Enrichment of aluminium in the near‐surface region of natural quartzite rock after aluminium exposure

Alkali–silica reaction (ASR) is an ongoing problem that causes damage to concrete constructions and reduces their durability. Therefore, minimizing this undesired reaction is of great interest for both safety and economic reasons. Additives containing high aluminium content are very effective in reducing the release of silica and enhancing the durability of concrete; however, the mechanism for this effect is still under discussion. In this study, an enrichment of aluminium in the near‐surface region was observed for natural quartzite rock after storage in Al (OH)3 and metakaolin as aluminium sources, from which we conclude that the formation of aluminosilicate sheets of a few nanometres inhibits the silica release; this hypothesis is supported by high‐resolution spectra of Al 2p, Si 2p and O 1s.

Alkali-silica reaction (ASR) is an ongoing problem that causes damage to concrete constructions and reduces their durability. Therefore, minimizing this undesired reaction is of great interest for both safety and economic reasons. Additives containing high aluminium content are very effective in reducing the release of silica and enhancing the durability of concrete; however, the mechanism for this effect is still under discussion. In this study, an enrichment of aluminium in the near-surface region was observed for natural quartzite rock after storage in Al (OH) 3 and metakaolin as aluminium sources, from which we conclude that the formation of aluminosilicate sheets of a few nanometres inhibits the silica release; this hypothesis is supported by high-resolution spectra of Al 2p, Si 2p and O 1s.

K E Y W O R D S
alkali-silica reaction, quartzite rock, X-ray photoelectron spectroscopy

| INTRODUCTION
Interactions between two or more phases are the basics of modern construction materials. Such interactions mainly occur through surface contact of the various phases of multicomponent systems. Therefore, the knowledge of surface properties is crucial for an understanding of these interactions.
Concrete is such a multicomponent system, consisting of the binder cement stone together with aggregates that are distributed throughout the matrix. Here, interactions between the binder and the aggregates are of great importance for achieving a dense structure and thus good strength properties. 1 On the other hand, surface properties of aggregates are responsible also for damaging reactions, which can lead to a destruction of the entire concrete structure 2 through reactions such as the alkali-silica reaction (ASR), which influences the durability of concrete; inhibiting this reaction is therefore an economically important goal. 3 One possibility to inhibit the ASR is to add so-called supplementary cementing materials (SCMs) to the concrete mixture, which are high-aluminium content additives that are very effective and have a great importance for concrete technology. 4 However, the inhibition mechanisms of SCMs are not completely understood. One effect arising from incorporation of SCMs is that the dissolution rate of silica from aggregates is strongly reduced by the presence of aluminium. This decreased release of silica can reduce the formation of undesired reaction products that are believed to lead to concrete damage. Changes of the surface proper- with an aluminium supply. For these investigations, amorphous silica (silica glass) was used. The silica substrate was split into four quarters, and each of the quarters has been treated differently. Chappex and Scrivener 8 described results of X-ray photoelectron spectroscopy (XPS) investigations with such glass materials and concluded that aluminium was incorporated into the silica framework at reactive sites and decreased the interfacial free energy, which slowed down the silica dissolution. It must be noted that differences on the surfaces with and without Al influence are visible with naked eye (see Figure 1), but only XPS was able to detect a very thin aluminium-containing layer on the surface of the silica glass. Si-Al reference samples were prepared by precipitating an aluminosilicate; sodium aluminate was added to a saturated potassium water glass solution to give a Si/Al ratio of 2.24. One sample was prepared at 20 C, and the other sample was warmed up at 80 C.

| Characterization
The XPS investigations were performed with an AXIS Ultra DLD pho-   Figure 2C. It can be seen that the amount of soluble silica, which comes from the quartzite grain, is reduced from 90 mg/L to approximately 60 mg/L. Important is that the Al concentration is near zero in this experiment, which means that the reduction of silica is not caused by an aluminosilicate formation precipitation process in the solution.
Due to the fact that many analytical methods failed so far [Grazing Incidence X-ray diffraction (GI-XRD), Fourier transform infrared spectroscopy (FTIR)], the XPS method was selected to investigate this assumed thin layer on the surface of the quartzite.

| XPS analysis
XPS analysis were performed on the pristine material and both Al (OH) 3 -and metakaolin-coated samples. Figure 3A shows  To obtain further insights into the chemical nature of the Al-and Si-containing compounds, high-resolution XPS measurements were performed to determine the binding energies of Si 2p, Al 2p and O1s photoelectrons ( Figure 4). Table 1 summarizes the results. For these experiments, both Al (OH) 3 and metakaolin as Al sources were used.
The aim of these measurements was to examine the formation of aluminosilicates; therefore, two aluminosilicate samples were measured as references.
As expected, for both treated reference samples, a higher amount of Al was found (see last two lines in Table 1   Note. The relative uncertainty range for the ratio is ±15%, for the binding energies including Δ ± 0.2 eV.