Indicaciones no aprobadas

On way to understanding the mechanism behind/theories proposed in the foregoing section, the following explanations could be given:

- Ad (1)(2) With IC introduction method used (see Section 3.2.4 for more details), some portion of concrete ingredients including binder becomes substituted with the gel-like containers. The main constituent of the latter is weakly bound water, especially when compared to state the same fluid would present in contact with cement. This fact is important for two reasons: a very high heat capacity of water [Ben 08a] and dependence of (specific) heat capacity of hardening concrete on unbound water content in concrete [Sch 05]. Taking both into account, a clear benefit from implementation of IC is therefore yielded especially in terms of concrete temperature evolution and is additional to dilution effect.

- Ad (3) The extra water for curing that is first separated from hydrating system will be subsequently released when favourable condition or trigger occurs, plausibly prior to setting. Should the concentration of hydration-relevant species (especially Ca2+) undergo decrease due to water migration from the water tanks to bulk/pore solution, greater amount of C3S must dissolve to precipitate hydrates and thus to give impulse

to hydration restart. In other words, there will be delay in time of attaining supersaturation level and thus hydrate formation. The general knowledge is that setting time is closely associated with concentration of Ca2+ in liquid phase [Uch 92]. Similarly, it is the ionic concentration in pores which governs the rate of heat evolution acc. to [Scr 11]. Thereby, upon reduction in the calcium concentration and

increasing distances between hydrating particles due to increase in effective w/c, prolongation of dormant period and retarded setting may be postulated based on [Lan 02] and [Uch 92][Ben 08a], respectively. On the other hand, the increase in water content may be useful in later stages of hydration as more particles may react. Past experience indeed shows more intense hydration (from which higher heat release may result) and higher degree of hydration upon increase of w/cm [Sch 05]. Important in respect to IC, for system of given total w/cm, retention of certain water amount and its gradual delivery instead of unrestricted availability to hydration may be the condition sufficient enough for considerably lower heat of hydration evolved acc. to Korpa et al. [Kor 08].

- Ad (4) Before hydration gains in velocity/rate and hydration truly restarts after period of negligible reactivity, the main process underway until supersaturation state is achieved is diffusion of calcium ions. Important in selfsame respect, many polymeric materials in general and hydrogels of ionic type in particular tend to bind various ionic species in presence of dissociated ionic functional groups e.g. K+ and Ca2+ in case of anionic SAP [Lam 05]. Hence, with SAP in concrete, there will be removal of calcium ions leading to their concentration depression in the liquid phase, from which a consequent hindered solid phase nucleation and growth/crystallization (C-S-H, portlandite, ettringite, e.g. [Lar 90]) and thereby retarded setting would result. This effect would be consistent with delay of setting caused by chelate production through the interaction of Ca2+ with the unadsorbed admixture remaining in the liquid phase as postulated by Uchikawa et al. [Uch 92]71. Upon binding of Ca2+ by polymeric admixtures, more time would be needed for the formation of protective membrane coating [Tho 81] (sometimes appearing as Ca-rich hydrate/surface layer on the clinker minerals [Wei et al., ref. 20 Ibid. Lan 02]), resulting in longer induction period. This interpretation and mechanism alone may still appear somewhat insufficient in view of potentially greater penetrability/permeability of retarding hydrate layer formed on surface of C3S in environment of decreased Ca2+ concentration [Kor 08].

Why resultant reduction of Ca2+ concentration does not promote acceleration of C3S

hydration and therefore hydration as such (including the process of diffusion) which,

71 _{It should be mentioned though that occurrence of complex formation phenomenon for PCE [Sow 15] has been}

with exceptions [Qin 07]72, is typically expected for low w/cm systems modified by Ca2+ consumers like silica fume and superplasticizer [Pin 99], this could be linked with specificity of SAP’s Ca-sink role. When anionic hydrogel contacts free calcium ions, there will be ionic bonding with carboxylate ions of polymer or an interaction with free atom valences. The additional cross-link formation and one of potential consequences leaves no or otherwise only few free valences of Ca2+ available. Adding to this little content of admixture, this implies that possibility of SAP to act as nucleation/precipitation site73 would be very limited. It may be plausibly restricted74 to formation/precipitation of calcium carbonate crystals (CaCO3) [Sno 12][Pou 13]

which as result of very early carbonation may imply initially accelerated hydration, as similarly to effect of CaCO3-containing admixture [Ram 96]. However, for good

efficiency of the process, carbonation also requires numerous favourable conditions to be fulfilled first, which is impossible, at least difficult, to imagine in sealed low w/c system with IC at early hydration stage, e.g. presence of well dissolved CO2,

intermediate relative humidity, and other [Ber 04]. Furthermore, on reestablishment of free calcium level in solution, a thicker and heavier protective coating around cement grains may form [Tho 81]. For hydration followed based on osmotic membrane model, this gives argument for retarded cement hydration.

Alternative explanation of the effect could be related to formation of multilayer assemblage of PCE on solid surfaces [Sow 15] but to reason other than conformation suggested in the reference. In gist, since the process of Ca2+ concentration regaining to level of system without modifying admixtures could be very rapid [Tho 81], after the first adsorption of PCE (and perhaps extractables), the hydration proceeds further: front of the hydrate products progresses from surface towards aqueous solution and covers the polymeric coating75. This delivers new solid surfaces/interfaces on which polymeric admixtures could adsorb, leading to reinitiating steric hindrance effect and extension of time needed to attain the percolation, setting etc.

