Association of dissolved aluminum with silica: Connecting molecular structure to surface reactivity using NMR

Abstract

We studied uptake mechanisms for dissolved Al on amorphous silica by combining bulk-solution chemistry experiments with solid-state Nuclear Magnetic Resonance techniques (²⁷Al magic-angle spinning (MAS) NMR, ²⁷Al{¹H} cross-polarization (CP) MAS NMR and ²⁹Si{¹H} CP-MAS NMR). We find that reaction of Al (1 mM) with amorphous silica consists of at least three reaction pathways; (1) adsorption of Al to surface silanol sites, (2) surface-enhanced precipitation of an aluminum hydroxide, and (3) bulk precipitation of an aluminosilicate phase. From the NMR speciation and water chemistry data, we calculate that 0.20 (±0.04) tetrahedral Al atoms nm⁻² sorb to the silica surface. Once the surface has sorbed roughly half of the total dissolved Al (∼8% site coverage), aluminum hydroxides and aluminosilicates precipitate from solution. These precipitation reactions are dependent upon solution pH and total dissolved silica concentration. We find that the Si:Al stoichiometry of the aluminosilicate precipitate is roughly 1:1 and suggest a chemical formula of NaAlSiO₄ in which Na⁺ acts as the charge compensating cation. For the adsorption of Al, we propose a surface-controlled reaction mechanism where Al sorbs as an inner-sphere coordination complex at the silica surface. Analogous to the hydrolysis of $\text{Al}({\text{OH}}_2)_6{}^{3+}$, we suggest that rapid deprotonation by surface hydroxyls followed by dehydration of ligated waters results in four-coordinate (>SiOH)₂Al(OH)₂ sites at the surface of amorphous silica.

Introduction

Aluminum plays a key role in mineral reactivity in a wide range of environments that include the formation of hydrous aluminosilicates from natural waters, preservation of deep sea sediments, and precipitation of geothermal scales. Doucet et al. (2001) found that hydroxyaluminosilicate (HAS) phases form by competitive condensation of silica on a hydroxy-aluminum template and that these HAS phases have similar ²⁷Al MAS NMR spectra to HAS found in soil horizons (Barron et al., 1982). The structural similarities between precipitated HAS and those found in soils suggest that the precipitation of HAS may regulate the bioavailability and thus toxicity of aqueous Al in natural waters (Doucet et al., 2001). As another example, Van Cappellen et al. (2002) examined the solubility of diatom cultures and siliceous sediments containing different amounts of Al and found that bulk solubilities typically decreased by an order of magnitude when the Al/Si increased by a factor of two. It was shown using X-ray Absorption Near Edge Spectroscopy (XANES) that Al in diatoms had been incorporated into 4-fold coordination (Gehlen et al., 2002). This suggests that incorporation of tetrahedral Al into the diatom framework during synthesis is most likely the reason that dissolution rates of biogenic opal in deep-sea sediments are reduced by over an order of magnitude relative to surface waters. Lastly, Carroll et al. (1998) measured silica precipitation rates for amorphous silica in complex geothermal brines at the Wairakei geothermal energy plant (New Zealand) and found that rates were much faster than in laboratory studies (pH 7–8, reaction time 48 h). They suggested that the accelerated silica precipitation rates found in the field were due to precipitation of Al-containing phases. Gallup (1997) reported an ²⁷Al MAS NMR spectrum for an Al-rich silica scale from Awibengkok geothermal field which showed only tetrahedrally coordinated Al (²⁷Al MAS NMR, ^[4]Al = 52.5 ppm; uncorrected for second-order quadrupolar shift), most likely due to aluminosilicate precipitation in which the Al had been supplied by the geothermal brine.

These examples illustrate the importance of understanding the connection between structure and reactivity. Due to the complex structural arrangement of natural systems, however, spectroscopic studies are often difficult to perform and interpret. Much of our understanding of metal cation sorption and precipitation comes from changes in bulk, aqueous measurements. While extremely useful, bulk-chemistry studies lack molecular-scale information which allows for the identification of reaction mechanisms that control solubility.

The motivation behind our study is to explain silicate surface reactivity in terms of molecular structure using NMR spectroscopy. We use NMR to identify the structural form of Al associated with amorphous silica over a pH range similar to that of natural waters. Because amorphous silicas are often used as a proxy for natural silicas, we anticipate the surface chemistry to be analogous to that of natural phases. The speciation information we obtain from NMR combined with the water chemistry data allow us to identify three reaction mechanisms that describe the solubility of Al and surface reactivity of amorphous silica.