Stabilization of solvent impregnated resins (SIRs) by coating with water soluble polymers and chemical crosslinking

Reactive & Functional Polymers 59 (2004) 1–7 REACTIVE & FUNCTIONAL POLYMERS www.elsevier.com/locate/react Stabilization of solvent impregnated resins (SIRs) by coating with water soluble polymers and chemical crosslinking A.W. Trochimczuk a,1, N. Kabay b, M. Arda c, M. Streat a,* a Department of Chemical Engineering, Loughborough University, Loughborough, Leicestershire LE11 3TU, UK b Department of Chemical Engineering, Ege University, Izmir 35100, Turkey c Department of Chemistry, Ege University, Izmir 35100, Turkey Received April 2003; received in revised form 18 November 2003; accepted 24 December 2003 Abstract Solvent impregnated resins (SIRs) have been stabilized by the formation of a surface coating consisting of crosslinked poly(vinyl alcohol). Amberlite XAD-4 and vinyl sulphone respectively proved to be the most effective matrix and crosslinking agent for the protective layer in the preparation of SIRs containing Aliquat 336 reagent. The derived materials have been characterized in terms of Aliquat 336 content, amount of poly(vinyl alcohol), sulphur elemental content in the protective layer as well as operational stability. It has been found that the Aliquat 336 loading of impregnated XAD-4 is 0.82 mmol of extractant/g: this falls to 0.55 mmol of extractant/g after coating with a layer of vinyl sulphone crosslinked poly(vinyl alcohol). Ó 2004 Elsevier B.V. All rights reserved. Keywords: Solvent impregnated resin; Aliquat 336; Chemical crosslinking; Coating; Protection barriers; Extraction; Chromium (VI) 1. Introduction Warshawsky [1] and Grinstead [2] were the first to describe the synthesis and applications of solvent impregnated resins (SIRs) in 1971. Thereafter, SIRs containing various liquid ionic extractants, e.g. amines (such as Kelex 100 – 7-(4-ethyl-1-methy* Corresponding author. Tel.: +44-1509-222-506; fax: +441509-223-923. E-mail address: m.streat@lboro.ac.uk (M. Streat). 1 On leave from Department of Chemistry, Institute of Organic and Polymer Technology, Wroclaw University of Technology, 50-370 Wroclaw, Poland. loctyl)-8-quinolinol [3]), phosphoric, phosphinic, thiophosphinic acids and esters (for example 2ethylhexyl phosphoric acid [4,5] and bis[2,4,4trimethylpentyl] monothiophosphinic acid [6]) or solvating extractants (tributyl phosphate, trioctyl phosphine oxide, octyl(phenyl)-N,N-diisobutylcarbamoylmethylphosphine oxide [7] and more recently dimethyl dibutyl tetradecyl-1,3-malonamide [8]) have been synthesised and investigated for a range of potential applications. Stabilization of the extractant within the organic matrix, i.e. inside a highly crosslinked polymeric resin imparts an important and essential advantage to these materials rendering them physically stable and suitable 1381-5148/$ - see front matter Ó 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.reactfunctpolym.2003.12.011 2 A.W. Trochimczuk et al. / Reactive & Functional Polymers 59 (2004) 1–7 for use in conventional packed column mode. An extensive review by Warshawsky in 1981 discusses the preparation and potential applications of SIRs [9]. Interest in the commercial development of SIRs and their application for the removal of various metal ions from aqueous solutions was stimulated by the fact that they combine the advantageous features of both liquid/liquid extraction and ionexchange in a conventional polymeric adsorbent material. These features include, (i) the specificity and selectivity of readily available extractants, (ii) the straightforward mechanism of interaction of metal ions with liquid/liquid extractants, (iii) lack of third phase formation, (iv) the possibility of treating unclarified solutions and/or adopting a continuous separation process. A secondary reason is that synthesis of SIRs is relatively simple and versatile. It is possible