Influence of silicon-dioxide nanoparticles in cementitious mortars: verification using x-ray diffraction, thermal analysis, physical, and mechanical tests

In recent years, nanotechnology has been applied to building materials, such as cementitious composites (e.g., mortar and concrete), to improve their properties. The aim of this study was to analyze the thermal, physical, and mechanical properties of mortars with and without silicon-dioxide (SiO2) nanoparticles. Experiments such as thermogravimetry and differential thermal analysis (TG-DTA), x-ray diffraction (XRD), fresh density, incorporated-air content, bulk density, capillary absorption, capillarity coefficient, flexural tensile strength, and compressive strength on prismatic specimens were performed on mortars and analyzed for different levels of nanosilica (nS). These levels were 1% and 3%, in addition to the reference mortar (0% nS). The TG-DTA curves showed an elevated content of chemically combined water and a lower content of calcium hydroxide (Ca(OH)2) in the 3% nS compositions, while the XRD curves presented a lower content of calcite and portlandite in the same mortar. These results indicate the fixation capacity of lime for the formation of calcium silicate hydrate (C-S-H), which is the primary cause of resistance in cementitious mortars. In addition, it was found that the use of nanosilica contributed to a fresh density increase of approximately 15%, which caused a minimum air-incorporated content decrease of 37% and a minimum bulk density increase of 10%. Higher densities resulted in a minimum water absorption reduction of 36%, owing to fewer pores in the mortars. Therefore, the capillarity coefficient decreased by a minimum of 41%. These nanoparticles also improved the minimum flexural tensile and compressive strengths by 88% and 158%, respectively, when using a 3% nS composition. These results can enable the use of lightweight aggregates in cementitious composites, improving their physical and mechanical characteristics and allowing greater reuse of these materials, including construction waste.


Introduction
Recent nanotechnology research has improved the properties of cement composites used in civil engineering construction [1]. Nanoparticles, also termed nanomaterials, have at least one critical dimension that can range from 1 to 100 nm (in the order of 10 −9 m) and are responsible for conferring new and interesting physical and chemical properties [2][3][4][5][6][7][8].
Nanomaterials increase the surface area of materials owing to their nanometric scale, resulting in a much higher reactivity than that of commonly used materials. This method has certain advantages in terms of reinforcement and efficiency. They also absorb or release heat more easily, and their melting temperatures are typically lower than those of solids [3,9].
Various studies evaluating conventional concrete and mortar using nanotechnology have concluded that it is possible to achieve compressive and flexural tensile strengths up to 50% higher depending on the nanomaterial content [4,5,10].
However, nanomaterials must adsorb water, which increases the consistency and reduces the fluidity of the mixture. This is because of the high specific area of the nanomaterial, which absorbs more water from the particle surface [11,12].
Studies have shown promising results for increasing the mechanical strength of materials using nanotechnology. However, there are few experimental results using materials from the Brazilian construction industry that have performed thermal analyses and physical properties [11,16].
Nanosilica has been frequently studied and contains ultrafine particles of silicon dioxide (SiO 2 ) in its crystalline (amorphous) form, which is a good pozzolanic material for addition in cementitious composites. This nanomaterial has dimensions of 10-50 nm and high purity (99% SiO 2 ), which contributes to higher particle packing capacities and lower concrete and mortar permeabilities [10,17,19].
At the microscale or nanoscale, silicon dioxide (SiO 2 ) can participate in the hydration process to produce hydrated calcium-silicate crystals (C-S-H) by reacting with calcium hydroxide (Ca(OH) 2 ) [10]. This is the most significant contribution of nanosilica to composites, which includes the production of secondary C-S-H gel. The portlandite or Ca(OH) 2 content in the composite is reduced by reactions with nanosilica, producing an additional dense product [6,10,16,17].
Thus, nS accelerates cement hydration via the nucleation effect and forms C-S-H in the matrix. Owing to the decreased voids content, this formation reduces the porosity, increases density, and reduces the watertransport properties. However, when the optimal nS content is exceeded, which is approximately 2%-3%, SiO 2 agglomeration can increase the porosity, contributing to an increase in the capillarity coefficient [1,19,20].
Thus, for high doses of nS, the newly formed hydration products have a greater volume, promoting internal tension between the paste and aggregates, which reduces stiffness. In addition, the high cost of nanoparticles may limit their application at high doses. Therefore, most studies have been limited to low doses [19].
Furthermore, issues regarding dosage problems remain and require further study. A balance between properties such as density, strength, stiffness, and durability has not been achieved in previous studies [1,20].
There is no ideal nS content defined for mortars and concrete; however, in a previous study [19] a content of 10% was used to replace the same mass of cement. Other studies used 5% as the maximum acceptable dosage for nS [21].
These different maximum acceptable dosages depend on the type of cement, size of the nanosilica, and whether the nanosilica comes in powder form or is already dispersed in the plasticizer.
In general, nanosilica promotes a better filling and packing effect in mixtures when added in optimum proportions, thereby reducing porosity and increasing density. When they are added above this ideal proportion, there is a reduction in the density of the mixtures owing to the high potential for agglomeration of nS, which contributes to increasing the segregation and viscosity, leaving a greater number of voids [20].
In addition to the nS content, previous studies have observed that dry nanosilica (nS) grains must be thoroughly dispersed in water and plasticizer additives to ensure better results. Nanosilica has a strong tendency for ionic adsorption in aqueous solutions; therefore, the formation of agglomerates is expected and plasticizing additives are used [27].
Within this context, the present study aims to present the results of an experimental study with conventional mortars, which varied the type of cement and content of nanosilica in the powder. The contents of 1% and 3% nS were used to replace the equivalent mass of cement. One sample without nanosilica (0%) was used as the basis for the comparative analysis.
The tests performed were thermal, physical, and mechanical, and the thermogravimetry and differential thermal analysis (TG-DTA) curves, x-ray diffraction (XRD) curves, fresh density, bulk density, air-incorporated ratio, capillary absorption, and capillarity coefficient were obtained. In addition, the flexural, compressive, and tensile strengths of the prismatic specimens were determined.

