Introduction

Geopolymers are synthetic inorganic polymers based on aluminosilicate or phosphate, dissolved in a strongly basic environment, produced according to a process defined as geopolymerization [1, 2]. Geopolymers can be considered ceramics consolidated by alkaline reaction as they respect the chemical and mechanical properties of ceramics [3, 4]. These materials are composite, consisting of a geopolymeric resin that acts as a glue for inert resources, thus improving chemical properties by functionalizing and reinforcing the reactive inorganic raw materials, in particular metakaolines [5]. Geopolymers are produced by molding without the use of additional finishes and used in many sectors such as construction, waste treatment, synthesis of biomaterials, [6,7,8]. Waste and industrial by-products of various kinds can be recycled as raw materials or inert fillers for this type of application at low processing temperatures, reducing production prices, thus making geopolymers eco-friendly and eco-sustainable materials [9]. Geopolymers can be exposed to high heat sources, due to a great tolerance to high temperatures because they are non-combustible, they do not develop gas or fumes, they are not toxic and do not explode as, contrary to traditional hydraulic cements, they do not contain water in structure. In recent years, these inorganic polymers have attracted growing attention because they have performances comparable to those of Portland cement, but are able to reduce CO2 emissions [10, 11]. Metakaolin (MK) is unique as it is the only entirely natural supplementary cementitious material, which is not a by-product of an industrial process. It is obtained by calcination of kaolin clays, a hydrated aluminosilicate mineral clay.

The aim of this work is to create a composite geopolymer using fly ash recycling in order to reduce the amount of waste to be treated or disposed in landfills, allowing both to decrease the environmental damage and to save the costs of transport and disposal of waste. New compounds, with improved physical and chemical properties, based on metakaolin with different percentages of flying ash (20, 50 and 70 mass%) were synthesized through the geopolymerization process in order to recycle waste materials. Fourier transform infrared spectroscopy (FTIR) and energy-dispersive X-ray fluorescence (ED-XRF) analysis was performed to follow the structural modification of the materials induced by the addition of ash. Thermo-mechanical analysis, namely dynamic mechanical analysis (DMA), differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA), was carried out to study the variation of physical properties of the prepared materials as a function of temperature.

Experimental

Materials

The kaolinite was calcinated at 700 °C for 4 h and then ground and sieved to < 70 µm, obtaining the metakaolin (MK) that was used as aluminosilicate precursor. In order to prepare an alkaline solution, the sodium silicate solution was used; it is a viscous liquid produced for the cement industry (Ingessil, Verona, Italy) with molar SiO2/Na2O = 3 and mass ratio = 2.89 (26.45 mass% of SiO2 and 9.14 mass% of Na2O), as well as a density of 1.38 g cm−3 and 39.44 Baume scales. The sodium hydroxide solution was obtained by dissolving pellets (99.6 mass%, Carlo Erba, Italy) in distilled water to reach 8 M concentration. The fly ashes were obtained by the calcination process of the glass processing scraps, which are used as source materials.

Synthesis method

After the glass working process, the waste that is obtained is not dry, having a consistency similar to that of mud because they are rich in H2O. In order to completely remove the water from these scraps, the mud was dried in a stove at 100 °C for 24 h. The ash obtained was ground in a fast laboratory ball mill and sieved through 110- and 30-µm filters in order to evaluate the effect of grain size [12,13,14]. Different percentages of fly ash (20, 50 and 70 mass%) and the last part of pure metakaolin were used to prepare the several geopolymers. This powder was subjected to alkaline activation, using a strongly basic solution consisting of sodium silicate (26.5 mass% SiO2, 10.6 mass% Na2O, 62.9 mass% H2O) in which NaOH pellets have been dissolved. The formulation of mixtures was designed to obtain Si/Al and Na/Al ratios of 2.16 and 1.01, respectively. The dough was poured into plastic molds, and the consolidation phase carried out at 50 °C for 24 h; the hardening phase was carried out at room temperature for 7, 14 and 28 days. This last curing time was investigated for samples containing glass with a fine grain size only. The whole process of synthesis is illustrated in Fig. 1.

