Exploring the impact of graphene oxide on mechanical and durability properties of mortars incorporating demolition waste: micro and nano-pore structure effects

2.1. Materials

For the production of the different compositions explored in this study, several components were used: Portland cement (CEM II/B-L 32.5 N) conforming to EN 197-1 (3131. EN 197-1. (2011) Cement - Part 1: Composition, specifications and conformity criteria for common cements, European Committee for Standardization.
); standard sand (i.e., natural river sand) conforming to EN 196-1 (3131. EN 197-1. (2011) Cement - Part 1: Composition, specifications and conformity criteria for common cements, European Committee for Standardization.
); and sand from CDW obtained from the ALCOREC plant in San Jose de la Rinconada (Seville, Spain). Both sands were previously sieved to obtain a maximum particle size of 1.25 mm.

Figure 1 shows the comparison between standard sand and recycled sand as received (without previous sieve). At first sight, it can be seen that demolition waste has larger particles than standard sand. The colour is different, with recycled sand having a cementitious colour and standard sand having a more intense yellowish colour.

2.2. Preparation of mortars

The dosage of each component was conducted by weighing them according to the established ratio shown on Table 2.

The water/cement ratio is higher in mortars with recycled aggregates, due to their low specific gravity. CDW absorbs water during the mixing process, and a high water/cement ratio is necessary to obtain a homogeneous mix (3232. Leiva, C.; Solís-Guzmán, J.; Marrero, M.; García-Arenas, C. (2013) Recycled blocks with improved sound and fire insulation containing construction and demolition waste. Waste Manag. 33 [3], 663-671. https://doi.org/10.1016/j.wasman.2012.06.011.
).

Cement and sand (standard or recycled) were weighed and mixed for 30 seconds in a laboratory mixer. Next, water was added and mixed with the solids for 2 minutes, until a homogeneous paste was obtained. When GO was added, the required GO at a concentration of 4 g/L (Graphenea) was previously stirred for 24 hours and sonicated for 15 minutes. All GO solutions had the same concentration during sonication. The sonicated solution was mixed with clean water (i.e., without GO) and then added to the solids. The dose of GO (0.0075% of solid content) was chosen according to a previous study (2626. Lv, S.; Liu, J.; Sun, T.; Ma, Y.; Zhou, Q. (2014) Effect of GO nanosheets on shapes of cement hydration crystals and their formation process. Constr. Build. Mater. 64, 231-239. https://doi.org/10.1016/J.CONBUILDMAT.2014.04.061.
).

The material was left in the mould for 24 hours so that it was hard enough for demoulding. After demoulding, the mortars were placed in water for 14 days to cure. At the end of this period, they were removed from the water and exposed to air for 13 days. After 28 days, they were ready for exploring their physical and mechanical properties.

2.3. Leaching study

Given that waste is used in mortars, a leaching test was necessary. According to European standards, construction materials should not emit hazardous substances, but, for example, in Spain, no national tests or limits are specified to evaluate the leaching properties of construction materials containing waste. This study was conducted in accordance with the EN-12457 (3333. EN 12457-4. (2003) Characterisation of waste - Leaching - Compliance test for leaching of granular waste materials and sludges. Part 4: One stage batch test at a liquid to solid ratio of 10 l/kg for materials with particle size below 10 mm (without or with size reduction), European Committee for Standardization.
) waste characterization standard. It is a conformity test for the leaching of granular waste and sludge including a one-stage batch test with a liquid-solid ratio of 10 l/kg for waste with a particle size of less than 10 mm. This test is the most common leaching test in Europe to classify wastes, and some countries, such as Italy (3434. IMD 186. (2006) Individuazione dei rifiuti non pericolosi sottoposti alle procedure semplificate di recupero ai sensi degli articoli 31 e 33 del decreto legislativo 05/02/1997. Gazzetta Ufficiale n. 115.
), use the results of this test to determine whether a residue can be used in building materials when compared to certain reference values. This test is a protocol used to accelerate the release of chemical species contained in a material in order to characterize its potential to be mobilized into the environment. The pollutants of greatest interest that can be mobilized by weathering and leaching due to rainfall are heavy metals, due to their high toxicity at low concentrations. Analysis in leachates was carried out through Inductively Coupled Plasma technique.