to obtain a whole range of materials tailored for particular ion(s) in solution using various combinations of polymeric supports and liquid extractants. Unfortunately, the main disadvantage of SIRs is the loss of extractant due to solubility in the aqueous phase. Leakage of the extractant from the polymeric support leads to a steady loss of sorptive capacity towards targeted ions thereby rendering SIRs ineffective after several cycles of application. Moreover, this leakage is not acceptable from an environmental point of view as the leachate is likely to contain toxic and odorous compounds (amines, thiophosphinic acids and esters, etc.) that will contaminate effluents. This problem is thought to be the main reason why SIR technologies have not evolved into large-scale application. It is surprising that so little effort has been devoted to research aimed at increasing the stability of the SIRs [10,11]. The factors influencing the stabilization of SIRs have been discussed by Muraviev in a recent review [12]. Two methods are recommended, i.e. to carry out several ionexchange cycles or alternatively keeping freshly prepared SIRs in boiling water for several hours. Both techniques attempt to remove the portion of extractant that loosely adheres to the structure of polymeric support, i.e. located in large pores and easily removed due to lack of interaction with the polymer. Recently, two papers have appeared in the open literature devoted to improvement of SIR operational stability [13,14]. In both these articles, SIRs were stabilized by post-impregnation formation of a protective barrier. In the former paper, a mixture of glycidyl methacrylate and N,N-methylenebisacrylamide monomers was polymerized on the surface of the SIR, containing di(2-ethylhexyl) phosphoric acid. The double, vinylidene bonds, necessary for the chemical attachment of the protective layer, were generated on the polymer surface in a separate reaction of phosphonate ester with phenyllithium. The authors observed high stability of the coated resins: 96% of Cu(II) ions were removed in each of five cycles, whereas removal efficiency of the uncoated SIR dropped to 11% in three cycles. In the latter paper, a different approach was taken and referred to as post-impregnation encapsulation. This process included precipitation of a linear poly(sulphone) on the surface of SIRs. The method was simple and consisted of removing excess extractant from the external surface of beads and placing the SIR in a solution of poly(sulphone) in dimethylformamide (DMF). Subsequent rinsing of the polymeric beads with water, a non-solvent for poly(sulphone), caused precipitation of the protective layer of poly(sulphone) on the surface of the SIR. This method leads to physical sorption of a film-forming polymer, i.e. poly(sulphone) on the surface of the bead. The authors observed a positive effect of encapsulation on the SIR stability. Post-impregnation coated resins retained a much higher level of Zn(II) and Cd(II) removal when compared to resins treated by conventional methods [14]. Both these methods, although innovative and efficient, suffer from serious disadvantages. The method described in [13] is a multi-step process and is therefore time consuming. In addition, it requires use of the phenyllithium (an expensive reagent) and a low temperature lithiation. The other method involving post-impregnation encapsulation with poly(sulphone) is much simpler but requires the soaking of the freshly prepared SIR in a DMF solution. Dimethylformamide is a good solvent for the majority of extractants, thus causing an immediate loss of the extractant phase A.W. Trochimczuk et al. / Reactive & Functional Polymers 59 (2004) 1–7 to the DMF solution during the encapsulation procedure. The current work aims to circumvent both these problems (i.e. complexity of chemical modification and the necessity of using organic solvents) by using