Thermal and Mechanical Properties of Cement Mortar Composite Containing Recycled Expanded Glass Aggregate and Nano Titanium Dioxide

: One of the growing concerns in the construction industry is energy consumption and energy e ﬃ ciency in residential buildings. Moreover, management of non-degradable solid glass wastes is becoming a critical issue worldwide. Accordingly, incorporation of recycled expanded glass aggregates (EGA) as a substitution for natural ﬁne aggregate in cement composites would be a sustainable solution in terms of energy consumption in the buildings and waste management. This experimental research aims to investigate the e ﬀ ects of EGA on fresh and hardened properties and thermal insulating performance of cement mortar. To enhance the mechanical properties and water resistance of the EGA-mortar, nano titanium dioxide (nTiO 2 ) was used as nanoﬁllers. The results showed an increase in workability and water absorption of the EGA-mortar. In addition, a signiﬁcant decrease in bulk density and compressive strength observed by incorporating EGA into the cement mortar. The EGA-mortar exhibited a low heat transfer rate and excellent thermal insulation property. Furthermore, inclusion of nTiO 2 increased compressive strength and water resistance of EGA-mortar, however, their heat transfer rate was increased. The results demonstrated that EGA-mortar can be integrated into the building envelop or non-load bearing elements such as wall partition as a thermal resistance to reduce the energy consumption in residential buildings.


Introduction
In the last few decades, demand for energy consumption in the residential building has risen and there is a high intention for reducing energy consumption in the buildings. The energy efficiency of the buildings has become increasingly critical with the rising costs of energy as well as increasing awareness on global warming effects [1,2]. Furthermore, waste management has become a critical issue. In fact, non-degradable wastes such as glass are unable to break down naturally which is developing environmental problems [3]. In Australia, about 1.1 Mt of glass waste was generated in the year of 2016-2017 from that 43% was stockpiled. In New South Wales, companies accepting the landfill levy to dump their glass waste in landfill or arranging to relocate the waste to other states where the landfill levy does not apply. Thus, use of waste glass in the large scale is a sustainable solution in terms of reduction of carbon footprint and saving the costs and energy. The construction industry is a potential sector for utilization of waste glass. In this regard, use of solid wastes for manufacturing the building materials with high thermal insulation properties is an effective approach toward sustainable Appl. Sci. 2020, 10, 2246 3 of 14 increment of compressive strength up to 40% [34,36]. Moreover, TiO 2 accelerates the formation of C-S-H gel by increasing the amount of crystalline Ca(OH) 2 at the early age of hydration [32,37]. Ma et al. [32] reported 37% and 44% increase in tensile and flexural strength respectively for the samples containing TiO 2 . In addition, the results indicated that the addition of TiO 2 could significantly refine the pores and shift them to the harmless pores. In the research conducted by Behfarnia et al. [31], it was observed that TiO 2 nanoparticles decreased the permeability of the cement matrix. In the research conducted by Khushwaha et al. [34] and Sorathiya et al. [33], the effect of various proportion of TiO 2 was studied. It was concluded that addition of TiO 2 up to 1% could significantly enhance the mechanical properties of concrete.
This research aims to develop a cement mortar with a lower heat transfer rate and insulating properties using substitution of a natural aggregate (NA) with EGA. In this study, the effect of incorporation of EGA and TiO 2 nanoparticles on workability, bulk density, water penetration, compressive strength, and heat transfer rate of the cement mortar were investigated. Infrared thermography (IRT) was used to measure the thermal insulation property of EGA cement mortar. The IRT technique has been utilized to evaluate the thermal energy storage performance of building materials in previous studies [38], however it has not been used for measuring the thermal insulating property of the EGA cement mortar. The conducted research is an additional step toward development of insulating building material and sustainable application of EGA in the construction industry.