2.4. Physical properties

2.5. Mechanical properties

Flexural strength was determined according to ASTM C-348-02 (3737. ASTM C348. (2021) Standard test method for flexural strength of hydraulic-cement mortars, ASTM International (ASTM).
). Three different parallelepipeds 160 × 40 × 40 mm were used and tested at a loading rate of 15 mm/min. The equipment used to calculate flexural strength was the same as that used for the compression test: the Suzpecar machine, model MEM102/50 t.

Compressive strength was measured at the age of 28 days following the procedure given in EN 196-1 (3838. EN 196-1. (2018) Methods of testing cement - Part 1: Determination of strength, European Committee for Standardization.
). The two pieces of each sample after the flexural test (6 samples) were subjected to compressive tests. This standard establishes that compressive strength is determined by applying a normal force to the surface of the sample and measuring the stress applied when breakage occurs. Compressive strength tests were performed with a Suzpecar machine, model MEM102/50 t. Six samples were broken down for each type of composition. In addition, speed was controlled by the displacement of the top face, which was tensioned at a rate of 0.5 mm/min.

2.5.1. Acid attack test

 ⌅

Resistance to sulfuric acid attack was measured using six samples after they had been cured for 28 days. Three were exposed to air and three were immersed in 1 molar sulfuric acid for 14 days. The acid volume used was 3 times the volume of the samples, all the surfaces were exposed to the acid attack as it can be seen in Figure 2. They were hanging and during the 15 days the total volume of acid was constant, since acid was renewed as water evaporated, keeping the acid volume constant. Samples were removed from their containers and their compressive strengths were measured after immersion (3939. Cerulli, T.; Pistolesi, C.; Maltese, C.; Salvioni, D. (2003) Durability of traditional plasters with respect to blast furnace slag-based plaster. Cem. Concr. Res. 33 [9], 1375-1383. https://doi.org/10.1016/S0008-8846(03)00072-3.
).

Figure 2.  Samples during de acid test.

Results are expressed as the ratio between the compressive strength of the mortars immersed in acid and that of the mortars exposed to air, as this calculation shows:

$\mathrm{Compressive\; strength\; variation}\left(\%\right)=\frac{C_{i-}{C}_{air}}{C_{air}}\bullet 100$  [5]

where C_i (MPa) is Compressive strength of mortars immersed in acid after 14 days and C_air (MPa) is the Compressive strength of non-immersed after 14 days.

3.1. Characterization of materials

The chemical composition using an X-ray fluorescence spectrometer of all components is shown on Table 3. SiO₂ was the component most present in recycled sand, although standard sand contained a higher proportion of it. CaO, Al₂O₃ and Fe₂O₃ were also among the main components of recycled sand. Given that recycled aggregate results from construction and demolition waste, its composition is similar to that of cement (without considering SiO₂).

The specific gravity of recycled sand is lower than that of standard sand. The specific gravity of recycled aggregate is lower than that of natural aggregates due to the mixture of materials (e.g., Portland cement, concrete, gypsum, bricks) that form CDW (4040. Arenas, C.; Luna-Galiano, Y.; Leiva, C.; Vilches, L.F.; Arroyo, F.; Villegas, R.; Fernandez-Pereira, C. (2017) Development of a fly ash-based geopolymeric concrete with construction and demolition wastes as aggregates in acoustic barriers. Constr. Build. Mater. 134, 433-442. https://doi.org/10.1016/J.CONBUILDMAT.2016.12.119.
).

The particle size distribution of standard and recycled sand was measured with a particle size analyzer (Mastersizer 3000, Malvern, UK) and is presented in Figure 3. The cumulative percentage is represented by continuous lines, while dotted lines represent the total percentage.

CDW had smaller particle size than natural river sand. Cement was the component with the smallest particle size, with an particle size interval between 0 and 150 μm. Although the particle size distribution is a key factor on the properties of the mortars, in this study, the different sands have no been subjected to any previous treatment (except the previous sieving at 1.25 mm for both) in order to compare the results as they were received. According to the technical data sheet the composition of GO consists of the following elements: C (49-56%), O (41-50%), S (2-4%), H (0-1%) and N (0-1%).

