Analysis of the production conditions of geopolymer matrices from natural pozzolana and fired clay brick wastes

Construction and Building Materials 215 (2019) 633–643

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Analysis of the production conditions of geopolymer matrices from natural pozzolana and fired clay brick wastes Guido Silva a, David Castañeda a, Suyeon Kim a, Alvaro Castañeda b, Bruno Bertolotti c, Luis Ortega-San-Martin b, Javier Nakamatsu b, Rafael Aguilar a,⇑ a b c

Engineering Department, Pontificia Universidad Catolica del Peru PUCP, Lima, Peru Science Department, Chemistry Section, Pontificia Universidad Catolica del Peru PUCP, Lima, Peru Research and Development Department, Compañia Minera Agregados Calcareos S.A (COMACSA), Lima, Peru

h i g h l i g h t s  Pozzolana and fired clay brick wastes were studied as possible geopolymer sources.  The 7th day compression strength was measured with different production conditions.  Compression strengths above 25 MPa were obtained when proper conditions were set.  Curing conditions were decisive for obtaining improved mechanical properties.

a r t i c l e

i n f o

Article history: Received 14 January 2019 Received in revised form 22 April 2019 Accepted 26 April 2019 Available online 3 May 2019 Keywords: Geopolymers Fired clay brick wastes Natural pozzolana Silica modulus Sodium oxide content Water content Curing conditions Compressive strength

a b s t r a c t The improvement of the mechanical properties of geopolymer matrices relies on the characteristics of the source materials as well as in proper optimization of the alkaline activating solution and curing conditions during geopolimerization. This study presents the optimization analyses carried out to determine the proper production conditions of Fired Clay Brick (FCB) and Natural Pozzolana (NP)-based geopolymers. The results indicate that high compressive strengths of up to 37 MPa and 26 MPa can be obtained for FCB and NP-based geopolymers, respectively when the proper production conditions are employed. The optimum alkaline solution for FCB consisted of Ms = 0.60, Na2O content of 8%, water/binder ratio = 0.27 with oven curing conditions between 65 and 80 °C for 7 days. On the other hand, NP-based geopolymers with the highest mechanical properties were obtained with an alkaline solution composed of a Ms = 1.08, 8% Na2O content and 0.52 of water/binder ratio cured in an oven at 65 °C for 7 days. The methodology for the optimization of production conditions of geopolymer matrices validated in this study demonstrated to provide consistent results and, therefore, could be applied for the analysis of the production process of geopolymers based on other aluminosilicate sources. Ó 2019 Elsevier Ltd. All rights reserved.

1. Introduction Fired clay brick constitutes one of the major components of the construction and demolition wastes, representing an average of 30% of the total amount of these wastes in the EU [1] and reaching higher amounts (up to 54%) in some countries such as Spain [2]. It is an urgent necessity to evaluate alternative applications of these wastes considering the large amounts produced per year (e.g. the US alone generated 534 millions of tons of construction and demolition wastes in 2014 [3], and the yearly production of the EU is estimated in 855 million tons [4]). The use of as the fired clay brick ⇑ Corresponding author. E-mail address: [email protected] (R. Aguilar). https://doi.org/10.1016/j.conbuildmat.2019.04.247 0950-0618/Ó 2019 Elsevier Ltd. All rights reserved.

wastes to develop alternative building materials has environmental and economic advantages since it avoids filling scarce landfills and also diminishes extraction and production of new raw materials [5]. In this line, green types of cement that use fired clay brick wastes have been developed by partial replacement of Ordinary Portland Cement (OPC) [6–8]. In [8], the use of up to 35% of Fired Clay Brick (FCB) powder is reported for the production of blended cements. Other studies have reported the use of crushed fired clay brick as coarse and fine aggregates in OPC-based concrete production with no loss of mechanical strengths compared to conventional concretes [7]. On the other hand, volcanic ash is one of the largest mineral resources available in the world. Natural Pozzolana (NP) is a type of volcanic ash (other examples are pumice, slag, and andesite)

634

G. Silva et al. / Construction and Building Materials 215 (2019) 633–643

with high silica and alumina content, which is abundant in volcanic regions around the world [9]. This non-metallic mineral has a wide range of applications, for instance, NP is used in agriculture to increase soil porosity. The usage of NP has also been reported for the production of ceramic materials, thermal insulators, absorbent materials and natural filters [9,10]. However, the use of NP as a lightweight aggregate for concrete production or as a mineral admixture in blended cement production is its most widespread and common application [9]. The extraction of NP in Peru alone is approximately 1 million metric tons per year [11]. Despite its high availability in the southern region of Peru, its industrial use is limited to the manufacture of blended cement. As a matter of fact, the maximum cement replacement by NP is around 30–35% because higher percentages cause lower compressive strengths at early curing stages [9,10]. Although blended cements represent an attractive option for using wastes and non-metallic minerals in the construction sector, there is another alternative based on alkaline activation to develop alternative cementitious materials from aluminosilicate sources (e.g. FCB and NP). This type of materials, called geopolymers or alkaline-activated materials, are reported to be eco-friendlier in comparison to OPC-based solutions in terms of global warming potential [12]. However, special attention has to be given to the composition of the alkaline activation when performing the life cycle analysis of these materials as the sodium silicate is still a harmful component in this sense and the use of heat during the curing process may increase its environmental impact [12,13]. FCB and NP have shown great potential as raw materials for geopolymer production [9,14–24]. Baronio and Binda [25] have claimed that FCB powder has pozzolanic activity potential due to the destruction of the crystalline network when structural hydroxyl groups of clay minerals are lost with high temperatures during production. On the other hand, Djobo et al. [26] have reported that NP powder is not highly reactive in alkaline media like other aluminosilicates (such as fly ash or metakaolin), however, this reactivity improves with higher alkaline concentrations and curing temperatures. The optimal alkaline activation of geopolymers varies widely depending on the type of raw material due to differences in particle size distribution, morphology, mineralogical and chemical composition. Reig et al. [14,18] and Allaverdhi et al. [22,23] optimized the formation of FCB and NP-based geopolymers by varying the alkaline solution parameters such as the silica modulus (Ms), the sodium oxide content (Na2O) or NaOH concentration, and also the water/binder ratio. On the other hand, the curing conditions showed to have a direct and significant influence on the mechanical properties of the geopolymers produced as in [26,27]. The objective of this research is to optimize the alkaline activating solution composition and the curing conditions to produce high strength geopolymers from FCB from the construction industry and NP from Peru.

