New elastomeric polymethylsiloxane membranes bearing cationic exchanging sites for anionic dyestuffs sensors

European Polymer Journal 56 (2014) 140–158

Contents lists available at ScienceDirect

European Polymer Journal journal homepage: www.elsevier.com/locate/europolj

New elastomeric polymethylsiloxane membranes bearing cationic exchanging sites for anionic dyestuffs sensors Hassen Touzi a, Yves Chevalier b,⇑, Rafik Kalfat c, Nicole Jaffrezic-Renault d a

Laboratoire des Interfaces et Matériaux Avancés, Faculté des Sciences de Monastir, Université de Monastir, 5000 Monastir, Tunisia Laboratoire d’Automatique et de Génie des Procédés, UMR 5007 CNRS, Université Lyon 1, 43 bd 11 Novembre, 69622 Villeurbanne, France c Laboratoire des Méthodes et Techniques d’Analyse, INRAP, BioTechpôle Sidi-Thabet, 2020 Sidi-Thabet, Tunisia d Institut des Sciences Analytiques, UMR CNRS 5180, Université Lyon 1, 43 bd 11 Novembre, 69622 Villeurbanne, France b

a r t i c l e

i n f o

Article history: Received 25 July 2013 Received in revised form 9 April 2014 Accepted 14 April 2014 Available online 26 April 2014 Keywords: Anionic dyestuffs Polymethylsiloxane Anion-exchange Impedance spectroscopy

a b s t r a c t Novel anion exchanging materials based on functional polymethylsiloxane elastomers have been prepared and used as sensitive membranes of electrochemical sensors made of SiO2/Si heterostructures coated with thin films of functional polymer. Their synthesis and their use in electrolyte/insulator/semiconductor (EIS) electrochemical devices are reported. The anion exchanging materials are cationic cross-linked polymers based on either poly(methylhydrosiloxane) (PMHS) or poly(methylhydrosiloxane-co-dimethylsiloxane) copolymer (PMHS-co-PDMS 50/50). Their synthesis has been carried out in two stages. A brominated derivative was firstly prepared by hydrosilylation reaction of undecenyl bromide with SiAH bonds of various polymethylsiloxane in the presence of Karstedt’s catalyst. The kinetics of the hydrosilylation reaction was investigated. The second stage is based on the formation of quaternary ammonium (A+N(CH2ACH3)3), pyridinium (A+NC5H5) and phosphonium (A+P(C6H5)3) groups by quaternarization reactions to brominated polymers. Full characterizations of the materials by IR, liquid 1H and 13C NMR and solid 29Si NMR spectroscopy are given. The sensitivity of EIS devices to different anionic dyestuffs species have been assessed for Acid Blue 25 (AB25), Acid Blue 74 (AB74) and Acid Yellow 99 (AY99) in aqueous solution. The shifts of flat band potential and the variations of capacitance in inversion mode were extracted from the impedance measurements as a function of the concentration of anionic dyestuffs species in solution. The electrical parameters of the EIS devices came from a non-specific adsorption of organic dyes and anion exchange phenomena that showed specificity with respect to the type of anion exchanging group. The membranes bearing pyridinium or phosphonium groups gave a Nernstian response towards anionic dyestuffs AB25 and AY99. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Residual dyestuffs are present in many industrial effluents, such those of textile and leather tanning; their toxic character is a major concern regarding human health and environment. Their removal from wastewaters rely on ⇑ Corresponding author. Tel.: +33 472431877. E-mail address: [email protected] (Y. Chevalier). http://dx.doi.org/10.1016/j.eurpolymj.2014.04.010 0014-3057/Ó 2014 Elsevier Ltd. All rights reserved.

several techniques and processes such as chemical precipitation [1,2], photodegradation [3,4], electrochemical degradation [5] and ion exchange [6,7]. Such processes are necessarily associated with appropriate analyses of the residual dyestuff concentration before and after the purification process of wastewaters. The measurement of dyestuffs concentrations is generally performed in an analytical laboratory by means of various methods such as capillary electrophoresis in aqueous and non-aqueous

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

media [8–10], thin-layer chromatography [11], HPLC [12–14], UV [15–17], fluorescence [18], mass spectrometry [19,20] and electrochemical detection [21–23]. It would be advantageous to replace such methods requiring sample collection and transfer to the analysis laboratory by in line measurements using chemical sensors. The development of chemical sensors sensitive to anionic dyestuff requires the design of sensitive membranes based on anion exchange materials. The same anion exchange materials may also be useful for the wastewater treatment by ion exchange. The present work is aimed at the design of such materials and their evaluation for the determination of anionic dyestuff concentration using electrochemical measurements. Ion exchange is the microscopic phenomenon involved in most analytical devices aimed at measuring the concentration of ionic species in solution [24–26]. Ion exchanging materials are required for that purpose, either as the membrane separating two compartments containing the analyte and a reference solution in conventional ionsensitive electrodes (ISE), or as thin films deposited at the surface of a transducer in chemical sensor applications [27,28]. In the later case, the ion exchange phenomenon taking place at the close vicinity of the transducer surface gives rise to an electrical or optical signal that is detected, amplified and converted into the concentration of analyte. In most instances, the electrochemical methods rely on measurements of the variation of either surface potential or electrical conductivity [29,30]. Such principles have been applied in our recent works dealing with the chemical sensing of anions by means of anion exchange inside a polymeric membrane (thin film) deposited at the surface of Si/SiO2 electrodes [31,32]. In particular, measurements of the impedance of an electrolyte/insulator/semiconductor (EIS) heterostructure as a function of applied polarization provided two parameters that appeared quite sensitive to anion concentrations in the solution. Firstly, the capacitance of the polymer coating where the anion exchange was taking place was depending on several parameters such as the dielectric constant and thickness of the membrane. This was depending on ion exchange because the type of anion inside the membrane influences both the polarizability of the material and its swelling by water. Secondly, the variation of the flat band potential with respect to the anion concentration was identical to the variation of the electrical potential drop inside the polymer membrane and in the ionic double layer at the membrane–solution interface [33]. Measurements of the flat band potential in such heterostructures incorporating semiconducting silicon are an easy way to investigate the surface potential in the presence of an ionic double layer. This type of impedance measurement itself can be used for the design of chemical sensors since it is sensitive to the concentration and type of anion [34]. Ion-sensitive field-effect transistor (ISFET) is an alternative technology which is working on the basis of similar principles (variation of surface potential) [35]. ISFETs can be miniaturized as robust devices used for in line measurements; their manufacture involves a rather complex technology however. Impedance measurements are readily available tools for investigating and evaluating different ion

141

exchanging membranes that could be used for the sensitization of ISFET’s in a subsequently development stage. Lastly, investigations of ion exchange phenomena, especially their specificity, are also of prime importance at the academic level because of their implication in so many fields of technological applications and life sciences [36]. Specific effects in ion exchange phenomena are often discussed with regards to the Hofmeister lyotropic series [37]. The Hofmeister series offers an interesting framework as this empirical rule appears of quite a general bearing [37–39]. Such a rule has been applied in particular to the case of conventional membrane-based ion-sensitive electrodes [40,41] because the underlying electrochemical phenomena involve ion transfer into the membrane, ion adsorption to the membrane/electrolyte surfaces and ion accumulation or depletion inside the electrical diffuse double layer. However the Hofmeister series does not provide a universal rationale for ion-sensitive electrodes since the reverse Hofmeister effect (anti-Hofmeister selectivity) has been observed in several instances [42,43]. Whether direct or reverse Hofmeister series are followed, the different ions sort in the same order. This issue has been addressed by looking carefully to the several elementary physicochemical phenomena included in the global framework known as Hofmeister series [44]. The thin films used as sensitive layers must be stable and strongly immobilized at the surface of the transducer. Polysiloxane materials are suitable candidates because of their several interesting properties, e.g. excellent heat resistance, low toxicity, biocompatibility, poor wettability, low surface tension, outstanding electrical isolating properties [45,46]. Organic functionalization of poly(methylhydrosiloxane) (PMHS) allows a fine tuning of physical and chemical properties of the resulting polysiloxanes [47]. Appropriate substitution on the polysiloxane backbone affords various materials such as cross-linked materials [48] conductive [49] and electroluminescent [50] polymers, non-linear optical materials [51], liquid crystalline ionomers [52], polymeric surfactants [53], and drug delivery systems [54]. The functionalization of linear polysiloxane is often performed by means of hydrosilylation and dehydrocoupling of SiAH bonds of a precursor poly(methylhydrosiloxane) [55,56] using platinum catalyst (especially Karstedt’s catalyst [57,58]). The present work deals with the elaboration of elastomeric polymethylsiloxane membranes bearing cationic functionalities working as anion exchanging groups. It is required that the cationic polymethylsiloxane is not soluble in water and adheres at the surface of the transducer made of silica. The elaboration of such a membrane consists in two stages: firstly, the synthesis of cationic elastomers; and secondly, the deposition of a thin film. The elaboration of a cationic silicone polymer starts from poly(methylhydrosiloxane) (PMHS) or the poly(methylhydrosiloxane)-co-poly(dimethylsiloxane) copolymer (PMHS-co-PDMS 50/50). The synthesis has been designed such that the deposited film of the non water-soluble polymer could be chemically grafted on silica and cross-linked in situ at the surface of the transducer. The synthesis and characterization of functional polymers bearing either quaternary ammonium, pyridinium of phosphonium

142

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

groups is reported in a first part. The elaboration of the different membranes and measurements of the electrical response of the devices coated with these membranes towards Acid Blue 25 (AB25), Acid Blue 74 (AB74) and Acid Yellow 99 (AY99) anionic dyestuffs are reported in a second part.