Fourier-transform infrared spectroscopy (FTIR) measurements were performed on a Nicolet 380 infrared spectrometer (Thermo Electron Corporation, USA). To perform the FTIR, 1 mg paste powder samples were mixed with 100 mg KBr to produce slices. The FTIR spectrum of graphene oxide is shown in Figure 4; the curve exhibits a sizable peak at 3216 cm^-1 in the high frequency region, which is related to the stretching mode of the O-H bond and reveals the existence of hydroxyl groups in the graphene oxide. The carboxyl group was assigned to the band at 1724 cm^-1. The stretching and bending vibration of the C=C groups of the water molecules adsorbed on the graphene oxide may have caused the peak at 1610 cm^-1. The C-O-H group was responsible for the peak at 1366 cm^-1. The peak at 1046 cm^-1was the vibrational mode of the C-O group, and the peak at 1173 cm^-1 represented C-O-C stretching.

The time and power of ultrasonication of GO to achieve a good dispersion of GO were very variable in previous studies. In some cases, samples were exposed for 5 minutes (4141. Li, X.; Korayem, A.H.; Li, C.; Liu, Y.; He, H.; Sanjayan, J.G.; Duan, W.H. (2016) Incorporation of graphene oxide and silica fume into cement paste: A study of dispersion and compressive strength. Constr. Build. Mater. 123, 327-335. https://doi.org/10.1016/J.CONBUILDMAT.2016.07.022.
), while in others they were exposed for about 3 hours (4242. Horszczaruk, E.; Mijowska, E., Kalenczuk, R.J.; Aleksandrzak, M.; Mijowska, S. (2015) Nanocomposite of cement/graphene oxide - Impact on hydration kinetics and Young’s modulus. Constr. Build. Mater. 78, 234-242. https://doi.org/10.1016/J.CONBUILDMAT.2014.12.009.
). For instance, in (2121. Mohammed, A.; Sanjayan, J.G., Duan, W.H.; Nazari, A. (2015) Incorporating graphene oxide in cement composites: A study of transport properties. Constr. Build. Mater. 84, 341-347. https://doi.org/10.1016/J.CONBUILDMAT.2015.01.083.
, 2828. Li, X.; Wang, L.; Liu, Y.; Li, W.; Dong, B.; Duan, W.H. (2018) Dispersion of graphene oxide agglomerates in cement paste and its effects on electrical resistivity and flexural strength. Cem. Concr. Compos. 92, 145-154. https://doi.org/10.1016/j.cemconcomp.2018.06.008.
) GO was not even ultrasonicated. Ultrasonication breaks GO sheets, producing smaller platelets. Therefore, the size of GO sheets decreases with increasing time and power. The two types of sand used in this study had a relatively large particle size (between 0.1 and 1 mm), which was expected to mark their mechanical performance (4343. Ríos, J.D.; Leiva, C.; Ariza, M.P.; Seitl, S.; Cifuentes, H. (2019) Analysis of the tensile fracture properties of ultra-high-strength fiber-reinforced concrete with different types of steel fibers by X-ray tomography. Mater. Des. 165, 107582. https://doi.org/10.1016/j.matdes.2019.107582.
). Therefore, a moderate sonication process was carried out to adequately determine the particle size of GO and the pore size of the matrix, because smaller pore sizes of GO do not affect the larger pores of the matrix. The GO solution was previously agitated for 24 hours. GO was sonicated for 15 minutes using an ULTRASONS 3000513 device with a power of 150 W to increase the dispersion of graphene oxide nanoparticles in water.

The distribution of graphene oxide is shown in Figure 5. It was measured with a high-definition digital particle size analyzer (Saturn DigiSizer II). Different peaks can be seen in Figure 5 due to the three dimensions of GO (i.e., thickness, width, and length). The two larger peaks - at 2.5 and 4 mm - correspond to the width and length and are quite wide, which means that a high level of dispersion was not reached. The two smaller peaks (at 0.8 and 1.6 mm) correspond to particles of different thicknesses and indicate the low GO-dispersion achieved.

