Physical properties of alkali activated aluminosilicates based on low-reactivity ceramics in room temperature conditions

Alkali activated aluminosilicates (AAA), or geopolymers, are widely studied materials because they are supposed to become a more sustainable alternative to materials based on Portland cement, such is especially concrete. AAA materials are generally produced by activation of an aluminosilicate precursor by an alkaline solution – usually solution of sodium silicate and sodium hydroxide. The capability of the prepared material to be used as construction material is commonly evaluated by means of its compressive strength. The present paper aims to broaden the AAA materials characterization to other physical properties such are porosity and thermal conductivity, since these measures are closely related to the engineering performance of the material. The waste ceramic dust was used as precursor while the solution of potassium silicate was an activator. The relationships between the above listed physical properties and obviously on the material composition were searched.


Introduction
Alkali activated aluminosilicates (AAA) are prepared from an aluminosilicate precursor and a liquid or solid alkaline activator.The highest potential application, at least in terms of volume, of AAA composite is found in civil engineering [1].The possible positive environmental effect of AAA concrete lies in the reduction of CO2 emissions compared to Portland cement production, especially when secondary raw materials are used [2,3].While the range of secondary raw materials used as alkaline activator is quite limited, as the precursor can be used number of aluminosilicate wastes.Some of them are conventional and widely studied in the past decades like ground granulated blast furnace slag (GGBFS) or coal flay ash (FA).Nevertheless with ongoing decline in coal energy and iron production in Europe it seems to be reasonable to pay attention to another secondary aluminosilicates [4].Within this range of potential precursors, one may mention another types of slag (from steel and non-ferrous metals metallurgy), red mud from alumina production, ashes from biomass combustion or number of sludges or sediments containing clay minerals [5,6].The last group of potential secondary precursors is likely the most voluminousconstruction and demolition waste (CDW).Here one can distinguish various fractionsin terms of size and especially in terms of composition (concrete, ceramics, mixture etc.).The alkaline activation of powdered concrete debris provides very poor strength [7] but the waste ceramics works very well [8].It is due to relatively high content of thermally activated clay minerals, i.e. the ceramic may be considered as a sort of "thermally activated clay" what is another broadly studied alkaline activation precursor (especially metakaolin) [9].But, there is one important difference between waste ceramic materials and intentionally calcined claysthe waste ceramic materials contain not only thermally decomposed clay minerals but also number of crystalline minerals which are "diluting" the amorphous matter coming from the clay dehydroxylation.This is important because the amorphous matters are more reactive in (among other processes) alkaline activation than crystalline minerals.Most of the conventional precursors are containing about 90 % of amorphous aluminosilicate matter but the ceramic wastes are much poorer in this term [10].It means that alkaline activation of a ceramic waste is challenging due to lower content of reactive species.The present paper can be understood as a continuation of research presented in [11]; there a waste ceramic precursor was activated by means of sodium-based alkaline solution (mixture of sodium silicate and sodium hydroxide) while in this paper, potassium-containing activator has been used.The goal was to evaluate feasibility of the ceramic activation by K-activator.

Experimental
The waste ceramic powder, generated in brick plant in Libochovice, Czech Republic, was used as precursor for alkaline activation.Its chemical composition (examined by XRF spectroscopy) is interesting, when compared by "common" red-clay ceramics, by relatively high content of CaO (table 1), what is due to the composition of the loess in Libochovice [12,13].The phase composition of the precursor (table 2) represents a typical red-clay ceramicsit contains quartz and number of silicate minerals.The" amorphous phase" represents the matter generated during the bricks firing from the clay minerals which undergo dehydration and dehydroxylation.The XRD results were acquired by help of PANalytical Aeris diffractometer equipped with CoKα tube.The amorphous content was determined by help of internal standard (20% ZnO); the data were evaluated by help of Profex software [14].The used precursorwaste ceramic powderis obtained in process of brick blocks grinding to a given height.It has thus rather broad granulometry since it contains both fine grinding particles as well as the larger shards (up to 4 mm).The grading curve of the precursor was published elsewhere [11].
One of partial goal of this study was to evaluate the effect of the precursor maximum particle size (all the particles bellow the given sieve remained in the precursor).Hence the precursor powder "as received" was sieved in order to remove coarser particles.Thus four grades of precursor were acquired denoted as follows: AR for "as received" (just the coarsest particles above 4 mm were removed), 0.125 for precursor sieved by sieve 0.125 mm and 0.5 and 1.0 sieved by the respective sieves.All these precursor powders were used for mixing of AAA composites according to recipe in table 3.This specific proportion of components resulted from the optimization experiment published in [15].The difference between previous studies [11,15] lies in fact that in this work, potassium-based activator was used.Specifically, the K 2 O•SiO 2 solution (water glass) containing 26 % by mass of K 2 O and 28.5 % by mass of SiO 2 was used.The particles of sand were 0-2 mm large.The mortars mixed according to table 3 were poured to 160 x 40 x 40 mm 3 molds and stored in laboratory conditions for 28 days (the quick waster evaporation was prevented by a plastic foil).Than flexural and compressive strength were determined.The basic physical properties were determined by helium pycnometry (density) and vacuum saturation method (bulk density and porosity).Selected thermal properties were measured by help of portable device Isomet 2104 based on heat pulse method.Specifically, thermal conductivity λ, specific heat capacity c and thermal diffusivity a were measured.