Experimental program 2.1. Mixing ratios
The constant parameters used in the mixture were a binder:aggregate ratio of 1:3 and a water:cement factor of 0.52, obtained through a consistency test [28]. This test was used to determine the ideal spread between 260 and 265 mm for coating mortars.
A plasticizing additive content of 1% in relation to the binder mass was adopted based on supplier specifications and previous studies [5,17,21,23].
The variables analyzed were the type of cement and the nanosilica content. For the binder, two commercialized cements with the same minimum compressive strengths were selected, that is, CP II F 40 and CP II E 40, to identify the differences in the results. For the nanosilica content, two contents (1% and 3%) were chosen to replace the cement mass, which corresponded to ARG 1% and ARG 3% mortars, respectively. The reference mortar (ARG 0% ) corresponds to no nanosilica and was used as a basis for comparison. A higher nS content was not considered because this leads to higher mix consistencies, which reduces fluidity.
Therefore, the reference mortars (ARG 0% ) did not use nanosilica, whereas the mortars ARG 1% and ARG 3% used 1% and 3% nS, respectively, equivalent to the mass of cement. The mixtures analyzed are listed in table 1.

Materials
Portland cement was directly obtained from the manufacturer. The specific densities were 3.08 g cm −3 (cement 1) and 3.13 g cm −3 (cement 2), derived from experimental tests [29]. Cement CP II F 40 has a compressive strength of 40.8 and 49.7 MPa at 7 and 28 d, respectively; whereas CP II E 40 cement has a strength of 36.6 and 46.5 MPa at 7 and 28 d, respectively, which were derived from tests conducted by the manufacturer [42].
For the aggregate, the sand was washed and placed in a kiln at a temperature of 100°C-105°C for 24 h. Thereafter, the sand was passed through a standard mesh sieve (#4) and stored in a closed container.
The grading curve of the sand (figure 1) reached its limits between the minimum and maximum values up to the N°30 (0.6 mm) sieve. However, from the N°50 (0.3 mm) sieve, the accumulated percentages approached the lower limit, returning to between the minimum and maximum values in the N°200 (0.075 mm) sieve.
Because the variation between the minimum limit and the values obtained in the N°50 (0.3 mm) and N°100 (0.15 mm) sieves were less than 4% [30], the entire grading curve was considered optimum. In addition, no more than 50% must be retained between two consecutive sieves [31], which was also observed, and no more than 25% must be retained between the N°50 and 100 sieves [31], yielding approximately 22.6%. Furthermore, the fine material was passed through the N°200 (0.075 mm) sieve should not exceeded 3% [24], which was 1.9%.
Using this curve, it was possible to obtain the maximum characteristic dimension (DMC) corresponding to the N°8 (2.36 mm) sieve and fineness module (MF), which corresponded to 2.27. The DMC indicated that the  sand was fine and that the fineness modulus was within the optimum zone for mortars, which were 2.05 and 3.05 [31].
The addition of plasticizer additives to the mixtures improved the cohesion and consistency of the mortars. Its bulk density was between 1.002 and 1.004 kg dm −13 according to the manufacturer's information. Water was obtained from a concessionaire for cold water supply in the city.
Nanomaterials are not currently sold in Brazil; therefore, they were imported from China for this study. It is a lightweight product with a density of 2.2-2.6 g ml −1 and sold as powdered silicon dioxide with 99% purity. This material has a spherical shape with an average diameter of 20 nm and surface area of 145-160 m 2 g −1 .