2. Materials and experimental plan 2.1. Description of the raw materials Fine powders of commercial FCB from the construction industry in Lima were obtained after a milling process consisting of three phases. First, the entire brick was placed in an impact crusher to obtain pieces smaller than 25 mm. After that, a rolling mill reduced the particle size to 0.5 mm or less, which were finally placed in a ball mill for 4 h. On the other hand, NP was provided by the Research and Development Department of the mining company COMACSA from a pit in Arequipa (Peru) and ground in a ball mill. The grinding process was performed until all NP powder passed through ASTM No. 100 sieve (150 lm aperture). A particle size analyzer, SediGraph 5100, was used to determine the particle size distribution envelope of both raw materials (Fig. 1) using the sedimentation method where the particle mass is measured directly via X-ray absorption. FCB powder exhibited a mean particle size of 19.66 um, d90 of 58.66 lm, d50 of 7.39 lm and d10 of

1.21 lm (Fig. 1a). For NP powder, the mean particle size was 11.19 lm, d90 of 19.71 lm, d50 of 10.22 lm and d10 of 4.52 lm. The specific gravity of the powders of FCB and NP was 2.40 g/cm3 and 2.78 g/cm3, respectively (Fig. 1b). The micromorphology of the powders of FCB and NP was investigated by scanning electron microscopy (SEM), using a Thermo Fisher Scientific Quanta 250 microscope. As shown in Fig. 2, both powders presented a very irregular and crystal-like shape (in contrast to the spherical shape of fly ash [28]). The chemical composition of the raw materials was obtained by X-ray fluorescence spectroscopy using an ARL OPTIM’X Spectrometer (see the summary of results in Table 1). Both, FCB and NP powders show a high and similar content of SiO2, 53.45% and 53.55%, respectively, however, their Al2O3 content differs, FCB almost doubles the amount of NP (20.52% and 10.81%, respectively). According to ASTM C618 [29], NP does not satisfy the chemical requirements for its use in blended concrete due to the sum of SiO2, Al2O3 and Fe2O3 is 65.34% and the minimum amount required is 70%. However, NP and FCB can be used as a good source material for the development of geopolymers due to their high content of SiO2 and Al2O3 [30]. The quantitative analysis of elemental composition (%) of FCB and NP powders performed by SEM equipped with an energy dispersive X-ray (EDX) confirmed the results from X-ray fluorescence spectroscopy. In order to gather the mineralogical composition of the FCB and NP, X-ray diffraction (XRD) analysis was performed using a Bruker D8 Discover DAVINCI XRD instrument equipped with a Lynxeye detector and a Cu Ka X-ray tube. The XRD data were collected for phase angles (2h) between 10° and 70°, with a 0.02° step and an integration time of 4 s. Crystalline phases were identified from PDF-2 database (2013 edition) using Bruker AXS Software Difracc.EVA 4. According to the X-ray diffractogram shown in Fig. 3a, FCB contains mainly quartz, SiO2, a mixture of feldspars, (Na, Ca, K)AlSi3O8, and hematite (iron oxide), as well as other lesser minerals such as goethite, FeO(OH) and actinolite, Ca2(Mg4.5-2.5Fe0.5-2.5) Si8O22(OH)2. Quartz and mixtures of different types of feldspars are the main minerals detected in the crystalline phase of NP (Fig. 3b). Furthermore, it is to be noted that NP is a mineral with very low crystallinity (see the wide maxima in the 2h region from 14° to 36°), the amount of which was estimated to be roughly 25% using the area analysis implemented in the Bruker AXS software Diffrac.TOPAS 5. 2.2. Characterization of geopolymer matrices Mechanical and chemical characterization of geopolymer matrices were conducted after 7 days of fabrication. For performing the mechanical characterization, the uniaxial compressive strength of FCB and NP-based geopolymers was studied following the guidelines of ASTM C109/C109M – 16a [31]. Silicon moulds b) were employed to produce 50 mm cubic samples (c), which were used for mechanical characterization. An electromechanical testing machine MTS model Exceed 45.105 controlled by displacement was used for all compression tests (a). The displacement rate of the load frame was set to 0.5 mm/min. Load cell displacements were recorded in all tests and were taken as global deformations. For physical characterization, density was calculated to evaluate the influence of all studied variables on the geopolymer samples. For chemical characterization, Fourier-transform infrared spectroscopy (FT-IR) was used by means of a Perkin Elmer model 100 FT-IR spectrometer. Samples were prepared as thin transparent pellets using KBr, while spectra were recorded in transmittance mode with 32 scans. IR spectra were obtained to follow the change in bonding and chemical environments around Al and Si atoms present in geopolymers made from FCB and NP. As geopolymer formation involves both dissolution and polymerization reactions bonds must be broken and formed during the process. It is also expected that the chemical environment around the Si and Al atoms change as oxygen atoms create new bridges between them, which should be reflected in their vibration bands [32,33]. 2.3. Preparation of samples Alkaline solutions were prepared by mixing sodium hydroxide pearls (99.27% purity, technical grade, New China Chemicals Co., Ltd.), sodium silicate solution (28% SiO2 + 8% Na2O + 64% H2O, technical grade, Abastecimientos Químicos Ciatex S.A.C) and distilled water. Then, the raw materials (Fig. 4a and d) and the alkaline solution were mixed in a mortar-mixing machine type STJBJ-5 (Zhejiang Tugong Instrument Co., Ltd.) for about 90 s after which a homogeneous paste was obtained (Fig. 4b and e). Geopolymer matrices were then placed into the cubic silicon molds and samples were subjected to mechanical vibration to remove trapped air bubbles. Afterwards, specimens (Fig. 4c and f) were cured in an oven at different temperatures and curing times. The curing process was completed when the samples were left at ambient temperature (20 °C) until mechanically tested (all samples were de-molded after the first day of curing and were tested after 7 days of production). For both geopolymer matrices, five variables were defined as key parameters in the alkaline activation: silica modulus (Ms), Na2O content, water/binder ratio (w/b), curing temperature and curing time. Ms is defined as the molar ratio between SiO2 and Na2O in the alkaline solution (made of sodium hydroxide, sodium silicate and water), while the variable defined as Na2O (wt. %) is the sodium oxide content in the alkaline solution with respect to the weight of the raw material in its dry condition. A four-stage experimental research plan was designed taking into count these variables to optimize the alkaline activating solution composition and the curing

635

G. Silva et al. / Construction and Building Materials 215 (2019) 633–643

Fig. 1. Granulometric curves envelope of ground raw material: (a) FCB and (b) NP.

