Properties Evolution of Some Hydraulic Mortars Incorporating Graphene Oxides

: In this experimental study, the mechanical and adhesion properties of several hydraulic lime mortars incorporating graphene oxide (GO)-based nanomaterials were evaluated. Four different composite mortar samples were prepared by adding different percentages of GO-based powders (functionalized), i.e., 1 wt.% GO, 5 wt.% GO, 5wt.% GO-Ag-GO-Fly ash, and 5 wt.% GO-ZnO-GO-TiO 2 into the reference mortar sample. The mortar specimens were analyzed through mechanical tests, FT-IR, and SEM. The behavior of selected mortars exposed to chemical attacks was also investigated. The results indicate that the addition of the functionalized GO-based powders leads to a signiﬁcant improvement in the mortar’s adhesion to the brick substrate (up to 80%) compared to the reference sample, especially in the case of the hydraulic lime mortar incorporating the mixture of GO-Ag and GO-Fly ash, which also showed good resistance to chemical attacks. are reduced compared to the reference sample, while their adhesion strength to the brick substrate increased. The best results were obtained by incorporating a mixture of GO-Ag and GO-Fly ash into the reference mortar sample. The M- GO-Ag and GO-Fly ash also showed good resistance to chemical attacks.


Introduction
It is known that the characteristics of the mortars used in cultural and historical building conservation and restoration works can be modelled by controlling several factors, such as the nature and the ratio of the binder and the aggregate, the addition of various additives, etc. [1,2].
During restoration works on a historical building, it is crucial to ensure the materials' compatibility, not only in terms of chemical and physical characteristics, but also regarding the level of mechanical strength [3].
In the last decade, Natural Hydraulic Lime (NHL) mortars have been extensively used for the restoration of historical structures, due to their enhanced compatibility with historical materials [4]. However, when choosing an NHL, it is not always the best solution to use the hardest lime. For instance, in the case of historical masonry, the mechanical strength could be less important than the chemical and physical compatibility of the materials [4]. An NHL 3.5 used for plastering will give the layer its sweating properties, allowing the control of the relative humidity in indoor spaces. Furthermore, due to its low modulus of elasticity, NHL 3.5 has the ability to work with the natural bricks in the masonry, compensating for the effects of thermal expansion and, thus ensuring greater durability [5]. However, the environmental conditions and degree of mortar exposure to atmospheric agents should be also considered.
Recently, it was found that various nanomaterials, such as nano-TiO 2 [6,7] and graphene [8,9] could be successfully used to improve the mechanical properties and durability of the mortars [10][11][12]. Nano-TiO 2 has been also proved an excellent photocatalytic material that imparts biocidal, self-cleaning, and smog-abating functionalities when added to cement-based materials [13,14].
Graphene oxide (GO) nanofoils, an oxidized derivative of graphene, present an extremely large, specific surface area, excellent mechanical properties, and thermal conductivity, while their surface contains a large number of active groups containing oxygen, such as hydroxyl, carboxyl, and epoxy groups [15]. Recent studies proved that graphene oxide (GO) nanofoils are highly effective for improving the mechanical properties of cement-based materials [16][17][18][19][20].
Special attention has also been paid to the development of nanocomposite materials that combine GO with different nanoparticles, such as Ag, Ag/ZnO, TiO 2 , SiO 2 , ZnO, etc. For instance, silver-doped graphene oxide materials have favorable properties of low strength, good dispersion, and improved mechanical strength [21]. El-Shafai et al. [22] also reported that the GO sheets functionalized with Ag/ZnO could improve the antibacterial and antifungal effect of the materials into which they are introduced [22], by inhibiting the pathogens' growth. Likewise, the doped ZnO powders containing small amounts of Ag revealed good antimicrobial properties, which allowed their application as pigments for indoor and outdoor painting [23].
Fly ash is a by-product of coal-fired thermal power plants that is used as fine aggregate (filler) in mortars, leading to considerable improvements in their strength [24]. Fly ash is made of mud-sized particles, which are generally spherical, usually between 10 and 100 microns in size. These small spheres of glass improve the fluidity and workability of the fresh mortar. Finesse is another important property that contributes to the pozzolanic reactivity of the fly ash [25]. Oltulu et al. [26] showed that the addition of nano-SiO 2 , nano-Al 2 O 3 and nano-Fe 2 O 3 powders into the cement mortar containing fly ash improved the mechanical and physical properties of the mortars by increasing their pozzolanic activity.
The aim of this study is to obtain four mortar variants, with additives based on graphene oxide/mixtures of functionalized graphene oxides, intended for the restoration of old buildings (churches, historic buildings), in order to improve their adhesion to the original masonry components which have deteriorated over time and must be restored. There are situations in which the use of a cement mortar has created problems, proving to be ineffective. Historical mortars are usually weak mortars that crumble to the touch. It is known that lime hydraulic mortars are more permeable to water vapor and also that the mortars with lower compressive strengths are more susceptible to higher permeability rates. The aim is for these graphene mortars to ensure good adhesion to defects or mortar joints that need to be repaired, ensuring adequate compatibility with the original mortar. Hydraulic lime mortars (NHL) harden over a much longer period of time than cement mortars, therefore it is recommended that the restored site be kept dry until it has hardened. The study also focused on determining the mechanical properties, looking at how the addition of graphene oxide additives can influence them in a positive or negative way. The obvious aim is to slightly improve these properties compared to the reference material, but without increasing them too much, so that compatibility with the original mortars can be maintained. The newly experimental mortars were characterized by mechanical tests, Fourier transform infrared spectroscopy (FT-IR), and scanning electron microscopy (SEM). The evaluated physical properties of the mortars include the compressive and flexural strength, as well as their adhesion to the brick substrate. The microstructural changes of the selected composite mortars exposed to chemical attacks were also studied.

