Influence of SiO2 nanoparticles on the microstructure, mechanical properties, and thermal stability of Portland cement nanocomposites

In the context of the microstructure of nano-SiO2 cement nanocomposites, information regarding the tricalcium silicate (C3S) and dicalcium silicate (C2S) phases during the hydration reaction remains inadequate. Therefore, in this research, the effects of different SiO2 nanoparticles contents on the microstructures, mechanical performance, and thermal stability of cement nanocomposites were investigated. The X-ray diffraction (XRD) results highlighted that the incorporation of nano-SiO2 enhanced the microstructure of the nanocomposites by decreasing the Ca(OH)2, C3S, and C2S peaks, leading to a higher production of calcium silicate hydrate (CSH) products. Moreover, the intensity of the unhydrated gypsum phase increased with the increase in the nano-SiO2 contents, which was suggestive of the retardation effect of nano-SiO2 on the aluminate phases and aluminate reaction. Furthermore, the addition of nano-SiO2 considerably enhanced the flexural strength and thermal stability of the nanocomposites compared to the control composites. The optimum ratio of nano-SiO2 was 1 wt%.


Introduction
Nanomaterials represent a novel class of materials that exhibit unique advantages at the nanoscale level. Recently, nanomaterials have been exploited to manufacture novel materials and composites in the most branches of materials science such as metals, polymers, and ceramics with superior physical, chemical, and biological properties compared to common or bulk materials [1][2][3][4][5][6]. Moreover, in biomedical applications, the use of nanomaterials such as nanodiamonds and zinc and magnesium nanoparticles in several biomedical procedures such as drug delivery, tissue scaffold and surgical implant preparation, and bio-imaging and therapeutic applications has received considerable attention [7,8].
In the applications of metal composite materials, the metal substrate and metal alloys (such as magnesium and its alloys) can be protected against corrosion by implementing coating techniques involving nanocomposites (polymers or conductive polymers containing nanoparticles) [9,10]. Furthermore, microparticles are often used to reinforce the metallic matrix; for example, the addition of 7 wt% TiB 2 micro-ceramic particles to reinforce metallic aluminum matrix composites can considerably improve the mechanical properties of such metallic composites [11].
Portland cement is a category of ceramic composite that consists of 86 wt% of two main crystalline compounds: tricalcium silicate (3CaO•SiO 2 ) or (C 3 S) and dicalcium silicate (2CaO•SiO 2 ) or (C 2 S). The reaction of C 3 S or C 2 S compounds with H 2 O to produce calcium silicate hydrate gel (C 3 S 2 H 8 ) and calcium hydroxide crystals (Ca(OH) 2 ) can be expressed as a hydration reaction [25,26]: (1) cementitious materials, SCM) to reduce environmental issues, as in general, concrete emits extremely high levels of CO 2 [27]. Such materials, which consist predominantly of micro-silica (micro-SiO 2 ), silica fume and fly ash as the pozzolanic materials, usually react chemically with the calcium hydroxide (CH) in cement composites or concrete to yield extra CSH products through a chemical pozzolanic reaction [28,29].
Notably, nano-SiO 2 has emerged as a progressive pozzolan to enhance the microstructure and mechanical performance of cement composites and concrete [30][31][32]. The chemical reaction of SiO 2 nanoparticles with Ca(OH) 2 crystals to yield CSH gel can be expressed as a pozzolanic reaction [33].
In addition, nano-SiO 2 is principally useful as a filler to densify the cement composite and enhance the microstructure and mechanical properties compared to those of the control cement or concrete composites [34]. Biricik and Sarier [35] reported that cement composites containing nano-silica outperformed cement composites containing silica fume or fly ash. However, research regarding the C 3 S and C 2 S peaks in the hydration reaction for cement nanocomposites containing nano-SiO 2 , as observed in X-ray diffraction (XRD) spectra, remains limited. Furthermore, the thermal stability has been not extensively examined through thermogravimetry analyses (TGA). Therefore, this study was aimed at clarifying the effects of the percentage content of nano-SiO 2 on the microstructure, mechanical properties, and thermal stability of nanocomposites.

Materials
Nano-SiO 2 with an average size of 18-25 nm and 99% amorphous silicon dioxide content ( Figure 1) was procured from Nanostructured and Amorphous Materials, Inc. (USA). Ordinary Portland cement type I (OPC) was used. The physical properties and chemical constitution of the OPC and nano-SiO 2 (NS) are summarized in Table 1.

Preparation of nanocomposites
Portland cement powder was partly replaced by nano-SiO 2 for 1%, 2%, and 3% weight percentages of cement (labeled 1NS, 2NS, and 3NS, respectively). Cement powders and NS were mixed as a dry binder for 10 min by using a Hobart mixer. The nanocomposites were prepared by via water to binder ratio of 0.48. According to ASTM C270 Standard [36], these nanocomposite pastes were mixed as wet mix for 5 min by using a Hobart mixer and subsequently case to the molds for a specific curing period. Details of the cement nanocomposites are listed in Table 2. For each mixture, five rectangular prism samples (7×2×1 cm 3 ) were preserved in water for 28 d, after which, the samples were immediately dried and tested.

