Synthesis and characterization of the SFAs
Figures 2 and 3 present the XRD patterns of the SFAs products with SiO2/Na2O ratio equal to 1 and 2, respectively, after thermal and microwave treatment.
The silica fume pattern presents a broad hump around 21o 2θ which is indicative of amorphous silicon oxide. In the case of products with molar ratio SiO2/Na2O=1, rregardless of the treatment method or the applied parameters, silica fume was converted successfully to the crystalline phase of Na2SiO3, similarly to the commercial product 1SS_COM (Fig. 2a, 3a). In the case of thermally treated samples, traces of natrite, Na2CO3, and thermonatrite, Na2CO3·H2O (< 4 % w/w) are also present in the final products, because of the sodium hydroxide carbonation. However, carbonation did not occur in the microwave treated samples since no carbonates were detected in their XRD patterns. This is most probably due to the limited air-exposure of these samples as the treatment lasts less than 20 min.
Moreover, SFAs with ratio SiO2/Na2O=2 (Fig. 2b, 3b) show high amorphousness like the commercial product 2SS_COM. A displacement of the amorphous peak of silica fume in higher degrees (29 o) is observed for both products prepared with SiO2/Na2O=2 (2SS_TT_330C_1h and 2SS_MT_M_2m). This is a sign of atoms’ reorganization and the formation of new amorphous phases. Table 4 presents the % w/w mineral composition of the solid activators as it was obtained from the quantitative XRD analysis along with their solubility in water.
Table 4
Mineral composition and water solubility of the solid activators.
Sample ID
|
Na2SiO3 (%)
|
Na2CO3 (%)
|
Na2CO3*H2O
(%)
|
Amorphous
(%)
|
Normalized solubility (%)
|
Commercial
|
|
|
|
|
|
1SS_COM
|
50.90
|
-
|
-
|
49.10
|
99.10
|
2SS_COM
|
0.00
|
-
|
-
|
100.00
|
99.20
|
Thermal Treated
|
|
|
|
|
|
1SS_TT_ 330C_0.5h
|
66.71
|
3.65
|
1.14
|
28.49
|
98.00
|
1SS_TT_ 330C_1h
|
77.35
|
1.91
|
1.75
|
18.98
|
95.81
|
1SS_TT_ 330C_2h
|
71.04
|
2.67
|
0.43
|
25.85
|
96.09
|
1SS_TT_ 330C_3h
|
72.29
|
|
0.16
|
25.98
|
96.73
|
1SS_TT_ 150C_1h
|
58.97
|
2.24
|
0.64
|
38.15
|
97.22
|
1SS_TT_ 250C_1h
|
69.48
|
3.20
|
0.43
|
26.89
|
96.89
|
1SS_TT_ 450C_1h
|
80.91
|
1.91
|
0.01
|
17.17
|
96.58
|
2SS_TT_ 330C_1h
|
1.76
|
0.00
|
0.00
|
98.24
|
76.40
|
Microwave treated
|
|
|
|
|
|
1SS_MT_M_2min
|
56.10
|
-
|
-
|
43.90
|
97.63
|
1SS_MT_H_2min
|
52.60
|
-
|
-
|
43.00
|
98.83
|
1SS_MT_L_12min
|
50.30
|
-
|
-
|
49.70
|
97.92
|
1SS_MT_M_12min
|
69.90
|
-
|
-
|
30.10
|
97.32
|
1SS_MT_H_12min
|
66.70
|
-
|
-
|
33.30
|
97.76
|
1SS_MT_M_20min
|
64.30
|
-
|
-
|
35.70
|
98.37
|
2SS_MT_M_2min
|
0.00
|
-
|
-
|
100.00
|
95.52
|
SF
|
-
|
-
|
-
|
100.00
|
7.93
|
* Normalized solubility refers to the SFAs soluble part excluding the soluble crystalline carbonate phases |
Concerning the SFAs with molar ratio SiO2/Na2O=1, it is obvious that products with higher crystallinity, compared with the commercial ones, were prepared. In addition, heat treatment is more efficient for the preparation of high crystalline samples (62-81 %), compared with the microwave treatment (50-70 %). Sodium silicate Na2SiO3 is the main crystalline phase, and in some cases the only one, that characterizes the SFAs composition.
The increase of the temperature from 150 to 450 oC, increases the content of crystalline Na2SiO3 by 27 %, in the thermally treated samples (Fig. 4a). In a similar manner, the crystalline content of Na2SiO3 in the microwave treated samples was enhcaced by 28 % when the power of the microwave source altered from 120 to 460 W (Fig. 4b). In both cases, further increase of temperature or power did not yield any remarkable change of the crystalline content.
