Langmuir 2006, 22, 6923-6930
6923
Effects of Electrolytes on Adsorbed Polymer Layers: Poly(ethylene oxide)-Silica System Charlie Flood* and Terence Cosgrove* School of Chemistry, UniVersity of Bristol, Cantock’s Close, Bristol BS8 1TS, U.K.
Ian Howell and Patricia Revell UnileVer Research, Port Sunlight Laboratory, Quarry Road East, Bebington, The Wirral, CH63 3JW, U.K. ReceiVed March 17, 2006. In Final Form: May 23, 2006 The effects of various electrolytes on the adsorption of poly(ethylene oxide) onto silica have been studied. The salts were the chlorides of Na+, Mg2+, Ca2+, and La3+. The methods used were adsorption isotherms, found using a depletion method with phosphomolibdic acid, photon correlation spectroscopy, and solvent relaxation NMR. All the salts increased the particle-polymer affinity and adsorbed amount according to the adsorption isotherms, and a linear relationship was found between the initial slope of the isotherms and the ionic strength of the solution. Final adsorbed amounts were approximately 0.4-0.5 mg m-2. The polymer layer thicknesses as found by PCS were of the same order as the radius of gyration of the polymer and increased with both the concentration and the valency of the salt due to increased adsorption. Solvent relaxation NMR showed that NaCl is too weak to have a noticeable effect on the polymer train layer, but the divalent salts clearly did increase both the strength of solvent binding close to the silica surface and the amount of PEO required to reach the maximum train density.
Introduction This study uses adsorption isotherms, dynamic light scattering (DLS), and solvent relaxation nuclear magnetic resonance (NMR) to focus on how metal ions affect the structure of physisorbed PEO at the silica-water interface. Surface modification using polymers and polymeric surfactants is an important field and has been the subject of a considerable amount of previous research,1-7 with applications including paints, detergents, and conditioners for hair, skin, and fabric. The adsorption of PEO at interfaces has been studied previously in some detail,8-10 particularly with silica,11-25 but in contrast, the effects of salt on the system have * Corresponding authors. E-mail:
[email protected] and
[email protected]. (1) Goddard, E. D. J. Colloid Interface Sci. 2002, 256, 228. (2) Somasundaran, P.; Krishnakumar, S. Colloids Surf., A: Physicochem. Eng. Aspects 1997, 123-124, 491. (3) Somasundaran, P.; Cleverdon, J. Colloids Surf. 1985, 13, 73. (4) Cosgrove, T.; Griffiths, P. C.; Lloyd, P. M. Langmuir 1995, 11, 1457. (5) Goodwin, J. Colloids and Interfaces with Surfactants and Polymers; Wiley: New York, 2004. (6) Atkins, P. W. Physical Chemistry, 6th ed.; Oxford: New York, 1998. (7) Colloid-Polymer Interactions; Farinato, R. S., Dubin, P. L., Eds.; Wiley: New York, 1999. (8) Mears, S. J.; Cosgrove, T.; Obey, T.; Thompson, L.; Howell, I. Langmuir 1998, 14, 4997. (9) Wesley, R. D.; Cosgrove, T.; Thompson, L.; Armes, S. P.; Baines, F. L. Langmuir 2002, 18, 5704. (10) Wesley, R. D.; Cosgrove, T.; Armes, S. P.; Billingham, N. C.; Baines, F. L. Langmuir 2000, 16, 4467. (11) Cosgrove, T.; Mears, S. J.; Obey, T.; Thompson, L.; Wesley, R. D. Colloids Surf., A: Physicochem. Eng. Aspects 1999, 149, 329. (12) Mears, S. J.; Cosgrove, T.; Thompson, L.; Howell, I. Langmuir 1998, 14, 997. (13) Nelson, A.; Jack, K. S.; Cosgrove, T.; Kozak, D. Langmuir 2002, 7, 2750. (14) Cosgrove, T.; Mears, S. J.; Thompson, L.; Howell, I. ACS Symp. Ser. 1995, 615, 196. (15) Otsuka, H.; Esumi, K.; Ring, T. A.; Li, J. T.; Caldwell, K. D. Colloids Surf., A: Physicochem. Eng. Aspects 1996, 116, 161. (16) Bury, R.; Desmazieres, B.; Treiner, C. Colloids Surf., A: Physicochem. Eng. Aspects 1997, 127, 113. (17) Fleming, B. D.; Wanless, E. J.; Biggs, S. Langmuir 1999, 15, 8719. (18) Otsuka, H.; Esumi, K. Langmuir 1994, 10, 45. (19) Shubin, V. Langmuir 1994, 10, 1093. (20) Esumi, K.; Meguro, K., J. Colloid Interface Sci. 1989, 129, 217.