Experimental tests
Tests were conducted to characterize the mortars in their fresh and hardened states, as well as to determine the flexural tensile and compressive strengths. Additionally, differential thermal analysis (DTA) and thermogravimetric (TG) analyses were conducted, as well as characterization of the pastes by x-ray diffraction (XRD), as detailed in table 2, which includes the number of specimens, reference standard, and specimen size.
For the determination of the mechanical strength, ages of 7 and 28 d were considered, in which the specimens were initially ruptured by flexural tensile and subsequently ruptured by compression. In this test, the standard deviation of the results reached a maximum value of 0.3 MPa for flexural tensile and 0.5 MPa for compression.
The TG-DTA tests were performed on the mortars using CPII E 40 cement because it required temperatures up to 1000°C. The TG-DTA curves were simultaneously obtained using the Shimadzu DTH 60 equipment at a heating rate of 15°C min −1 and a maximum temperature of 1000°C. The analysis was performed under an inert atmosphere of nitrogen (N 2 ) with a constant flow rate of 50 ml min −1 . Calcined alumina was used as the reference specimen. The specimen holder was a cylindrical alumina crucible with a diameter of 5.0 mm and height of 2.5 mm. The measurements were performed using approximately 40 mg of each specimen.

Mixing procedure
Initially, the nanomaterial was mixed with water and a plasticizer additive for 1 min in an electric disperser at 10,000 rpm (figure 2) [10,18,19] to optimize the results.
Subsequently, mortar was prepared according to the procedure described below, which was based on a standard reference [36]. However, this standard reference [36] does not provide the time of insertion of the nanomaterial and the plasticizer additive; therefore, it was adapted from [10,18,19].  (1) Prepare 2.5 kg of anhydrous material, which refers to sand and cement.
(5) Mix all materials for 60 s at high power (speed).
(7) Add 25% remaining water mass to mixture 3 and mix for 60 s at low power (speed).
For the bulk density, capillary absorption, flexural tensile strength, and compressive strength tests, the specimens were removed from the molds after 48 h and immersed in a water tank until the test day.
Qualitatively, the DTA curves indicate that there are no differences in mortar decomposition temperatures, which are 50°C-200°C and 400°C-500°C. Endothermic peaks occurred at approximate temperatures 115°C and 470°C. At 28 d, the mortar with 3% nS exhibited the lowest heat loss at approximately 150°C. As discussed below, this is related to the higher pozzolanic activity in this composition, based on the thermogravimetric curve (TG).
The thermogravimetric curves (TG) of ARG 0% , ARG 1% , and ARG 3% mortars are presented in figure 4. Qualitatively, there was a weight reduction throughout the analyzed temperature range, owing to the loss of chemically combined water (AQC) up to 428°C. In addition, the remaining water was from the calcium hydroxide (Ca(OH) 2 ) produced up to 498°C, and the release of volatile carbon dioxide (CO 2 ) was from calcium carbonate (CaCO 3 ) decomposition up to 1000°C. Overall, the material lost 25%-27% of its mass. This depends on the composition and leaves a residual mass.
From the TG-DTA curves, it was possible to obtain the content of the hydrated compounds produced, content of portlandite remaining in the hydrated matrix, and content of calcium carbonate, based on the methodology of [38], as indicated in table 3.
Although the residual weights are close, it is not possible to compare them because the volatile components change. Thus, weight loss must be compensated through a normalization factor for a residual weight of 100% [41]. Thus, the weight losses were incremented by a corresponding value for the comparative analysis.
The chemically combined water content (AQC) was higher in ARG 3% , as shown in figure 5, indicating greater cement hydration and pozzolanic activity. According to [38], the presence of fine particles favors cement nucleation, precipitating hydrated compounds on the added particles, and consequently increasing hydration.
The water from calcium hydroxide was also lower for this mortar, as was the remaining lime content, owing to the lower amount of cement and higher amount of active addition. The lower alkaline reserve also evidences the lime-fixing capacity in the formation of C-S-H, as a result of the high specific area and amorphous character of the nanosilica. By characterizing the sample using XRD, it was possible to observe lower portlandite and calcite peaks in the paste with 3% nS ( figure 6). This is also due to the lower amount of cement in this paste in view of the mass replacement by nanosilica, but the lower remaining alkaline reserve may indicate a higher content of C-S-H.
In this way, the SiO 2 nanoparticles influenced the formation of additional C-S-H, which was higher in mortars with 3% nanosilica. Thus, this result indicates that this mortar with 3% nS presents hydration growth due to the formation of additional C-S-H, which influences the physical and mechanical properties of these composites and may bring new and interesting characteristics. Therefore, these analyses were relevant for identifying the contents of these composites and justifying the results presented below.