Materials
The materials used in the study to fabricate cement mortar composite were ordinary Portland cement (OPC), natural aggregate (NA), recycled expanded glass aggregate (EGA), superplasticizer (SP), and nano titanium dioxide (nTiO 2 ). OPC from Boral Australia Co. and in accordance with AS3972 was used as a binder and Sikament NN was used as a superplasticizer (SP) in the mix, which meets all requirements as per AS1478.1 for high range water reducing admixture. EGA with a particle size of 0.25-4 mm from EGT Co. is shown in Figure 1. The specifications of EGA are compliant with EN and DIN standards. Figure 2 shows a SEM image of the utilized EGA in this study. Table 1 demonstrates the physical, mechanical, and thermal properties of the EGA.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 3 of 14 S-H gel by increasing the amount of crystalline Ca(OH)2 at the early age of hydration [32,37]. Ma et al. [32] reported 37% and 44% increase in tensile and flexural strength respectively for the samples containing TiO2. In addition, the results indicated that the addition of TiO2 could significantly refine the pores and shift them to the harmless pores. In the research conducted by Behfarnia et al. [31], it was observed that TiO2 nanoparticles decreased the permeability of the cement matrix. In the research conducted by Khushwaha et al. [34] and Sorathiya et al. [33], the effect of various proportion of TiO2 was studied. It was concluded that addition of TiO2 up to 1% could significantly enhance the mechanical properties of concrete. This research aims to develop a cement mortar with a lower heat transfer rate and insulating properties using substitution of a natural aggregate (NA) with EGA. In this study, the effect of incorporation of EGA and TiO2 nanoparticles on workability, bulk density, water penetration, compressive strength, and heat transfer rate of the cement mortar were investigated. Infrared thermography (IRT) was used to measure the thermal insulation property of EGA cement mortar. The IRT technique has been utilized to evaluate the thermal energy storage performance of building materials in previous studies [38], however it has not been used for measuring the thermal insulating property of the EGA cement mortar. The conducted research is an additional step toward development of insulating building material and sustainable application of EGA in the construction industry.

Materials
The materials used in the study to fabricate cement mortar composite were ordinary Portland cement (OPC), natural aggregate (NA), recycled expanded glass aggregate (EGA), superplasticizer (SP), and nano titanium dioxide (nTiO2). OPC from Boral Australia Co. and in accordance with AS3972 was used as a binder and Sikament NN was used as a superplasticizer (SP) in the mix, which meets all requirements as per AS1478.1 for high range water reducing admixture. EGA with a particle size of 0.25-4 mm from EGT Co. is shown in Figure 1. The specifications of EGA are compliant with EN and DIN standards. Figure 2 shows a SEM image of the utilized EGA in this study. Table 1 demonstrates the physical, mechanical, and thermal properties of the EGA.      Crushed gravel with the maximum size of 4.0 mm and density of 2800 kg/m 3 was used as NA. The NA was subjected to the particle size distribution test to precisely replicate the distribution of NA for the replacement of EGA by volume in the cement mortar. The size distribution testing was completed in accordance with AS1012 and the results are found in Figure 3.  Crushed gravel with the maximum size of 4.0 mm and density of 2800 kg/m 3 was used as NA. The NA was subjected to the particle size distribution test to precisely replicate the distribution of NA for the replacement of EGA by volume in the cement mortar. The size distribution testing was completed in accordance with AS1012 and the results are found in Figure 3. Moreover, the mercury intrusion porosimetry (MIP) test was undertaken to measure the porosity as well as pore size distribution of the EGA. The MIP test results of the EGA are revealed in Figure 4. Moreover, the mercury intrusion porosimetry (MIP) test was undertaken to measure the porosity as well as pore size distribution of the EGA. The MIP test results of the EGA are revealed in Figure 4.
Nanoparticles titanium dioxide (nTiO 2 ) purchased in the powder form from US Research Nanomaterials, Inc. Table 2 demonstrates the properties of the nTiO 2 as indicated by the manufacturer.  Nanoparticles titanium dioxide (nTiO2) purchased in the powder form from US Research Nanomaterials, Inc. Table 2 demonstrates the properties of the nTiO2 as indicated by the manufacturer.