Fig. 2. SEM micrographs of ground raw materials: (a) FCB and (b) NP.

Table 1 Chemical composition of FCB and NP powders by X-ray fluorescence spectroscopy. Oxide content (% wt.)

Counts/ 1000

FCB NP

SiO2

Al2O3

Fe2O3

K2O

MgO

CaO

Na2O

Others

53.45 53.55

20.52 10.81

7.80 0.98

2.63 4.07

1.85 0.55

1.62 1.43

1.50 2.30

10.63 26.31

18 16 14 12 10 8 6 4 2 0

Counts/1000

Raw material

10

20

30 40 50 2θ (degrees) (a)

60

70

8 7 6 5 4 3 2 1 0 5

15

25 35 45 2θ (degrees) (b)

55

65

Fig. 3. X-ray diffractograms: (a) FCB and (b) NP. Q: Quartz (SiO2); H: Hematite (Fe2O3); P: Plagioclase (albite/anortite,(Ca,Na)Al2Si2O8); A: Actinolite (Ca2(Mg4.5-2.5Fe0.5-2.5) Si8O22(OH)2); M: Microcline (KAl2Si2O8); C: Calcite (CaCO3); K: Orthoclase/Microcline (KAl2Si2O8).

636

G. Silva et al. / Construction and Building Materials 215 (2019) 633–643

Fig. 4. Production process of geopolymers: (a) FCB; (b) FCB-based geopolymer paste; (c) FCB-based geopolymer cubic sample; (d) NP; (e) NP-based geopolymer paste and (f) NP-based geopolymer cubic sample. conditions for geopolymers production (the optimization was carried out comparing the resultant compressive strength). Table 2 shows all fifty-one series of geopolymer matrices considered in this study (three specimens were tested to uniaxial compression in each series). The first stage was related to delimitate a range of Ms values for each geopolymer matrix. In the second stage, the objective was to determine the best Na2O content and Ms. The effect of water content was investigated in the third stage by means of the weight ratio between water (namely the combination of distilled water and water in the sodium silicate solution) and the dry raw material. Finally, the fourth stage involved the study of the effect of the curing temperature and curing time on the development of mechanical resistance of both geopolymers matrices.

3. Results and discussion 3.1. Analysis of the effects of silica modulus (Ms) Several authors have demonstrated the importance of defining an adequate Ms in the production of various types of geopolymers [18,23,34]. The existence of an optimum Ms is related to a SiO2/ Na2O molar ratio appropriated to form a highly crosslinked aluminosilicate network, which reduces the presence of unreacted silica [35]. The influence of silica modulus (Ms) on the 7th-day compressive strength was assessed keeping constant all the remaining variables. In this first stage, for FCB-based geopolymers, the Na2O content, water/binder ratio, oven curing temperature and oven curing time were fixed at 8%, 0.30, 65 °C and 1 day. For NP-based geopolymers, the aforementioned parameters were 10%, 0.40, 65 °C and 1 day, respectively. Density values for all samples of FCB and NP were around 2.0 and 1.8 g/cm3, respectively, regardless of the Ms indicating that density is not affected by the SiO2/Na2O molar ratio. On the other hand, Figs. 5a and 6b show the 7th-day compressive strength results of FCB and NP-based geopolymers for Ms ratios of 0.50, 0.75, 1.00, 1.25 and 1.50, respectively. The results indicate an inverse relationship between the compressive strength and the value of Ms for FCB-based geopolymers. The 7th-day compressive strength increased from 0.6 to 1.9 MPa when Ms was reduced from 1.50 to 0.50 (Fig. 5a). For NP-based geopolymers, the highest compressive strength (1.1 MPa) was found to corre-

spond to a Ms of 1.25. It can be observed that there is an increasing trend from Ms value of 0.5 to 1.25 (Fig. 5b). After this peak, the compressive strength decreased by about 50% with a Ms of 1.50. The results indicate that the optimum Ms values for FCB and NP-based geopolymers are around 0.50 and 1.00–1.25, respectively. Therefore, a more accurate evaluation of this parameter in a smaller range was further carried out for the final definition of the optimum Ms value for each type of geopolymer.

3.2. Analysis of the effects of the Na2O content and optimization of the silica modulus (Ms) Several authors have found that there is an optimum value of Na2O content for FCB-and NP-based geopolymers. According to Komnitsas [20] and Allaverdhi [22], the optimum value for the FCB ones is 8%. Likewise, Allahverdi [15] reported an optimum Na2O content of 10% for a pumice type pozzolana geopolymer, for which the compressive strength was observed to increase with Na2O content. According to these works, the optimum Na2O concentration for geopolymer formation is the one that is able to balance electrical charges in the Si and Al tetrahedral chemical structures, accelerates the geopolymerization reaction and gives higher compressive strengths [23,36]. It is to note that higher Na2O content might result in the presence of unreacted alkali which usually lowers the mechanical properties, as observed in FCB-based geopolymers [20]. To evaluate the effects of the Na2O content in the geopolymer matrix production and to determine the optimum value of Ms, density and 7th-day compressive strength were evaluated in samples with different mixtures and oven conditions. For FCB-based geopolymer samples, the water/binder ratio and the oven curing time were kept constant at 0.29 and 7 days, respectively. For NPbased geopolymer samples, the water/binder ratio was 0.40 while the oven curing time was 2 days. The oven curing temperature was 65 °C for both geopolymer matrices. FCB-based geopolymer samples were produced with Ms values of 0.50, 0.55 and 0.60 (all around 0.50 as reported in the previous section) and different Na2O contents (6, 8 and 10%). Na2O contents evaluated for the

637

G. Silva et al. / Construction and Building Materials 215 (2019) 633–643 Table 2 Experimental plan for the optimization of FCB and NP-based geopolymers production. Stage

N°

Raw Material

Ms (mol/mol)

Na2O (wt. %)

Water/binder (wt./wt.)