72 _{Similar exception exists in system without superplasticizer, see Langan et al. [Lan 02], for instance.}

73 _{Potential role of nucleation site e.g. for the C-S-H phase from solution [Sil 06] has been reported for many}

polymers. On the other hand, this role is typically common to much smaller polymers, perhaps only of size of extractables or smaller (< 1 µm).

74 _{Because both SAP and superplasticizer are carrying functional groups, complexing effect of Al}3+ _{is possible}

(see also discussion of theory 6). This can result in precipitates of an amorphous solid, in agreement with effect of organic admixtures by Diamond [Dia 72]. Therefore, a certain simplification might be made.

75 _{Although first hydrates after precipitation on solid surfaces could be subsequently dispersed in the aqueous}

- Ad (5) Not only gain of water (see theory 3), but also its early loss may constitute of composition of bulk solution/pore solution. By absorbing portion of mixing water by the hydrophilic polymer, notably less water can directly contribute to hydration process of cement. This consequently results in less cement taking part in hydration process, meaning limited hydrolysis of cement phases [Kor 08] and dissolution of supplementary cement ingredients such as free lime or gypsum [Uch 92] which are decisive for the concentration of Ca2+. In turn, and similarly to important adsorption property of silica fume in low w/cm system [Lan 02], some delay in hydration may be expected, particularly prolongation of dormant period and reduced rate of hydration heat during acceleration period.

Similar explanation should be also applied to concentration of alkalis present in pore solution. In agreement with [Dan 62], on lowering the basic w/c, the amount of alkali released from calcium silicates (and likely other sources) notably decreases, thereby limiting rate of heat of hydration and related hydration-acceleration effect. The counterargument in own study is however little amount of alkalis as such as related to cement used.

- Ad (6) Owing to high concentration of Ca2+ ions, the content of SAP complexing them must be also sufficiently high. In mixtures with IC, this is unlikely the case given that for IC only small amount of polymeric agent is needed. On the other hand, chelate formation in the presence of carboxylic acid groups needn’t be only limited to calcium species but should be also extended for various especially multivalent cations including aluminate, ferrite and silicate ions [You 72]. Studies on SAP showed in fact that some of the ionic crosslinks between certain ions and anionic groups in the polymer network e.g. caused by Al3+ [Zhu 15] can be much stronger or even replace the one already formed due to complexing of Ca2+. Should the action between IC agent and the species take place, early precipitation of hydration products may be prevented. This could lead to more time required before hydration barriers are set up and consequently hydration delay, in accordance with hypothesis of Young [You 72]. On the other hand, Al3+ can exert poisoning effect on growth of C-S-H nuclei [Scr 11], meaning its removal will lead to earlier start of acceleration period of alite. This relates to the fact that Al-free C-S-H can readily grow unlike the calcium alumino-

silicate hydrate which precipitates rivalry when Al3+ is available, see ref. 37 ibid. [Scr 11] for more details.

Important to point out/acknowledge, as own EDX investigations on pure SAP treated with pore solutions of various compositions showed, other ions including potassium (K+) and sulphate ions (SO42-) can be adsorbed simultaneously by the polymer to form

a complex on its surface and perhaps inside the polymer network. During hydration process, this may decrease their already low concentration in pore solution, and if so, affect the progress of chemical reactions. By complexing the former, many phenomena otherwise accompanying their presence and in detail reviewed in [Jaw 78] would not occur. This would be unfavourable in view of C3S and C3A early hydration

acceleration potential of alkalis. In comparison, by complexing the latter, there will be decreasing concentration of sulphate ions in pore solution, without or with limitation of their adsorption on the reactive sites [Scr 11], causing faster dissolution of both- sulphate carriers (gypsum, …) and C3A [Jan 13]. In theory, the resultant higher

availability of Ca2+ concentration, either due to induced higher rate of alite dissolution or simply from dissolution of sulphate carriers, may trigger faster activation of start of silicate reactions, i.e. acceleration effect. However, this fails on presence of superplasticizer which, when dissolved in mixing water, is readily adsorbed, unlike the release of SO42- into liquid solution being more gradual process. That is to say, the

rate-controlling effect of sulphate ions on hydration of C3A [Jol 98] is likely overtaken

by the superplasticizer. In view of additional role ̶ complexation of Ca2+ ̶ the retardation is a result to be awaited.

Larbi and Bijen [Lar 90] found increased amount of K+ and SO42- in liquid phase of

system with organic admixtures which was proposed to be related to interaction between polymers’ charged groups and ions released by cement during hydration, including Ca2+ and OH-. Similarly, in agreement with rule of optimum SO3 content

described in [Jaw 98], for given C3A content and gypsum in cement but

simultaneously upon decreasing alkali content in system with IC/SAP, it may be expected that excessive SO42- content becomes available. Different than in [Lar 90] is

however that in system with IC/SAP this content may yet undergo changes due to potential chelate formation phenomenon as discussed above. The overall effect of the alkali and SO42- seems therefore very complex and depends likely on more parameters