Concurrent Aggregation and Deposition of TiO 2 Nanoparticles in a Sandy Porous Media NATALIA SOLOVITCH,* ,† J ´ ERO ˆME LABILLE, † J ´ ERO ˆME ROSE, † PERRINE CHAURAND, † DANIEL BORSCHNECK, † MARK R. WIESNER, ‡ AND JEAN-YVES BOTTERO † CEREGE, UMR 6635 CNRS/Aix Marseille University, Europo ˆle Me ´diterrane ´en de l′Arbois, 13545 Aix-en-Provence, Cedex 04, France. GDRI- ICEINT, and Department of Civil and Environmental Engineering, Duke University, PO Box 90287, Durham, North Carolina 27708-0287 Received January 18, 2010. Revised manuscript received May 11, 2010. Accepted May 25, 2010. The possibility of simultaneous particle aggregation and deposition in a porous medium was examined for the case of TiO 2 nanoparticles (NPs). While potential for particle aggregation is typically assumed to be negligible in porous media due to favored interactions with porous media surfaces (collectors), we show that nanoscale particle dimensions may favor aggregation kinetics, thus altering the transport and retention of these materials in saturated porous media. When surface chemistry favors nanoparticle-nanoparticle attachment ( R pp ) over nanoparticle-collector attachment ( R pc ), the rate of particle aggregation within pores may be comparable to that of deposition at ratios of collector to nanoparticle surface areas as high as 40. Aggregation of NPs in the porous media enhances NP deposition, however aggregates that are not removed will sample a smaller portion of the available pore network within the column due to size exclusion. Introduction Manufactured nanoparticles, varying in diameter from 1 to 100 nm, are widely and increasingly used in commercial nanomaterials due to novel properties that can improve the functionality of a range of commercial products. Normal use and aging of nanomaterials, accidental release, or inap- propriate disposal, are among the potential sources of the nanoparticles to the environment (1-4). The transport of nanoparticles or nanomaterial residues in porous media is of interest due to both potential for introduction of nanoparticles to aquifers and the need to understand the removal capabilities of engineered filters in water and wastewater treatment. Nanoparticle deposition and mobility in porous media is strongly dependent not only on pore size and organization, but also on the physicochemical parameters of solution chemistry (ionic strength, pH, presence of natural organic matter), nanoparticle surface properties, and flow rate. Those parameters that affect the interactions between the nano- particles and the solid media, also control particle aggrega- tion, which may subsequently influence the balance between free migration of particles and deposition (5-17). While Thomas Camp first proposed that particles may flocculate in the pore spaces of filters (18), relatively little consideration has been given to the potential for simulta- neous aggregation and deposition in porous media since deposition is often assumed to dominate aggregation. Indeed, when viewed as competing parallel reactions dependent on particle volume fraction, aggregation in the pore space of filters is predicted to occur at much slower rates than deposition since the volume fraction of the porous medium is typically an order of magnitude larger than the suspended particle volume fraction. However, from the perspective of surface area concentrations, nanoscale particles, even at moderate mass concentrations, present large surface areas for particle-particle contacts (aggregation). For example, the particle surface area of a 50 mg L -1 suspension of 30 nm nanoparticles with a specific density of 4, and filling a 350 µm pore is only a factor of 0.2 times smaller than the surface area within the pore (considered as a sphere). This ratio increases to a value of unity for a particle diameter of 6 nm. In column studies (12), high concentrations (1-6gL -1 ) of 20 nm nanoparticles of nanoscale zerovalent iron appear to favor aggregation within the porous medium, while aggrega- tion was largely absent at a lower concentration (30 mg L -1 ). Thus, simultaneous consideration of aggregation and depo- sition would appear to be warranted for nanoparticle suspensions. The relationship between aggregate size, structure, and deposition are likely to be complex. While DLVO theory (19-21) predicts decreased attachment to a porous medium with larger particle diameter, in some cases, the opposite trend has been observed (9, 20-22). In yet other studies, no effect of size was observed (19). Such discrepancies might be due to nonidealities such as surface roughness on the particles collectors (porous medium), hydrodynamic interactions, nonhomogeneity of surface charges, dynamics of colloidal interactions, physical straining or trapping of particles (10, 19-22). Simultaneous aggregation and the deposition of aggregates in porous media further complicate this picture. The transport of particle aggregates will certainly differ from that of individual nanoparticles. Indeed the permeability of porous aggregates decreases drag coefficients compared to equivalent nonporous solid spheres (23). Moreover the hydrodynamic driving of nonporous particles increases when size decreases, due to weight loss. Both of these tendencies tend to favor further aggregation and deposition of a porous aggregate, compared to the individual nonporous nanopar- ticles constituting it. In this work we consider the aggregation and deposition of nanoparticles in a porous medium in the context of TiO 2 nanoparticles. Among all manufactured nanomaterials, TiO 2 - based nanomaterials have attracted great attention due to their photocatalytic, and UV-absorbance properties, with applications that include photovoltaics, self-cleaning sur- faces, water treatment, and sunscreens. Significant releases of manufactured TiO 2 nanoparticles into aquatic environ- ments have been calculated in modeling work (3) and confirmed experimentally (24). Moreover, TiO 2 has been recently classified in the Group 2B of the potentially carcinogenic materials (IARC 2006, volume 93) by the International Agency For Research On Cancer, which high- lights the need for new evaluation of its safe use. Only few studies have examined the mobility of TiO 2 nanoparticles in porous media (5, 13, 14). They generally considered the effect of aggregation as insignificant. The * Corresponding author e-mail: solovitch-vella@cerege.fr. † Aix Marseille University. ‡ Duke University. Environ. Sci. Technol. 2010, 44, 4897–4902 10.1021/es1000819  2010 American Chemical Society VOL. 44, NO. 13, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 4897 Published on Web 06/04/2010