the precipitation of water soluble poly(vinyl alcohol) onto the solvent impregnated resins with the subsequent crosslinking of poly(vinyl alcohol) layer with vinyl sulphone. Aliquat 336 has been chosen as an extractant for the purpose of this study. This compound is used widely in liquid–liquid extraction processes and has already found some application in the preparation of solvent impregnated resins. For example, Aliquat 336 immobilized in Amberlite XAD-7 resin has been used in the separation of tetravalent actinides [15]. The same combination of extractant and a polymeric resin has also been used in the development of a separation method for platinum group metals [16]. It is known that Aliquat 336 is effective for the removal of Cr(VI) from industrial effluents [17– 20]. Also, there are reports of the use of Aliquat 336 immobilized in hollow fibers [21] or in macroporous resins [22] for chromate removal. However the stability of solvent impregnated materials was not investigated in either of these case studies. 3 (that may contain some ethyl styrene) and XAD-7 is a highly crosslinked methacrylate resin. 2.2. Impregnation procedure Typical procedure involves immersing 3 g of dry resin in 50 ml of 1 M solution of Aliquat 336 in hexane and shaking at ambient temperature for 17 h. The polymer beads were subsequently separated by filtration using a Buchner funnel and washed with distilled water. Impregnated beads were then air-dried for several hours and subsequently vacuum-dried at room temperature. Formation of the protective layer on the surface of impregnated polymers is achieved by immersing these polymers in a solution of poly(vinyl alcohol). In a typical procedure, 3 g of vacuum-dried impregnated beads are immersed in 50 ml of 3% wt/v PVA 98–99% hydrolysed and shaken for 17 h. After this time, 10 ml of 1 M potassium chloride solution was added and shaking continued for 24 h. The solution was removed using a water aspirator and the beads were re-suspended in 20 ml of 1 M sodium carbonate to which was added 2 ml of vinyl sulphone after waiting 1 h. The contents of the flask were shaken for 24 h and then the beads were separated on Buchner funnel and washed with a large excess of distilled water. 2. Experimental 2.1. Materials 2.3. Methods Aliquat 336, polymeric resins Amberlite XAD-4 and XAD-7, poly(vinyl alcohols) MW ¼ 31,000– 50,000 and 124,000–186,000 both having degree of hydrolysis 98–99% as well as crosslinking agents: vinyl sulphone and 1,4-butanediol diglycidyl ether were purchased from Aldrich. The polymeric resins were extracted with ethanol in a Soxhlet extractor for 8 h and subsequently dried. All other reagents, including potassium chromate (Fisher Chemicals) were used as obtained from the supplier. Amberlite XAD-4 and XAD-7 are commercial products manufactured by Rohm and Haas (Philadelphia) and the precise chemical compositions are not published. XAD-4 is a highly crosslinked styrene/divinyl benzene copolymer The efficiency of impregnation was measured by changes of the sample weight and by nitrogen elemental analysis, performed using a Buchi set of equipment for the Kjeldahl method. The sulfur content was measured using elemental analysis. The percentage of SIR in swollen polymer was determined by the centrifugation method. Thus, ca. 2 ml of swollen polymeric beads were centrifuged at 3000 rpm for 3 min in a fritted-glass bottom vial to remove excess surface water and weighed before and after vacuum drying. The percentage of SIR was then calculated as wd =ws , where ws is the weight of swollen sample after centrifugation and wd is the weight of sample after drying. 