Oven curing temperature (°C)

Oven curing time (days)

Total curing time (days)

I Analysis of the effects of silica modulus (Ms)

1 2 3 4 5 6 7 8 9 10

FCB

0.50 0.75 1.00 1.25 1.50 0.50 0.75 1.00 1.25 1.50

8

0.30

65

1

7

10

0.40

65

1

7

11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

FCB

0.50

6 8 10 6 8 10 6 8 10 6 8 10 6 8 10 6 8 10

0.29

65

7

7

0.29

65

7

7

0.29

65

7

7

0.40

65

2

7

0.40

65

2

7

0.40

65

2

7

0.28 0.27 0.26 0.25 0.40 0.52 0.64

65

7

7

65

7

7

1 3 1 3 7 1 3 7 1 3 1 3 7 1 3 7

7

II Analysis of the effects of Na2O content and optimization of the value of silica modulus (Ms)

III Optimization of water/binder ratio

IV Optimization of oven curing conditions

NP

0.55

0.60

NP

1.08

1.17

1.25

29 30 31 32 33 34 35

FCB

0.60

8

NP

1.08

8

36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

FCB

0.6

8

0.27

65

0.6

8

0.27

80

0.6

8

0.27

95

1.08

8

0.52

65

1.08

8

0.52

80

1.08

8

0.52

95

NP

NP-based geopolymers were also 6, 8 and 10% while Ms values were 1.08, 1.17 and 1.25 (all between 1.00 and 1.25 as reported before). The density results in both matrices indicate that for a constant Na2O, the density of geopolymers remained almost constant, regardless of the Ms value. However, for a constant Ms and varying conditions of Na2O, the densities in both geopolymers increase significantly with higher Na2O content. For instance, density values of 1.71, 1.91 and 2.03 g/cm3 were obtained in FCB-based geopolymers for a Na2O content of 6%, 8% and 10% with a Ms of 0.60, respectively. In the case of NP-based geopolymers, a slightly increasing trend in density values as the percentage of Na2O rises was also observed (1.63, 1.73 and 1.76 for 6, 8 and 10%, respectively). Figs. 6a and 7b illustrate the influence of the Na2O content for a given Ms on the 7th-day compressive strength of FCB and NP-based geopolymers, respectively. As shown, a Na2O content of 8% gave

7

7

7 7

7

the highest compressive strength for all Ms values for both geopolymer matrices. In the case of FCB-based geopolymers (Fig. 6a), the highest compressive strength (21.3 MPa) was obtained when Na2O content raised from 6% to 8% with a Ms value of 0.60. Nevertheless, when the geopolymer matrix was prepared with a Na2O content of 10%, the compressive strength was reduced to 7.3 MPa. For NP-based geopolymers (Fig. 6b), the highest compressive strength was 3.7 MPa, which was obtained when the Na2O content was 8% and the Ms value was 1.08. Higher and lower Na2O contents caused a decrease in the 7th-day compressive strength (2.5 and 2.3 MPa for Na2O contents of 6 and 8%, respectively). Fig. 7 shows the FTIR spectra of FCB and NP-based geopolymers with different Na2O contents. Fig. 8a corresponds to FCB and FCBbased geopolymers while Fig. 8b corresponds to NP and NP-based geopolymers. Even though the spectra of the geopolymers

G. Silva et al. / Construction and Building Materials 215 (2019) 633–643

2.5

Compressive strength (MPa)

Compressive Strength (MPa)

638

2.0 1.5 1.0 0.5 0.0 0.5

0.75

1 Ms (a)

1.25

1.2 0.9 0.6 0.3 0.0

1.5

0.50

0.75

1 Ms (b)

1.25

1.5

Fig. 5. Effect of silica modulus (Ms) on the 7th-day compressive strength: (a) FCB and (b) NP-based geopolymers.

Ms = 0.50

25

Ms = 0.55

Ms = 0.60

20 15 10 5

5.0 Compressive strength (MPa)

Compressive strength (MPa)

30

Ms = 1.08 Ms = 1.25

4.0

Ms = 1.17

3.0 2.0 1.0 0.0

0 6

8 10 Na2O content (%) (a)

6

8 10 Na2O Content (%) (b)

Fig. 6. Effect of Na2O content on the 7th-day compressive strength for a given Ms: (a) FCB and (b) NP-based geopolymers.

produced are very similar to the one for unreacted FCB, some changes can be noticed. The band around 1079 cm 1 in the FCB spectrum corresponds to T–O–Si asymmetric stretching (where T can be Al or Si) in the unreacted material. Some of these bonds are broken (during depolymerization and dissolution) and some new ones form (during polymerization and gelation) during geopolymer formation. In the initial stages of the reaction, the alkaline activating solution induces depolymerization, which involves the breaking of some of these bonds and the formation of ionic species or moieties with non-bridging oxygen atoms. It has been reported that for low Al content materials, higher concentrations of metal hydroxides are required [37]. During polymerization, new T–O–Si bonds are formed. This has been explained by Lee and van Deventer [32] when studying the geopolymerization of heterogeneous amorphous aluminosilicates. They also stated that bands at 1203, 1167 and 1117 cm 1 correspond to satellites of the band at 1079 cm 1. These researchers explained that during the geopolymerization reactions, the shift of the band at 1079 cm 1 to lower wavenumbers is due to the redistribution of the different T–O–Si structures, and presumably increase in Al proportion producing weaker bonds. These changes are apparent in Fig. 8a with the reduction of the transmission band at 1079 cm 1 and the appearance of bands at 1056, 1039 and 1027 cm 1, with increased Na2O. This reduction of the transmission band is also observed in Fig. 8b, where the band at 1041 cm 1 (corresponding to T–O–Si asymmetric stretching) decreases to 1015, 1014 and then to 1013 cm- 1 with the increase in Na2O content. Depolymer-