Results and discussion
Figure 1.Density, bulk density and porosity of AAA composites.
The differences in (matrix) density of mortars prepared with different precursors are just negligible (figure 1).Certain trend may be observed in values of bulk density and porosity; the increasing particle size causes the slight increase of porosity.The porosity is, not only in building materials, traditionally linked mechanical, thermal and durability properties.Among these, the mechanical properties are of primary importance in civil engineering when the AAA are supposed to replace the OPC based concrete.The compressive strength of geopolymer concrete lies within a broad range between 30-80 MPa (obviously some extremes may be also found in literature) [16], what corresponds to most of the OPC-concrete production.The flexural strength is between 2-12 MPa, what is again comparable with conventional concrete.The mortars prepared from waste ceramic precursor reached after 28 days of curing compressive strength 39-51 MPa (figure 2), what may be understood reasonable range for construction purposes.Specifically the highest compressive strength was measured for the finest precursor "K 0.125" and the growing coarseness caused the decrease of strength.The flexural strength basically followed the same pattern.The performed experiments does not allow to specify the reason for such behaviour but one may assume that coarser precursor features lower reactivity than the finer; such results was obtained be help of isothermal calorimetry in [11].The results of thermal properties investigation (table 4) shows that there is certain effect of increasing porosity to thermal conductivity and diffusivity but the measured values are very close to each other.More interesting is the comparison of values in table 4 for AAA with the same measures obtained for concrete [17].The comparison indicates that Ordinary Portland Cement based concrete has lower thermal conductivity/diffusivity whileinterestinglyit has had also lower porosity.Together with the lower specific heat capacity of conventional concrete, it implies that thermal behavior of alkaliactivated material is rather different from the "common concrete".It is likely caused by different nature of alkali activated materials where the binder phase is formed from K-A-S-H/C-A-S-H gel while in cementitious system, the C-S-H gel is the binding phase.As it was mentioned in Introduction, the present results can be considered as continuation of studies published in [11].There Na-based activator (of the same composition) was used for activation of the same precursors.It may be stated that the strength followed the same pattern on both systems but the K-activator reached strength higher by about 20% compared to Na-activator.To answer what is the reason of this behavior, is not easy.From the literature is known that K-activator is causing less Compressive Flexural ordered structure of alkali-activated matter compared to Na + [18,19].This leads to higher porosity and lower strengthat least in case of activated fly ash [19].It means the waste ceramic powder behaves in the opposite manner.The explanation of this observation is motivation for the further research.

Conclusions
The performed experiments revealed that waste ceramic powder with low content of amorphous matter (28%) can be activated by potassium based alkaline activator.The influence of particle size distribution on the physical properties of activated product was investigated.There is not a stronglypronounced effect of the precursor granulometry on the density and porosity of the activated mortar, but influence of granulometry on the strength may be considered as significant.The coarser precursor reduced both compressive and flexural strength, likely due to the lower reactivity of coarser particles of precursor.When the efficiency of K + activator with Na + one is compared, potassium-system provides about 20% higher strength.The reason is not clear, but it is known that kind of cation in AAA is influencing the microstructure and porosity of the matter.This influence of potassium may be negative, as in fly ash activation, but also positive, as it is reported in the present paper.Definitely, the explanation of cation importance is worth of study.

Figure 2 .
Figure 2. Compressive and flexural strength of AAA composites.Note the auxiliary axis for flexural strength.

Table 1 .
Chemical composition of the used waste ceramic precursor (% by mass).

Table 2 .
Phase composition of the used waste ceramic precursor (% by mass).

Table 3 .
Composition of AAA composites (in g).

Table 4 .
Composition of AAA composites (in g).