Sample Preparation
The mixes had a water to cement ratio of 0.4 and a sand to cement ratio of 3:1. Two set of mixes were prepared: the first set of mixes were fabricated by partial and full replacement of NA with EGA without inclusion of nTiO2. The designed mixes with 0%, 50%, and 100% replacement percentage of EGA implied with CS, E50, and E100, respectively. The second set of mixes was fabricated by partial and full replacement of NA with EGA and incorporation of 1% nTiO2. The designed mixes with incorporation of TiO2 and the EGA replacement percentage of 0%, 50%, and 100% defined as CT, E50T, and E100T, respectively.
To fabricate the mixes, the dry materials (cement and NA/EGA) were placed in the mixer and mixed on the low speed for 2.0 min. In the case of CT, E50T, and E100T mixes, the nTiO2 were sonicated for 15 min in the solution of water and superplasticizer (SP) [39]. Then the dispersed nTiO2/SP/water solution was added slowly to the mix and the materials were mixed for another 5 min. The mixes cast in 70 × 70 × 70 mm 3 cubes and demolded after 24 h. The samples were cured in the fog room at a constant temperature of 23 °C and in accordance with AS1012.8. Table 3 demonstrates the mix proportion of the samples. The abbreviations for labeling each mix are defined in a way that the letters C and E representing control sample and mortar sample containing EGA respectively and number after the letters presents the percentage of NA replacement with EGA into the mixture. The letter T demonstrates the presence of TiO2 in the mix. For instance, the E50T mixture represents the sample that contains 50% EGA and TiO2.

Sample Preparation
The mixes had a water to cement ratio of 0.4 and a sand to cement ratio of 3:1. Two set of mixes were prepared: the first set of mixes were fabricated by partial and full replacement of NA with EGA without inclusion of nTiO 2 . The designed mixes with 0%, 50%, and 100% replacement percentage of EGA implied with CS, E50, and E100, respectively. The second set of mixes was fabricated by partial and full replacement of NA with EGA and incorporation of 1% nTiO 2 . The designed mixes with incorporation of TiO 2 and the EGA replacement percentage of 0%, 50%, and 100% defined as CT, E50T, and E100T, respectively.
To fabricate the mixes, the dry materials (cement and NA/EGA) were placed in the mixer and mixed on the low speed for 2.0 min. In the case of CT, E50T, and E100T mixes, the nTiO 2 were sonicated for 15 min in the solution of water and superplasticizer (SP) [39]. Then the dispersed nTiO 2 /SP/water solution was added slowly to the mix and the materials were mixed for another 5 min. The mixes cast in 70 × 70 × 70 mm 3 cubes and demolded after 24 h. The samples were cured in the fog room at a constant temperature of 23 • C and in accordance with AS1012.8. Table 3 demonstrates the mix proportion of the samples. The abbreviations for labeling each mix are defined in a way that the letters C and E representing control sample and mortar sample containing EGA respectively and number after the letters presents the percentage of NA replacement with EGA into the mixture. The letter T demonstrates the presence of TiO 2 in the mix. For instance, the E50T mixture represents the sample that contains 50% EGA and TiO 2 .

Experimental Tests
The flow table test was undertaken on the fresh cement mortar samples in accordance to the AS2701 to measure the mixtures workability and consistency. Moreover, the density of the mixture was determined via the density test according to AS2701. To measure the water penetration of the specimens, the water absorption test was conducted in accordance with AS1012.21 at the age of 28 days. The compressive test was undertaken on the cube specimens with the size of 70 mm × 70 mm × 70 mm and in accordance with AS1012.9 at the age of 7, 14, and 28 days. For each test, three samples were tested and the average including the error bar were reported.
In this study, the thermal insulation property and heat transfer rate of cement mortar containing EGA was evaluated by measuring the surface temperature distribution using infrared thermal imaging camera. For this purpose, the specimens with the dimension of 70 mm × 70 mm × 30 mm were prepared and kept at about 27 • C for a few hours to allow all samples to achieve the same initial temperature. Then the samples were exposed to a heat source and the surface temperature distribution of the other side was captured by an infrared thermal camera for 15 min (Testo 872, Testo Australia). The thermal test was repeated for three times for each sample. Figure 5 illustrates a schematic diagram of the thermal test.