2.2. Methods Infrared spectra from KBr pellets were obtained using a Bio-Rad 6000 FTS spectrometer. Absorbance was recorded in the wave number range 400–4000 cm1. Liquid state 1H and 13C NMR spectra were recorded in CDCl3 and DMSO-d6 (d in ppm from TMS) with a Bruker Advance 500 spectrometer. 29 Si CP-MAS NMR spectra were recorded at 99.35 MHz on Bruker ASX 500 spectrometer operating in a static field of 11.7 T. The spinning frequency, recycle delay and contact time were 6 kHz, 5 s and 5 ms respectively for all measurements. Chemical shifts were referenced to the external tetramethylsilane (TMS). The a.c. capacitance measurements on functionalized Si/SiO2/gel-membrane hetero-structures were performed in an electrochemical cell using a three electrodes potentiostatic set up. A saturated calomel electrode (SCE) and platinum wafer were used as a reference and counter electrodes respectively. Electrical measurements were assured by impedance spectrometer (VoltaLab 40 Radiometer Analytical) working under an alternating tension of amplitude 10 mV and 10 kHz frequency [59].

2. Experimental Syntheses of all functionalized polysiloxanes were carried out under argon flow so as to avoid moisture and oxygen from the air and to avoid hydrolyses and polycondensation reactions of hydrosilane groups of poly(hydromethylsiloxane) PMHS and hydromethylsiloxane–dimethylsiloxane copolymer PMHS-co-PDMS 50/50.

2.1. Reagents and materials Undecenyl bromide 95%, platinum(0)-1,3-divinyl1,1,3,3-tetramethyldisiloxane complex (PDTD) 3% in xylene (Karstedt’s catalyst) and hexadiene, were purchased from Sigma–Aldrich. The three different commercially available textile dyestuffs Acid Blue 25 (AB25), Acid Blue 74 (AB74) and Acid Yellow 99 (AY99) were supplied by Sigma–Aldrich. The characteristics and chemical structures of these dyestuffs are given in Table 1 and Fig. 1a respectively. The following polymers purchased from ABCR were used in the preparation of anion exchange membranes: poly(hydromethylsiloxane) (PMHS, HMS-991) of number-average molar mass 2262 g mol1, hydromethylsiloxane–dimethylsiloxane copolymer containing 50% of SiAH and 50% of SiACH3 units (PMHS-co-PDMS 50/50, HMS-501) (M = 2544 g mol1) (Fig. 1b), and trimethylsilyl-terminated polydimethylsiloxane (PDMS, AB112154) (M = 2000 g mol1, viscosity = 20 cSt) as a reference non-functional silicone material. All other reagents and solvents, dimethylformamide (DMF), chloroform, triethylamine, pyridine, triphenylphosphine and sodium acetate 99% NormapurÒ were supplied by Prolabo. Silica gel Kieselgel 60 PF254 (particle size 0.63–0.2 mm) was purchased from Merck. The silicon/silica substrates Si/SiO2 were supplied by LAAS-CNRS (Toulouse). It consists of p-type silicon wafers with crystal orientation h1 0 0i. A thin silica layer was generated on the surface of silicon plates by thermal oxidation. The ohmic contact was realized by deposition of an aluminum film on backside of the wafers. Reactive silanol sites were generated at the silica surface by reaction with concentrated sulfochromic solution and thorough rinsing with bidistilled water. Final drying at 110 °C during 2 h removed the adsorbed water molecules.

2.3. Hydrosilylation of PMHS polymer and PMHS-co-PDMS 50/50 copolymer PMHS polymer contains 35 hydrosilane groups SiAH which is considered as an average degree of polymerization. A part of the SiAH groups was used for coupling anion exchanging groups, and the remaining part was kept nonreacted for further cross-linking of polymeric chains and grafting to silica surface of silicon/silica substrates. The same reactions have been carried out on pyrogenic silica gel as a model silica surface allowing easier spectroscopic analyses. The cross-linking of the various polymers chains was carried out by hexadiene cross-linking agent. The full three-dimensional network was grafted on the surfaces of the pyrogenic silica gel and silicon/silica substrates by the formation of siloxane bridges SiAOASi just after cross-linking. The hydrosilylation reaction of vinyl compounds by PMHS using transition metal complexes as catalyst has been described in several works [60]. In the present work, only 10/35 units of SiAH groups were used in the hydrosilylation reaction of PMHS with undecenyl bromide. 3 g of PMHS (46.4 mmol of SiAH) were reacted with 3 g of undecenyl bromide (12.9 mmol of AC@CH2). 150 lL of Karstedt’s catalyst (PDTD: platinum(0)-1,3-divinyl1,1,3,3-tetramethyldisiloxane) was added under strong stirring and the reaction was performed in bulk at 60 °C during 4 h. The increase of the amount of catalyst and

Table 1 Characteristics of the selected dyestuffs. Dyestuffs (C.I. Name)

Abbreviation

kmax (nm)

Molar mass (g mol1)

Purity (%)

Acid Blue 25 Acid Yellow 99 Acid Blue 74

AB25 AY99 AB74

600 450 608

416.39 496.35 466.36

45 40 98

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

143

Fig. 1. Chemical structures of the selected dyestuffs (a); poly(methylhydrosiloxane) and poly(methylhydrosiloxane)-co-poly(dimethylsiloxane) copolymer (b).

the heating time had no effect on the conversion. The resulting polymer PMHSBr was recovered as a viscous liquid after evaporation of the solvent under reduced pressure. Fig. 2a shows the chemical reaction corresponding to hydrosilylation reaction of PMHS. 1 H and 13C NMR (in CDCl3, d in ppm from TMS as a reference) and IR (KBr, cm1) spectra of the mixture of PMHS and undecenyl bromide before hydrosilylation reaction showed the following peaks: 1H NMR: 0.12 (u, Si(CH3)3, s, 1.8 H), 0.20 [w, (OASiH(CH3)AO), d(3JHAH = 1 Hz), 10.5 H], 1.3 (d-i, m, 12 H), 1.85 (j, m, 2 H), 2.05 (c, m, 2 H), 3.4 (k, t(3JHAH = 6.8 Hz, 2 H), 4.7 (l, SiH, s, 3.5 H), 4.95 (a, m, 2 H) and 5.8 (b, m, 1 H). 13C NMR: 0.61 [u, (OASi(CH3)AO], 1.34 (w, SiA(CH3)3, 27.9 (i), 28.5 (h), 28.65 (d), 28.82 (c), 29.12 (e–g), 32.57 (k), 33.52 (j), 113.9 (a) and 138.79 (b). IR: 646 (CABr), 768–892 (SiAC), 915 (CH@CH2), 1045– 1100 (SiAOASi), 1260 (SiACH3), 1408 (SiACH3), 1637 (C@C), 2167 (SiAH), 2965 (CAH) and 3074 (@CH2). 1 H and 13C NMR (in CDCl3, d in ppm from TMS as a reference) and IR (KBr, cm1) spectra of the functional polymer PMHSBr after hydrosilylation reaction showed the following peaks: 1H NMR: 0.12 (Si(CH3)3, s, 1.8 H), 0.17 (OASiR(CH3)AO, s, 7.5 H), 0.20 [(OASiH(CH3)AO, d, 3 H)],

0.55 (a1, m, 1.65 H), 1.3 (c1-i, m, 14 H), 1.85 (j, m), 2 (b1, m), 3.4 (k, t, 2 H) and 4.7 (l, SiH, s, 2.5 H). 13C NMR: 0.5– 0.8 (OASi(CH3)AO), 1.25 (SiA(CH3)3, 16.52 (a1), 16.95 (b1), 27.84 (i), 28.43 (h, d), 28.95 (e), 29 (f), 29.21 (g), 32.5 (c1), 32.85 (k) and 33.6 (j). IR: 646 (CABr), 768–892 (SiAC), 1045–1100 (SiAOASi), 1260 (SiACH3), 1408 (SiACH3), 2167 (SiAH) and 2965 (CAH). The copolymer (PMHS-co-PDMS 50/50) contains 50% of SiAH groups and 50% of SiACH3 groups. In this case, only 10/17 units of SiAH groups were used in hydrosilylation reaction. The 7 remaining units will be used thereafter for the cross-linking. Then, 3 g of PMHS-co-PDMS 50/50 (20.6 mmol of SiAH) were reacted with 2.7 g of undecenyl bromide (11.6 mmol) in the presence of PDTD complex as catalyst. The mixture was stirred at 60 °C during 2 h. The functional polymer PMHSBr-co-PDMS 50/50 was recovered as a fluid liquid. 2.4. Preparation of functional polymers bearing quaternary ammonium sites Functional polymer bearing quaternary ammonium sites was synthesized by quaternarization reaction using

144

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

Fig. 2. Hydrosilylation reaction of PMHS by undecenyl bromide (a); Quaternarization reaction of PMHSBr to obtain polymethylsiloxane containing quaternary ammonium sites (b). Numbers used for the assignment of NMR spectra are given in blue. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

an excess of triethylamine N(Et)3. 4 g of PMHSBr were dissolved in 5 mL of chloroform (the slow dissolution took 10 h). 2 mL of triethylamine N(Et)3 was added and stirred at 60 °C during 5 days. The new resulting polymer PMHSNEt was recovered as a brown solid after evaporation of chloroform and excess triethylamine under reduced pressure at 70 °C. Fig. 2b shows the quaternarization reaction of PMHSBr with the triethylamine to yield PMHSNEt. 1 H and 13C NMR (in DMSO-d6, d in ppm from TMS as a reference) and IR (KBr, cm1) spectra of the resulting polymer PMHSNEt after quaternarization reaction showed the following peaks: 1H NMR: 0.05–0.20 (Si(CH3)3 and OASiH(CH3)AO), 0.51 (a1, m), 1.24 (q and s, m), 1.16 (c1-i, m), 1.55 (j1, m), 1.93 (b1, m), 2.95 (r, m), 3.12 (k1, m), 3.24 (n, m), 3.34 (water) and 4.67 (l, SiH, s). 13C NMR: 1–2 (Si(CH3)3 and OASiH(CH3)AO), 7.45 (q), 8.97 (s), 16.52 (a1), 16.95 (b1), 21.24 (j1), 25.87 (i1), 29 (d–h), 32.39 (c1), 45.9 (r), 52.13 (n) and 56.21 (k1). IR: 768–892 (SiAC), 1045–1100 (SiAOASi), 1260 (SiACH3), 1408 (SiACH3), 1490 (CH3 of ammonium), 2167 (SiAH), 2490 and 2676 (CH2 in ammonium) and 2965 (CAH). The same protocol was used for the quaternarization reaction of PMHSBr-co-PDMS 50/50 by an excess of triethylamine N(Et)3. The dissolution in chloroform was fast and