Densities and air-incorporated content
The fresh densities of the ARG 0% , ARG 1% , and ARG 3% mortars are presented in figure 7. The fresh density obtained for mortar ARG 1%,C1 was approximately 5% higher than that of ARG 0%,C1 , while the density of mortar ARG 3%,C1 was approximately 16% higher than that of mortar without nanomaterials. Thus, the effect of nS is greater when a 3% nanomaterial is used in mortars with cement 1. For mortars with cement 2, it was possible to   reach a 15% higher density than that of the reference mortar when 1% nS was added, indicating better accommodation of the grains, which can influence the mechanical strength.
The incorporated-air content of these mortars is also shown in figure 8, which indicates that there is a decrease in voids with the addition of the nanosilica. This parameter is inversely related to the fresh density, which means that a higher density can decrease the incorporated-air content.
This reduction was approximately 11.5% and 37% for ARG 1%,C1 and ARG 3%,C1 mortars at 28 days in relation to the reference mortar. For the ARG 1%,C2 and ARG 3%,C2 mortars, this reduction was approximately   33% and 39%, respectively, in relation to the reference mortar. Overall, the use of 3% nS to replace the cement weight can reduce the void size by approximately 38%, which can contribute to an increase in the durability and resistance.
In the hardened state, the bulk densities also increased for both cements with 3% nS, approximately 17% and 10% for mortars with cement 1 and 2, respectively, as shown in figure 9.
These results are also related to the formation of additional C-S-H in the mortars with 3% nS, as verified by the TG-DTA and XRD curves. Overall, the fresh and bulk densities increased for both cements, while the content of the incorporated air decreased because the nanosilica particles filled the voids in the mixture and became denser.
These physical properties can also influence the capillary absorption and mechanical strength, as presented below, which are related to the durability of the material.
In general, denser and less porous cementitious materials have greater durability and are being developed for applications in civil construction because of its application to aggressive environments.
Understanding these physical characteristics and achieving better results can support the development of more durable cementitious composites.

Capillary absorption and coefficient
The average absorption of the three specimens for each composition, measured at 10 and 90 min according to the reference standard, is shown in figure 10.
At 10 min, there was no change between the absorptions with different contents of nanosilica. However, a reduction for both cements was verified at 90 min − 36% and 40% for mortars using cement 1 and 2, respectively. Moreover, mortars with the addition of blast-furnace slag present a lower capillary absorption than cement with the addition of mineral filler, which suggests that this cement makes capillarity difficult owing to better filling of the voids.
These results are consistent with the results for the fresh density, bulk density, and incorporated air. Higher densities imply fewer pores, and thus influence capillary absorption.  Consequently, the average capillarity coefficient (C) also decreased by approximately 41% with an increase in the nanosilica content for mortars using type 1 cement, as shown in figure 11. For mortars using cement 2, the reduction was approximately 83%, with lower absolute values.
Furthermore, mortars using cement CP II E 40 (cement 2) presented lower absorption and coefficient values than mortars with cement 1 (CP II F 40), indicating a lower water permeability in cement with blast-furnace slag.
When these results were compared with the densities and incorporated-air content, it was verified that cement 2 (CPII E 40) presented higher fresh and bulk densities, as well as a lower content of incorporated air, influencing the voids used for water percolation. Therefore, the cement exhibited lower absorption and capillarity results.
These results also indicate that an increase up to 3% in the nS content can reduce the capillarity in mortars, as well as the use of cement with blast-furnace slag.
These results enabled the identification of the type of cement and nS content that contribute to greater durability, considering that some of the defect in civil engineering construction are caused by the constant presence of water. In addition, it can contribute to the reuse of construction waste in cementitious composites, thereby increasing the durability of these materials in applications.