Experimental Tests
The flow table test was undertaken on the fresh cement mortar samples in accordance to the AS2701 to measure the mixtures workability and consistency. Moreover, the density of the mixture was determined via the density test according to AS2701. To measure the water penetration of the specimens, the water absorption test was conducted in accordance with AS1012.21 at the age of 28 days. The compressive test was undertaken on the cube specimens with the size of 70 mm × 70 mm × 70 mm and in accordance with AS1012.9 at the age of 7, 14, and 28 days. For each test, three samples were tested and the average including the error bar were reported.
In this study, the thermal insulation property and heat transfer rate of cement mortar containing EGA was evaluated by measuring the surface temperature distribution using infrared thermal imaging camera. For this purpose, the specimens with the dimension of 70 mm × 70 mm × 30 mm were prepared and kept at about 27 °C for a few hours to allow all samples to achieve the same initial temperature. Then the samples were exposed to a heat source and the surface temperature distribution of the other side was captured by an infrared thermal camera for 15 min (Testo 872, Testo Australia). The thermal test was repeated for three times for each sample. Figure 5 illustrates a schematic diagram of the thermal test.  Figure 6 shows the flow table test and the flow table test results are presented in Table 4. The flow table test values were determined by averaging the diameters of each mixes test. All mixes showed flow values in the range of 140-215 mm, without segregation or bleeding. The results revealed that addition of EGA increased the workability of cement mortar up to 26.6% and 41.25% for the E50 and E100 mixes respectively compared to the control mix (CS). The increment trend in workability despite the decreasing on the amount of superplasticizer in E50 and E100 mixes is contributed to the smooth surface and spherical shape of EGA [40][41][42]. Adding to this, the increase  Figure 6 shows the flow table test and the flow table test results are presented in Table 4. The flow table test values were determined by averaging the diameters of each mixes test. All mixes showed flow values in the range of 140-215 mm, without segregation or bleeding. The results revealed that addition of EGA increased the workability of cement mortar up to 26.6% and 41.25% for the E50 and E100 mixes respectively compared to the control mix (CS). The increment trend in workability despite the decreasing on the amount of superplasticizer in E50 and E100 mixes is contributed to the smooth surface and spherical shape of EGA [40][41][42]. Adding to this, the increase in the flow values can be due to the increase in the amount of entrapped air voids. Furthermore, the workability of CT, E50T, and E100T mixes increased by 5.35%, 30.4%, and 53.7% respectively in comparison to the CS, E50, and E100 mixes respectively, which is attributed to the induction of the microbubble in the water solution during the sonication process and consequently increased in small air voids in the mixes. in the flow values can be due to the increase in the amount of entrapped air voids. Furthermore, the workability of CT, E50T, and E100T mixes increased by 5.35%, 30.4%, and 53.7% respectively in comparison to the CS, E50, and E100 mixes respectively, which is attributed to the induction of the microbubble in the water solution during the sonication process and consequently increased in small air voids in the mixes.

Density
The density of the samples was measured, and the results are demonstrated in Figure 7. The measurement revealed the density of 2354, 1769, and 987 Kg/m 3 for C.S, E50, and E100 respectively. It shows the density of E50 and E100 decreased 30% and 65% respectively in comparison to the CS, which is attributed to the very low density of EGA and its porous structure. In addition, the densities of CT, E50T, and E100T were 2%, 3%, and 6% higher than CS, E50, and E100 respectively. It can be concluded that the increase in density was attributed to the lower porosity in the cement matrix due to the incorporation of nTiO2. It is noteworthy that E100 with density of 987 kg/m 3 was classified as a lightweight mortar that can be used for production of lightweight concrete. Figure 8 illustrates the cross section of CS and E100 samples.

Density
The density of the samples was measured, and the results are demonstrated in Figure 7. The measurement revealed the density of 2354, 1769, and 987 kg/m 3 for C.S, E50, and E100 respectively. It shows the density of E50 and E100 decreased 30% and 65% respectively in comparison to the CS, which is attributed to the very low density of EGA and its porous structure. In addition, the densities of CT, E50T, and E100T were 2%, 3%, and 6% higher than CS, E50, and E100 respectively. It can be concluded that the increase in density was attributed to the lower porosity in the cement matrix due to the incorporation of nTiO 2 . It is noteworthy that E100 with density of 987 kg/m 3 was classified as a lightweight mortar that can be used for production of lightweight concrete. Figure 8 illustrates the cross section of CS and E100 samples.