Treatment duration effect on Na2SiO3 crystalline content of SFAs was not so pronounced in relation to the temperature or power effect. In particular, a treatment duration increase, from 0.5 to 1 h, in thermally treated samples led to a 14 % increase of the crystalline Na2SiO3 (Fig. 4a). Similarly, an alteration of treatment duration, from 2 to 12 min, in microwave treated samples led to a 20 % increase in the crystalline content (Fig. 4b). Higher treatment duration, for both treatment methods, did not favor higher crystalline contents.
Concerning the effect of SiO2/Na2O molar ratio, products with SiO2/Na2O=1 obtained a high crystalline content (>50 %) while those prepared with SiO2/Na2O=2 are almost totally amorphous, following the structure of the commercial sodium silicates (1SS_COM and 2SS_COM).
Table 4, also presents the normalized solubility of the activators, calculated by excluding the carbonates’ content. Results showed that commercial sodium silicates 1SS_COM and 2SS_COM are totally soluble in water. Silica fume, the raw material of the SFAs, is almost entirely insoluble in water, since it contains more than 90 % SiO2 (7.9 %). Moreover, all SFAs with molar ratio SiO2/Na2O=1, regardless the way they have been treated, are highly soluble in water (> 95 %), similarly to the commercial product 1SS_COM, indicating that their amorphous phase consists of highly soluble compounds and, of course, the successful conversion of silica fume to new products. As concerns SFAs with molar ratio SiO2/Na2O=2, the 2SS_TT_330C_1h sample exhibited a lower water solubility (76.4 %), indicating lower conversion yield. On the contrary, the sample 2SS_MT_M_2min showed a high solubility, close to 95 %, which may be attributed to higher conversion of silica to soluble silicate compounds.
Table 5
FTIR band assignments (Brooker and Bates 1971; de Man and van Santen 1992; Ratcliffe and Irish 2002; Guo et al. 2010; Rashid et al. 2012; Kioupis et al. 2018a; Zanoletti et al. 2018; Ryu and Lee 2018)
Frequencies (cm− 1)
|
Assignments
|
1630
|
– OH bending vibrations (H2O)
|
1450
|
O-C-O stretching vibrations
|
Silica Fume
|
1100
|
Si–O–Si asymmetric stretching vibrations
|
800
|
Si–O–Si symmetric stretching vibrations
(inter-tetrahedral Si-O-Si)
|
475
|
O–Si–O bending vibrations
|
1SS
|
1035
|
Si–O–Na asymmetric stretching vibrations
|
970
|
Si–O–Si asymmetric stretching vibrations
|
880
|
Si–O–Na symmetric stretching vibrations
|
710
|
Si–O–Si symmetric stretching vibration
|
590
520
|
in-plane Si–O–Si stretching mode coupled with
O–Si–O and Si–O–Si bending modes
|
2SS
|
1005
|
Si–O–Na asymmetric stretching vibrations
|
880
|
Si–O–Na symmetric stretching vibrations
|
750
|
O–Si–O stretching
|
Figure 5a presents the FTIR spectra of the SFAs (SiO2/Na2O=1) prepared in this study. For comparison, the spectrum of the commercial sodium silicate (1SS_COM) as well as that of silica fume are presented in the same figure. Silica fume exhibits the typical vibrations of Si-O bonds. In particular, the absorption peaks at 1100 and 804 cm-1 are assigned to asymmetric and symmetric stretching vibrations of Si – O – Si bonds while that at 470 cm-1 is attributed to the bending vibrations of O – Si – O bonds. The band at 1630 cm-1 is related to the -OH bending vibrations and it indicates the existence of absorbed water. Concerning the 1SS_COM, the absorption peaks at 1036, 967, 885, 711, 590 and 515 cm-1 are attributed to varied IR vibrations of Si-O-T (T: Si or Na) and O-Si-O bonds on the structure of sodium silicates (Table 5). Furthermore, the band at 1450 cm-1 is assigned to the vibrations of carbonate ions and it is linked to the presence of sodium carbonate phase. It is obvious that irrespective the processing conditions, the reactants have successfully been transformed to sodium silicates. In particular, all the SFAs exhibit identical infrared spectra to that of the commercial product (Fig. 5). Furthermore, the characteristic broad peak of silica fume is not detected in the spectra of the activators, indicating the total conversion of silica fume to sodium silicates. The formation of the sodium silicate structure is majorly indicated by the shift of the Si – O – Si band at 1100 cm-1 to lower wavenumbers through the enrichment of Si – O – T bonds by Na; “trident” absorption at 1040, 970 and 880 cm-1. These results are in accordance with solubility rates of 1SSs (Table 4).
Figure 5b presents the FTIR spectra of the 2SS activators (SiO2/Na2O=2), prepared in this study, along with the spectra of the commercial sodium silicate (2SS_COM) and silica fume. The 2SS_COM exhibits absorption peaks at 1005 and 880 cm-1 which are assigned to the asymmetric and symmetric stretching vibrations of Si-O-Na bonds, respectively (Table 5). In addition, the band at 755 cm-1 could be linked to the stretching of O – Si – O bonds. The spectrum of 2SS_COM is relatively featureless revealing the amorphous nature of this material, in accordance with the XRD analysis. As in the case of 1SS_COM, the band at 1450 cm-1 is related to the presence of sodium carbonate phases.