the reaction lasted 2 days. The recovered polymer was a brown solid PMHSNEt-co-PDMS 50/50. 2.5. Preparation of functional polymers bearing pyridinium or phosphonium sites Functional polymer having pyridinium groups as anionic exchange sites was prepared by quaternarization using an excess of pyridine. 2 mL of pyridine was added to a solution of 2 g of PMHSBr dissolved in 10 mL of chloroform under stirring and reacted at 60 °C for 20 h. The elimination of chloroform and excess of pyridine were carried out at 80 °C under reduced pressure. The resulting polymer PMHSPy was recovered as dark brown solid after drying. Fig. 3a shows the quaternarization reaction of PMHSBr with the pyridine that yielded PMHSPy. 1 H and 13C NMR (in DMSO-d6, d in ppm from TMS as a reference) and IR (KBr, cm1) spectra of the resulting polymer PMHSPy after quaternarization reaction showed the following peaks: 1H NMR: 0.0–0.25 (Si(CH3)3 and OASiH(CH3)AO), 0.51 (a1, m), 1.26 (c1-i, m), 1.92 (b1 and j2, m), 4.63 (k2, m), 4.67 (l, SiH, s), 8.19 (m, meta, t), 8.63 (p, para, t) and 9.15 (o, ortho, d). 13C NMR: 1–2 (Si(CH3)3 and OASiH(CH3)AO), 16.88 (a1), 17.11 (b1), 22.62 (j2),

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

145

Fig. 3. Quaternarization reactions of PMHSBr to obtain polymethylsiloxanes containing pyridinium sites (a) and phosphonium sites (b). Numbers used for the assignment of NMR spectra are given in blue. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

25.95 (i2), 29.34 (e–h), 31.23 (d), 33.12 (c1), 61.06 (k2), 128.52 (m, meta), 145.23 (p, para) and 145.95 (o, ortho). IR: 768–892 (SiAC), 1045–1100 (SiAOASi), 1260 (SiACH3), 1408 (SiACH3), 1491 (C@N), 1639 (C@C), 2167 (SiAH) and 2965 (CAH), 3018 (CAH arom). The phosphonium sites were prepared by quaternarization reaction of PMHSBr by triphenylphosphine. A mixture of 2 g of PMHSBr and 1.5 g triphenylphosphine P(C6H6)3 in 10 mL of chloroform was stirred at 80 °C for 7 days. The evaporation of the chloroform was carried out by heating at 80 °C under a primary vacuum, but the final product still contained triphenylphosphine. Then, the purification of the resulting polymer PMHSPPh was obtained by precipitation in diethyl ether, filtration and washing the final polymer three times with toluene to yield a white solid. Fig. 3b shows the quaternarization reaction of PMHSBr with triphenylphosphine that yielded PMHSPPh. 1 H and 13C NMR (in DMSO-d6, d in ppm from TMS as a reference) and IR (KBr, cm1) spectra of the resulting poly-

mer PMHSPPh after quaternarization reaction showed the following peaks: 1H NMR: 0.0–0.25 (Si(CH3)3 and OASiH(CH3)AO), 0.49 (a1, m), 1.2 (c1-i, m), 1.45 (j3, m), 1.92 (b1, m), 3.64 (k3, m), 4.67 (l, SiH, s), 7.7–7.95 (arom, m). 13C NMR: 1–2 (Si(CH3)3 and OASiH(CH3)AO), 16.44 (a1), 16.67 (b1), 20.08 (k3, d(1JPAC = 50.0 Hz)), 22.00 (j3 d(2JPAC = 25.0 Hz), 28.90 (e–h), 29.80 (i3, d(3JPAC = 16.9 Hz)), 32.25(c1), 32.60 (d), 118.54 (v, quart, d(1JPAC = 85.0 Hz)), 130.17 (m1, meta, d(3JPAC = 12.5 Hz)), 133.57 (o1, ortho, d(2JPAC = 8.8 Hz)) and 134.76 (p1, para). IR: 693 (PAC), 768–892 (SiAC), 1045–1100 (SiAOASi), 1260 (SiACH3), 1408 (SiACH3), 1428 (PACH2), 1594 (PAAr), 2167 (SiAH) and 2965 (CAH), 3048 (CAH arom). 2.6. Cross-linking and grafting of different resulting functional polymers on pyrogenic silica gel 1 g of each PMHSNEt, PMHSPy, PMHSPPh and PMHS-coPDMS 50/50 functional polymer was dissolved in 10 mL

146

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

dimethylformamide DMF and was kept overnight at 70 °C until complete dissolution. 0.2 mL hexadiene cross-linking agent and 150 lL of Karstedt catalyst PDTD were added to each solution. The mixtures were stirred at 50 °C for 30 min to form a gel suspension just before grafting on the pyrogenic silica gel and deposition on the silicon/silica substrates. The cross-linking of different polymeric chains was produced by hydrosilylation reaction between remaining ASiH groups of functional polymers and vinyl groups [61,62] of hexadiene. The PMHSNEt, PMHSPy, PMHSPPh and PMHS-coPDMS 50/50 functional polymers were grafted on pyrogenic silica gel in order to characterize them by 29Si CP-MAS NMR. A sufficient quantity of commercial pyrogenic silica gel was dried at 150 °C in primary vacuum during 2 h. 1 g of this pyrogenic silica gel was soaked in 5 mL of each solution of cross-linked functional polymer in DMF so as to generate a thin layer of grafted and cross-linked functional polymer at the surface of silica. One drop of water was added to activate the hydrolysis of the remaining ASiH groups of the functional polymers and the polycondensation with the silanol groups of the silica gel and the four solutions were maintained under strong stirring at ambient temperature for 3 h. Then, the grafted pyrogenic silica gel was separated from the reaction medium by sedimentation under natural gravity and washed three times with DMF in order to eliminate the non-grafted functional polymers. The residual DMF solvent was evaporated at 70 °C for several days. Fig. 4 is a sketch of the membrane material after cross-linking and grafting of resulting functional polymers bearing cationic sites attached to the surface of pyrogenic silica gel.

2.7. Functionalization of silicon/silica wafers The silica surface of wafer Si/SiO2 was treated with pyranha mixture and sulfochromic solution and washed with bidistilled water, so as to create reactive silanol sites. The physisorbed water molecules were removed by drying at 150 °C during 2 h. The deposition of cross-linked PMHSNEt, PMHSPy, PMHSPPh and PMHS-co-PDMS 50/50 functional polymers was processed by the dip-coating technique. Several parameters of the dip-coating process influence the homogeneity and thickness of the membrane such as the soaking and climbing speeds and the time spent in solution [63]. In this deposition technique, the two speeds were fixed at 0.5 min s1 and the wafer was kept 300 s in 5 mL of gel suspension before moving it up. The residual solvent on the membrane was evaporated at room temperature at 70 °C for 2 days. The control samples with non-functional PDMS were prepared in the same way by dip-coating in a 1% solution of PDMS in DMF. 2.8. Impedance measurements The impedance measurement was performed at 10 kHz frequency and 10 mV amplitude of alternating voltage. Out-of-phase impedance was recorded for bias voltages V varying from 500 to 2700 mV/SCE. The background electrolyte used for test solutions in electrochemical cell was 10 mM sodium acetate at pH 6.8. Anionic dyestuffs exchange measurements were carried out for Acid Blue 25 (AB25), Acid Blue 74 (AB74) and Acid Yellow 99 (AY99). The electrical response of the device to a given anionic dyestuff X consists in a variation of DVFB as a function of the concentration of X expressed through pX = Log[X].

Fig. 4. Cross-linked and grafted functional polymethylsiloxanes bearing different cationic sites on the silica surface.

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

147

3. Results and discussion 3.1. Synthesis and spectroscopic characterizations of functional polymers bearing cationic sites 3.1.1. Synthesis scheme The synthesis started by grafting alkyl bromide pendent chains to the silicone backbone; the farther reaction of this functional polymer with triethylamine, pyridine or triphenylphosphine yielded the three cationic derivatives of polymethylsiloxane. Cross-linking of the polymer was performed by reaction of the difunctional hexadiene with residual SiAH groups. The preparation of the siliconebased cross-linked cationic polymers faced several synthesis issues. The SiAH functions were used for both purposes of attachment of the cationic groups to the silicone backbone and cross-linking. Therefore, the grafting process of undecenyl bromide was limited to 30% of the available SiAH groups of the starting polymer reagent. This could be achieved either by limiting the reaction duration such that the conversion corresponded to the willed grafting degree, or by running the reaction of a sub-stoichiometric amount of undecenyl bromide to full conversion. The second option was much more reliable and repeatable. A long enough reaction time was necessary for ensuring full conversion and side reactions had to be avoided, in particular hydrolysis of SiAH by moisture. The reaction of the bromo derivative of the polymer with triethylamine, pyridine or triphenylphosphine was a SN2 reaction that yielded a polar compound. A polar solvent is suitable for such a reaction [64]. Unfortunately, the silicone polymer reagents were not soluble in the polar solvents that are generally used for such reactions. Chloroform was a satisfactory compromise as a solvent that ensured dissolution of the polymer and a fast enough reaction rate. The reaction durations in chloroform were much longer that in usual polar solvent. Cross-linking was performed by using a cross-linking agent (the di-vinyl compound hexadiene). It could have been performed by means of hydrolysis and condensation: silanol groups (SiAOH) are generated by hydrolysis of SiAH, and their condensation into SiAOASi bonds creates cross-linkages. Since the condensation reaction releases water, water acts as a catalyst and a slight amount of moisture is enough for reaching a full conversion of the SiAH groups. This reaction was very slow however, so that its implementation was difficult. Cross-linking with hexadiene and the Karstedt’s catalyst was fast and repeatable.