3.2. Leaching study

Table 4 shows the concentration of leached heavy metals according to EN-12457 (3333. EN 12457-4. (2003) Characterisation of waste - Leaching - Compliance test for leaching of granular waste materials and sludges. Part 4: One stage batch test at a liquid to solid ratio of 10 l/kg for materials with particle size below 10 mm (without or with size reduction), European Committee for Standardization.
) of standard sand and recycled sand. The results were also compared to the limits set by European landfill regulations (4444. Council Directive 1999/31/EC of 26 April (1999) On the landfill of waste. Official Journal L. 182, 16/07/1999 P. 0001 - 0019. European Commission (1999) http://data.europa.eu/eli/dir/1999/31/oj.
), which establish three categories of waste: inert, non-hazardous and hazardous. Recycled waste (and standard sand) can be considered as inert waste because both meet the limits set in the regulations.

In Italy, EN 12457-1 (3333. EN 12457-4. (2003) Characterisation of waste - Leaching - Compliance test for leaching of granular waste materials and sludges. Part 4: One stage batch test at a liquid to solid ratio of 10 l/kg for materials with particle size below 10 mm (without or with size reduction), European Committee for Standardization.
) results are used to determine whether waste can be used in construction materials. According to the limits of the Italian Ministerial Decree (3434. IMD 186. (2006) Individuazione dei rifiuti non pericolosi sottoposti alle procedure semplificate di recupero ai sensi degli articoli 31 e 33 del decreto legislativo 05/02/1997. Gazzetta Ufficiale n. 115.
), CDW could be used as they do not exceed any limits, but Standard Sand exceeds the limit set for Zn.

3.3. Physical properties

3.3.1. Density

 ⌅

Figure 6 shows the density of the different compositions with standard sand and CDW waste, with and without the addition of GO. Density is related to two parameters: particle size distribution and the specific gravity of the two types of sands. A higher particle size distribution produces higher porosity between the particles. If the specific gravity is lower, it results in higher porosity inside the particles. SS has a slightly higher particle size and a higher specific density (see Figure 3 and Table 1), so standard sand mortars have a higher density than recycled sand mortars.

Figure 6.  Density of mortars with different fine aggregates and GO addition.

Regardless of the material mortars are made of, their density decreases as the water/cement ratio rises. This is because when a high water/cement ratio is used, non-reacted water evaporates during the last period of curing and creates a higher number of pores. When CDW aggregate is used, part of the water is stored inside the CDW aggregate producing a lower density diminution.

As shown by Figure 6, the addition of GO did not affect the mortar density because the amount of GO added was very low. Nevertheless, the effect of GO on the pore size distribution of mortars was very significant (Figure 7). By adding GO, large pores decreased in size and were divided into smaller pores. Between 100 and 1000 μm, the number of large pores decreased, and a larger number of pores between 10 and 100 μm appeared. The peak clearly displaced to finer pores with the addition of GO, and pore volume decreased significantly. The same thing happened to the peak around 0.1 μm: thanks to the addition of GO, the number of pores of that size decreased.

Figure 7.  Porosimetry of mortars with CDW and the same water/cement ratio.

To analyze the influence in pores smaller than 0.1 μm, gas adsorption measurements were performed (Figure 8). As can be seen, there was an increase in the number of nanopores between 60 and 100 nanometers when GO was added. This increase was due to the previous reduction of pores by the addition of 0.1 μm GO. If GO with a low particle size had been used, the largest pores in the mortar would not have undergone any changes.

Figure 8.  Nanopores of mortars with CDW and the same water/cement ratio.

3.4. Mechanical properties

Figure 9 shows the flexural strength of mortars with different compositions after the 28-day curing period. The strength of all mortars decreased as the amount of water increased. This happened because of the growth in porosity.

There was a considerable difference in the flexural strength values obtained between the two materials. Standard sand showed better results than CDW with the same water/cement ratio. The addition of GO improved the flexural strength of recycled waste. The improvement was greater for CDW.

