2SO4. From the results presented in Table 3 it can be deduced that, under the experimental conditions, the concentration of H2SO4 has little influence on Cr(III) permeation coefficient, whereas the percentage of chromium (III) stripped also remained near constant with the increase of sulphuric acid concentration in the stripping phase. On the other hand, it was observed that when the organic phase contained no modifier, third phase formation is observed in the pseudo-emulsion tank when phases are separated. This is an undesiderable situation will void the practical use of a given system, thus a modifier (such as n-decanol) was added to the organic phase until third phase disappeared. The results of this investigation are resumed in Table 4. 3.2.2. Influence of the diluent The membrane diluent or organic phase diluent chosen as a water-resistant barrier in any liquid membrane process may influence the membrane performance. The diluent should retain the carrier to the maximum extent and yet at the same time retain a relatively high concentration of water to aid the transfer of hydrated species without loss of carrier to the aqueous phase. Liquid membrane performance is primarily dependent of intrinsic membrane diluent properties. In Table 5, the effect of the diluent on Cr(III) transport is given and n-decane and n-heptane (aliphatic diluents) are the ones giving the best permeability values compared with aromatic diluents such as toluene and cumene. 9 3.2.3. Influence of carrier concentration To analyze the effect of the carrier concentration in the organic phase, experiments were performed at 0.01 g L-1 Cr(III) and 0.1 M NaOH in the feed phase. The results (Table 6) show that permeability increases with increasing extractant concentration up to 10 % v/v, and the transport rate is therefore limited by diffusion through the aqueous film on the feed side of the membrane in this region. The decreasing in PCr, at higher carrier concentration, can be attributed to an increase in organic phase viscosity, and thus, increasing the membrane resistance to transport. 3.3. Influence of feed phase composition on permeability of Cr(III) 3.3.1. Influence of metal concentration The influence of the metal concentration on the permeation coefficient was also investigated and the results are presented in Table 7. It can be seen that chromium transport through the hollow fibers is increased as the initial metal concentration in the feed solution decreases. Thus, in this range of metal concentrations, the permeation process should be controlled by diffusion of metal species. 3.3.2. Influence of NaOH concentration To analyse the influence of the NaOH concentration in the feed phase, a set of experiments was carried out at various NaOH concentrations, keeping other variables constant. The results are shown in Table 8. It can be seen that Cr(III) permeability increases when the NaOH concentration in the feed phase is decreased up to 0.1 M, and then remained constant. This trend is in accordance with solvent extraction investigations, in which the chromium (III) distribution ratio increases as the NaOH in the aqueous solution decreases (from 0.5 to 0.1 M) [8]. 3.4. Evaluation of mass transfer coefficients and diffusional parameters Equation (16) was used in order to evaluate aqueous mass transfer coefficient (ki), membrane mass transfer coefficient (km) and diffusivity. By plotting 1/PCr as a function of [Cl-]aq/Kext[TOMA+Cl-]org for different extractant concentrations from alkaline media (0.1 M NaOH), one should obtain a straight line with slope ri/rlmkm and an ordinate to 10 calculate 1/ki. The results of this plot gives a= 1066+7.5·107b (r2= 0.989) and the values of ki and km, are thus 9.3x10-4 and 1.2x10-8 cm s-1, respectively. The first term of equation (16) represents the local value of the total resistance. This resistance is in turn the sum of the mass transfer resistances inside the fiber and across the fiber wall. It had been previously shown the corresponding values of ki and km. Therefore, on the right-hand side of equation (16), the first terms is the reciprocal of ki, that is 1066 s cm-1, whereas the second terms of such equation ranges from 1.7x103 to 17x103 s/cm. The overall resistance or 1/PCr in the experiments, calculated from equation (11), was observed to be 2.5x103 to 18x103 s cm-1, as compared to an overall resistance value of 2.7x103 to 18x103 s cm-1 estimated from equation (16) (details in Table 9), which shows that the resistance due to the membrane become dominant as the concentration of the carrier in the organic decreases. In this respect, the fractional resistance of each step of the overall process Ri0 and Rm0, can be calculated as: R i0 Ri Ri Rm (17) R 0m Rm Ri Rm (18) where Ri and Rm are the mass transfer resistances due to feed and membrane. As can be seen from Table 9, and under the present experimental conditions, the values of %Ri0 and %Rm0 were 5.8-38.2 and 61.8-94.2. This clearly indicated that the ratecontrolling step was membrane diffusion specially at lower extractant concentrations, though the contribution of aqueous feed resistance become more important at higher carrier concentration. The calculated value of the effective diffusion coefficient, defined as [9]: D eff k m d org was 1.1x10-10 cm2 s-1. 