ization increases with more Na2O since it favors the breakage of T– O–Si bonds and the formation of ionic species with non-bridging oxygen atoms, including AlO–4 anions with their corresponding Na+ counterions. The lack of sharpness in the bands has been attributed to disorder in the chemical structure, which is expected as geopolymers are mostly amorphous materials [38]. This is also supported by the gradual disappearance of the shoulder observed at 567 cm 1 in the raw material, which corresponds to Si–O and Al–O bonds present in glasses with certain long-range structural order, such as rings, tetrahedral or octahedral structures [32], that are destroyed during the polymerization. As a conclusion from the second stage of the experimental plan, the optimum alkaline solution for FCB was determined to have a Na2O content of 8% with a Ms of 0.60, this produced the highest compressive strength of 21.29 MPa after 7 days of oven curing time. In the case of NP-based geopolymers, the optimum alkaline solution had a Na2O content of 8% and Ms of 1.08, which gives a 7th-day compressive strength of 3.7 MPa with 2-day oven curing time. 3.3. Optimization of water/binder ratio Several authors reported the importance of the water content for the production of various types of geopolymers [14,18,23,34,39]. Indeed, Reig et al. [18] found that the water content was a key parameter for achieving high compressive

639

1200

800

600

451 435

695 695

451

695

776

451 447

780

1015

463 1041

463 1000

780

1013 1014

460 460

567 567

797 778

693

881

1079 1039 1079 1056

778

% Transmittance

459

578 579

694 695

881

1027

1080

694

10% Na2O=10%, Ms=1.08, w/b=0.40, 7d at 65°C

797 778

8% Na2O=8%, Ms=1.08, w/b=0.40, 7d at 65°C

Na2O=10%, Ms=0.60, Na2O=10% (Mixw/b=0.29, N°14) 7d at 65°C

798 778

Na2O=6%, Ms=1.08, w/b=0.40, 7d at 65°C 6%

Na2O=8%, Ms=0.60, Na2O=8% (Mix w/b=0.29, N°13) 7d at 65°C

881

Unreacted NPNP Unreacted

Na2O=6%, Ms=0.60, Na2O=6% (Mix w/b=0.29, N°12) 7d at 65°C

797 778

Unreacted FCBFCBP Unreacted

1079

1165 1165 1165

1165

% Transmittance

G. Silva et al. / Construction and Building Materials 215 (2019) 633–643

400

1200

1000

800

600

400

Wavenumber (cm-1)

Wavenumber (cm-1)

(a)

(b )

30

40 35 30 25 20 15 10 5 0

Compressive strength (MPa)

Compressive strength (MPa)

Fig. 7. FT-IR spectra: (a) Unreacted FCB and FCB- based geopolymers with different Na2O contents and (b) Unreacted NP and NP-based geopolymers with different Na2O contents.

0.25 0.26 0.27 0.28 0.29 Water/binder ratio (w/b) (a)

25 20 15 10 5 0 0.40 0.52 0.64 Water/binder ratio (w/b) (b)

Fig. 8. Effect of the water/binder ratio on the 7th-day compressive strength: (a) FCB and (b) NP-based geopolymers.

strengths. Allahverdi [15,23] has suggested that when the water content decreases, the total volume of pores and the formation of shrinkage cracks during drying also decrease, resulting in a geopolymer matrix with enhanced mechanical properties. However, lower water contents than the optimum caused a decreased in the compressive strength probably due to the presence of unreacted particles and poor cohesiveness if there is not enough liquid phase to wet all the particles. In this third stage of the experimental plan, the influence of the water content on the 7th day compressive strength of geopolymer matrices was evaluated. The water content, here named as water/ binder ratio, is expressed as the weight ratio between water in the alkaline sodium silicate solution and the dry raw material used in the mixture. Oven curing conditions were kept constant in this stage for both types of matrices (65 °C for 7 days). For FCB-based geopolymers, Na2O content and Ms were fixed to 8% and 0.60, respectively. Water/binder ratio was reduced from 0.29 (used in stage 2) until 0.25 with steps of 0.01. For NP-based

geopolymers, the optimum values of Na2O content and Ms were also used (8% and 1.08, respectively). Water/binder ratio was studied from 0.40 to 0.64 with steps of 0.12. Lower values of water/binder ratio than the lowest ones studied for each geopolymer were discarded because of poor workability of the mixtures. Density measurements after 7 days of curing for both geopolymer matrices led to the conclusion that there is no significant change when the water/binder ratio varies for a constant Ms ratio and Na2O content. On the contrary, as shown in Fig. 8, the water/binder ratio did exhibit a significant influence in the compressive strength of both geopolymer matrices. The highest compressive strength achieved in this second stage for FCB-based geopolymers (see Fig. 8a) was 35.2 MPa when the water/binder ratio was 0.27. The results evidenced also the existence of an optimum content of water since values lower than 0.27 (lower amount of water) exhibited low compressive strengths. On the other hand, the highest compressive strength for NP-based geopolymers (see Fig. 8b) was 25.8 MPa and

G. Silva et al. / Construction and Building Materials 215 (2019) 633–643

was achieved for a water/binder ratio of 0.52. The results indicate a significant increment of the compressive strength when the water/ binder ratio was reduced up to 0.52. It can also be observed that the compressive strengths for the water/binder ratio of 0.40 and 0.52 are very similar. These results lead to the conclusion that a similar behavior under compression loads can be obtained with both values even if there is an increased in workability in the case of a water/binder ratio of 0.52. These results indicate that the optimum water/binder ratio depends on the raw material. In addition, the formation of cracks due to shrinkage during drying was strongly reduced when optimum values of water/binder ratio were employed for both FCB and NP-based geopolymer matrices. From this stage of the experimental plan, it can be concluded that an optimum alkaline activation for FCB-based geopolymers occurs with an 8% Na2O content, a 0.60 Ms ratio and a water/binder ratio of 0.27. On the other hand, for the NP-based geopolymers, the optimum values are 8%, 1.08 and 0.52, respectively. 3.4. Optimization of oven curing conditions

Compressive strength (MPa)