3.1.2. IR spectroscopy IR spectra recorded from KBr pellets provided qualitative analyses of the reactions (Fig. 5). The IR spectrum of the mixture of undecenyl bromide and PMHS recorded before reaction showed several characteristic bands. The bands at 2965 and 2167 cm1 were assigned to asymmetric stretching vibrations m(CAH sp3) and the stretching vibrations m(SiAH) for PMHS respectively. The bands at 892–768, 1045–1100, 1260 and 1408 cm1 were attributed to stretching m(SiAC), asymmetric stretching m(SiAOASi), symmetric deformation d(SiACH3) and

Fig. 5. IR spectra of the starting and final materials. PMHS (a), PMHSBr (b), PMHSNEt (c), PMHSPy (d), and PMHSPPh (e).

asymmetric deformation vibration of SiACH3 respectively. Other important bands in the spectrum were attributed to undecenyl bromide: the bands at 646, 915, 1637 and 3074 cm1 were assigned to stretching m(CABr), strong deformation d(CH@CH2), weak m(C@C) stretching and stretching vibration (@CH2) respectively. After hydrosilylation reaction of PMHS, the three bands at 915 (CH@CH2), 1637 (C@C), 3074 (@CH2) cm1 have disappeared in the spectrum, which confirmed that the hydrosilylation reaction had occurred. The spectrum of resulting polymer PMHSNEt after quaternarization reaction gave three supplementary bands with respect to the spectrum of PMHSBr at 1490, 2490, 2676 cm1 attributed to deformation vibration d(CH3 sp3) of quaternary ammonium, deformation and stretching vibrations of methylenes on quaternary ammonium respectively. Moreover, the disappearance of band at 646 cm1 for CABr groups confirmed the formation of PMHSNEt. The IR spectrum of the PMHSPy polymer displayed three supplementary bands with respect to the spectrum of PMHSBr at 1491, 1639 and 3018 cm1 attributed to stretching vibrations m(C@N, C@C and CAH) of aromatic rings of pyridinium groups. The spectrum of PMHSPPh showed four supplementary bands at 693, 1428, 1597 and 3048 cm1. The bands at 693 and 1428 cm1 were attributed to stretching vibrations m(PAC) and deformation vibration d(PACH2A) respectively. The bands at 1597 and 3048 cm1 were assigned to stretching vibrations m(PAAr) and m(CAH) of aromatic rings respectively. 3.1.3. Liquid 1H and 13C NMR The 1H NMR spectra before and after hydrosilylation were compared (Fig. 6a and b). The peaks a at 4.95 ppm and b at 5.8 ppm corresponding to (CH2a@CHbA) protons of the vinyl group of undecenyl bromide in the spectrum before hydrosilylation reaction have disappeared once

148

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

Fig. 6. 1H NMR spectra before (a) and after (b) hydrosilylation reaction of PMHS and after quaternarization reaction (c) to obtain PMHSNEt polymer.

the reaction have been completed. Two peaks have appeared in the spectrum after hydrosilylation reaction: a1 at 0.55 ppm and b1 at 2.0 ppm assigned to protons of ethyl groups in position a and b with from the silicon atom (SiACH2a1ACH2b1A). The peak c at 2.05 ppm corresponding to proton of CH2 groups adjacent to the vinyl group was observed in the spectrum before hydrosilylation reaction and has also disappeared at the end of the reaction. Consequently, another peak c1 of the same proton of CH2 groups has appeared at 1.3 ppm after hydrosilylation reaction. Finally, the integral of the peak l at 4.7 ppm corresponding to proton of the SiAH groups was lower in the final polymer PMHSBr than in PMHS; their ratio corresponded to the stoichiometry of the starting reagents (CH@CH2/ SiH = 10/35). The 1H NMR spectrum of the polymer PMHSNEt resulting from the quaternarization reaction (Fig. 6c) showed two news peaks j1 at 1.55 ppm and k1 at 3.12 ppm corresponding to protons of methylene groups in position b and a from the quaternary ammonium group respectively have appeared in the spectrum after the reaction, together with the two peaks q at 1.24 ppm and n at 3.24 ppm attributed to protons of methyl and methylene groups of quaternary ammonium respectively. The two peaks k at 3.4 ppm and j at 1.85 ppm assigned to protons of methylene groups (ACH2jACH2kABr) in position a and b from the bromine atom respectively have disappeared. The other peaks a1, b1, c1 and l (SiAH) remained unchanged. Finally, two peaks r at 2.95 ppm and s at 1.24 ppm have appeared. They were ascribed to protons of methylene and methyl groups of triethylammonium cation electrostatically bound to deprotonated silanol groups. Such SiAO,+NH(CH2rACH3s)3 groups resulted from the hydrolysis of SiAH into SiAOH by moisture followed by proton transfer from SiAOH to triethylamine. The consequence of the formation of ASiAO,NH+ groups might be the subsequent cross-linking of the polymer layer by reaction of SiAOA with SiAOH before any

cross-linking process is applied. The polymers remained soluble during all subsequent chemical modifications, so that any early cross-linking has remained of limited extent. Since the polymer layer was cross-linked at the end of the functionalization, the possible early cross-linking has no consequence on the final material. The 13C NMR spectra before and after hydrosilylation reaction of PMHS and undecenyl bromide, and after quaternarization reaction (yielding the PMHSNEt polymer) are presented in Fig. 7. The peaks a at 113.9 ppm and b at 138.79 ppm assigned to (CaH2@CbHA) carbons of the vinyl groups of undecenyl bromide were present in the spectrum before hydrosilylation reaction and had disappeared at the end of the reaction. Two new peaks a1 at 16.52 ppm and b1 at 16.95 ppm corresponding to carbons of methylene groups in position a and b from silicon (SiACa1H2ACb1H2A) have appeared in spectrum after hydrosilylation reaction (Fig. 7b). The peak c at 28.82 ppm corresponding to CH2 groups adjacent to vinyl groups has also disappeared after completion of the reaction, while the peak c1 of the same CH2 groups appeared at 32.5 ppm after hydrosilylation reaction. Finally, all other peaks found in the spectrum before hydrosilylation reaction of PMHS (Fig. 6a) remained unchanged: in particular the peaks i (27.9 ppm), k (32.57 ppm) and j (33.52 ppm) assigned to carbons of methylene groups (ACiH2ACjH2ACkH2ABr) in positions c, a and b to bromide groups. The 13C NMR spectrum of final polymer PMHSNEt after quaternarization reaction (Fig. 7c) showed three new peaks j1, i1 and k1 at 21.24, 25.87 and 56.21 ppm corresponding to carbons of methylene groups in positions b, c and a from the quaternary ammonium group respectively have appeared after quaternarization reaction. The formation of ammonium groups was associated with the disappearance of three other peaks i, j and k assigned to carbons of methylene groups (ACiH2ACjH2ACkH2ABr) in positions c, a and b from the bromine atom and the

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

Fig. 7.

13

149

C NMR spectra before (a) and after (b) hydrosilylation reaction of PMHS and after quaternarization reaction (c) to obtain PMHSNEt polymer.

appearance of two other peaks q at 7.45 ppm and n at 52.13 ppm attributed to carbons of methyl and methylene groups of quaternary ammonium respectively. Finally the peaks resulting from the formation of the triethylammonium salt by hydrolysis of SiAH groups were also observed in the 13C NMR spectrum: the peaks s at 8.97 ppm and r at 45.9 ppm corresponding to carbons of methyl and methylene groups of +NH(CrH2ACsH3)3 attached to SiAO groups. The 13C NMR spectrum of the PMHSPy (Fig. 8b) exhibited three peaks j2, i2 and k2 at 22.62, 25.95 and 61.06 ppm corresponding to carbons of CH2 groups in positions b, c and a relative to the pyridinium group respectively. Such formation was associated by the disappearance of the three peaks i (27.9 ppm), k (32.57 ppm) and j (33.52 ppm) assigned to carbons of methylene groups (ACiH2ACjH2ACkH2ABr) in positions c, a and b relative to bromide in the spectrum of PMHSBr. Moreover, the appearance of three peaks m, p and o at 128.52, 145.23 and 145.95 ppm corresponding to the carbons of the aromatic ring of the pyridinium groups in positions meta, para and ortho respectively were observed in the 13C NMR spectrum of the PMHSPy polymer. The resulting polymer PMHSPPh obtained after quaternarization reaction of PMHSBr by the triphenylphosphine was characterized as the others polymers by 1H NMR (Fig. 9a). The appearance of two new peaks j3 and k3 at 1.45 and 3.64 ppm respectively was noticed in the spectrum. These peaks are assigned to protons of CH2 groups in positions b and a relative to the phosphonium group. The two peaks k and j of the corresponding CH2 groups

in PMHSBr have disappeared. Moreover, a broad peak between 7.7 and 7.95 ppm corresponding to the protons of phenyl of phosphonium groups in positions meta, para and ortho was present in the spectrum of the final polymer. The other peaks a1, b1, c1, i and SiAH remained unchanged in the spectrum. The 13C NMR spectrum of the PMHSPPh polymer (Fig. 9b) showed three new peaks k3, j3 and i3 at 20.08, 22.00 and 29.80 ppm respectively which corresponded to carbons of CH2 groups in positions c, a and b from the phosphonium group. As a consequence, the peaks i, j and k corresponding to carbons of CH2 groups present in the spectrum of PMHSBr have disappeared after completion of the quaternarization reaction. The four new peaks v, m1, o1 and p1 respectively at 118.54, 130.17, 133.57 and 134.76 ppm was assigned to quart, meta, ortho and para carbon atoms of phenyl groups of phosphonium respectively. 3.1.4. Kinetics of the hydrosilylation reaction The mechanism of the hydrosilylation reaction is rather complex and the kinetics of the reaction depends on several parameters including the functional groups present on the grafted lateral chains. It has been claimed that every case is specific and deserves a detailed investigation [65]. Additionally, grafting to a polymer backbone makes the reaction kinetics dependent on the conversion since the steric hindrance of bulky lateral chains prevent further reaction at high grafting densities [66,67]. The kinetics of the hydrosilylation reaction of PMHS and undecenyl bromide was determined by 1H NMR spectroscopy analysis.