Investigation of the microstructures
XRD patterns of the nanocomposites and control composites were acquired using a D8 Advance Diffractometer (Bruker, Germany) with a Cu Ka (λ = 1.5406 Å) source. The scanning of 2-theta (2θ) from 7°to 70°w as performed at a rate of 0.5 • /min. Scanning electron microscope (SEM) images of the morphology and microstructure of each sample were obtained using the NEON 40ESB, ZEISS instrument (high-resolution imaging of 2.5 nm at 1 kV).

TGA
TGA was conducted using a TGA/differential scanning calorimetry analyzer (Mettler Toledo, 1 STAR) to evaluate the thermal stability of the mixtures. The tests were conducted in an argon atmosphere with a range of temperature from 35 to1000°C and the heating rate was 10°C/min.

Sample porosity (P%)
Along with ASTM C20 Standard [37], the porosity value was determined using the Archimedes' principle: m s is the saturated weight (weight of the sample saturated with water, measured in air), and m d is the dry weight. m i is the suspended weight (weight of the sample saturated with water, measured while suspended in water). All the weights were specified in grams (g).

Mechanical properties: flexural strength
A three-point bending flexural test ( Figure 2) was implemented on a universal testing machine (LLOYD, 50 kN capacity) to measure the flexural strength. The test of five rectangular prism specimens (7×2×1 cm 3 ) of each type (control composites and nanocomposites) was estimated, and the displacement speed of 0.5 mm/min was used. The flexural strength can be evaluated as follows: where P m ,S, B, and Wdenote the maximum force or load, span, width, and thickness of the specimen, respectively.  (ii) dicalcium silicate (C 2 S; PDF 00-033-0302), with two main 2θ peaks corresponding to 32.14 • and 32.59 • ; (iii) tricalcium silicate (C 3 S; PDF 00-049-0442), with three main 2θ peaks corresponding to 29.29 • , 32.12 • , and 32.46 • . A certain overlap could be between the peaks of the C 2 S and C 3 S phases, as shown in Figure 4(b). The spectra presented in Figure 3 indicated that the incorporation of SiO 2 nanoparticles in the cement paste composites significantly modified the peak intensities of different phases of the nanocomposites compared to C, with the most notable change corresponding to 1NS. As shown in Figure 4(a), the reduction in the Ca(OH) 2 peaks indicated that nano-SiO 2 consumed additional CH through the pozzolanic reaction to form additional CSH products [38,39]. Moreover, as shown in Figure 4(b), the reduction in the C 2 S and C 3 S peaks indicated that the presence of SiO 2 nanoparticles accelerated the hydration reaction to produce more CSH products [40,41]. Thus, the microstructures of nanocomposites become denser than those of plain cement composites, as confirmed through the formation of additional CSH products. This enhancement indicated that nano-SiO 2 functions as a dispenser in the cement matrix, and the SiO 2 nanoparticles serve to promote the pozzolanic reaction.

The microstructural examination by XRD analysis
Moreover, the XRD patterns indicated an increase in the intensity of gypsum peaks (2θ peaks corresponding to approximately 11.67 • , as indicated in PDF Card 04-015-4421) with the increase in the SiO 2 nanoparticles content. Portland cement clinker contains tricalcium aluminate (C 3 A) and gypsum (calcium sulfate hydrate CaSO 4 ·2H 2 O). C 3 A reacts with gypsum rapidly at the early age in the hydration reaction; however, the presence of SiO 2 NS likely delays this aluminate reaction, thereby reducing the consumption of gypsum. Thus, in the observed phenomenon, it was likely that a portion of gypsum had not reacted, and the aluminate reaction between C 3 A and gypsum was still in progress [42,43]. This phenomenon has also been reported by other researchers. Hou et al. [44] studied the effect of 1 wt% nano-SiO 2 on the hydration process of C 3 A-gypsum and C 3 A-C 3 S-gypsum systems and obtained similar results: The intensity of the gypsum peak, as observed in the XRD pattern for the C 3 Agypsum system with 1 wt% nano-SiO 2 , was higher than that in the pattern for the C 3 A-gypsum system without 1 wt% nano-SiO 2 . Moreover, the authors derived a similar conclusion for C 3 A-C 3 S-gypsum systems containing 1 wt% nano-SiO 2 : The mass fraction obtained through the quantitative XRD (QXRD) of gypsum was higher than that of the plain specimen after 12 d. The authors considered that the SiO 2 nanoparticles adsorbed on the surface of C 3 A owing to the electrostatic interaction, which led to the retardation effect of the NS on the aluminate phases and aluminate reaction or hydration. However, the reaction between C 3 A and either gypsum or water was still in progress at 28 d, and in later age, this reaction could likely enhance the microstructure and mechanical properties of the nanocomposite.