4. Conclusions (19) 11 The PEHFSD technology was found to be a feasible process for the simultaneous separation and concentration of Cr(III) from alkaline solutions using trioctyl methylammonium chloride as mobile carrier. By the use of the technology, saturatiomn of the carriers does not occur, as it is continuously regenerated in the pseudo-emulsion, and gives better performance. For the system investigated in this study, metal permeation can be described by a time independent overall permeation coefficient, with the aqueous and membrane mass transfer coefficients calculated as 9.3x10-4 and 1.2x10-8 cm s-1, respectively; whereas the respective mass transfer resistances were estimated as 1066 and 1.7-17x103 s cm-1. Acknowledgement To Agencia CSIC (Spain) for support. References [1] M. Casadevall, A.Kortenkamp, in: B. Sarkar (Ed.), Heavy Metals and the Environment, Marcel Dekker, New York, 2002, p. 271. [2] B. Wionczyk, W. Apostoluk, W.A. Charewicz, Hydrometallurgy 82 (2006) 83. [3] A. Urtiaga, M.J. Abellan, J.A. Irabien, I. Ortiz, J. Membr. Sci. 257 (2005) 161. [4] D. He, X. Luo, C. Yang, M. Ma, Y. Wan, Desalination 194 (2006) 40. [5] J.V. Sonawane, A. Kumar Pabby, A.M. Sastre, J. Membr. Sci. 300 (2007) 147. [6] F.J. Alguacil, M. Alonso, F.A. Lopez, A. Lopez-Delgado, Chemosphere 72 (2008) 684. [7] A. Kumar Pabby, S.C. Roy, J.V. Sonawane, F.J. Alguacil, A.M. Sastre, in: A.K. Pabby, S.S.H. Rizvi, A.M. Sastre (Eds.), Handbook of Membrana Separation, CRC Press, Boca Raton, 2008 p. 1057. [8] B. Wionczyk, W. Apostoluk, Hydrometallurgy 78 (2005) 116. [9] A.M. Urtiaga, I. Ortiz, E. Salazar, J.A. Irabien, Ind. Eng. Chem. Res. 31 (1992) 877. 12 Table 1 Characteristics of the unit used in the investigation nº fibers 10000 polymeric material polypropylene active interfacial area 1.4 m2 fiber i.d. (di) 24x10-3 cm fiber o.d. (do) 30x10-3 cm fiber thickness (dorg) 3x10-3 cm fiber length (L) 15 cm porosity (ε) 30 % pore size 3x10-6 cm tortuosity (τ) 3 Table 2 Influence of feed flow on permeability of Cr(III) Feed flow (cm3 min-1) PCrx104 (cm s-1) 50 0.9 100 1.9 200 4.0 300 3.4 400 2.8 Feed phase: 0.01 g L-1 Cr(III) and 0.1 M NaOH. Pseudo-emulsion phase: 10 % v/v TOMACl and 5 % v/v n-decanol in n-decane + 0.5 M H2SO4. 13 Table 3 Effect of H2SO4 concentration on permeability and % stripping of Cr(III) H2SO4 (M) PCrx104 (cm s-1) % Stripping (after 1 h) 0.5 4.0 55.1 2 3.8 53.4 Feed phase: 0.01 g L-1 Cr(III) and 0.1 M NaOH. Pseudo-emulsion phase: 10 % v/v TOMACl and 5 % v/v n-decanol in n-decane + H2SO4 solution. Feed flow: 200 cm3 min-1. Pseudo-emulsion flow: 100 cm3 min-1. Table 4 Effect of modifier on permeability of Cr(III) and performance of the system n-decanol (% v/v) PCrx104 (cm s-1) Remarks - n.d. Third phase formation. Unsuitable for use 2.5 n.d. Third phase formation. Unsuitable for use 5 4.0 Two phases. Suitable for use Feed phase: 0.01 g L-1 Cr(III) and 0.1 M NaOH. Pseudo-emulsion phase: 10 % v/v TOMACl and modifier in n-decane + 0.5 M H2SO4. Flows as in Table 3. 14 Table 5 Influence of the diluent of the organic phase on Cr(III) transport Diluent PCrx104 (cm s-1) n-heptane 3.4 n-decane 4.0 cumene 1.9 toluene 1.7 Feed phase: 0.01 g L-1 Cr(III) and 0.1 M NaOH. Pseudo-emulsion phase: 10 % v/v TOMACl and 5 % v/v n-decanol in decane + 0.5 M H2SO4. Flows as in Table 3. Table 6 Chromium (III) permeabilities at various TOMACl concentrations TOMACl (% v/v) PCrx104 (cm s-1) 1 0.56 2.5 1.1 5 2.5 10 4.0 20 0.80 Pseudo-emulsion phase: TOMACl and 5 % v/v n-decanol in n-decane + 0.5 M H2SO4. Flows as in Table 3. Table 7 Influence of metal concentration on Cr(III) transport [Cr(III]0 (g L-1) PCrx104 (cm s-1) 0.01 4.0 0.05 2.3 0.1 1.6 0.5 0.83 Feed phase: Cr(III) and 0.1 M NaOH. Others experimental conditions as in Table 3. 15 Table 8 Influence of NaOH concentration in the feed phase on chromium transport [NaOH] (M) PCrx104 (cm s-1) 0.5 0.61 0.25 1.7 0.1 4.0 0.05 4.0 Feed phase: 0.01 g L-1 Cr(III) and NaOH. Others experimental conditions as in Table 3. Table 9 Mass transfer resistance due to feed and membrane and contribution of the respective fractional resistances 1/PCr 1/ki ri[Cl-]aq/rlmkmKext[TOMACl]org 1/PCr (s cm-1) (s cm-1) (s cm-1) (s cm-1) exp. 18000 %Ri0 %Rm0 Eq. (16) 1066 17250 18316 5.8 94.2 9000 6825 7891 13.5 86.5 4000 3450 4516 23.6 76.4 2500 1725 2791 38.2 61.8 16 Fig. 1. View of PEHFSD operated in recycling mode for recovery of Cr(III) from aqueous alkaline media. (1) Hollow fiber membrane contactor. (2) Aqueous feed solution into tube side. (3) Pseudo-emulsion phase into shell side. Fig. 2. Schematic view of transport mechanism of Cr(III) from alkaline media with TOMACl by PEHFSD with single hollow fiber module. 17 1 2 3 Figure 1 18 Figure 2