Portland cement (IRAM 50000) (21) was used to prepare paste and mortar specimens. This cement is equivalent to CEM I of EN 197:1(22) containing up to 5% of limestone filler as minor constituent. The chemical composition was provided by the producer and the mineralogical composition was calculated by Bogue´s formulas based on clinker chemical composition. The specific gravity and Blaine fineness are 3.14 and 315 m²/kg, respectively (Table 1).

Three DCW originated by different industrial processes were used: mud-brick (MB), industrial ceramic block (CB) and sanitary ceramic ware (SW). DCWs were obtained in coarse aggregate particle size. Their particle size was further reduced to passing 4.75-mm sieve (#4) and, finally, ground in ball mill until 100% pass through 75 μm sieve (#200). For ground DCWs, the Blaine specific surface area (BSS) (23) and the particle size distribution (PSD) were determined. The PSD was determined using Malvern Mastersizer 2000 laser particle size analyzer, and the d₉₀, d₅₀, and d₁₀ diameters were calculated. The chemical composition of different ceramic wastes is given in Table 1 and it was determined by Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP-AES) analysis in an external laboratory (ALS, Argentine). The mineralogical composition of DCWs was identified by X-ray diffraction (XRD) and FT-IR spectroscopy (FT-IR). XRD analysis was performed using a Philips PW 3710 diffractometer operating with Cu Kα radiation at 40 kV and 20 mA and the FT-IR spectra were performed on NICOLET Magna 500 using KBr technique.

Nine PBCs were prepared replacing 8, 24 and 40% of each DCW by mass of Portland cement. The PBCs are identified using the replacement level and the type of ceramic waste used (CB, MB and SW). Table 2 shows the nomenclature used in this study.

Pozzolanic activity was determined by Frattini test according to the procedure described in EN196-5 at 2, 7 and 28 days (24). This test evaluates the pozzolanic activity of DCW measuring its reaction with the Ca(OH)₂ released during the PC hydration. The result is considered positive when the points are located below the calcium hydroxide solubility isotherm at 40 °C.

Mortars were made with standard siliceous sand using a sand/cement ratio of 3 and a water-to-cementitious material ratio (w/cm) of 0.50. It was mixed using the standard procedure. The workability of mortar was determined using the flow table test according to ASTM C1437 (25). For each mortar batch, the flowability was determined as the diameters of spread mortar measured along four lines in mm. Mortars were cast in 40 x 40 x 160 mm specimens moulds and then compacted in two layers, smoothed and covered with plastic sheet to prevent water loss at 20 ± 2 °C. After 24 h, specimens were removed from the molds and cured in water until the test age. From different blended cements, the compressive strength was determined on prisms at 2, 7 and 28 days. The reported results are the average of six compression tests. The compressive strength index (CS Index) was calculated as the ratio between the compressive strength of the blended cement and the strength of the PC at the same test age.

To study the progress of hydration, blended cement pastes were prepared using a w/cm of 0.50 and cured in sealed plastic bags in water-bath at 20 °C during 2, 7 and 28 days. At this time, fragments of paste samples were carefully ground to particle size lower than 45 µm (#325) and the crystalline hydration phases were identified using XRD analyses.

Determination was performed on Philips PW 3710 diffractmeter operating with CuKα radiation at 40 kV and 20 mA using carbon monochromator.

The XRD patterns and FT-IR spectra od DCWs are shown in Figure 1. The mineralogical composition of the ceramic wastes obtained by XRD (Figure 1-a) includes quartz (SiO₂), anorthite (CaAl₂Si₂O₈) and hematite (Fe₂O₃) for CB and MB, while SW is composed by quartz (SiO₂) and mullite (Al₆Si₂O₁₃). The XRD patterns of DCWs show a small but evident diffuse band ranging from 20 to 30, 2θ deg., indicating the presence of amorphous materials (26). According to the XRD data, DCWs contain weak crystalline minerals and amorphous phase. Judged by the XRD diffuse band intensity, the proportion of non-crystalline phases is broader in the SW sample.

FTIR spectra of DCWs (Figure 1-b) confirm the phase identification by XRD. The quartz in all wastes is identified through the strong band at 460 cm⁻¹, corresponding to deformation of Si-O. Stretching bands of Si-O are placed at 695 cm⁻¹ (m), doublet 780–800 cm⁻¹ (s) and 1080 cm⁻¹ (vs) and 1170 cm⁻¹ (s) (27). In addition, CB and MB samples present assignable bands to feldspars (anorthite in this case) at 1100 cm⁻¹ range (1095, 948 cm⁻¹), which deform and widen the quartz band, and the additional band at 571 cm⁻¹ is also observed.

The physical characteristics of all materials (density, particle size distribution -d₁₀, d₅₀, and d₉₀- and the BSS) are given in Table 3. According to Table 3, three DCW have lower density, larger specific surface area and particle size distribution with large volume (d₁₀ and d₅₀ are low) of particles, smaller than the corresponding to Portland cement. This is evident for MB and SW samples.