Fig. 1
figure 1

Flowchart of synthesis

Full size image

Chemical, spectroscopic and structural characterization

The chemical structure of the synthesized powders was investigated by Fourier transform infrared spectroscopy using a Prestige 21 Shimadzu (Kyoto, Japan) instrument equipped with a DTGS detector, used to study the interactions between the different components. The spectra of the materials were recorded in the 400–4000 cm−1 region, with resolution of 4 cm−1 (45 scans), by using KBr pelletized disks containing 2 mg of each sample and 198 mg of KBr. The FTIR spectra were then processed by Prestige software (IR solution, software version 1.50, Shimadzu, Kyoto, Japan).

The chemical composition of the materials was ascertained by means of energy-dispersive X-Ray fluorescence spectroscopy using a Shimadzu energy-dispersive X-Ray fluorescence spectrometer EDX-720 equipped with 50 W Rh target X-ray tube, a high-energy resolution Si(Li) detector and five primary X-ray filters.

The solid residues obtained from the thermal analysis were investigated by FTIR using a Perkin-Elmer Spectrum 100 spectrometer with an universal ATR sampling accessory. Spectra were recorded at r.t. from 600 to 4000 cm−1, with a resolution of 4.0 cm−1, directly on compounds, without any preliminary treatment.

Microstructural analysis

The geopolymer powders were analyzed, as produced after alkaline activation, with a SEM EVO (EVO-SEM, Zeiss, Cambridge, UK). All of the samples were gold-sputtered up to a thickness of 20 nm by means of an Emitech K-550 sputter coater (Ashford Kent, UK). An accelerating voltage of 15 kV was used to collect the micrographs.

pH and conductivity analysis

In order to indirectly evaluate ion release from the different materials, pH and conductivity analysis were carried out in water, with a solid/liquid ratio of 1/10, over a period of 24 h under stirring.

Chemical analysis of fly ash and eluates

The chemical analysis of glasses was performed by inductively coupled plasma atomic emission spectroscopy (ICP-AES Varian Liberty 200). The sample was dissolved in HNO3 solution. The metal content of all samples was investigated by microwave-assisted wet digestion using a 3:1 mixture of HNO3 and HCl. The determination was made using an inductively coupled plasma emission spectrometer (ICPE-9000 Shimadzu, Tokyo, Japan).

Thermal analysis

Dynamic mechanical analysis was carried out on a dynamic mechanical thermal analyzer (TRITEC2000 by Triton Technology, Leicestershire, United Kingdom) by single cantilever geometry. The geopolymers, after 28 days of hardening, were tested in their powder form using the pocket DMA approach, a technique used for testing powders in the pharmaceutical field [15] and for polymer blends [16]. The geopolymers obtained from the syntheses were finely micronized in powder with an average dimension of 30 µm. Then 0.35 g of geopolymer powder was weighted in a standard stainless steel pocket purchased from Triton and pressed to obtain a uniform thickness. The test was carried out according to the following protocol: The sample was stabilized at 25 °C and then heated up to 210 °C at 5 °C min−1; the samples were cooled down naturally and re-heated up to 210 °C at 5 °C min−1. Similar techniques were also reported by Carlier et al. [17] for organic polymers under the name supported DMA. This kind of technique allows for direct evaluation of thermal transitions from E’ and tan δ traces. However, the absolute values of E’ and tan δ for the polymer are influenced by the presence of the metal pocket, and thus, the real values should be analyzed out considering the assembly as a sandwich material. The tan δ versus temperature was plotted.

A Shimadzu DSC-60 (Kyoto, Japan) was used for the calorimetric investigation. The enthalpy and temperature calibrations of the equipment were made according to a well-established procedure [18]. Samples of about 6.0 × 10−3 g, held in sealed aluminum crucibles, and a heating rate of 10 °C min−1 were used for measurements. Two consecutive DSC scans were carried out for each sample; the first one starting from r.t. up to 250 °C, while the second one starting from r.t. up to 400 °C.

