Global quantitative monitoring of the ion exchange balance in a chloride migration test on cementitious materials with mineral additions

Abstract

The concrete pore solution changes during chloride transfer due to the species leaching and the dissolution/precipitation of compounds. The latter could affect the material porosity and the chloride diffusion coefficient. In order to quantify these changes, the global movement of species was monitored on blended cement materials during a chloride migration test. Compartment solutions and sample pore solutions were analyzed in order to quantify the ion exchange balance. Two migration test protocols were carried out using (i) NaOH, KOH and NaCl, and (ii) synthetic seawater and synthetic pore solution. The use of slag reduces the free chloride in the porosity by about 9 times. For the Portland cement material, a precipitation of 0.3%wt of Na, 0.6%wt of K and 1%wt of Cl participated in a porosity clogging of 60%. Finally, the materials with mineral additions reduce the portlandite dissolution during chloride transfer and the calcium leaching phenomena up to 26%.

Introduction

Concrete and cement-based materials, with or without mineral additions, are porous materials that can be penetrated by external aggressive species. Among these species, we quote carbon dioxide, sulfates, de-icing or marine salts, etc. The latter present the principal risk that might affect the durability of reinforced concrete structures in marine areas. Indeed, chlorides penetrate in the cover concrete, under a concentration gradient and/or liquid pressure gradient, reach the rebar and generate their corrosion that can causes several degradations leading to a local or general ruin of the structure [[1], [2], [3]].

In the last decades, chloride transfer has been widely studied because of its important economic issues [[4], [5], [6], [7], [8], [9], [10]]. This phenomenon has been the subject of several experimental and numerical studies that were intended to propose methods and tools for predicting chloride transfer and structure service life. In fact, several methods to determine the chloride diffusion coefficient were proposed: some of them are based on the steady state migration test [[11], [12], [13]], and some consider non-steady state tests [[14], [15], [16]]. The latter allow a rapid determination of the chloride diffusivity and avoid a large part of leaching phenomena during chloride migration. Nevertheless, such methods do not consider the chemical interactions and the precipitation phenomena. Moreover, it has been shown that chloride transfer is influenced by electrochemical interactions in the layer of C-S-H (electrical double layer [[17], [18], [19], [20]]) and by chemical interactions with the anhydrous cement (C₃A and C₄AF [21,22]). The latter generate a formation of Friedel's and Kuzel's salts in the porosity. Furthermore, chloride diffusion or migration into the pore solution and species leaching could generate the portlandite dissolution [[23], [24], [25], [26]]. Sutter et al. [25] highlighted the consumption of portlandite to form new unstable compounds such as Friedel's salt and the hydrated calcium oxychloride (3CaO CaCl₂ 15H₂O). These phenomena affect the microstructure of concretes, pore solution and their chloride diffusion coefficients measured by the migration test [[27], [28], [29]]. In fact, Andrade et al. [30], and Castellote et al. [[31], [32], [33], [34]] studied the specie evolutions in the two compartments of the migration cell and the leaching of calcium and silicon accelerated by the electrical field in order to monitoring concrete degradations. Moreover, Buckley et al. [35] and Kyle et al. [36] extracted and analyzed pore solutions of a wide range of mortars and cement pastes containing sodium chloride in the mixing water. With 7.5% of chlorides by mass of cement, free chloride content of the pore solution reached a maximum of about 16%. This level is assumed to represent the saturation concentration of the solution (with pH of about 13.5). These experimental investigations did not consider: (i) all the monovalent and divalent ion exchanges of the pore solution together during the diffusion or migration of chlorides, nor (ii) the mineral additions in cementitious materials studied.

Finally, Tran et al. [37,38] developed a chloride transfer model taking into account the portlandite dissolution, salt precipitation (Friedel's and Kugel's ones) and the kinetic control to predict chloride binding in concrete. The comparison between simulation results and experimental ones underlined the need to take into account, in transport modeling, the actual pore solution composition of materials and the hydrates/pore solution interactions. However, there is a lack of data in the literature on these interactions and their effects on the chemical composition of the concrete pore solution.

In this paper, we quantify the ion exchange balance during the chloride migration test. The investigations allow studying the hydrates/pore solution interactions, multispecies leaching and the composition changes of the pore solution of blended cement pastes with blast furnace slag, fly ash and silica fume submitted to the migration test (considering all ions likely to be involved in transport during the migration test). The data will allow to better model electrochemical process and microstructure changes during transfer phenomena.

For this purpose, cement pastes were submitted to chloride migration tests. An investigation was performed on the three zones of the migration cell: the upstream compartment, the sample and the downstream compartment. On the one hand, pore solutions of materials were extracted and analyzed before and after the migration test, and on the other hand, compartments solutions were regularly analyzed. In addition, scanning electron microscopy (SEM) analyses for the materials tested were performed. The investigation allows the monitoring of diffusion/leaching/precipitation phenomena and then quantifying pore solution changes resulting from these processes. Finally, an alternative protocol of the chloride migration test was used for realistically simulate the chloride transfer through the cementitious materials exposed to seawater. Data of the chloride diffusion coefficient, free chlorides in the pore solution and leaching indicators (calcium content) provided by this modified protocol of the migration test were compared to those of the usual one based on NaOH, KOH and NaCl. This comparison should indicate the appropriateness of using this modified method which should avoid too much leaching and consider more precise interactions of the pore solution ions.