Figure 2 shows the results of Frattini test for the PBCs at 2, 7 and 28 days. At 2 days, results for all blended cements were above the calcium solubility isotherm, showing pozzolanic activity did not occur. For the three DCWs, the [CaO] increased when the level of replacement in blended cement increased, showing the stimulation effect on the hydration of Portland cement phases. In addition, it can be observed that the dilution effect caused by the replacement of PC by DCWs produced a decrease of [OH⁻] when the DCW replacement level increased.

At 7 days, the reduction of [OH⁻] and [CaO] showed that PBCs with high replacement level (40% for all DCWs) have a pozzolanic activity and the test-points are located below the saturation curve. For 24CB blended cement, the point is on the saturation curve, indicating the early reaction of this DCW. At 28 days, all blended cements presented proper pozzolanic activity.

According to Frattini test results, MB has a slower pozzolanicity than SW, although both have similar specific surface area and particle size distribution. It can be exclusively attributed to the surface characteristics of the particles and the glass content.

The results of mortar flow tests for different blended cements are shown in Figure 3. It can be observed that the flow of mortar slightly decreases as the DCW replacement increases in blended cement.

The flow of mortar using 8CB (128%) was approximately the same than that of the control mortar (PC = 127%). For mortars 24CB and 40CB, the flow decreases 14 and 22% when compared to the control, respectively. For blended cements 8MB and 8SW, the flow was higher than that of control and their reduction was 3 and 10% for mortars 40% MB and 40SW, respectively.

The high flow reduction for CB blended cements is attributable to the large BSS (see Table 3) of this ceramic waste when compared with the specific surface of MB or SW ceramic wastes. For blended cement with low replacement (8%), the increase of flow compared with the control mortar may be attributable to the reduction of water fraction used to fill the inter-grains voids due to the better packing of blend that compensates the water needed to wet the large specific surface of DCWs. When the percentage of DCW increases, the flow decreases due to the increase mortar viscosity caused by the high specific surface area and the large volume occupied by solid fraction in paste due to the lower density of DCWs.

For each ceramic waste, the evolution of compressive strength (CS) of blended cements up to 28 days is shown in Figure 4. Additionally, the figure includes the evolution CS Index. For all DCWs, the compressive strength increases with the curing age and it decreases when the replacement level increases.

At 2 days, CS of all blended cements was less than that of the control mortar and its reduction increased when increasing the replacement level due to the dilution effect caused by the incorporation of ceramic waste. Dilution effect reduces the volume of PC available to react at early age increasing the effective w/c (0.54, 0.66 and 0.83 for 8, 24 and 40% replacement, respectively) that could not be compensated by the stimulation of PC hydration.

At 7 days, the CS of blended cements with 8% replacement was similar to that corresponding to PC, indicating that the stimulation of PC hydration compensates the dilution effect. On the other hand, the CS of mortars with 24% and 40% of replacement is lower than control, but its reduction is lower than the percentage of DCW replacement.

At 28 days, the CS of blended cement was lower than the control. For 8 and 24% replacement, the compressive strength is comparable to the control mixture, but it was low for blended cement with 40% replacement.

For PBCs containing 8, 24 and 40% DCW, the CS Index was higher than 0.92; 0.74 and 0.60 at 7 and 28 days, respectively. In all cases, the CS Index was higher than the replacement level in the PBCs. At 28 days, the CS index ranged from 0.93 to 1.00 for 8 and 24% DCW indicating that blended cements could be classified at the same strength class.

The study of the hydration progress of PC and PBCs by XRD at 2, 7, and 28 days showed a similar behavior. The hydration phases were identical at the same hydration times and the variation of intensity of peaks was only observed due to the different percentages of DCW replacement. As a consequence, Figure 5 only shows the XRD patterns of PC and PBC with 24% replacement of different ceramic wastes at 2, 7 and 28 days.

At 2 days, the PC showed the presence of calcium hydroxide (CH) and ettringite (Ett). In all blended cements, the intensity of Ett and CH peaks had a similar value to PC despite the dilution effect caused by ceramic addition. These observations reveal that stimulation of PC hydration occurred due to the filler effect, the heterogeneous nucleation and the large amount of free water in the system (28). Quartz (Q) is identified in PBCs from ceramic waste.

At 7 days, the Ett and CH were accompanied by hemicarboaluminate (Hc), and the incipient formation of monocarboaluminate (Mc). The Hc and Mc formation was promoted by the calcium carbonate (CC) present in PC as a minor component.

At 28 days, the intensity of CH and Ett peaks decreased in PBCs. The peak intensity of Mc increased and the intensity of Hc peaks decreased. The reduction of CH peak was attributed to the pozzolanic reaction generated by the addition of ceramic wastes, according to the Frattini test results. Similar trends were observed for the three materials.