Finally, oven curing conditions were investigated in terms of temperature and curing time using the optimized alkaline solutions found in the previous stages for each geopolymer matrix. These two conditions, curing temperature and curing time, affect some of the processes that occur during geopolymerization: diffusion and reaction rates (favored by higher temperatures) and water evaporation (that comes not only from the activating solution but is also produced during the reactions). Therefore, a balance between these two conditions must be achieved in order to allow the formation of an extended network. The studied oven curing times for FCB and NP-based geopolymers were 1, 3 and 7 days while the curing temperatures were 65 °C, 80 °C and 95 °C. As expected, for a given temperature, the density decreased with longer oven curing time in both geopolymer matrices. A similar behavior in density was observed when the temperature increased for a fixed curing time. For instance, density values of FCB-based geopolymer samples oven-cured for 7 days were 1.91, 1.83 and 1.74 g/cm3 when the curing temperature was 65 °C, 80 °C and 95 °C, respectively. In the case of NPbased geopolymers, an oven curing time of 7 days at 65 °C, 80 °C and 95 °C resulted in densities of 1.61, 1.55 and 1.42 g/cm3, respectively. The reduction of density could be attributed to the greater loss of water by evaporation with increasing temperature and longer oven curing times.

45 40 35 30 25 20 15 10 5 0

65°C

80°C

95°C

The influence of the curing conditions on the 7th-day compressive strength of FCB and NP-based geopolymers is illustrated in Figs. 9a and 10b, respectively. As shown in Fig. 9a, the results of FCB-based geopolymer samples cured at 65 °C and 80 °C are very similar and evidence a linear relationship of compressive strength and the oven curing time. At 95 °C, however, there is an important change in behavior for times longer than 3 days: no improvement of compressive strength is observed. These results suggest that, at 95 °C, polymerization reactions stop after 3 days probably as a result of a rapid loss of the liquid phase due to higher water evaporation. This affects negatively the geopolimerization since reactive species cannot react if there is no diffusion in the mixture [20]. For lower temperatures, it is clear that longer curing times result in higher compressive strengths which reach nearly 37 MPa for 7 days. Similar to FCB-based geopolymers, the results for NP-based geopolymers (see Fig. 9b) indicate that the compressive strength is influenced by oven curing time and curing temperature. For curing at 65 °C, the largest increase in compression resistance is seen after 3 days of curing. For curing at 80 °C, the largest increase in strength occurred in the first 3 days (a nearly 700% fold was registered). When curing was performed at 95 °C, the compression strength increased significantly in the first 3 days of curing. However, after 7 days at this temperature, the strength did not change suggesting that polymerization had also stopped at 3 days. The results indicate that the best curing conditions for NP-based geopolymers are 65 °C for 7 days, which produces a material with 26 MPa of compression strength. Failure modes were further studied to understand the influence of different oven curing conditions. As shown in Fig. 10, failure modes shifts from ductile to a very brittle when oven curing time is increased in both geopolymer matrices. Moreover, samples with longer curing times evidenced a higher modulus of elasticity. As shown, an abrupt loss of resistance after reaching peak stress is exhibited in samples cured in an oven for 7 days (this is typical behavior of brittle cementitious materials). It seems that, unlike fly ash-based geopolymers, FCB and NP-based geopolymers need more time to react and, therefore, to develop higher mechanical properties. This result is in accordance with other studies of geopolymer samples formed with FCB as the source material, that were also cured for 7 days in the oven to produce high strength geopolymer matrices and mortars [14,18,20]. Fig. 11 shows the FTIR spectra of the FCB and NP geopolymers with different oven curing times, Fig. 11a corresponds to FCB and FCB-based geopolymers while Fig. 11b corresponds to NP and NP-based geopolymers. Fig. 11a presents the reduction of the band

30 Compressive Strength (MPa)

640

65°C

80°C

95°C

25 20 15 10 5 0

1

3 7 Oven curing time (days) (a)

1

3 7 Oven curing time (days) (b)

Fig. 9. Effect of oven curing time on the 7th-day compressive strength for a given oven curing temperature: (a) FCB and (b) NP-based geopolymers.

641

G. Silva et al. / Construction and Building Materials 215 (2019) 633–643

35

1d at 80°C + 6d at 22 °C 3d at 80°C + 4d at 22 °C 7d at 80°C

40 35

Compressive stress (MPa)

Compression stress (MPa)

45

30 25

20 15 10 5 0

1d at 65°C + 6d at 22 °C 3d at 65°C + 4d at 22 °C 7d at 65°C

30 25

20 15 10 5

0 0

0.05

0.1

0

Strain (mm/mm)

0.05 0.1 0.15 Strain (mm/mm)

(a)

0.2

(b)

Fig. 10. Compression stress vs. strain curves and failure modes for a given oven curing time: (a) FCB and (b) NP-based geopolymers.

Unreacted FCBP

1200

456

779 781 780

881

1036

1000 800 600 -1 Wavenumber (cm ) (a)

1041

463

567

463

1036

780

451 455

881

455 455

694

693

% Transmittance

694 694

455

797 778 797 778 797 778 797 778

1031

1078 1039 1078

1081 1059 1032 1010 1079

1165 1165 1165

1165

% Transmittance

7d at 80°C

1036

3d at 80°C + 4d at 22°C

881

Unreacted NP 1d at 65°C + 6d at 22°C 3d at 65°C + 4d at 22°C 7d at 65°C

1d at 80°C + 6d at 22°C

400 1200

1000

800

Wavenumber (b)

600

400

(cm-1)

Fig. 11. FT-IR spectra: (a) Unreacted FCB and FCB- based geopolymers cured for different oven time periods and (b) Unreacted NP and NP-based geopolymers cured for different oven time periods.

at 1079 cm 1 and the increase of the band at 1031 cm 1, which shows that curing time at 80 °C affects also the distribution of the T–O–Si bonds in the products. Fig. 11b also presents a similar shift. It can be seen that the band at 1041 cm 1 in NP moves to 1036 cm 1 in the cured products. This band is the same for all NP geopolymers at different oven curing times, suggesting that the distribution of the T–O–Si bonds is similar for all the products. The shoulder at around 570 cm 1 shows a loss of long-range order structures after geopolymerization according to previous studies [40,41]. From deconvolution analysis of IR spectra of fly ash geopolymers, Lee and van Deventer [32] assigned a band at 1102–1105 cm 1 to tetrahedral SiO4 structures, as a prove of polymerization. In the spectra shown in Fig. 11, a subtle shoulder can be seen around this region, which is more noticeable in the sample cured in the oven for 7 days.