been largely overlooked, even though common household products are often used with water containing a significant level of metal ions. There is currently interest in whether divalent cations have any unique effects, and so here we compare the chlorides of four ions with three different levels of charge; Na+, Ca2+, Mg2+, and La3+. Polymers and polyelectrolytes are often added to colloidal dispersions in order to stabilize the particles and hence prevent them from aggregating. In general, polymers will tend to sterically stabilize a colloid if there is a large adsorbed amount and a high layer thickness, provided there is no desorption during particle contact and the adsorbed polymer is in a good solvent. Copolymers are often better stabilizers than homopolymers because the solvent can be chosen to be poor for one-half the polymer, so that it adsorbs, but good for the other half, so that it forms an extended, stabilizing layer. Spectroscopic techniques have been used to show that polymer conformations at surfaces are influenced by factors including the nature of the solvent as well as the ionic strength and pH of the solution. The molecular weight of the polymer and the porosity of the substrate are also important parameters. With regard to this particular system it has been found that, in general, as PEO is nonionic, its interactions with surfaces are dominated by hydrogen bonding, much like many nonionic surfactants. It is able to adsorb onto silica fairly readily, with adsorbed amounts typically in the range 0.4-1.0 mg m-2.8-14,26 An important consideration when studying colloidal systems in the presence of electrolytes is the possibility of coagulation, which can be quantified using the Schultz-Hardy rule. This states that polyvalent ions have a greater effect on the coagulation (21) Sjo¨berg, M.; Bergstro¨m, L.; Larsson, A.; Sjo¨stro¨m, E. Colloids Surf., A: Physicochem. Eng. Aspects 1999, 159, 197. (22) Guo, W.; Uchiyama, H.; Tucker, E. E.; Christian, S. D.; Scamehorn, J. F. Colloids Surf., A: Physicochem. Eng. Aspects 1997, 123-124, 695. (23) Arnold, G. B.; Breuer, M. M. Colloids Surf. 1985, 13, 103. (24) Duffy, D. C.; Davies, P. B. Langmuir 1995, 11, 2931. (25) Szczypa, J.; Chibowski, S. Colloids Surf. 1981, 3, 393. (26) Wesley, R. D.; Cosgrove, T.; Thompson, L. Langmuir 1999, 15, 8376.
10.1021/la060724+ CCC: $33.50 © 2006 American Chemical Society Published on Web 06/27/2006
6924 Langmuir, Vol. 22, No. 16, 2006
Flood et al.