150

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

Fig. 8. The 1H (a) and

Fig. 9. The 1H (a) and

13

C (b) NMR spectra of PMHSPy bearing pyridinium sites after quaternarization reaction of PMHSBr with pyridine.

13

C (b) NMR spectra of PMHSPPh bearing phosphonium sites after quaternarization reaction of PMHSBr with triphenylphosphine.

The area of the peak b at 5.8 ppm corresponding to proton (CH2@CHbA) of the vinyl groups of undecenyl bromide decreased as a function of time (Fig. 10a). The complete disappearance of this peak was reached after 240 min. At same time, the area of the peak a1 at 0.55 ppm assigned to protons of methylene groups in position a from silicon (SiACH2a1A) has increased up to the maximum in spectrum at 240 min. The area of the peak l at 4.7 ppm corre-

sponding to proton of the SiAH groups in the final polymer PMHSBr has decreased accordingly. The degree of advancement of the hydrosilylation reaction n was calculated from the areas of peak l at 4.7 ppm assigned to proton of SiAH groups and peak b at 5.8 ppm corresponding to protons of vinyl groups (CH2@CHbA) relative to the constant area of the peak k at 3.4 ppm of protons of CH2Br groups. The different samples were collected

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

151

Fig. 10. Progress of the hydrosilylation reaction of PMHS with undecenyl bromide monitored by 1H NMR analysis (a) and advancement degrees nb,k (s) and nSiAH,k (j) as a function of time (b). The inset (c) shows the linear relationship of ln(1  n) as a function of time.

from the reaction mixture at several times up to 240 min. The advancement degree n as a function of time t is shown in Fig. 10b.

nb;k ðtÞ ¼ 1 

Ib ðtÞ=Ik ðtÞ Ib ð0Þ=Ik ð0Þ

with Ib ð0Þ=Ik ð0Þ ¼ 0:5

ð1Þ

Ib(t) is the area of peak b at 5.8 ppm of vinyl groups at time t and Ik(t) is the area of peak k at 3.4 ppm of CH2Br groups at time t. nSiAH;k ðtÞ ¼ 2½ISiAH ð0Þ=Ik ð0Þ  ISiAH ðtÞ=Ik ðtÞ with ISiAH ð0Þ=Ik ð0Þ ¼ 1:75 ð2Þ

ISiAH(t) is the area of peak l at 4.7 ppm of SiAH at time t. In the framework of the Chalk–Harrod mechanism of hydrosilylation reaction [68], it has been shown that the rate-limiting step of the hydrosilylation reaction of PMHS with an alkene is the reaction of the SiAH groups with the activated alkene species formed by complexation of the vinyl group and the platinum metal of the Karstedt catalyst [69,53]. A simplified view of the complex overall mechanism is as follows. Since the concentration of Karstedt catalyst is much lower than that of alkene, the concentration of activated alkene species is nearly constant, and the reaction rate is therefore independent of the concentration of alkene. As consequence, the overall reaction shows first order kinetics with respect to the concentration of SiAH. The reaction rate is

v¼

dn 0 ¼ k ½CH2 @CHc dt

with c ¼ 1

ð3Þ

Integration of Eq: ð3Þ yields 0

ln½1=ð1  nÞ ¼  lnð1  nÞ ¼ k t

ð4Þ

The plot of ln (1  n) as a function of time (Fig. 10c) was linear, showing that the reaction was first order with respect to SiAH and zeroth order with respect to undecenyl bromide. The apparent rate constant was k0 = 102 min1. 3.1.5. Solid 29Si NMR of resulting polymers bound to pyrogenic silica gel surface 29 Si CP-MAS NMR spectra (a), (b), (c), (d) and (e) of various system of silica gel, silica gel/PMHSBr, silica gel/ PMHSNEt and silica gel/PMHSPy and silica gel/PMHSPPh respectively are given in Fig. 11. All the spectra exhibited two main resonance peaks at 101 and 110 ppm which were assigned to the Q3OH silanol groups present on the surface of silica gel and to the silicon atoms of the internal silica network (units Q4) respectively. The peak of low intensity appearing at 92 ppm was attributed to geminal silanols Q2OH. The comparison of the relative intensities of peaks at 101 and 110 ppm in each spectrum (b), (c) and (d) excepting spectrum (e) clearly showed that the systems silica gel/PMHSBr, silica gel/PMHSNEt and silica gel/PMHSPy contained less Q3OH and Q2OH sites than silica gel relative to the sites Q4 [70]. These results confirmed the grafting of PMHSBr, PMHSNEt and PMHSPy on the silica gel surface. The spectra (b), (c), (d) and (e) of various system of silica gel, silica gel/PMHSBr, silica gel/PMHSNEt, silica gel/ PMHSPy and PMHSPPh/silica gel respectively exhibited four supplementary peaks at 22, 38, 56 and 57 ppm which were assigned to units D2, D2H, T2OH and T3 respectively (Fig. 4). The sites D2H were initially present in commercial poly(hydromethylsiloxane) (PMHS). The presence of the units D2 [71] in the spectrum (b) confirmed the occurrence of the hydrosilylation reaction of PMHS with undecenyl bromide to form PMHSBr and the

152

Fig. 11.

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

29

Si CP-MAS NMR spectra of silica gel (a), PMHSBr/silica gel (b), PMHSNEt/silica gel (c), PMHSPy/silica gel (d), and PMHSPPh/silica gel (e).

cross-linking of different polymeric chains by hexadiene cross-linking agent. The detection of units T2OH [72] suggested the presence of silanols groups non-condensed on different functional polymethylsiloxane after hydrolysis by ambient moisture. The sites T3 were due to the grafting reaction of all these functional polymethylsiloxanes to the silica gel surface. 3.2. Electrochemical measurements The anionic dyestuffs exchange phenomena have been investigated from measurements of the electrical response of the Si/SiO2 devices coated with triethylammonium, pyridinium or phosphonium cationic sites in the substituted membrane (PMHSNEt, PMHSNEt-co-PDMS 50/50, PMHSPy or PMHSPPh) and with the non-functional PDMS as a function of the concentration of various anionic dyestuffs X: Acid Blue 25 (AB25), Acid Blue 74 (AB74) and Acid Yellow 99 (AY99). The background electrolyte was sodium acetate 10 mM and pH = 6.8. Thus, the functionalized wafers used as electrodes were immersed in the solution of sodium acetate inside the electrochemical cell for stabilization for 24 h prior starting the measurements. The stabilization was controlled after impedance measurement as a function of tension of polarization (C–V) repeated each 2 h. The device was stabilized when subsequently observed capacitance curves were superimposed. The bromide counter-ions Br of cationic sites in the coated polymer were exchanged with the acetate anions AcO of the electrolyte during the stabilization and exchanging equilibrium. Another process that has to be waited for is swelling of the polymer layer by water. Therefore, the measured anion exchange process was the replacement of acetate ions AcO by anionic

dyestuffs X inside the polymer layer. Such exchange process was dependent on the anionic dyestuffs concentration [X]. The variations of dyestuffs concentrations in electrochemical cell were obtained by successive additions of small aliquots from two solutions of 102 and 103 mol L1 concentrations. The two parameters, flat band potential shift DVFB and variation of the capacitance in inversion regime DC are shown in Figs. 12 and 13 as a function of Log[X] for each anionic dyestuff and for the different functionalized polymer electrodes. 3.2.1. The behavior of functional membranes with respect to the control non-functional PDMS membrane The EIS structures coated with the non-functional PDMS membrane gave a significant electrical response to the presence of dyes. Such phenomenon was ascribed to the adsorption of the organic dyes from their aqueous solution to the surface of the hydrophobic silicone material. Adsorption by hydrophobic interactions brought charged molecules to the surface of the electrodes and caused an electrical signal by means of a mechanism that was not an anion exchange. Such hydrophobic interactions were probably also operating with the functional materials. The non-functional PDMS membrane displayed a significant sensitivity to the dyes because such organic cations could bind to the hydrophobic PDMS. The DVFB values of this control sample were smaller than for the functional cationic membrane in all instances however. This meant that the origin of the adsorption of the dyes to the sensitive membranes was the combination of two contributions: (i) a simple adsorption of surfactant-like organic molecules at the interface between water and the hydrophobic silicone material; and (ii) the anion exchange of acetate anions for the anionic dyes. The amount of dyes bound to the

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

Fig. 12. Variation of the flat band potential DVFB as a function of pX for the different dyestuffs: AB25 ( ), AB74 ( PMHSNEt, PMHSNEt-co-PDMS 50/50, PMHSPy, PMHSPPh, and PDMS membranes.

membrane is not the sum of that bound by hydrophobic interactions and that bound by anion exchange however. On the same footing, the electrical potential created by ions adsorption is not the sum of the electrical potentials coming from each contribution. Therefore it is not correct to subtract the DVFB values of the non-functional PDMS from that of the cationic PMHS materials. Adsorption of ionic organic dyes to hydrophobic surfaces is a well-documented phenomenon. This is the driving force for dyeing synthetic fabrics such as polyamides

153

) and AY99 ( ) for the different

and polyesters [73–75]. Ionic organic dyes behave as surfactants when their molecular structure has well-separated hydrophobic parts and polar ionic groups. They are known to self-aggregate in a similar way to surfactants [76,77] and they mix favorably with surfactants as mixed micelles [78–80]. The strong lowering of surface tension of aqueous solution of ionic organic dyes demonstrates their surface-active properties [80]. Adsorption of organic anions (ss-DNA, organophosphate pesticides) to PDMS materials has been observed as a cause of troubles in the