Materials
The basic mixture of the reference mortar (M) was prepared using a commercial natural hydraulic lime (NHL 3.5) with a 2% dependence on methacholine (Metastar 402 from Imerys Minerals) as the binder (Table 1). It was supplied by Calix Blanca, Italcementi, Bologna, Italy. The NHL 3.5 natural hydraulic lime is suitable for preparing mortars used for various applications, including masonry, rendering, plastering or interior and exterior finishing. The mechanical characteristics, porosity, and low salt content of NHL 3.5 ensure its compatibility with traditional construction materials (i.e., brick and stone masonry, earth, and vault constructions, etc.), which makes it suitable for conservation, restoration, and maintenance works for cultural and historical buildings and monuments. FTIR analysis was performed in order to determine if there were changes in the absorption bands with the introduction of graphene oxide additives in the composition of the mortars, or depending on their concentration. It is also possible to make a comparison of the experimental mortars with those used in the original plaster, by inputting evidence of the absorption peaks, attributed to the vibrations from the specific functional groups of the components.
The ATR-FTIR spectra of the hardened mortar samples were recorded using an FTIR spectrophotometer (Jasco Europe srl, Cremella, Italy) equipped with a total attenuated reflectance (ATR) attachment with horizontal ZnSe crystal (Jasco PRO400S). The FTIR spectra were measured with a resolution of 4 cm −1 , in the spectral range of 4000-500 cm −1 , and the scans were repeated 100 times.

Mechanical Tests
The durability of mortar is a significant parameter to ensure that successful restoration ensues and should be assessed either through laboratory studies [27] or in-situ tests [28].
In order to better identify the differences between the five types of mortars tested, the results were subjected to an Anova one-way statistical analysis, using Origin2019b graphing and analysis (OriginLab). The Tukey test was also used to better understand the differences between the five mortars. The null hypothesis is that the means of all levels are equal. The results were considered significant for p ≤ 0.05. The study groups (Table 1) are the five different mortar recipes, with comparisons being made between all of them for each mechanical test, and each group being represented by n = 20 results.
The preparation of the samples was subjected to the flexural and compressive strength of the hardened mortars and was carried out in accordance with SR EN 1015-11:2002 [29].
The compressive strength tests of all the mortar mixes were performed on 40 mm × 40 mm × 40 mm cubical mortar specimens after 28 days of curing. The specimens were compressed using a Lloyd Instruments mechanical testing machine (Ametek LR5K Plus Twin Column Bench Mounted) using the BS EN 1607-1997 test [30]. The maximum load cell was 5 kN, and the load on the samples was applied at a speed of 0.5 mm/min and a uniform force of 2000 N. For each type of recipe, 20 tests were recorded.
The flexural strength tests were performed on 160 mm × 40 mm × 40 mm mortar specimens, measured using three-point bending tests on three beams. The samples were subjected to a force up to fracture using the Lloyd Instruments mechanical testing machine at a loading rate of 50 N/s, using the ASTM C 348-14 [31] test. For each type of recipe, 20 tests were recorded.
The determination of the adhesive strength of the tested mortars was performed according to SR EN 1015-12: 2001 [32], using tensile pellets attached (5 cm in diameter) to the investigated mortars with a universal adhesive (Bison Epoxy, 5 min-2 component epoxy adhesive). A 10 cm brick was used as a base surface for the adhesive strength tests. The mortars were applied in a 20 mm layer on solid wet bricks, which were kept in a vertical position during the application. The adhesion strengths were estimated by the pull-off method, at a temperature of 25 • C, 55% humidity and a peel speed of 10 mm/min, using the Lloyd Instruments Mechanical Testing Machine. For each type of mortar recipe, 20 substrate adhesion tests were recorded.

