Journal of Membrane Science 241 (2004) 235–248 Colloidal adhesion to hydrophilic membrane surfaces Jonathan A. Brant, Amy E. Childress ∗ Department of Civil Engineering, University of Nevada, Reno, NV 89523, USA Received 22 October 2003; received in revised form 21 April 2004; accepted 26 April 2004 Abstract Colloidal adhesion to membrane surfaces is an important parameter in determining membrane fouling propensity and in optimizing membrane cleaning strategies. It has previously been demonstrated that acid–base interactions can significantly affect colloid–membrane interaction as a colloid approaches a membrane surface, however, the effect of acid–base interactions on adhesion has received less attention. In this investigation, the approach and adhesion of a silica and polystyrene colloid was measured on three commercially available hydrophilic water treatment membranes using an atomic force microscope and the colloid probe technique. It was found that the hydrophobic polystyrene colloid adhered more weakly to each membrane compared to the hydrophilic silica colloid. These results could not be resolved through classic DLVO analysis alone and were in direct contrast to the expected interaction based on the strong hydrophobic character of the polystyrene colloid. However, the results could be explained by considering the magnitude of the surface’s electron-acceptor (γ + ) and electron-donor (γ - ) components. It is hypothesized that through hydrogen bonding with surface γ + and γ - groups, structured water layers exist to varying extents at the surfaces of the silica colloid and the hydrophilic membranes, and that their removal results in the formation of strong adhesive bonds between reciprocal γ + and γ - groups. Furthermore, even when surface roughness is substantial, γ + and γ - groups appear to play some role in determining the magnitude of the measured adhesion. The lack of such groups on the polystyrene colloid, and thus the lack of hydrogen bonding capacity, was responsible for its weaker adhesion with the membranes. © 2004 Elsevier B.V. All rights reserved. Keywords: Fouling; Reverse osmosis; Thermodynamics; Adhesion; AFM 1. Introduction Adhesion of colloidal particles to membrane surfaces is a significant problem that must be overcome in order to reduce the effects of membrane fouling. Colloid deposition and attachment (adhesion) result in the formation of a cake layer that may not be easily removed during hydraulic clean- ing. Removal of colloidal cake layers may be facilitated by hydrodynamic shear and changes in aqueous solution chemistry [1–4]. Recent models describe the detachment of colloidal particles from collector surfaces as being depen- dent on both the hydrodynamic and thermodynamic prop- erties of the system [2–5]. An accurate assessment of both hydrodynamic and thermodynamic interactions is necessary to minimize colloid attachment to membrane surfaces and to facilitate colloid detachment from membrane surfaces. In order to reduce membrane fouling, membrane surfaces must ∗ Corresponding author. Tel.: +1 775 784 6942; fax: +1 775 784 1390. E-mail address: amyec@unr.edu (A.E. Childress). provide an unfavorable environment for adhesion to occur, thus minimizing the critical hydrodynamic force [6], or the minimum force required for particle detachment from the surface [1]. Furthermore, a comprehensive understanding of the thermodynamic mechanisms necessary for particle detachment from a surface allows for the optimization of membrane cleaning processes by providing an avenue in which the critical hydrodynamic force may be reduced through adjustment of feed water chemistry (e.g., pH, ionic strength, and surfactant addition). It is widely accepted that adhesion is controlled by short-range interfacial interactions [7–9]. The summation of these interfacial interactions is the thermodynamic work of adhesion (W 132 ), which quantifies the energy available for adhesion based on the energetic properties of two in- teracting surfaces [9–11]. Although many researchers have found qualitative agreement between W 132 values and di- rect adhesion force measurements, calculated values are in many cases one to two orders of magnitude less than measured values [9,12]. This is undoubtedly due to the complex nature of the mechanisms affecting adhesion and 0376-7388/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.memsci.2004.04.036