Sample porosity
The porosities of the C and nanocomposite samples are listed in Table 3. The addition of SiO 2 nanoparticles in the cement matrix significantly decreased the porosity of the nanocomposites compared to that of the C samples. The porosity of 1 NS nanocomposites was 23.6% smaller than that of the control paste. This enhancement highlighted that the microstructure of nanocomposites, especially those with 1 wt% nano-SiO 2 , become more highly packed and compact. This enhancement could be attributed to the filling effect of NS [45][46][47].  However, the addition of more nano-SiO 2 led to an increased porosity, which indicated that the voids in the matrix grew due to the agglomerations of the high nano-SiO 2 contents. Oltulu and Sahin [48] reported that the incorporation of nano-SiO 2 in excess of the optimum ratio of 1.25 wt% could lead to an increase in the pore volume of nanocomposite mortars owing to the agglomeration of the NS.

Flexural strength
The flexural strengths of the samples are presented in Figure 5. The values for 1NS, 2NS, 3NS, and plain composites were 7.23, 6.64, 5.99, and 5.12 MPa, respectively. In particular, the flexural strengths of 1NS, 2NS, and 3NS were 41%, 30%, and 17% higher than that of the basic composite, respectively. This strength enhancement could be attributed to two factors, namely, the filler effect and pozzolanic reaction. In terms of the filler effect, the NS filled the voids in the porous CSH microstructures owing to their small size. Consequently, the CSH microstructure became denser. Furthermore, nano-SiO 2 likely supported the pozzolanic reaction, which increased the amount of CSH in the nanocomposites [49,50]. Notably, 1 wt% NS was the optimal ratio to achieve a high nanocomposite strength. Nazerigivi and Najigivi [51] conducted flexural strength tests and demonstrated that cement composites with 2% (15 nm) and 1.5% (80 nm) blends of nano-SiO 2 exhibited a higher strength in comparison with those of the other specimens. However, when the content of nanoparticles exceeded this critical optimal   ratio, the strength reduced significantly owing to the agglomeration and low dispersion of the NS [52,53].

SEM-based analysis of the sample morphology
SEM micrographs of the nano-SiO 2 nanocomposites and control composite are shown in Figure 6(a-d). The comparative study for the SEM images of the specimens was focused on three factors, pertaining to the CSH gel, small pores mainly observed in the previous sections, and Ca(OH) 2 crystals. The microstructure of the control composite, as presented in Figure 6(a), had several pores as discussed in previous sections, few CSH products, and more Ca(OH) 2 crystals in comparison with those observed in the nanocomposite images. Furthermore, micro-cracking was observed. The addition of NS in the matrix of the composites improved the microstructure. As shown in Figure 6(b), the microstructure of 1NS nanocomposites was more compact with fewer voids and additional CSH gels because of the filler effect and pozzolanic activity. However, if the content of NS exceeded the critical or optimal ratio, agglomeration and low dispersion effects were observed in the microstructures of the nanocomposites [54,55]. For example, 3NS nanocomposites ( Figure 6(b)) displayed more pores than 1NS nanocomposites.

Examination of the thermal stability via TGA
The thermal stability (weight loss (%)) of the nanocomposites and control composites was examined by performing a TGA. The TGA curves of the nanocomposites and control samples are displayed in Figure 7.
The TGA curves of the samples exhibited three diverse decomposition stages or weight loss percentages. The first decomposition (35-230°C) was attributed to the dehydration (loss of water) of CSH products. The second decomposition (400-510°C) was related to the decomposition of Ca(OH) 2 . The third decomposition (670-780°C) corresponded to the decomposition of CaCO 3 [56][57][58]. Overall, the nanocomposites demonstrated a higher thermal stability than that of control composites from 35-1000°C. This improvement could be attributed to the dense and compact microstructure of the nanocomposites owing to the effect of the fillers and pozzolanic activity, which led to the formation of additional CSH gel. Furthermore, 1NS nanocomposites displayed a higher thermal stability than that of the nanocomposites in all the decomposition stages. Notably, the agglomeration and low dispersion of nano-SiO 2 at high contents affected the thermal stability, as in the case of the 3NS nanocomposites.

Conclusions
Overall, the flexural strength and thermal stability of the nanocomposites were significantly enhanced with the incorporation of SiO 2 nanoparticles and nanocomposites with 1 wt% NS led to the highest strength and thermal stability. The XRD analysis and SEM images confirmed that the microstructures of the cement nanocomposites were denser compared to those of the control composites. The pozzolanic activity of nano-SiO 2 helped reduce the Ca(OH) 2 , C 3 S, and C 2 S peaks, which led to the formation of further CSH products. Moreover, the filler effect of nano-SiO 2 helped reduce the number of pores in the microstructures of the nanocomposites. The agglomeration problem of nano-SiO 2 at higher contents requires further research. However, it was observed that the intensity of the unhydrated gypsum phase in the XRD patterns increased with an increase in the SiO 2 nanoparticle contents, which was suggestive of the retardation effect of nano-SiO 2 on the aluminate phases and hydration. Nano-SiO 2 has prospective applications in building materials in the form of ceramic nanocomposites and fiber-reinforced concrete.