Mechanical strength
The average results of the flexural tensile strength of the three specimens for each mixture are shown in the figure 12, which were taken at 7 and 28 d. In this test, the absolute maximum standard deviation for these results was evaluated according to [35] and resulted in a value of 0.3 MPa.
The results showed an increase in strength with the inclusion of nanosilica in the mixture, primarily at 7 d. Mortars with 3% nS achieved strengths approximately 88% higher when cement 1 was used and 112.5% higher when cement 2 was used. Therefore, there is a possibility of satisfactory strength gain with an increase in the nanomaterial content in the mixture at 28 d.
When the cement results were compared, binder CP II F 40 achieved better results, with values between 32% and 50% higher.  For the compressive strength, the average results of six specimens per mixture also showed an increase in strength with the inclusion of nanosilica, mainly at 7 d, as shown in figure 13.
In this test, the maximum absolute standard deviation (sd) per composition was also analyzed according to [35], which was 0.5 MPa. When this sd was exceeded, the discrepant individual result in relation to the average of the six values was removed, and a new average and standard deviation were obtained with the remaining five individual results.
The mortars with cement 1 and 3% nS reached higher strengths of approximately 158% in relation to the reference mortar, whereas the mortars with cement 2 and 3% nS achieved a higher strength of approximately 168%. Therefore, the mechanical strength at 28 d was approximately 2.5 times greater with 3% nS in the mixture, which is a very satisfactory result.
When the results achieved by the cements were compared, binder CP II F 40 also achieved better results, with values ranging from 30%-39% higher than that with binder CP II E 40. These results may be associated with the compressive strength of the cement with the filler itself, which is higher at 28 d, according to the manufacturer.
In view of the results, a 1% mass replacement of cement by nS contributed to overall improvement. However, the 3% nS content resulted in significant changes in the mechanical strength of the material, and the minimum content requires for further experiments.
These experimental tests could enable the use of lightweight aggregates in cementitious composites because these aggregates reduce mechanical strength. Nanomaterials can compensate for this decrease by filling voids, which consequently increases strength, thus becoming viable lightweight aggregates. In addition, strength gain could imply greater use of waste in the construction industry, which could reduce the consumption of natural resources [37].
Thus, these contributions to mechanical strength also indicate that 3% nS could be used in further studies with lightweight aggregates in cementitious composites, considering that the decreased strength caused by these aggregates can be compensated by using nanomaterials. Therefore, they contribute to environmental and developmental sustainability.

Conclusion
In this study, two nS contents (1% and 3%) in mortars were analyzed in relation to a reference mortar (0% nS) with two types of cement.
The main conclusions are as follows: • The granular packing efficiency owing to the addition of nS increased the compressive and flexural tensile strengths.
• Moreover, this increase was significant after the addition of 3% nS. However, 1% nS did not significantly change the mortar strength.
• XRD and TG-DTA were used to identify the formation of additional C-S-H gels, which contributed to an increase in strength.
• The addition of nS increased both fresh and bulk mortar densities. • Instead of capillary absorption, the capillary coefficient decreased as the nS content increased. This indicated that the voids were filled, hampering water migration.
The results confirmed that the density increased, permeability decreased, and the mechanical strength improved, owing to the addition of nanomaterials to cementitious materials. Therefore, compensating for the reduction in strength owing to the use of lightweight aggregates or construction waste in construction materials is an interesting and viable solution. Thus, greater efficiency can be provided to the civil construction industry through the reuse of these lightweight materials using nanosilica.