Water Absorption
The water absorption test was completed on all mixes and the results are shown in Figure 7. The water absorption of 4.19% obtained for control sample (CS) however a higher water absorption rate was obtained for the mixes containing EGA. The water absorption of E50 and E100 mixes were 7.47% and 14.74% respectively, which shows a 78% and 252% increase in the permeability of the matrix, compared to the control sample. The increase in water absorption is due to the high porosity of EGA in comparison to NA. The results revealed that the water penetration increased by increasing the EGA content. The water absorption rate obtained for the E100 (values of approximately 14%) was higher than the acceptable range (<10%) [43,44]. The addition of nTiO 2 reduced the water absorption value by 28%, 17%, and 2% for samples containing 0%, 50%, and 100% EGA respectively. The decrease in water absorption upon the inclusion of nTiO 2 coincides with previous studies [45] and aligns with the density results. The reduction in water absorption was attributed to the filling effect of nTiO 2 and reducing the porosity of the cement matrix. It is worthy to note that the sonication process resulted in tiny bubbles of air uniformly distributed in the mortar. These small bubbles are like entraining air that improves the workability of the mixes. Indeed, nTiO 2 acted as nanofillers in mortar and improved the resistance to water penetration of the cement composite [46].

Water Absorption
The water absorption test was completed on all mixes and the results are shown in Figure 7. The water absorption of 4.19% obtained for control sample (CS) however a higher water absorption rate was obtained for the mixes containing EGA. The water absorption of E50 and E100 mixes were 7.47% and 14.74% respectively, which shows a 78% and 252% increase in the permeability of the matrix, compared to the control sample. The increase in water absorption is due to the high porosity of EGA in comparison to NA. The results revealed that the water penetration increased by increasing the EGA content. The water absorption rate obtained for the E100 (values of approximately 14%) was higher than the acceptable range (<10%) [43,44]. The addition of nTiO2 reduced the water absorption value by 28%, 17%, and 2% for samples containing 0%, 50%, and 100% EGA respectively. The decrease in water absorption upon the inclusion of nTiO2 coincides with previous studies [45] and aligns with the density results. The reduction in water absorption was attributed to the filling effect of nTiO2 and reducing the porosity of the cement matrix. It is worthy to note that the sonication process resulted in tiny bubbles of air uniformly distributed in the mortar. These small bubbles are like entraining air that improves the workability of the mixes. Indeed, nTiO2 acted as nanofillers in mortar and improved the resistance to water penetration of the cement composite [46].