154

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

Fig. 13. Variation of the capacitance DC as a function of pX for the different dyestuffs: AB25 ( ), AB74 ( co-PDMS 50/50, PMHSPy and PMHSPPh membranes.

course of microcontact printing and manufacture of molecularly imprinted materials [81–82]. Based on the chemical structure of the dyes (Fig. 1), the Acid Blue 25 is the most hydrophobic dye because it bears only one sulfonate group and also contains amino groups that are partly protonated at pH 6.8. Acid Yellow 99 is a more polar molecule because of the presence of several polar substituents, and Acid Blue 74 is even more polar because it bears two sulfonate groups. The electrical response of the PDMS coating due to the adsorption of the dyes followed this order of polarity of the dyes. Hydrophobic interactions of the dyes with the silicone lead to adsorption at the water–silicone interface and should cause anion exchange. Thus, hydrophobic adsorption should cause a release of acetate anions for the silicone layer remains electrically neutral. In other words, the hydrophobic interactions increase the interactions that are responsible for the anion exchange. As a rough approach, the free energy of dye binding might be the sum of a hydrophobic contribution and an ion exchange contribution that consists in short-range interactions between the cationic sites and the anionic dyes. The short-range interactions include electrostatic and other types of interactions (dispersion, polar, acid–base); they are specific with respect to the types of dyes and cationic groups. On the contrary, hydrophobic interactions are not

) and AY99 ( ) for different PMHSNEt, PMHSNEt-

specific. The electrical potential shift caused by dye binding can be specific because the short-range interactions include specific interactions. It is difficult to assess the relative contributions of the specific and non-specific interactions from measurements of the full electrical consequence of dye binding. Although the phenomena are not additive, a comparison of the DVFB values of the four functional PMHS and the control PDMS provides a qualitative estimate of the contribution of specific interactions with the cationic groups of the membranes. The PMHSNEt and PMHSNEt-co-PDMS 50/50 displayed the smallest difference with respect to PDMS, revealing a weak interaction of the triethylammonium group with the dyes. Even the differences with respect to PDMS were not significant for the most polar AB74 dye. Adsorption might go beyond the ion exchange phenomenon that keeps the electrical charge constant. Indeed, once the anion exchange reached completion, it is still possible to bind supplementary dye molecules by hydrophobic interactions with the non-functionalized parts of the silicone backbone. 3.2.2. Effects of ion exchange on analytical and electrical characteristics For the four functionalized membranes and the control PDMS material, the flat band potential DVFB decreased

155

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

against the concentration of each dyestuff AB25, AY99 and AB74 (Fig. 12). The variation of DVFB was very few at low concentrations of X, indicating that neither ion exchange in the cationic membrane between acetate counter ions and dyestuffs, nor hydrophobic adsorption had significantly taken place. A stronger variation DVFB was observed at a concentration of X above a threshold concentration that was identified as the detection limit of X. Above the detection limit where a significant variation of DVFB was noticed, the measured electrical parameters depended on the affinity of the anionic dyestuffs for each functionalized cationic membranes. A low detection limit meant a high affinity of anionic dyestuffs for the cationic sites. The full flat band potential variation corresponded to the electrical effect of the counter-anions exchange process gone to completion. The different parameters reported in Table 2 are the limit of detection expressed as its pX = Log[X], the variation of flat band potential DVFB at full dyestuff exchange and the slope of the curve DVFB vs pX in its most linear part (at its inflexion point). The values of the slopes of DVFB against Log[X] are usually compared to the Nernst law pertaining to conventional ISE’s as reference. This slope is 59 mV/decade of mono-charged anions AB25 and AY99, and 29.5 mV/decade for the di-charged anion AB74. However, there is no theoretical foundation for expecting such functionalized polymer electrodes to follow the Nernst law. The values of such slopes are often identified as the sensitivity of the analytical device. All characteristics are gathered in Table 2 for the three dyestuffs and the various membranes. The sensitivity, variations of flat band potential and capacitance (DVFB and DC), and detection limit are related together: Table 2 shows the strong correlation between these four quantities since the characteristics pertaining to the three dyestuffs appeared sorted in the same order. The only PMHSPPh membrane showed up quite a peculiar behavior. A low detection limit indicates a high affinity of the anion to cationic sites. High values of the sensitivity, DVFB, and DC mark a high sensitivity of electrical phenomena to ion exchange inside

the membrane. The three later quantities are not obviously related to the affinity of ions to the cationic sites; in particular DVFB and DC are variations once anion exchange has reached completion. It appeared from the experimental data that a high affinity of anions was correlated with a high sensitivity of the electrical parameters. Possibly, a high affinity of the anions caused large structural reorganizations inside the membranes upon anion exchange. The affinity of the dyestuffs and their influence on the electrical properties of the membranes was in the order AB25 > AY99 > AB74. According to their chemical structure, the hydrophobic character of the 3 dyes is indeed in this order. Indeed, the number of cycles in the AB25 molecule is larger than for AY99 and AB74. The AB74 dye is more hydrophilic than AY99 because the former molecule contains two anionic sulfonate groups. The sensitivity in the order of lipophilic character of the anions is the main effect at the origin of the Hofmeister effect [37]. Incidentally, the Nernst law was followed in 5 instances out of 12, almost half the cases. There was only a slight dependence with respect to the polymer type. There was no significant difference between the analytical parameters (detection limits and sensitivities) of PMHSNEt, PMHSNEt-co-PDMS 50/50, showing that the density of cationic sites was not a crucial parameter. The type of cationic site was of higher relevance: the anionic dyestuffs had a higher affinity for the pyridinium than for the triethylammonium sites, an effect that might be caused by supplementary p-stacking interactions operating between aromatic rings of the dyes and pyridinium. The case of the triphenylphosphonium group is specific because the order of affinity of the 3 dyes is different of the pyridinium and triethylammonium. The high affinity of the AY99 dye for triphenylphosphonium sites was explained by the formation of p-(arene)-complexes [83] between the chromium of AY99 and aromatic cycles of the triphenylphosphonium groups supplementing anionic dyestuffs exchange. Ion-specific interactions such as the present one are the origin of the anti-Hofmeister effects observed for ion-sensitive electrodes [42].

Table 2 Electrical characteristics of different functional polymethylsiloxane membranes and the reference PDMS material with respect to exchange of AcO for X and adsorption, as inferred from experiments as a function of dyestuff concentrations. Slope (mV/pX)

DC (nF cm2)

60 39 29

32 15 10

0.52 0.37 0.23

5.4 5.3 5.0

58 42 24

23 20 9

0.34 0.26 0.10

AB25 AY99 AB74

5.2 5.0 4.8

123 78 38

59 34 19

0.92 0.48 0.25

PMHSPPh

AB25 AY99 AB74

4.8 5.0 4.7

63 122 35

27 59 18

0.47 1.11 0.20

PDMS

AB25 AY99 AB74

5.0 4.6 4.3

50 32 18

18 14 9

Membrane bearing cationic sites

Dyestuff

Detection limit (pX)

PMHSNEt

AB25 AY99 AB74

5.6 5.0 4.7

PMHSNEt-co-PDMS 50/50

AB25 AY99 AB74

PMHSPy

DVFB (mV) at full dyestuff exchange

156

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

3.2.3. Effects anion adsorption and exchange on the flat band potential The variation of flat band potential reflects the potential drops taking place in the solution and electrochemical device upon ion absorption inside the membrane and ion adsorption at the interfaces. Since the anion exchange of anion dye for acetate reached completion, the membrane remained electrically neutral upon the anion exchange process. As consequence anion exchange inside the bulk of the membrane should not cause a variation of potential. Relevant potential drops are at interfaces. It is not expected that the silica-polysiloxane membrane interface is a site of ion adsorption because it is not in contact with the solution. The main effect is expected at the membrane solution interface. A similar conclusion was reached in an analysis of the ion selectivity of membrane-based ion-sensitive electrodes [84]. Electrostatic effects in the diffuse electrical double layer are constant because the ionic strength was fixed constant by the sodium acetate 10 mM background electrolyte. The predominant electrostatic effects are presumed occurring in the Stern layer of the membrane-electrolyte interface. On this basis, a higher affinity of the anion to the cationic sites of the membrane cause a tighter binding of anion to the cationic surface sites, and the flat band potential varies to more negative values. The correlation between analytical parameters (detection limit and sensitivity) and DVFB makes sense with regards to this mechanism.

3.2.4. Effects anion exchange and adsorption on the capacity The behavior of variation of the electrical capacity of the devices as a function of the concentration of X anionic dyestuffs was not similar to the sigmoidal behavior of the flat band potential (Fig. 13). Thus, DC of all AB25, AY99 and AB74 anionic dyestuffs in various PMHSNEt, PMHSNEt-co-PDMS 50/50, PMHSPy or PMHSPPh cationic membranes showed a slight increase for concentrations of X just above the detection limit and strong decrease for larger concentrations of X before it finally reached a constant value at large [X]. The values of DC given in Table 2 correspond to full anion exchange. The variation of electrical capacity corresponded to that of two capacities in series: the membrane and the electrical double layer in the solution near the membrane surface. Both the modifications of their thickness and dielectric permittivity upon ion exchange contributed to the variation of DC. Several physicochemical phenomena influence the capacity of the membrane. The membrane thickness plays on two parameters: the size of the anionic dyes incoming in place of the same acetate anion, and the swelling by water upon substitution of anions of different lipophilic characters. The dielectric permittivity is also under the influence of the same parameters, the various influences act contradictorily, so that it is difficult to give a clear-cut discussion of the variations of capacity. Any discussion would be largely speculative. A full investigation of the electrochemical behavior with impedance measurements as a function of frequency and modeling with the help of equivalent circuits [85] would be necessary in order to be able to reach definite conclusions.