4 A.W. Trochimczuk et al. / Reactive & Functional Polymers 59 (2004) 1–7 3. Results and discussion In our approach, we have stabilized the SIR by modifying the post-impregnation technique by using a hydrophilic (polyvinyl alcohol) rather than a hydrophobic polymer (e.g. polysulphone). This produces an adsorptive layer of hydrophilic polymer on the external surface of freshly prepared SIR. During this procedure the solvent impregnated resin is only contacted with the aqueous solutions and this prevents extractant loss to the aqueous phase. The majority of extractants, including Aliquat 336, are insoluble or only marginally soluble in water. Table 1 shows the nitrogen content of the SIRs prepared by Aliquat 336 impregnation into Amberlite XAD-4 and XAD-7 resins. As can be seen, samples R4 and R7, i.e. Amberlite XAD-4 and XAD-7 impregnated with Aliquat 336, display nitrogen content of 0.63 and 0.97 mmol/g, respectively. These values correspond to ca. 254 and 392 mg of extractant per gram of impregnated material. It can be noted that the more polar polymeric matrix – Amberlite XAD-7 retains a greater amount of liquid extractant than relatively more hydrophobic Amberlite XAD-4. Secondly, a reagent capable of reacting with two hydroxyl groups is added to the aqueous suspension of SIR already coated with poly(vinyl alcohol) in order to ensure good physical stability of the protective layer and to gain some control over the permeability of the poly(vinyl alcohol) layer. This reagent, e.g. vinylsulphone, reacts rapidly and efficiently with poly(vinyl alcohol) at about pH 12 causing chemical crosslinking [23]. This renders the protective barrier on the SIR surface stronger and might even prevent further migration of the extractant from within the internal pore structure of the SIR towards the external aqueous phase. The resin coating scheme and crosslinking is presented in Fig. 1. However, both coating with PVA and crosslinking of the surface protective layer causes a decrease in nitrogen content of the derived material due to the net weight gain of surface protected SIR. Table 1 shows that the XAD-7 based SIR (sample B) possesses a nitrogen capacity of 0.76 mmol/g upon coating and crosslinking. It is 0.23 mmol/g or 24% less than uncoated resin R7. A more hydrophobic SIR (sample A) has a nitrogen content of 0.45 mmol/g. This is 0.18 mmol/g or 28% less than R4 (0.63 mmol/g). Nitrogen loss was more significant when PVA of higher molecular weight was used (samples C and E). This is attributed to the greater amount of 124–186 kD PVA adsorbed onto these SIRs. An additional experiment was carried out in order to check that the nitrogen capacity loss is caused by the adsorption of PVA and the associated net weight gain. In these experiments, samples R4 and R7 were boiled in distilled water (1 g resin in 100 ml water) for 5 h and subsequently separated by filtration and dried in vacuum at room temperature. The nitrogen content of these samples changed to 0.58 and 0.86 mmol/g, respectively. It can be seen that SIRs did not lose as much of their nitrogen capacity (partition of Aliquat 336 between resin and hot water) under these harsh conditions as was observed during PVA coating and crosslinking. This confirms that the lowering of nitrogen content in coated samples is a result of weight gain and not caused by migration of Aliquat 336 to the aqueous solutions of PVA Table 1 Characteristics of the obtained SIRs Symbol Polymeric matrix PVA molecular weight (kD) Crosslinker Nitrogen content (mmol/g) Sulfur content (mmol/g) A B C D E R4 R7 XAD-4 XAD-7 XAD-4 XAD-4 XAD-4 XAD-4 XAD-7 31–50 31–50 124–186 31–50 124–186 NA NA VS VS VS BDE BDE NA NA 0.45 0.76 0.37 0.55 0.51 0.63 0.97 2.12 1.89 1.95 NA NA NA NA A.W. Trochimczuk et al. / Reactive & Functional Polymers 59 (2004) 1–7 5 Fig. 1. Scheme of the preparation of coated solvent impregnated resins. and sodium carbonate. All three samples crosslinked with vinylsulphone have a similar sulphur content of