The spectra of the prepared 2SSs show clear differences in relation to that of the commercial product (2SS_COM). The thermally treated product (2SS_TT_150C_1h) exhibits a displacement of the wide absorption peak (1000 cm-1) to higher wavenumbers (1050 cm-1) while the band at 880 cm-1 is almost undetectable. This observation indicates the partial transformation of the silica fume to silicate compounds since a lower portion of Si – O – Na bonds have been formed in the final product. This speculation is in good accordance with the solubility of the products since 2SS_TT_150C_1h sample shows a reduced solubility in water indicating that unreacted silica is still present in the product. The spectrum of the microwave treated sample (2SS_MT_M_2m) is quite similar with the spectrum of the thermally treated sample.
Synthesis and characterization of geopolymers
In order to evaluate the potential of the prepared samples, they were applied as activators for the geopolymerization of fly ash.
Its chemical composition was determined through X-ray fluorescence spectroscopy (XRF, Malvern Panalytical, Epsilon 1 Model) and is shown in Table 1. The raw materials’ particle size distribution was determined through Malvern Mastersizer Micro (Fig. 1). Fly ash was previously ground to a mean particle size (d50) of approximately 20 µm, which is a typical fineness of fly ash when used in the construction industry.
The SFAs were selected in terms of the lowest energy consumption during their preparation. Therefore, the xSS_TT_150C_1h and xSS_MT_M_2m samples (x=SiO2/Na2O=1 or 2) were used for the synthesis of geopolymers (Table 4). For comparison reasons, one and two-part geopolymers were also prepared, using the commercial products. It is worth-mentioning that the SFAs required quantities were calculated according to the assumption that they are totally consisted of sodium silicate phases.
An initial evaluation step of the successful production of building materials concerns the mechanical properties’ measurement. In Figure 6, the uniaxial compressive strength of the prepared geopolymers is presented (average of three specimens). The geopolymer synthesis through the traditional liquid activator (G2P) achieves a compressive strength of 62.3 MPa. The replacement of the activation solution with commercial solid activators was successful, as expected10, since products with identical mechanical strength were obtained irrespectively of the SiO2/Na2O ratio of the activator (Fig. 6). The further replacement of the solid commercial activators by the SFAs showed promising results. More particularly, the geopolymers incorporating the 1SS_TT_150C or 1SS_MT_M_2m activators achieved the 90 to 95% of the compressive strength of the reference synthesis (G2P). The fact that the thermally treated activator achieves slightly lower mechanical strengths can be attributed to the presence of carbonate phases which lowers the reactivity of the material.
Concerning the geopolymer synthesis with 2SSs activators, the prepared geopolymers exhibit remarkably lower compressive strength in comparison to the G2P. Both G1P_2SS_TT_150C and G1P_2SS_MT_M_2m have much lower compressive strength than geopolymers with the commercial activators. These results are in accordance with the characterization analysis (solubility studies and FTIR) of these SFAs showing that silica haw ton totally transformed in sodium silicate compounds.
Figure 7 presents backscattered images of the geopolymers and EDS analysis of selected points. The geopolymers present the typical microstructure of fly ash based geopolymers, exhibiting a heterogenous structure consisting of a dense matrix and unreacted fly ash material. In particular, the fly ash cenospheres are partially or totally dissolved, depending on the progress of geopolymerization reaction.
Comparing the geopolymer materials, it can be seen that the products prepared with SFAs of SiO2/Na2O=1, G1P_1SS_TT_150C (Fig. 7b) and G1P_1SS_MT_M (Fig. 7c), exhibit similar microstructure with that prepared with the commercial activator (G1P_1SS_COM, Fig. 7a). Furthermore, the stoichiometry of the geopolymer matrix obtained by EDS analysis is almost identical for the aforementioned samples and close to Si:Al:Na ~ 21:10:9.
The G1P_2SS_COM (Fig. 7d) sample presents comparable microstructure to G1P_1SS_COM (Fig 7a) while the prepared aluminosilicate gel contains a slightly higher amount of sodium (Si:Al:Na ~ 20:9:11). The sample prepared with thermally treated SFA (Fig. 7 e and f) shows a completely different morphology. The structure seems to be porous containing hollows of high diameter (> 200μm) while it lacks cohesiveness revealing that a low portion of aluminosilicate gel has been formed. This observation confirms the findings of the alkali silicates preparation experiments that showed an extremely low conversion yield of silica fume to silicates when SiO2/Na2O equaled to 2. The sample’s stoichiometry by EDS analysis was calculated to Si:Al:Na ~ 8:3:26 indicating that the unreacted sodium of waste-based activator have covered superficially the fly ash material.