of colloids; in fact the critical coagulation concentration (ccc) is inversely proportional to the square of the valency, z.27
ccc(mol‚L-1) ∝
ψ04 z2
(1)
This assumes that the coagulating particles have low surface potentials, ψ0. Without this assumption the ccc would depend inversely on the sixth power of the valency, but in reality this is not observed because surfaces with high potentials do not usually coagulate. In any case, it is clear that coagulation depends strongly on the amount and type of electrolyte added. Specific ion effects were first considered by Hofmeister in the late 19th century, who developed a series for cations and anions based on their relative tendencies to encourage the salting out of proteins. The series has been shown to apply to many phenomena relating to the coagulation of colloids, but although charge, geometry and polarizability have all been shown to play a role, the origins of Hofmeister effects have never been fully identified, possibly because conventional theories tend to underestimate the degree of solvent structure. Current thinking on the subject was discussed thoroughly in a recent special issue of Current Opinion in Colloid and Interface Science, which was devoted entirely to specific ion effects.28 Articles of particular interest within this issue review the topics of ‘Water and Ions at Interfaces’29 and ‘Ion Binding to Interfaces’30 The Hofmeister series suggests that for the chlorides used in our study trivalent La3+ ions should be the most destabilizing, followed by Mg2+, Ca2+, and finally the most weakly hydrated cation Na+. This order may however be sensitive to pH, temperature, and changes in the anion. The phase behavior of aqueous PEO with salts in the absence of particles must also be considered, and this is one aspect that has been studied previously in detail.31-36 These papers show that dilute solutions of PEO have cloud points close to 100 °C, but these can be reduced significantly by salt addition. In general, salts containing ions of high charge density tend to reduce the cloud point more markedly. Systematic measurements of the cloud point of PEO were carried out by Florin et al. in the presence of five alkali metal chlorides and four different potassium halides.32 All the chlorides reduced the cloud point by similar amounts, except lithium chloride, which was only about onehalf as effective. Potassium iodide only had a relatively small effect, but on ascending group 17 the reduction in cloud point became more pronounced to the extent that a 1 wt % PEO solution containing 1.5 mol kg-1 potassium fluoride is unstable at room temperature. The authors proposed a model with enhanced water structure and decreased salt concentration around each chain. They suggested that the salt-free region was a result of hydration of ions near the polymer and that this effect was sufficient to account for the variation in cloud point. The difference between the salts was explained in terms of their relative abilities to penetrate the region close to the polymer chains. This can be (27) Hunter, R. J. Foundations of Colloid Science; Oxford University Press: New York, 2001. (28) Curr. Opin. Colloid Interface Sci. 2004, 9, 1-2. (29) Netz, R. R. Curr. Opin. Colloid Interface Sci. 2004, 9, 192. (30) Wennerstrom, H. Curr. Opin. Colloid Interface Sci. 2004, 9, 163. (31) Bailey, F. E.; Callard, R. W. J. Appl. Polym. Sci. 1959, 1, 56. (32) Florin, E.; Kjellander, R.; Eriksson, J. C. J. Chem. Soc., Faraday Trans. 1 1984, 80, 2889. (33) Pang, P.; Englezos, P. Colloids Surf., A: Physicochem. Eng. Aspects 2002, 204, 23. (34) Boucher, E. A.; Hines, P. M. J. Polym. Sci.: Polym. Phys. Ed. 1976, 14, 2241. (35) Napper, D. H. J. Colloid Interface Sci. 1970, 32, 106. (36) Napper, D. H. J. Colloid Interface Sci. 1970, 33, 384.
correlated to the more recent findings of Pang and Englezos, who suggest that any factor, such as electrolyte, able to perturb the oriented solvent layer around a polymer chain is likely to favor chain-chain interactions and hence induce phase separation.33 They also note that structured solvent is entropically unfavorable, an effect which dominates at high temperature and hence contributes to the temperature dependence of phase separation. In another study by Boucher and Hines34 magnesium and calcium chloride were used to show that divalent ions generally reduce the cloud point of PEO more than monovalent cations. Finally, Pincus and co-workers discussed the phase behavior of PEO as a function of temperature and pressure37 as well as considered counterion-induced collapse of polyelectrolytes.38 The effect of trivalent ions on PEO has so far been overlooked in the literature, as largely have mixtures of PEO, salts, and surfactants. While the conditions used in this study are chosen to be far below the cloud point of PEO, the presence of salt still affects the solubility of the polymer and hence its adsorption isotherms. The substrate used in this study is colloidal silica, which is made using the Stober process, a method for growing silica spheres with relatively low polydispersity.39 The dispersions are prone to flocculation at pH values of ∼4-6. In alkaline solutions, the hydroxyl groups at the surface tend to be deprotonated. This leads to formation of an electrical double layer which stabilizes the particles. At pH values less than 4, the hydroxyl groups are protonated and, via hydrogen bonding, surrounded by a structured water layer. The energy required to disrupt this layer is sufficient to prevent coagulation. In this study the silica stock solution was dialyzed before use and the pH was then measured at 8.1 ( 0.1.