2 mmol/g of resin. The appearance and morphology of the external protective barrier was examined using optical microscopy and scanning electron microscopy. Fig. 2(a) and (b) shows that adsorption and subsequent chemical crosslinking of PVA on the SIR surface changes the morphology of the beads. Fig. 2(a) shows an uncoated SIR obtained from Amberlite XAD-4 and Aliquat 336. The beads are homogeneous with a smooth surface appearance. The surface of the same beads has a distinct ÔskinÕ as observed under the microscope after adsorption of PVA and chemical crosslinking with divinyl sulphone (Fig. 2(b)). The thickness of the surface protective layer can be estimated at 10–15 lm. Further examination under an electron scanning microscope reveals the structure of a bead crosssection. However, it must be noted that the beads used in the electron microscope experiments had been dried under vacuum and this treatment changes the hydration of the poly(vinyl alcohol) layer and consequently changes its thickness and appearance. Nevertheless, the cross-section illustrated in Fig. 3 shows that an external PVA polymer layer is present on the surface of SIR beads after adsorption and crosslinking. A preliminary screening of SIRs based on both types of polymeric resin support, i.e. hydrophobic and hydrophilic Amberlite XAD-4 and XAD-7, was carried out to select the most promising support phase. These experiments consisted of passing 500 and 750 bed volumes of 0.1 mM solution of HCl through SIRs columns and the results are shown in Table 2. The uncoated resins (R4 and R7) leached a greater amount of Aliquat 336 than Fig. 2. Optical microscope pictures of uncoated (a) and coated SIR (b) (magnification 100). the coated samples and the SIR based on more hydrophilic Amberlite XAD-7 was the least stable (sample B). This is attributed to the polar nature of XAD-7 and hence its ability to retain water, that in turn can causes greater leaching of extractant into the external aqueous solution. Coated resins are more stable regardless of the molecular weight of PVA and type of low molecular crosslinker used to form a protective surface 6 A.W. Trochimczuk et al. / Reactive & Functional Polymers 59 (2004) 1–7 Fig. 3. SEM picture of coated SIR cross-section. Table 2 Stability test results Resin Initial nitrogen (mmol/g) Nitrogen after 500 bed volumes (mmol/g) Nitrogen after 750 bed volumes (mmol/g) R4 R7 A B C D 0.63 0.97 0.45 0.76 0.37 0.55 0.50 0.60 0.40 0.54 0.32 0.46 0.48 0.52 0.40 0.43 0.31 0.46 79.9% 61.8% 88.2% 71.5% 87.2% 84.5% 75.6% 53.5% 89.1% 57.1% 85.0% 83.4% barrier. Resin A was chosen for all subsequent experiments. Another set of SIRs was studied to check their ability to retain the liquid extractant. These resins consisted of Amberlite XAD-4 and 31–50 kD PVA but were crosslinked using a variable amount of vinylsulphone. The properties of these materials are given in Table 3. The SIR with the highest amount of crosslinker (sample A) also contains the highest amount of sulphur 2.12 mmol/ Table 3 Characteristics of the SIRs obtained from XAD-4, 31–50 kD PVA and variable proportions of divinyl sulphone as crosslinking agent Symbol A A1 A2 Initial nitrogen (mmol/g) Initial sulphur (mmol/g) Nitrogen after four cycles mmol/g % 0.480 0.452 0.532 2.120 1.020 0.310 0.408 0.450 0.444 85.0 96.6 83.5 g. The SIR crosslinked with half the amount of VS used in the case of resin A (sample A1) has a S content of 1.02 mmol/g, whilst the lowest crosslinked SIR (sample A2) has a S content of only 0.31 mmol/g. It is concluded that sulphur content is roughly proportional to the amount of VS used to crosslink the adsorptive layer of PVA. Consequently, one can assume that all the VS reacted with the OH groups of PVA and did not migrate A.W. Trochimczuk et al. / Reactive & Functional Polymers 59 (2004) 1–7 to Aliquat 336 in the pores of SIRs. The operational stability of these resins is presented in the last column of Table 3 and shows that it is only the least crosslinked SIR (sample A2) that loses a significant amount of Aliquat 336, whereas both samples A and A1 are relatively stable over several cycles. 