Compressive Strength
The experimental test for compressive strength was carried out at different curing ages of 7, 14, and 28 days. Figure 9 shows the impact of EGA and TiO 2 inclusion on the compressive strength of the mortar composites at different ages. It is observed that the inclusion of the EGA significantly decreased the compressive strength of cement mortar. The results of 28-day compressive strength demonstrated that 50% replacement of NA with EGA reduced the strength about 65.8% in compare to the control sample (C.S). In addition, it was observed that as the EGA content increased from 50% to 100%, the compressive strength dropped dramatically from 26.25 to 8.20 MPa at the age of 28 days. It is noteworthy that the compressive strength was still in the acceptable range and similar or higher than reported results in the literature [21,25]. Namsone et al. [25] reported the 28-day compressive of 5.7 MPa for a foamed matrix using EGA and obtained the compressive strengths of 6.68-12.49 MPa for the EGA cement mortar. Indeed, the samples containing 100% EGA without nTiO 2 had the lowest compressive strength out of all the mixes. Furthermore, the results indicated a normal increasing trend for the compressive strength for CS, E50, and E100 mixes as the curing process progresses. However, the mixes containing nTiO 2 revealed a relatively different strength development tend. It was revealed that CT, E50T, and E100T mixes reached to 84.6%, 87.2%, and 77.2% of maximum strength within 7 days of curing while for samples without nTiO 2 (CS, E50, and E100 mixes) it happened at 14 days of curing. This behavior was attributed to the addition of nTiO 2 into the cementitious materials, which resulted in an accelerated rate of hydration process. A similar attribute has been reported in previous studies that when nTiO 2 is uniformly distributed throughout the matrix, the hydration process and formation of C-S-H gel is accelerated, which results in early strength [32,47,48]. In the other set of mixes, the effect of nTiO 2 inclusion on the compressive strength of mixes was investigated after a different curing time. The compressive strength results of E50T and E100T mixes at 28 days showed the similar trend. It was observed that the addition of EGA significantly decreased the compressive strength and the strength significantly dropped as the EGA content increased however, inclusion of nTiO 2 compensated some part of the compressive strength. The average compressive strength at 28 days of CT, E50T, and E100T mixes were 76.72, 29.70, and 11.4 MPa respectively, which shows 1.7%, 13.1%, and 39.0% enhancement in comparison to CS, E50, and E100 mixes respectively. It can be concluded that nTiO 2 acts as nanofillers in specimens and recovers their pore structure by decreasing voids and pores in the composite matrix [46]. and 28 days. Figure 9 shows the impact of EGA and TiO2 inclusion on the compressive strength of the mortar composites at different ages. It is observed that the inclusion of the EGA significantly decreased the compressive strength of cement mortar. The results of 28-day compressive strength demonstrated that 50% replacement of NA with EGA reduced the strength about 65.8% in compare to the control sample (C.S). In addition, it was observed that as the EGA content increased from 50% to 100%, the compressive strength dropped dramatically from 26.25 to 8.20 MPa at the age of 28 days. It is noteworthy that the compressive strength was still in the acceptable range and similar or higher than reported results in the literature [21,25]. Namsone et al. [25] reported the 28-day compressive of 5.7 MPa for a foamed matrix using EGA and obtained the compressive strengths of 6.68-12.49 MPa for the EGA cement mortar. Indeed, the samples containing 100% EGA without nTiO2 had the lowest compressive strength out of all the mixes. Furthermore, the results indicated a normal increasing trend for the compressive strength for CS, E50, and E100 mixes as the curing process progresses. However, the mixes containing nTiO2 revealed a relatively different strength development tend. It was revealed that CT, E50T, and E100T mixes reached to 84.6%, 87.2%, and 77.2% of maximum strength within 7 days of curing while for samples without nTiO2 (CS, E50, and E100 mixes) it happened at 14 days of curing. This behavior was attributed to the addition of nTiO2 into the cementitious materials, which resulted in an accelerated rate of hydration process. A similar attribute has been reported in previous studies that when nTiO2 is uniformly distributed throughout the matrix, the hydration process and formation of C-S-H gel is accelerated, which results in early strength [32,[47][48]. In the other set of mixes, the effect of nTiO2 inclusion on the compressive strength of mixes was investigated after a different curing time. The compressive strength results of E50T and E100T mixes at 28 days showed the similar trend. It was observed that the addition of EGA significantly decreased the compressive strength and the strength significantly dropped as the EGA content increased however, inclusion of nTiO2 compensated some part of the compressive strength. The average compressive strength at 28 days of CT, E50T, and E100T mixes were 76.72, 29.70, and 11.4 MPa respectively, which shows 1.7%, 13.1%, and 39.0% In summary it can be concluded that the compressive strength and water absorption of concrete are highly influenced by the density of the mix. The results revealed an interrelationship between density and compressive strength. It was observed that the compressive strength dropped by decrement of the sample's density (CS, E50, and E100 mixes). Similarly, an increase in density for CT, E50T, and E100T mixes resulted in an increase in compressive strength compared to CS, E50, and E100 mixes respectively. Moreover, the results demonstrated an inverse relation between density and water absorption. It was found that water absorption increased by decreasing the density of the mixes in case of CS, E50, and E100. However, the water absorption decreased by integration of TiO2 into the mixes (CT, E50T, and E100T) and increment of density due to a lower porosity of the matrix.