Theory Adsorption Isotherms. Isotherms were determined using a depletion method. This involved allowing adsorption to occur and then removing the silica, coated with polymer, by centrifugation. The concentration of polymer remaining in the supernatant solution was determined using the phosphomolibdic acid (PMA) method, and from this the adsorbed amount was calculated. The technique is based on a method discovered by Shaffer and Critchfield in which PEO is reacted with an excess of PMA and the resulting precipitate is weighed.40 This was developed further by Stevenson, who redissolved the precipitate in concentrated sulfuric acid and measured the absorption of the resulting solution at 520 nm.41 The modern preferred method involves removing the precipitate by centrifugation and then measuring the absorption of the supernatant solution using UV spectroscopy at 216 nm, a wavelength at which the unreacted phosphomolibdic acid absorbs strongly. PEO is reacted initially with barium ions in the presence of acid, resulting in a complex with a 2+ charge. This is reduced by PMA in a redox reaction which yields an oxidized precipitate of PMA which no longer absorbs at 216 nm.42 PCS. Dynamic light scattering, or photon correlation spectroscopy (PCS), has yielded some particularly interesting results on PEO-particle systems. The technique utilizes light scattered due to the Brownian motion of diffusing particles. Analysis of the spectral width of the scattered light using a correlation function allows calculation of particle size and, in the case of polymer layers, hydrodynamic thickness.27 (37) Bekiranov, S.; Bruinsma, R.; Pincus, P. Phys. ReV. E 1997, 55, 1. (38) Schiessel, H.; Pincus, P. Macromolecules 1998, 31, 7953. (39) Stober, W.; Fink, A.; Bohn, E. J. Colloid Interface Sci. 1968, 26, 62. (40) Shaffer, C. B.; Critchfield, F. H. Anal. Chem. 1947, 19, 32. (41) Stevenson, D. G. Analyst 1954, 79, 504. (42) Dale, P. Ph.D. Thesis, University of Bristol, 2004.
Effects of Electrolytes on Adsorbed Polymer Layers
Langmuir, Vol. 22, No. 16, 2006 6925
NMR. Solvent relaxation NMR was used to provide insight into the portion of the adsorbed polymer layer close to the particle surface. Plots of the specific relaxation rate of the solvent, R2sp, against the silica concentration are in some ways comparable to isotherms, providing information on the density of the polymer train layer and a measure of the total adsorbed amount. The study of silica/PEO/salt interactions had to be preceded by a series of calibrations involving the various components alone and in pairs. The technique takes advantage of the fact that protons within water molecules bound at an interface have a shorter spin-spin magnetic relaxation time, T2b, than the time for those free in solution, T2f. The T2f value for protons within free water molecules is around 2 s, but at interfaces spin-spin relaxation processes are efficient enough to give T2b values typically MgCl2 . NaCl. Adsorption isotherms showed that the presence of salt perhaps slightly increased the final adsorbed amount but most noticeably increased the affinity of the isotherms. Interestingly, a linear relationship was identified between the initial slope of the isotherms and the ionic strength of the solutions. After flocculation problems were overcome, PCS results showed that salts increased the thickness of the adsorbed layer, most probably primarily due to an increased adsorbed amount originating from the reduced solubility of the polymer. Layer thickness increased with increasing salt concentration and valency, with CaCl2 and MgCl2 giving very similar results. PCS and solvent relaxation NMR results indicated that ionic strength was not the only factor in play and that ion binding
6930 Langmuir, Vol. 22, No. 16, 2006
probably plays a role for polyvalent cations if not for Na+ which did not affect the polymer train density. Ca2+, Mg2+, and La3+ were all found by NMR to increase the strength of binding of solvent molecules within the train layer and increase the amount of polymer required to reach the ‘break point’, beyond which the train density increases no further. Ca2+ had a slightly greater effect on train density than Mg2+, but in other ways the effects of the two salts were similar. In summary, ionic strength has the biggest influence on the salt effects identified, but specific ion effects must also be
Flood et al.
considered in order to fully explain the results. The increased layer thicknesses and adsorbed amounts observed help to oppose the electrostatic screening effect of salts which could otherwise induce coagulation at lower ionic strengths. Acknowledgment. We thank Unilever and the EPSRC for funding and Youssef Espidel for his help with the NMR experiments. LA060724+