4. Conclusion We have shown that it is possible to produce relatively stable SIRs that contain Aliquat 336 as extractant by a post-impregnation coating and crosslinking procedure. This procedure involves adsorption of poly(vinyl alcohol) from an aqueous solution onto the surface of extractant-impregnated beads and crosslinking with vinyl sulfone. The method of stabilization of SIRs presented here is versatile since it is possible to change the molecular weight of PVA and the amount of vinylsulphone used for crosslinking of the protective PV barrier. A detailed study of coated SIRs used to remove chromate anions from aqueous solutions is discussed in the second and third paper in this series [24,25]. These papers discuss and compare the adsorption isotherms, kinetic parameters and column performance of coated SIRs for chromate ion removal from aqueous solutions. Acknowledgements This work was supported by an EPSRC Visiting Fellows Grants to A. Trochimczuk and N. Kabay. N. Kabay and M. Arda also acknowledge the British Council, Turkey for financial support. We would also like to thank Mr. A. Milne, Mr. D. Smith, and Dr. D.J. Malik for their kind assistances during the laboratory studies. We thank Mr. F. Page for the SEM image and Mr. M. Kerry for the microscope photos. References [1] A. Warshawsky, South African Patent Application 71/ 5637, 1971. 7 [2] R.R. Grinstead, Report by the Dow Chemical Co., on contract no 14-12-808 to the Water Quality Office of US Environmental Protection Administration, January 1971. [3] G. Cote, D. Bauer, S. Esteban, Solvent impregnating resins containing Kelex 100: aqueous solubility of Kelex 100 and distribution equilibria of germanium (IV), in: J.A. Greig (Ed.), Proceedings of IEXÕ96, Cambridge, UK, p. 363. [4] R.-S. Juang, S.-H. Lee, J. Chem. Tech. Biotechnol. 66 (1996) 153–159. [5] R.-S. Juang, M.-L. Chen, Sep. Sci. Technol. 32 (1997) 1017–1035. [6] N. Kabay, M. Demircioglu, H. Ekinci, M. Yuksel, M. Saglam, M. Streat, React. Funct. Polym. 38 (1998) 219– 226. [7] E.P. Horwitz, R. Chiarizia, M.L. Dietz, H. Diamond, D.M. Nelson, Anal. Chim. Acta 281 (1993) 361–372. [8] P.K. Mohapatra, S. Sriram, V.K. Manchanda, L.P. Badheka, Sep. Sci. Technol. 35 (2000) 39–55. [9] A. Warshawsky, in: J.A. Marinsky, Y. Marcus (Eds.), Ion Exchange and Solvent Extraction, 8, Marcel Dekker, New York, 1981, p. 229. [10] D.N. Muraviev, E. Hogfeldt, React. Polym. 8 (1988) 97– 102. [11] J.L. Cortina, A. Warshawsky, in: J.A. Marinsky, Y. Marcus (Eds.), Ion Exchange and Solvent Extraction, 13, Marcel Dekker, New York, 1997. [12] D. Muraviev, Solv. Extr. Ion Exch. 16 (1998) 381–457. [13] S.D. Alexandratos, K.P. Ripperger, Ind. Eng. Chem. Res. 37 (1998) 4756–4760. [14] D. Muraviev, L. Ganthous, M. Valiente, React. Funct. Polym. 38 (1998) 259–268. [15] E.P. Horwitz, M.L. Dietz, R. Chiarizia, H. Diamond, S.L. Maxwell III, M.R. Nelson, Anal. Chim. Acta 310 (1995) 63–78. [16] A. Makishima, M. Nakanishi, E. Nakamura, Anal. Chem. 73 (2001) 5240–5246. [17] M.I. Ortiz, B. Galan, A. Irabien, J. Membr. Sci. 118 (1996) 213–221. [18] E. Salazar, M.I. Ortiz, A.M. Urtiaga, J. Irabien, Ind. Eng. Chem. Res. 31 (1992) 1523–1529. [19] B. Galan, A.M. Urtiaga, A.I. Alonso, J.A. Irabien, Ind. Eng. Chem. Res. 33 (1994) 1765–1770. [20] M.I. Ortiz, M. Galan, A. Irabien, Ind. Eng. Chem. Res. 35 (1996) 1369–1377. [21] T. Vincent, E. Guibal, Ind. Eng. Chem. Res. 40 (2001) 1406–1411. [22] N. Kabay, M. Arda, B. Saha, M. Streat, React. Funct. Polym. 54 (2003) 103–115. [23] D.S. Pepper, Practical protein chromatography, in: A. Kenney, S. Fowell (Eds.), Methods in Molecular Biology, vol. 11, 1992, p. 173 (Chapter 11). [24] N. Kabay, M. Arda, A. Trochimczuk, M. Streat, React. Funct. Polym. 59 (2004) 9 (this issue, doi:10.1016/j.reactfunctpolym.2003.10.006). [25] N. Kabay, M. Arda, A. Trochimcuk, M. Streat, React. Funct. Polym. 59 (2004) 15 (this issue, doi:10.1016/j.reactfunctpolym.2003.12.012).