A Shimadzu DTG-60 simultaneous DTA-TG apparatus (Kyoto, Japan) was used for thermogravimetric analysis. Temperature, heat flow and mass calibrations were carried out following the procedure suggested by Shimadzu and reported elsewhere [19]. Geopolymer samples of about 12 × 10−3 g, placed in a 40 μL alumina open pan, were used for the thermal degradations. TGA scans were carried out in the temperature range 30–800 °C, in inert (flowing nitrogen 0.02 L min−1) and oxidative environment (static air), by setting a heating rate of 10 °C min−1. In order to correct the error in the mass determination due to the reduction in the buoyant force with increasing temperature, we adopted the blank method suggested by the ICTAC Kinetics Committee [20]. The sample mass was then recorded as a function of the temperature and reported as the percentage of un-degraded sample, (1 − D)%, where D = (W0 − W)/W0, and W0 and W were the masses at the starting point and during scanning.

All the values, considered in this work, were averaged from those of three runs, with a maximum difference between the average and the experimental values within ± 1 °C.

Results and discussion

Spectroscopic and structural characterization

FTIR spectra of geopolymers at different content of fly ash are reported in Fig. 2 together with those of pure fly ash and metakaolin. The bands at 3445 and 1640 cm−1, due to –OH stretching and bending vibrations, change position and shape for the formation of hydrogen-bonded solvent molecules (H2O) and hydrogen-bonded –OH groups related to Si atoms [21,22,23]. In particular, bending –OH vibration band at 1640 cm−1 in pure metakaolin spectra (curve b) was displaced by about 10 cm−1 (1650 cm−1) both in the hybrids spectra. Furthermore, in all spectra are present the typical peaks of silica matrix and polymer. The asymmetric Si–O stretching and symmetric Si–O stretching are observed at 1045 cm−1 with a shoulder at 1100 cm−1. The spectra (b, c, d, e and f) present the strongest vibration typical of aluminosilicates, which are assigned to internal vibrations of Si–O-Si and Si–O-Al, and an approximate relationship between the frequency of the absorption bands and the ratio of Si/Al in the alluminosilicate framework was observed: The higher the Al inclusions, the lower the wavelength [24,25,26]. The spectrum of MK also revealed sharp peaks in the range 3622–3693 cm−1, ascribed to the stretching of hydroxyl groups and of Al(VI)–OH bonds in the kaolinite residues [27]. The presence of the band around 800 cm−1 (curve b), due to the stretching of the Al(IV), is also visible in the spectra d, e and f with decreasing intensity. Even the typical sodium carbonate group at 1430 cm−1, very intense in the curve c and d, decreases as the percentage by mass of the built-in flying ash increases (curve e and f). That is due to the passage of water through the bulk during the drying of the sample, which generates a high concentration of sodium close to the surface, where it forms carbonates after the reaction with atmospheric CO2.

Fig. 2
figure 2

Representative FTIR spectra of (a) fly ash, (b) pure metakaolin, (c) GPMK, (d) MK/FA20 mass%, (e) MK/FA50 mass%, (f) MK/FA70 mass%

Full size image

The chemical composition of the powders and their purity have been asserted by the analysis and XRF, and the results are summarized in Table 1. No impurities were detected in the powders obtained after the geopolymerization process, while in the MK there were natural impurities, even if in small quantities. The color of the powders confirmed the result of the analysis XRF: MK was pinkish (mainly because of the presence of Fe2O3), while all the powders were light gray.

Table 1 Chemical composition (mass%) of (MK) pure metakaolin; (C0) GPMK; (C1) MK/FA20 mass%; (C2) MK/FA50 mass%; (C3) MK/FA70 mass%
Full size table

The chemical analysis (Table 2) of all samples was evaluated by ICP analysis. The results of chemical analysis (Table 2) confirmed that the fly ash was rich in Si. In geopolymers, the metals diminish as the percentage of fly ash increases. These metals have been incorporated into the materials. For this reason, the materials obtained with the geopolymerization are less toxic than the raw materials used.

Table 2 Chemical analysis test (expressed in mg L−1) of fly ash; (C0) GPMK; (C1) MK/FA20 mass%; (C2) MK/FA50 mass%; (C3) MK/FA70 mass%
Full size table

Chemical characterization

The geopolymerization process of all samples was evaluated as a function of the polymerization time, using pH and conductivity analysis. Figure 3 shows that as the polymerization time increases, the pH decreases. This result is due to the decrease in free ions in the network, corresponding to a higher degree of geopolymerization [28]. For all synthesized materials, the results are confirmed. In the geopolymers containing a large percentage by mass of fly ash, it is possible to observe the slight increase in the initial pH, followed by a similar decrease in the samples with a lower percentage by mass of fly ash, which occurs as a function of the aging time. This suggests that the composition of raw materials affects the properties of the geopolymer. The different behavior of the samples containing a high percentage by mass of fly ash is also evidenced by the conductivity values. In fact, from Fig. 4 lower values of conductivity are observed when the built-in fly ash decreases. These results are due to the smaller amount of ions in the eluted solution.

