Discussion A reply to the discussion by Mingshu Tang of the paper ‘‘Laboratory assessment of alkali contribution by aggregates to concrete and application to concrete structures affected by alkali–silica reactivity’’ $ M.A. Be ´rube ´ a, * , J. Duchesne a , J.F. Dorion a , M. Rivest b a De ´partement de ge ´ologie et de ge ´nie ge ´ologique, Universite ´ Laval, Sainte-Foy, QC, Canada G1K 7P4 4P5 b Direction Inge ´nierie, Hydro Que ´bec, 855 Ste-Catherine Est, 11th Floor, Montre ´al, QC, Canada H2L 4P5 Received 25 March 2003 Professor Tang questions whether or not the alkalies released by the aggregates are as aggressive with respect to ASR as alkalies released by the cement. This question constitutes the main concern of his discussion and is very critical for the design of new concrete structures. In actual fact, the aggressive ion species towards the aggregates are the OH ions rather than the alkali ions. Professor Tang certainly does not contest that aggregates can release alka- lies in the concrete pore solution, such as his question could be rewritten as it follows: ‘‘Are the alkali ions released by the aggregates increase by about the same amount the OH ions, then the pH, in the concrete pore solution?’’ Unfortu- nately, in our study, this concentration or the pH was not measured in the test solutions. However, it is our belief that the answer is yes to the above question. The mechanisms invoked hereafter are already known by Professor Tang; however, we consider important to recall them for the benefit of the readers of his discussion. In normal concrete, after a few days, the pore solution is largely dominated by Na, K, and OH ions in the form of soluble alkali hydroxides (NaOH, KOH), while all other ions (Ca +2 , SO 4 À 2 , etc.) are normally present as traces in solution. For instance, in the pore solution of a concrete made with a water to cement ratio of 0.5 and a cement containing 1% Na 2 O e , according to Diamond [1], the [OH] concentration is almost equal to the alkali (Na + +K + ) concentration and around 0.74 N (pH c 13.85) after 28 days and even more in the long term as a result of more complete cement hydration, which decreases the amount of pore water. So, the 0.7 N NaOH and KOH solutions used in the extraction tests described in our paper must be consid- ered as realistic in terms of chemical composition and pH, with the exception that solid portlandite or Ca(OH) 2 is not present, which constitutes an important source of Ca +2 and OH À ions in real concrete. Based on a number of studies such as those by Choquette et al. [2] and Ouali [3], it is a matter of fact that some common silicate minerals present in aggregates (e.g., silica minerals, feldspars, micas, clay minerals, etc.) react in highly alkaline and basic solutions and release different ionic species (Na + ,K + ,H 2 SiO 4 À 2 , Al(OH) 4 À , Ca +2 , etc.) with time. In the total absence of calcium in solution, these reactions consume OH ions, then the pH decreases, but the anion species released by the aggregates (e.g., H 2 SiO 4 À 2 , Al(OH) 4 À ) may remain in solution while satisfying to ionic equilibrium with the cations released (e.g., Na + ,K + ), and this as long as ion saturation is not achieved. For instance, in the study of Choquette et al. [2], several rock- and soil-forming silicate minerals suffered dissolution in a pure 1 N NaOH solution, while new silicate hydrates were not observed. However, in the presence of calcium in solution, even as traces, CSH, CASH, and ASR products as well, which contain calcium, silicium, and alkalies, will readily precipitate because their product of solubility is very low [2,4,5]. As discussed in more details in our reply to a discussion by Mr. Cajun Shi of our paper, also published in this issue, the formation of the above hydrates was clearly demonstrated by a number of previous studies [4,6,7] involving immersion of mineral or aggregate par- ticles in saturated Ca(OH) 2 solution (with solid lime in excess). However, the formation of such hydrates was even observed in pure KOH and KOH solutions [3,8], which suggests that calcium can be also released by aggregates. For sure, the new hydrates formed may entrap certain but limited amounts of alkalies, like the cement hydrates can do [9]. Assuming that not all alkalies released by the aggregates are entrapped in these new hydrates, the alkali concentration 0008-8846/$ – see front matter D 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.cemconres.2003.10.001 $ Cem. Concr. Res. 32 (2002) 1215– 1228. * Corresponding author. Tel.: +1-418-656-3930; fax: +1-418-656- 7339. E-mail address: berube@ggl.ulaval.ca (M.A. Be ´rube ´). Cement and Concrete Research 34 (2004) 903 – 904