4. Conclusion In this study, EIS devices immobilized by new anionexchange membranes for anionic dyestuffs have been prepared and evaluated. The functionalization of two polymethylsiloxanes (PMHS and PMHS-co-PDMS 50/50) by grafting various cationic exchanging groups was performed in two steps via hydrosilylation reaction in the presence of Karstedt’s catalyst and quaternarization reaction in order to yield quaternary ammonium, pyridinium and phosphonium sites. The different functionalized polymethylsiloxane chains were cross-linked and deposited at the surface of the silica transducer by the dip-coating technique. The chemical grafting of the polymer to the surface of the SiO2/Si heterostructure has been proved by solidstate 29Si NMR spectroscopy. The electrical measurements disclosed a specific sensitivity to anionic dyes that followed the Hofmeister series. One exception to the later behavior came from a specific interaction of a chromiumbased dye AY99 with the triphenylphosphonium groups. The analytical and electrical parameters appeared strongly correlated, so that the anions with stronger affinity gave the larger electrical signal. The range of detection limits of 105–104 mol L1 is quite interesting regarding the application to analysis of dyes in wastewaters. It is also in a favorable range for using the anion-exchanging materials for the purification of wastewater before their disposal into the environment. The selection of the type of cationic group for anion exchange allows modifying the selectivity. On this basis, it is expected that a fine analysis of dyes mixtures would be possible by using a device made of several parallel sensors sensitized with different cationic sites associated with a suitable signal processing.

References [1] Janin A, Zaviska F, Drogui P, Blais J-F, Mercier G. Selective recovery of metals in leachate from chromated copper arsenate treated wastes using electrochemical technology and chemical precipitation. Hydrometallurgy 2009;96:318–26. [2] Zhu M-X, Lee L, Wang H-H, Wang Z. Removal of an anionic dye by adsorption/precipitation processes using alkaline white mud. J Hazard Mater 2007;149:735–41. [3] Monteagudo JM, Durán A, Aguirre M, San Martín I. Photodegradation of Reactive Blue 4 solutions under ferrioxalate-assisted UV/solar photo-Fenton system with continuous addition of H2O2 and air injection. Chem Eng J 2010;162:702–9. [4] Köller G, Möder M, Czihal K. Peroxidative degradation of selected PCB: a mechanistic study. Chemosphere 2000;41:1827–34. [5] Fan L, Zhou Y, Yang W, Chen G, Yang F. Electrochemical degradation of aqueous solution of Amaranth azo dye on ACF under potentiostatic model. Dyes Pigments 2008;76:440–6. [6] Rengaraj S, Moon S-H. Kinetics of adsorption of Co(II) removal from water and wastewater by ion exchange resins. Water Res 2002;36:1783–93. [7] Cardoso VA, de Souza AG, Sartoratto PPC, Nunes LM. The ionic exchange process of cobalt, nickel and copper(II) in alkaline and acid-layered titanates. Colloid Surf A: Physicochem Eng Aspects 2004;248:145–9. [8] Huang H-Y, Shih Y-C, Chen Y-C. Determining eight colorants in milk beverages by capillary electrophoresis. J Chromatogr A 2002;959:317–25. [9] Peláez-Cid A-A, Blasco-Sancho S, Matysik F-M. Determination of textile dyes by means of non-aqueous capillary electrophoresis with electrochemical detection. Talanta 2008;75:1362–8. [10] Fakhari AR, Breadmore MC, Macka M, Haddad PR. Non-aqueous capillary electrophoresis with red light emitting diode absorbance

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

[11]

[12]

[13]

[14]

[15] [16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24] [25] [26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

detection for the analysis of basic dyes. Anal Chim Acta 2006;580:188–93. Laing DK, Hartshorne AW, Bennett DC. Thin layer chromatography of azoic dyes extracted from cotton fibres. J Forensic Sci Soc 1990;30:309–15. Scherpenisse P, Bergwerff AA. Determination of residues of malachite green in finfish by liquid chromatography tandem mass spectrometry. Anal Chim Acta 2005;529:173–7. Dávila-Jiménez MM, Elizalde-González MP, Peláez-Cid AA. Adsorption interaction between natural adsorbents and textile dyes in aqueous solution. Colloid Surf A: Physicochem Eng Aspects 2005;254:107–14. Zhao X, Hardin IR. HPLC and spectrophotometric analysis of biodegradation of azo dyes by Pleurotus ostreatus. Dyes Pigments 2007;73:322–5. Tsatsaroni EG, Eleftheriadis IC. UV-absorbers in the dyeing of polyester with disperse dyes. Dyes Pigments 2004;61:141–7. Blatny P, Fischer C-H, Rizzi A, Kenndler E. Linear polymers applied as pseudo-phases in capillary zone electrophoresis of azo compounds used as textile dyes. J Chromatogr A 1995;717:157–66. Hamai S, Sato K. Capillary electrophoretic and spectrophotometric investigations of the complexation of Methylene Blue with 2naphthol-6-sulfonate and 1,2-naphthoquinone-4-sulfonate in solution. Dyes Pigments 2003;57:15–20. Dasgupta PK, Genfa Z, Li J, Boring CB, Jambunathan S, Al-Horr R. Luminescence detection with a liquid core waveguide. Anal Chem 1999;71:1400–7. Doerge DR, Churchwell MI, Gehring TA, Pu YM, Plakas SM. Analysis of malachite green and metabolites in fish using liquid chromatography atmospheric pressure chemical ionization mass spectrometry. Rapid Commun Mass Spectrometry 1998;12:1625–34. Holcˇapek M, Jandera P, Prˇikryl J. Analysis of sulphonated dyes and intermediates by electrospray mass spectrometry. Dyes Pigments 1999;43:127–37. Matysik F-M. Potentialities of electrochemical detection in conjunction with non-aqueous capillary electrophoresis. Electrochim Acta 1998;43:3475–82. Matysik F-M. Studies on water-resistant dye compounds by means of nonaqueous capillary electrophoresis with electrochemical detection. Electroanalysis 1999;11:1017–21. Dossi N, Piccin E, Bontempelli G, Carrilho E, Wang J. Rapid analysis of azo-dyes in food by microchip electrophoresis with electrochemical detection. Electrophoresis 2007;28:4240–6. Jaffrezic N, Souteyrand E, Martelet C, Cosnier S, Labbé P, Pijolat C, editors. Les capteurs chimiques, CMC2, Écully; 1996. Bailey PL. Analysis with ion-selective electrodes. London: Heyden; 1976. Le Goff T, Braven J, Ebdon L, Scholefield D. High-performance nitrateselective electrodes containing immobilized amino acid betaines as sensors. Anal Chem 2002;74:2596–602. Bakker E, Bühlmann P, Pretsch E. Carrier-based ion-selective electrodes and bulk optodes. 1. General characteristics. Chem Rev 1997;97:3083–132. Bühlmann P, Pretsch E, Bakker E. Carrier-based ion-selective electrodes and bulk optodes. 2. Ionophores for potentiometric and optical sensors. Chem Rev 1998;98:1593–687. Khan AA, Baig U. Polyacrylonitrile-based organic–inorganic composite anion-exchange membranes: preparation, characterization and its application in making ion-selective membrane electrode for determination of As(V). Desalination 2012;289:21–6. Park J-S, Choi J-H, Woo J-J, Moon S-H. An electrical impedance spectroscopic (EIS) study on transport characteristics of ionexchange membrane systems. J Colloid Interface Sci 2006;300:655–62. Sakly N, Touzi H, Ben Ouada H, Jaffrezic-Renault N, Marie E, Chevalier Y. Electrical characterization of a new polymeric ionexchanging membrane for the chemical detection of anions. Mater Sci Eng C 2002;21:15–23. Touzi H, Sakly N, Kalfat R, Sfihi H, Jaffrezic-Renault N, Rammah MB, et al. Grafting of anion exchanging groups on SiO2/Si structures for anion detection in waters. Sensor Actuator B: Chem 2003;96: 399–406. Touzi H, Chevalier Y, Kalfat R, Ben Ouada H, Zarrouk H, Chapel J-P, et al. Elaboration and electrical characterization of silicone-based anion-exchange materials. Mater Sci Eng C 2006;26:462–71. Sakly N, Chevalier Y, Ben Ouada H, Jaffrezic-Renault N. Surface modification of Si/SiO2 by polymeric anion-exchanging membrane:

[35]

[36]

[37] [38]

[39] [40]

[41] [42]

[43]

[44]

[45] [46] [47]

[48]

[49]

[50]

[51]

[52]

[53]

[54]

[55]

[56] [57]

[58]

[59]