Infrared Thermography
In order to evaluate the thermal insulating properties of the cement composites, the infrared thermography (IRT) experiment was carried out on all the samples: CS, E50, E100, CST, E50T, and E100T. Figure 10 illustrates the thermal images of surface temperature distribution of the samples captured by the IRT camera at different heating times. According to the relationship between the color and temperature value, it can be suggested that the heat-transferring rate and thermal conductivity of cement composites were significantly decreased with the inclusion of EGA. The thermal images clearly demonstrate a different temperature distribution in the control sample (CS mix) and the samples containing EGA (E50 and E100 mixes). The results show that the temperature increased rapidly in the CS however, a noticeable slower heat transfer rate observed for samples incorporated with EGA (E50 and E100). The data also revealed a drop in the heat transfer rate as the EGA content increased. For instance, after 15 min the average surface temperature in the CS sample reached 55 • C while the average surface temperature in the E50 and E100 samples reached 52.7 and 48.7 • C respectively, which shows a temperature difference of 2.3 • C and 6.0 • C for E50 and E100 respectively. Moreover, the results demonstrated the heat transfer rate of 1.75, 1.60, and 1.35 • C/min for CS, E50, and E100 respectively that shows a lower rate for samples containing EGA (E50 and E100) in comparison to the control sample (CS). This observation was attributed to a high porosity and low thermal conductivity of EGA. Indeed, by incorporation of EGA the air void is replaced with sand, which has a high thermal conductivity. EGA has a thermal conductivity of 0.07 W/mK that is much less than that of sand (Expanded Glass Technologies). Consequently, by replacing the NA with EGA the heat transfer of the cement composite was reduced. mix) and the samples containing EGA (E50 and E100 mixes). The results show that the temperature increased rapidly in the CS however, a noticeable slower heat transfer rate observed for samples incorporated with EGA (E50 and E100). The data also revealed a drop in the heat transfer rate as the EGA content increased. For instance, after 15 min the average surface temperature in the CS sample reached 55 °C while the average surface temperature in the E50 and E100 samples reached 52.7 and 48.7 °C respectively, which shows a temperature difference of 2.3 °C and 6.0 °C for E50 and E100 respectively. Moreover, the results demonstrated the heat transfer rate of 1.75, 1.60, and 1.35 °C/min for CS, E50, and E100 respectively that shows a lower rate for samples containing EGA (E50 and E100) in comparison to the control sample (CS). This observation was attributed to a high porosity and low thermal conductivity of EGA. Indeed, by incorporation of EGA the air void is replaced with sand, which has a high thermal conductivity. EGA has a thermal conductivity of 0.07 W/mK that is much less than that of sand (Expanded Glass Technologies). Consequently, by replacing the NA with EGA the heat transfer of the cement composite was reduced.  Figure 10. Infrared thermography images of different samples. Table 5 demonstrates the average temperature differences for each mix. The thermal charging results for the samples inclusion nTiO2 showed a different trend to the first set of mixes (mixes without nTiO2). It was observed that incorporation nTiO2 increased the heat transfer rate, which is undesired in terms of thermal insulation properties. The thermal images demonstrated that inclusion of nTiO2 into the composite increased the heat transfer rate compared to the samples without nTiO2.  Table 5 demonstrates the average temperature differences for each mix. The thermal charging results for the samples inclusion nTiO 2 showed a different trend to the first set of mixes (mixes without nTiO 2 ). It was observed that incorporation nTiO 2 increased the heat transfer rate, which is undesired in terms of thermal insulation properties. The thermal images demonstrated that inclusion of nTiO 2 into the composite increased the heat transfer rate compared to the samples without nTiO 2 . For example, after 15 min the average surface temperature in CS, E50, and E100 samples reached to 55 • C, 52.7 • C, and 48.7 • C respectively. While average surface temperature in the CT, E50T, and E100T samples reached to 56.8 • C, 55 • C, and 49.1 • C respectively that shows an increase in the temperature difference of 1.6 • C, 2.3 • C, and 1.4 • C respectively. Furthermore, the results demonstrated the heat transfer rate of 1.87, 1.76, and 1.38 • C/min for CT, E50T, and E100T respectively that indicates higher rate than the samples without nTiO 2 . It can be concluded that nTiO 2 acts as a filler and changes the pore structures of the cement composite and consequently the thermal charging performance of the matrix. Therefore, in terms of thermal properties, NA substitution with EGA improves the thermal insulation properties of cement composites. This positive effect is attributed to lower thermally conductive and higher porosity of EGA compared to NA.

Conclusions
This experimental work investigated the physical properties as well as the thermal insulation property of cement mortar containing EGA and TiO 2 . The findings revealed that incorporating EGA into the mortar composite causing a significant decrease in density and compressive strength, which was attributed to the porous nature and low compressive strength of EGA. The results also demonstrated that as the EGA content increased, the workability and water absorption of cement composite increased. It is found that the increase in water absorption was due to the high porosity of EGA in comparison to NA. However, the beneficial effect of the EGA was the decrease in the heat-transferring rate of the cement composite, which indicates the feasibility of a potential reduction in energy consumption in buildings. Moreover, the results demonstrated that inclusion of TiO 2 into the cement composite partially compensated the water absorption and loss in compressive strength. However, it was revealed that addition of nTiO 2 into EGA-cement composites increased the heat transfer rate of the cement matrix and insulation properties as nTiO 2 acts as nanofillers and changes the pores structure in the cement matrix. It can be concluded that in terms of thermal behavior, substitution of NA with EGA decreases the heat transfer rate and consequently improves the thermal insulation properties of the cement mortar.