As a conclusion from the fourth stage of the experimental plan, an oven curing at 65 °C and 80 °C for 7 days were defined as the best curing conditions to produce high strength FCB-based geopolymers, achieving a 7th-day compressive strength around 35 MPa. In the case of NP-based geopolymers, samples cured at 65 °C for 7 days in an oven presented the highest 7th-day compressive strength (26 MPa).

4. Conclusions In this research, alternative cementitious materials with good mechanical properties and a high potential for structural purposes have been developed using FCB wastes and NP from volcanic origin. Compressive strengths of 36.6 ± 3.4 MPa and

642

G. Silva et al. / Construction and Building Materials 215 (2019) 633–643

25.8 MPa ± 0.3 MPa were achieved for FCB and NP-based geopolymers, respectively. The optimization of the production process of FCB and NP-based geopolymers was carried out by analyzing the influence of five key parameters: Na2O content, Ms, water/binder ratio, curing temperature and curing time. Based on the experimental results, the following conclusions can be drawn: 1. The Na2O content and the Ms ratio affect the compressive strength of the resulting geopolymers. Variations in the Na2O content significantly affect the 7th-day compressive strength. An 8% Na2O content resulted in the highest compressive strength for both geopolymer matrices, lower and higher values of Na2O content produce materials of lower compressive strengths. In addition, density values of FCB and NP-based geopolymers increase as the Na2O content increases. Furthermore, it was demonstrated that the optimum Ms values for FCB and NP-based geopolymers were different, 0.60 and 1.08, respectively. 2. A decrease in the water/binder ratio from 0.29 to 0.27 led to a significant increase in the compressive strength of FCB-based geopolymers. However, lower values of water/binder ratio resulted in lower compressive strengths which may be due to the presence of unreacted FCB in the resulting product. In the case of NP-based geopolymers, the compressive strength was very similar despite an increase in workability for water/binder ratio values of 0.40 and 0.52. 3. Oven curing time and temperature are the parameters that influenced the most the compressive strength of both geopolymers matrices. Unlike other silico-aluminate sources such as fly ash, it is apparent that both, FCB and NP, need more time to cure at high temperatures, this is, to polymerize and reach good mechanical properties. Experimental findings indicate that higher compressive strengths in FCB and NP-based matrices are achieved with moderate heating (65 °C-80 °C and 65 °C, respectively) with 7 days of curing times. Conflicts of interest The authors declare that there are no known conflicts of interest. Acknowledgements This work was supported by CONCYTEC and SENCICO PERU under the project: ‘‘GeoBloque: Desarrollo de bloques de construcción ultraligeros con geopolímeros” (Contract No. 105-207FONDECYT). Additional support was received from CONCYTEC (J108-2016) under the ERANet-LAC project (ELAC2015/T020721): ‘‘Development of eco-friendly composite materials based on geopolymer matrix and reinforced with waste fibers”. The authors are very grateful to Compañía Minera Agregados Calcáreos S.A (COMACSA) for providing raw materials and the use of its facilities and equipment. The authors would like to mention CAM-PUCP for facilitating the use of the XRD equipment. Finally, Guido Silva and David Castañeda acknowledge the support from CONCYTEC (Contract No 232-2015-FONDECYT) for their fellowship funding. References [1] S. Böhmer, C. Neubauer, E. Schachermayer, B. Winter, B. Walter,‘‘Aggregates Case Study: Data Gathering. Final Report Referring To Contract N° 1507872007 F1SC-AT,” Vienna, 2008. [2] Ministerio de Fomento de España, Actualización Del Catálogo De Residuos Utilizables En Construcción, 2010. [3] U.S. Environmental Protection Agency (EPA), Advancing Sustainable Materials Management: 2014 Fact Sheet, Washington DC, 2016. [4] S. Ghosh, S. Ghosh, Construction and demolition waste, Sustain. Solid Waste Manage. (2015) 511–547.