Fig. 3
figure 3

Representative pH curve of (C0) GPMK; (C1) MK/FA20 mass%; (C2) MK/FA50 mass%; (C3) MK/FA70 mass%

Full size image
Fig. 4
figure 4

Representative conductivity curve of (C0) GPMK; (C1) MK/FA20 mass%; (C2) MK/FA50 mass%; (C3) MK/FA70 mass%

Full size image

Thermal analysis

All the samples were tested under oscillatory mode while heating from 25 to 210 °C at 5 °C min−1. The results, from the first heating scan, are summarized in Fig. 5. The samples C2 and C3 show a clear tan δ peak centered between 75 and 85 °C, while the formulations C0 and C1 show a smoother tan δ peak at about 110–125 °C. Figure 6 shows the tan δ behavior for the formulation C2 tested at two different frequencies (1 and 10 Hz). The tan δ showed a shift to higher temperature when tested at 10 Hz rather than 1 Hz. This behavior is usually observed for relaxation phenomena like glass transition observed in polymers [29, 30]. The storage modulus (E’) showed a clear decay starting from 50 °C. Similar behavior is shown by polymers when relaxation phenomena like glass transition occur. The sample analyzed was hardened at room temperature for 28 days; therefore, the transitions measured by DMA can be explained as devitrification of the curing network which is formed during hardening [31].

Fig. 5
figure 5

DMA curves of (C0) GPMK; (C1) MK/FA20 mass%; (C2) MK/FA50 mass%; (C3) MK/FA70 mass%

Full size image
Fig. 6
figure 6

DMA curves of MK/FA50 mass% (C2) tested at 1 and 10 Hz

Full size image

The sample C2, after cooling at room temperature, was analyzed with a second ramp temperature and the tan δ peak shifted to 205 °C (Fig. 7). After the first heating, the sample displayed a clear increase in the transition peak. This behavior is because of the structural changes occurring during the hardening reaction induced by the heat increases, and it is the result of the partial curing that occurred on the sample hardened at room temperature only. Similar behavior was observed by C3 formulation, while C0 and C1 formulations did not display such a clear change in the temperature range analyzed. As reported by Rahier et al. [32], the geopolymer mixtures follow a polymerization route that resembles the curing of organic thermosets.

Fig. 7
figure 7

DMA curves of (C2): first heating scan vs second heating scan

Full size image

DSC analysis seems to confirm the behavior observed during the dynamic mechanical analysis. Apart from a peak due to the mass loss caused by the presence of structural water in the geopolymer composition [33], that was less intense and shifted to lower temperature by increase in the fly ash content in the formulation, C2 and C3 samples showed a clear irreversible endothermic phenomenon at 147 and 150 °C, respectively (Fig. 8). The same phenomena, at slight lower temperature, for C0 and C1 samples were probably covered by the water loss, as testified by the shoulder present in the main peak at 138 and 140 °C, respectively. A second scan was carried out for each geopolymer, after cooling, that confirm the irreversible nature of the endothermic transition and confirm the presence of the relaxation phenomena at higher temperatures (Fig. 9) with the same trend, namely 179, 186, 205 and 207 °C for samples C0, C1, C2 and C3, respectively.