157

effect on interfacial morphology and electrochemical properties. Mater Sci Eng C 2008;28:923–31. Kolytcheva NV, Petrukhin OM, Filipieva NV, Müller H. Long-lived potentiometric sensors with poly(vinylchloride) and photopolymerized polyacrylate matrix anion sensitive membranes, applied on field effect transistors and silver coated ceramic substrates. Sensor Actuator B: Chem 1998;48:491–500. Ohshima H, Baiju H, Hyono A. A theory of overshoot phenomena in an ion-exchange membrane system. Colloid Surface B: Biointerfaces 2004;39:53–6. Collins KD, Washabaugh MW. The Hofmeister effect and the behaviour of water at interfaces. Quart Rev Biophys 1985;18:323–422. Cacace MG, Landau EM, Ramsden JJ. The Hofmeister series: salt and solvent effects on interfacial phenomena. Quart Rev Biophys 1997;30:241–77. Kunz W, Lo Nostro P, Ninham BW. The present state of affairs with Hofmeister effects. Curr Opin Colloid Interface Sci 2004;9:1–18. Sollner K, Shean GM. Liquid ion-exchange membranes of extreme selectivity and high permeability for anions. J Am Chem Soc 1964;86:1901–2. Morf WE. The principles of ion-selective electrodes and of membrane transport. Amsterdam: Elsevier; 1981 [chapter 11]. Hodinárˇ A, Jyo A. Contribution of membrane components to the overall response of anion carrier based solvent polymeric membrane ion-selective electrodes. Anal Chem 1989;61:1169–71. Bakker E, Malinowsica E, Schiller RD, Meyerhof ME. Anion-selective membrane electrodes based on metalloporphyrins: the influence of lipophilic anionic and cationic sites on potentiometric selectivity. Talanta 1994;41:881–90. Parsons DF, Boström M, Maceina TJ, Salis A, Ninham BW. Why direct or reversed Hofmeister series? Interplay of hydration nonelectrostatic potentials, and ion size. Langmuir 2010;26:3323–8. Kircheldorf HR, editor. Silicon in polymer synthesis. Berlin: Springer; 1996. Jones RG, Ando W, Chojnowski J, editors. Silicon-containing polymers. Dordrecht: Kluwer Academic Publishers; 2000. Boutevin B, Guida-Pietrasanta F, Ratsimihety A. Side group modified polysiloxanes. In: Jones RG, Ando W, Chojnowski J, editors. Siliconcontaining polymers. Dordrecht: Kluwer Academic Publishers; 2000. p. 79–112 [chapter 3]. Coqueret X, El Achari A, Hajaiej A, Lablache-Combier A, Loucheux C, Randrianarisoa L. Some aspects of the reactivity of photodimerizable esters grafted onto silicone main chain polymers. Makromol Chem 1991;192:1517–34. Zhu Z, Einset AG, Yang C-Y, Chen W-X, Wnek GE. Synthesis of polysiloxanes bearing cyclic carbonate side chains. Dielectric properties and ionic conductivities of lithium triflate complexes. Macromolecules 1994;27:4076–9. Belfield KD, Chinna C, Najjar O. Synthesis of novel polysiloxanes containing charge transporting and second-order nonlinear optical functionalities with atom economical constructs. Macromolecules 1998;31:2918–24. Nohr RS, Macdonald JG. New biomaterials through surface segregation phenomenon: new quaternary ammonium compounds as antibacterial agents. J Biomater Sci Polym Ed 1994;5:607–19. Tong B, Zhang B, Hu J, Dai R, Deng Y. Synthesis and characterization of side-chain liquid-crystalline ionomers containing quaternary ammonium salt groups. J Appl Polym Sci 2003;90:2879–86. Fitremann J, Moukarzel W, Mauzac M. Glycosilicones. In: Boileau S, Boury B, Ganachaud F, editors. Silicon based polymers. Springer; 2008. p. 181–202. Noomen A, Hbaieb S, Parrot-Lopez H, Kalfat R, Fessi H, Amdouni N, et al. Emulsions of b-cyclodextrins grafted to silicone for the transport of antifungal drugs. Mater Sci Eng C 2008;28:705–15. Putzien S, Nuyken O, Kühn FE. Functionalized polysilalkylene siloxanes (polycarbosiloxanes) by hydrosilylation catalysis and synthesis. Prog Polym Sci 2010;35:687–713. Lukevics E, Dzintara M. The alcoholysis of hydrosilanes. J Organomet Chem 1985;295:265–315. Zhang C, Laine RM. Hydrosilylation of allyl alcohol with [HSiMe2OSiO1.5]8: octa(3-hydroxypropyldimethylsiloxy)octasilses quioxane and its octamethacrylate derivative as potential precursors to hybrid nanocomposites. J Am Chem Soc 2000;122:6979–88. Lappert FM, Scott FPA. The reaction pathway from Speier’s to Karstedt’s hydrosilylation catalyst. J Organomet Chem 1995;492:C11–3. Kalfat R, Ben Ali M, Mlika R, Fekih-Romdhane F, Jaffrezic-Renault N. Polysiloxane-gel for ion sensitive membrane. Int J Inorg Mater 2000;2:225–31.

158

H. Touzi et al. / European Polymer Journal 56 (2014) 140–158

[60] Maciejewski H, Wawrzyn´czak A, Dutkiewicz M, Fiedorow R. Silicone waxes-synthesis via hydrosilylation in homo-and heterogeneous systems. J Mol Catal A: Chem 2006;257:141–8. [61] Tronc F, Lestel L, Boileau S. Polycondensation using hydrosilylation: a tool for preparing tailor-made polysiloxanes with anchoring groups. Polymer 2000;41:5039–46. [62] Guo D-J, Han H-M, Wang J, Xiao S-J, Dai Z-D. Surface-hydrophilic and protein-resistant silicone elastomers prepared by hydrosilylation of vinyl poly(ethylene glycol) on hydrosilanes-poly(dimethylsiloxane) surfaces. Colloid Surf A: Physicochem Eng Aspects 2007;308: 129–35. [63] Nandi BK, Uppaluri R, Purkait MK. Effects of dip coating parameters on the morphology and transport properties of cellulose acetate– ceramic composite membranes. J Membrane Sci 2009;330:246–58. [64] Reichardt C, Welton T. Solvents and solvent effects in organic chemistry. Weinheim: Wiley-VCH; 2011 [chapter 5]. [65] Binet C, Dumont M, Fitremann J, Gineste S, Laurent E, Marty J-D, et al. Hydrosilylation of polymethylhydrosiloxanes in the presence of functional molecules such as amines, esters or alcohols. In: Boileau S, Boury B, Ganachaud F, editors. Silicon based polymers. Springer; 2008. p. 135–51. [66] Kowalewska A, Stan´czyk WA, Boileau S, Lestel L, Smith JD. Novel polymer systems with very bulky organosilicon side chain substituents. Polymer 1999;40:813–8. [67] Noomen A, Penciu A, Hbaieb S, Parrot-Lopez H, Amdouni N, Chevalier Y, et al. Grafting b-cyclodextrins to silicone, formulation of emulsions and encapsulation of antifungal drug. In: Boileau S, Boury B, Ganachaud F, editors. Silicon based polymers. Springer; 2008. p. 163–79. [68] Chalk AJ, Harrod J. Homogeneous catalysis. II. The mechanism of the hydrosilation of olefins catalyzed by group VIII metal complexes. J Am Chem Soc 1965;87:16–21. [69] Antic´ VV, Antic´ MP, Govedarica MN, Dvornic´ PR. Kinetics and mechanism of the formation of poly[(1,1,3,3-tetramethyl disiloxanyl)ethylene] and poly(methyldecylsiloxane) by hydrosilylation. J Polym Sci A: Polym Chem 2007;45:2246–58. [70] Lin J, Chen H, Ji Y, Zhang Y. Functionally modified monodisperse core–shell silica nanoparticles: silane coupling agent as capping and size tuning agent. Colloid Surf A: Physicochem Eng Aspects 2012;411:111–21. [71] Ben Ali M, Kalfat R, Sfihi H, Ben Ouada H, Chovelon J-M, JaffrezicRenault N. Cyclodextrin-polymethylhydrosiloxane gel as sensitive membrane for heavy ion sensors. Mater Sci Eng 1998;C6:53–8.

[72] Zhai S-R, Song Y, Zhai B, An Q-D, Ha C-S. One-pot synthesis of hybrid mesoporous xerogels starting with linear polymethylhydrosiloxane and bridged bis-(trimethoxysilyl)ethane. Microporous Mesoporous Mater 2012;163:178–85. [73] Rattee JD, Breuer MM. Physical chemistry of dye adsorption. London: Academic Press; 1974. [74] Espinosa-Jiménez M, Padilla-Weigand R, Ontiveros-Ortega A, PereaCarpio R. Thermodynamic characterization of the adsorption process of Mordant Black 17 dye onto polyamide fabric. Macromol Mater Eng 2001;286:302–8. [75] Yang Y, Huda S. Comparison of disperse dye exhaustion, color yield, and colorfastness between polylactide and poly(ethylene terephthalate). J Appl Polym Sci 2003;90:3285–90. [76] Rohatgi KK, Singhal GS. Nature of bonding in dye aggregates. J Phys Chem 1966;70:1695–701. [77] Sivaraja Iyer SR, Singh GS. Studies on the aggregation of dyes in aqueous solutions. Kolloid-Z Z Polym 1970;242:1196–200. [78] Savarino P, Viscardi G, Carpignano R, Barni E. Amphiphilic acid dyes. Dyeing properties and interactions with surface-active systems. Colloid Surface 1989;35:251–60. [79] Barni E, Savarino P, Viscardi G. Dye–surfactant interactions and their applications. Acc. Chem. Res. 1991;24:98–103. [80] Guo LN, Arnaud I, Petit-Ramel M, Gauthier R, Monnet C, Le Perchec P, et al. Solution behavior of dye–surfactant associations. J Colloid Interface Sci 1994;163:334–46. [81] Thibault C, Séverac C, Mingotaud A-F, Vieu C, Mauzac M. Poly(dimethylsiloxane) contamination in microcontact printing and its influence on patterning oligonucleotides. Langmuir 2007;23:10706–14. [82] Binet C, Bourrier D, Dilhan M, Estève D, Ferrère S, Garrigue J-C, et al. First approach to the use of liquid crystal elastomers for chemical sensors. Talanta 2006;69:757–62. [83] Spannenberg A, Arndt P, Baumann W, Rosenthal U. The molecular structure of Cr[(CH2)2PMe2]3: dimethylphosphonium-bis-methylide chromium compounds as inner-phosphonium-alkyl-ate- or 2phospha-allyl-complexes. J Organomet Chem 2003;683:261–6. [84] Wojciechowski K, Kucharek M, Wróblewski W, Warszyn´ski P. On the origin of the Hofmeister effect in anion-selective potentiometric electrodes with tetraalkylammonium salts. J Electroanal Chem 2010;638:204–11. [85] Demoz A, Verpoorte EMJ, Harrison DJ. An equivalent circuit model of ion-selective membrane insulator semiconductor interfaces used for chemical sensors. J Electroanal Chem 1995;389:71–8.