[5] M. Nehdi, A. Khan, Protective system for buried infrastructure using recycled tire rubber-filled cement mortars, Spec. Publ. 219 (2004) 99–114. [6] F. Puertas, I. García-Díaz, M. Palacios, M. Gazulla, M. Gómez, M. Orduña, Clinkers and cements obtained from raw mix containing ceramic waste as a raw material. characterization, hydration and leaching studies, Cem. Concr. Compos. 32 (3) (2010) 175–186. [7] F. Pacheco-Torgal, S. Jalali, Reusing ceramic wastes in concrete, Constr. Build. Mater. 24 (5) (2010) 832–838. [8] N. Ay, M. Ünal, Use of waste ceramic tile in cement production, Cem. Concr. Res. 30 (3) (2000) 497–499. [9] J.N.Y. Djobo, A. Elimbi, H.K. Tchakouté, S. Kumar, Volcanic ash-based geopolymer cements/concretes: the current state of the art and perspectives, Environ. Sci. Pollut. Res. 24 (5) (2017) 4433–4446. [10] A. Diaz, J. Ramirez, Estudio Geológico-Económico De Rocas Y Minerales Industriales De Arequipa Y Alrededores., INGEMMET. Boletín, Serie B: Geología Económica., no. 22, p. 158, 3 mapas, 2010. [11] A. Diaz, J. Ramirez,‘‘Compendio De Rocas Y Minerales Industriales En El Perú.,” INGEMMET. Boletín, Serie B: Geología Económica, no. 19, p. 482, 44 mapas, 2009. [12] G. Habert, J.B. D’Espinose De Lacaillerie, N. Roussel, An environmental evaluation of geopolymer based concrete production: reviewing current research trends, J. Clean. Prod. 19 (11) (2011) 1229–1238. [13] L.K. Turner, F.G. Collins, Carbon dioxide equivalent (CO2 -E) emissions: a comparison between geopolymer and OPC cement concrete, Constr. Build. Mater. 43 (2013) 125–130. [14] L. Reig, M. Tashima, L. Soriano, M. Borrachero, J. Monzó, J. Payá, Alkaline activation of ceramic waste materials, Waste Biomass Valorization 4 (4) (2013) 729–736. [15] A. Allahverdi, K. Mehrpour, E. Najafikani, Investigating the possibility of utilizing pumice-type natural pozzonal in production of geopolymer cement, Ceramics - Silikaty 52 (1) (2008) 16–23. [16] D. Bondar, C.J. Lynsdale, N.B. Milestone, N. Hassani, A.A. Ramezanianpour, Effect of type, form, and dosage of activators on strength of alkali-activated natural pozzolans, Cem. Concr. Compos. 33 (2) (2011) 251–260. [17] L. Reig, L. Soriano, M. Borrachero, J. Monzó, J. Payá, Influence Of Calcium Aluminate Cement (CAC) On Alkaline Activation Of Red Clay Brick Waste (RCBW), Cem. Concr. Compos. 65 (2016) 177–185. [18] L. Reig, M. Tashima, M. Borrachero, J. Monzó, C. Cheeseman, J. Payá, Properties and microstructure of alkali-activated red clay brick waste, Constr. Build. Mater. 43 (2013) 98–106. [19] N. Rakhimova, R. Rakhimov, Alkali-activated cements and mortars based on blast furnace slag and red clay brick waste, Mater. Des. 85 (2015) 324–331. [20] K. Komnitsas, D. Zaharaki, A. Vlachou, G. Bartzas, M. Galetakis, Effect of synthesis parameters on the quality of Construction and Demolition Wastes (CDW) geopolymers, Adv. Powder Technol. 26 (2) (2015) 368–376. [21] M. Zawrah, R. Gado, N. Feltin, S. Ducourtieux, L. Devoille, Recycling and utilization assessment of waste fired clay bricks (Grog) with granulated blastfurnace slag for geopolymer production, Process Saf. Environ. Protect. 103 (2016) 237–251. [22] A. Allahverdi, E. Najafi, Construction wastes as raw materials for geopolymer binders, Int. J. Civil Eng. 7 (3) (2009) 249–257. [23] A. Allahverdi, K. Mehrpour, N. Kani, Taftan pozzolan-based geopolymer cement, Int. J. Eng. Sci. 19 (3) (2008) 1–5. [24] E.N. Kani, A. Allahverdi, Effect of chemical composition on basic engineering properties of inorganic polymeric binder based on natural pozzolan, Ceram. – Silikaty 53 (3) (2009) 195–204. [25] G. Baronio, L. Binda, Study of the pozzolanicity of some bricks and clays, Constr. Build. Mater. 11 (1) (1997) 41–46. [26] J.N.Y. Djobo, A. Elimbi, H.K. Tchakouté, S. Kumar, Reactivity of volcanic ash in alkaline medium, microstructural and strength characteristics of resulting geopolymers under different synthesis conditions, J. Mater. Sci. 51 (22) (2016) 10301–10317. [27] C. Atisß, E. Görür, O. Karahan, C. Bilim, S. Ilkentapar, E. Luga, Very high strength (120 MPa) class F fly ash geopolymer mortar activated at different NaOH amount, heat curing temperature and heat curing duration, Constr. Build. Mater. 96 (2015) 673–678. [28] P. Duxson, J.L. Provis, Designing precursors for geopolymer cements, J. Am. Ceram. Soc. 91 (12) (2008) 3864–3869. [29] ASTM International, ASTM 618-17a, Standard Specification For Coal Fly Ash And Raw Or Calcined Natural Pozzolona For Use In Concrete, West Conshohocken, PA, 2017. [30] J. Davidovits, Geopolymers and geopolymeric materials, J. Therm. Anal. 35 (2) (1989) 429–441. [31] ASTM International, ASTM C109/C109M–16a, Standard Test Method For Compressive Strength Of Hydraulic Cement Mortars (Using 2-In . Or [ 50Mm ] Cube Specimens), West Conshohocken, PA, 2016. [32] W.K.W. Lee, J.S.J. Van Deventer, Use of infrared spectroscopy to study geopolymerization of heterogeneous amorphous aluminosilicates, Langmuir 19 (21) (2003) 8726–8734. [33] M. Catauro, F. Papale, G. Lamanna, F. Bollino, Geopolymer/PEG hybrid materials synthesis and investigation of the polymer influence on microstructure and mechanical behavior, Mater. Res. 18 (4) (2015) 698–705. [34] M. Yadollahi, A. Benli, R. Demirboʇa, The effects of silica modulus and aging on compressive strength of pumice-based geopolymer composites, Constr. Build. Mater. 94 (2015) 767–774.

G. Silva et al. / Construction and Building Materials 215 (2019) 633–643 [35] J.L. Provis, R.M. Harrex, S.A. Bernal, P. Duxson, J.S.J. Van Deventer, Dilatometry of geopolymers as a means of selecting desirable fly ash sources, J. NonCrystalline Solids 358 (16) (2012) 1930–1937. [36] V. Barbosa, Sintese E Caracterização De Polissialatos, Tese De Doutorado, Instituto militar de engenharia - IME, 1999. [37] C.A. Rees, J.L. Provis, G.C. Lukey, J.S.J. Van Deventer, Attenuated total reflectance fourier transform infrared analysis of fly ash geopolymer gel aging, Langmuir 23 (15) (2007) 8170–8179. [38] Y. Zhang, W. Sun, Z. Li, Infrared spectroscopy study of structural nature of geopolymeric products, J. Wuhan Univ. Technol., Mater. Sci. Ed. 23 (4) (2008) 522–527.

643

[39] C. Lampris, R. Lupo, C. Cheeseman, Geopolymerisation of silt generated from construction and demolition waste washing plants, Waste Manage. 29 (1) (2009) 368–373. [40] C. Panagiotopoulou, S. Tsivilis, G. Kakali, Application of the taguchi approach for the composition optimization of alkali activated fly ash binders, Constr. Build. Mater. 91 (2015) 17–22. [41] M. Vafaei, A. Allahverdi, Influence of calcium aluminate cement on geopolymerization of natural pozzolan, Constr. Build. Mater. 114 (2016) 290–296.