Fig. 8
figure 8

DSC curves, for the first heating, of (C0) GPMK; (C1) MK/FA20 mass%; (C2) MK/FA50 mass%; (C3) MK/FA70 mass%

Full size image
Fig. 9
figure 9

DSC curves, for the second heating (after cooling), of (C0) GPMK; (C1) MK/FA20 mass%; (C2) MK/FA50 mass%; (C3) MK/FA70 mass%

Full size image

Finally, thermogravimetric analysis was carried out, both for evaluating the resistance to the thermal degradation of these materials and for confirming that the DSC peaks around 100 °C can be attributed to mass (water) loss. TGA degradation was carried out in inert (nitrogen flow) (Fig. 10) and oxidative (static air) (Fig. 11) atmospheres, confirming the start of mass loss with the values of the first endothermic DSC peak in both investigated environments. After this degradation, the TGA curve proceeds without further mass loss and the formation of a stable residue remains constant in mass up to 800 °C, testifying the great thermal resistance of these materials. The stable residue, as expected, decreases with the increase in ash content in the formulation. The residues derived from the degradation at 800 °C were analyzed by FTIR spectroscopy, and the corresponding spectra are reported in Figs. 12 and 13. No differences in the residues of the different geopolymers have been found, showing all the investigated samples IR bands at about 960–980 and 700 cm−1 to be attributable to the presence of silica [34, 35].

Fig. 10
figure 10

TGA curves, in flowing nitrogen atmosphere, of (C0) GPMK; (C1) MK/FA20 mass%; (C2) MK/FA50 mass%; (C3) MK/FA70 mass%

Full size image
Fig. 11
figure 11

TGA curves, in static air atmosphere, of (C0) GPMK; (C1) MK/FA20 mass%; (C2) MK/FA50 mass%; (C3) MK/FA70 mass%

Full size image
Fig. 12
figure 12

FTIR spectra of the residues, obtained at 800 °C in flowing nitrogen, of (C0) GPMK; (C1) MK/FA20 mass%; (C2) MK/FA50 mass%; (C3) MK/FA70 mass%

Full size image
Fig. 13
figure 13

FTIR spectra of the residues, obtained at 800 °C in static atmosphere, of (C0) GPMK; (C1) MK/FA20 mass%; (C2) MK/FA50 mass%; (C3) MK/FA70 mass%

Full size image

Scanning electron microscopy (SEM) analysis

The hardened geopolymers (28 days) were analyzed by SEM. The sample C0 shows a uniform distribution of lamellar structures longer than 5 µm, and some granular structures with sizes variable from 5 to 20 µm (Fig. 14). The sample C1 shows the lamellar structures but smaller granular structures than the sample C0 as a result of the fly ash addition. Lamellar structures similar to those observed here are reported in the literature [36, 37]. With the increase in the amount of fly ash content the lamellar structure is reduced and the sample appears more compact (Fig. 15).

Fig. 14
figure 14

SEM analysis of (C0) GPMK on the left; (C1) MK/FA20 mass% on the right: low magnification (500 ×) up; high magnification (5000 ×) down

Full size image
Fig. 15
figure 15

SEM analysis of (C3) MK/FA70 mass% (magnification 500 ×)

Full size image

Conclusions

Geopolymerization is a possible technology for the management of large amount of potentially hazardous or abundant wastes. In particular, waste glass, obtained from the glass processing, acts as filler, not showing high reactivity in the geopolymeric matrix but also not hindering the geopolymerization of the matrix. After 28 days at room temperature, the samples synthesized using different percentages in mass of fly ash showed a good geopolymerization. This was confirmed by the FTIR analysis, which showed the formation of bonds between both components. The reduction of free ions in the network as the drying period increases was demonstrated with a decrease in the pH value and conductivity. The chemical composition, analyzed with ICP and XRF, highlighted the great presence of silicates in fly ash and metakaolin alumina. The geopolymers showed a proportional mixed composition.

The transitions observed by DMA were attributed to the devitrification of the curing network, which is formed during hardening at room temperature for 28 days. These results were also confirmed by DSC analysis that suggest the irreversible nature of the endothermic transition and confirm the presence of the relaxation phenomena at higher temperature in all samples.

Finally, to evaluate the resistance to the thermal degradation and to confirm the mass (water) loss observed in DSC analysis, thermogravimetric analysis was carried out. Furthermore, the residues derived from the degradation at 800 °C were analyzed by FTIR spectroscopy, showing the same nature for all the samples attributable to the presence of silica.

The SEM analysis showed us a reduction in the length of the lamellae with the increase in incorporated fly ash, thus demonstrating a greater compactness of the sample.

Although the results may show a potential use of these compounds containing different percentages of fly ash as building materials, future analyses will be carried out to evaluate their mechanical properties as well.