The mechanical properties of mortars are very dependent on their microstructure. The corresponding SEM images of the mortars were also examined to determine the link between mechanical strength and microstructure. SEM images of the microstructure of mortars containing recycled aggregate with and without GO are shown in Figure 10. When the mortar did not contain GO, many needle- and bar-like crystals emerged on the fracture surface; they were disorderly stacked cement hydration crystals of ettringite and/or gypsum (Figure 9A). it is not easy to distinguish both. During the initial stage of hydration, ettringite decomposes to produce monosulfate, and at the same time, sulfate ions can be adsorbed in calcium silicate hydrates. Then, at ambient temperature, the sulfate ions adsorbed in calcium silicate hydrates leach out and react with monosulfate to form ettringite (4545. Ando, Y.; Shinichi, H.; Katayama, T.; Torii, K. (2022) Microscopic observations of sites and forms of ettringite in the microstructure of deteriorated concrete. Mater. Construcc. 72 (346), e283. https://doi.org/10.3989/mc.2022.15521.
). Since GO is not easy to found at the presented doses. It has been postulated (46-4746. Basquiroto de Souza, F.; Shamsaei, E.; Sagoe-Crentsil, K.; Duan, W. (2022) Proposed mechanism for the enhanced microstructure of graphene oxide-Portland cement composites. J. Build. Eng. 54, 104604. https://doi.org/10.1016/j.jobe.2022.104604.
47. Sharma, S.; Kothiyal, N.C. (2015) Influence of graphene oxide as dispersed phase in cement mortar matrix in defining the crystal patterns of cement hydrates and its effect on mechanical, microstructural and crystallization properties. RSC Adv. 65, 52642-52657. http://doi.org/10.1039/C5RA08078A.
) that GO sheets act as 2D platforms to guide the formation of 2D calcium silicate hydrated microplates with a dense nanostructure form a 3D network that can mechanically reinforce the cement paste at the nanoscale. GO regulates the morphology of calcium silicate hydrates and induces the formation of compact, flower-like C-S-H crystals, so flexural strength was greater (2424. Lv, S.; Ma, Y.; Qiu, C.; Sun, T.; Liu, J.; Zhou, Q. (2013) Effect of graphene oxide nanosheets of microstructure and mechanical properties of cement composites. Constr. Build. Mater. 49, 121-127. https://doi.org/10.1016/j.conbuildmat.2013.08.022.
, 4848. Long, W.J.; Wei, J.J.; Xing, F.; Khayat, K.H. (2018) Enhanced dynamic mechanical properties of cement paste modified with graphene oxide nanosheets and its reinforcing mechanism. Cem. Concr. Compos. 93, 127-39. https://doi.org/10.1016/J.CEMCONCOMP.2018.07.001.
). GO gives a filler effect, supplies more nucleation sites at the time of the hydration process, and regulates the development of cement hydration crystals (4949. Wang, M.; Wang, R.; Yao, H.; Farhan, S.; Zheng, S.; Du, C. (2016) Study on the three dimensional mechanism of graphene oxide nanosheets modified cement. Constr. Build. Mater. 126, 730-739. https://doi.org/10.1016/j.conbuildmat.2016.09.092.
).

Figure 11 shows the effect of water/cement ratio, type of sand and GO addition on compressive strength; its variation was like flexural strength.

According to CEDEX, the Spanish National Public Works Research Centre (88. CEDEX. Construction and demolition waste. Waste usable in construction. November 2014. Retrieved from https://www.cedexmateriales.es/upload/docs/es_RESIDUOSDECONSTRUCCIONYDEMOLICIONNOV2014.pdf.
), when 100% fine aggregates are replaced by recycled mortars instead of standard aggregates, compressive strength can decrease by 33%. In this case, the experiment showed a 36% loss in compressive strength using recycled aggregate.

The addition of GO to standard sand and recycled sand mortars resulted in an improvement of approximately 20% in both samples. The action of GO benefitted from enhanced compressive strength. This material reduced the size of internal pores and reinforced its internal structure, leading to a decrease of large pores and an improvement of compressive strength.

The improvement in compressive strength due to the addition of small amounts of GO has been shown by many studies (2727. Li, W.; Li, X.; Chen, S.J.; Liu, Y.M.; Duan, W.H.; Shah, S.P. (2017) Effects of graphene oxide on early-age hydration and electrical resistivity of Portland cement paste. Constr. Build. Mater. 136, 506-514. https://doi.org/10.1016/j.conbuildmat.2017.01.066.
, 4848. Long, W.J.; Wei, J.J.; Xing, F.; Khayat, K.H. (2018) Enhanced dynamic mechanical properties of cement paste modified with graphene oxide nanosheets and its reinforcing mechanism. Cem. Concr. Compos. 93, 127-39. https://doi.org/10.1016/J.CEMCONCOMP.2018.07.001.
, 4949. Wang, M.; Wang, R.; Yao, H.; Farhan, S.; Zheng, S.; Du, C. (2016) Study on the three dimensional mechanism of graphene oxide nanosheets modified cement. Constr. Build. Mater. 126, 730-739. https://doi.org/10.1016/j.conbuildmat.2016.09.092.
). The introduction of small amounts of GO - as little as 0.0075% by weight - increased compressive strength by 15-33%. The polyhedron-like crystal hydration products formed a compacted structure and had greater compressive strength (Figure 9).

As can be seen in the figure, all mortars exceeded a compressive strength of 20 MPa. Thus, according to the EN 998-2 (5050. EN 998-2. (2018) Specification for mortar for masonry - Part 2: Masonry mortar. European Committee for Standardization.
) specifications for masonry mortars, they would be classified as class M20.

3.4.1. Acid attack test

 ⌅

The properties related to CDW durability have been studied previously (5151. Gómez-Cano, D.; Arias-Jaramillo, Y.P.; Bernal-Correa, R.; Tobón, J.I. (2023) Effect of enhancement treatments applied to recycled concrete aggregates on concrete durability: A review. Mater. Construcc. 73 [349], e308. https://doi.org/10.3989/mc.2023.296522.
), the foreign agents move through concrete by flowing through the porous system and diffusion and sorption, which introduces corrosion hazards. Figure 12 shows the variation in compressive strength in a mortar after immersion in acid for 14 days, compared to a mortar that was not immersed for the same period.

Figure 12.  Compressive strength ratio of acid-immersed and non-immersed mortars.

Acid attack resistance decreases with higher water/cement ratios and with the addition of recycled aggregate because open void porosity increases (see Table 5). Sulphuric acid attacks the matrix, producing gypsum inside the pores, which causes pore spalling and results in worse mechanical properties than cement. As CDW is made up of cement and concrete waste, the reactivity of recycled aggregate increases, decreasing acid resistance; by contrast, standard aggregate is composed mainly of SiO₂, which is not reactive, so only the cement matrix is affected by the acid attack.

Figure 13 shows a mortar made from recycled fine aggregate with a water/cement ratio of 0.6 and added GO after 14 days of immersion and two days of drying. The sample is covered by a white layer due to the formation of CaSO₄∙H₂O in the acid attack.

Figure 13.  Outer appearance of recycled sand mortar after immersion in acid.

When GO was added, the pore size distribution was lower, which prevented the entry of water (and acid); only gypsum was formed outside and the interior matrix remained unchanged (see Figure 13).

A recycled sand mortar with a water/cement ratio = 0.6 with GO is presented above (Figure 14. A) after the compression test. The sample had a white layer covering the surface, and the inside of the mortar had the colour of cement, which indicated that the attack was only superficial due to the addition of GO. Figure 14. B shows the standard sand mortar with a water/cement ratio = 0.5 without GO when the acid had penetrated the core of the mortar, after the compression test, although it is not as marked as in the outer zone of both samples. Due to its high-water absorption capacity, the acid was able to penetrate more easily, and gypsum formed on the inside, contributing to the breakdown of the material as a result of spalling.

Figure 14.  A) Mortar with recycled aggregate and water/cement ratio = 0.6 with GO and B) mortar with standard sand and water/cement ratio = 0.5 without GO.