Scanning Electron Microscopy (SEM)
For SEM investigations, the 28-day cured mortar samples taken from fractured mortar specimens after mechanical testing were used. The morphological and structural characterization of the fragments' surfaces was performed using an Inspect S electron microscope (FEI Hillsboro, OR, USA) at low vacuum, with an acceleration voltage of 30 kV and magnifications of 1000×, 5000×, 10000×.

Experimental Studies for the Evaluation of Mortars Exposed to Corrosive Attacks
In order to evaluate the behavior of the mortars in various external environmental conditions, chemical attacks were applied on selected mortar samples, i.e., the reference mortar (M) and the mortar sample containing GO-Ag and GO-Fly ash (M-GO-Ag-GO-Fly ash). Three different substances that simulate acidic rain and external corrosive factors were used for chemical attacks, as follows: rainwater, acetic acid, and a mixture of strong acids.
The morphological changes on the surface of the mortars following chemical attack at different pH values were investigated using SEM. The initial structures of the selected mortars were also analyzed.
For this study, mortar specimens of 40 × 40 × 80 mm dimension were prepared and cured for 28 days, at room temperature. One surface of the specimens was exposed to chemical attack, while the rest of the surface was sealed with epoxy resin.
The samples were stored at room temperature for 7 days. Each day, 2 mm of the free surfaces were immersed in the corrosive solutions for 30 min and then analyzed using SEM.    As seen in Figure 1, the FT-IR spectra of all graphene-containing mortars are similar to that of the reference sample. Four distinct groups of absorption bands can be identified in the ranges of 3642-3644 cm −1 , 3422-3446 cm −1 , 1419-1423 cm −1 and 1000-1083 cm −1 . An absorption peak at around 3640 cm −1 is noticeable in all the spectra from Figure 1. It could be attributed to the stretching vibrations of the hydroxyl group from Ca(OH) 2 . The wide band at 3436 cm −1 corresponds to the stretching vibrations of the O-H bond, which can also be attributed to the oxygen groups found both on the surface and at the edge of the GO sheets [8,33]. The third band could be assigned to the carbonatite phases, while the fourth one corresponds to the Si-O stretching vibrations. The small variations in the range of 1000-1083 cm −1 could be attributed to the silicon compounds in the sand [34]. However, the peak at approximately 1083 cm −1 is common to both silicates and aragonite, due to the Si-O asymmetric stretching vibrations and the symmetric CO 3 stretching, respectively [35]. The peaks at approximately 470 and 780 cm −1 could be ascribed to the silicates [36].

Mechanical Properties
It is known that the values of compressive strength can vary in a wide range, depending on both the composition and type of mortar. For lime mortars, the compressive strength lies between 0.1 and 1.0 N/mm 2 [37].
The results obtained from the compressive tests in our study are presented in Table 3. As shown in Figure 2, the highest value of compression tensile strength was obtained by the reference mortar, followed by the mortar containing 1% GO. The lowest compression strength was observed in the mortar sample containing ZnO and TiO 2 . The statistical Anova test revealed significant differences among the five groups of samples. However, the compressive strength values obtained from the samples M-GO-5%, M-GO-Ag-GO-Fly ash and M-GO-ZnO-GO-TiO 2 are rather similar.  The values of the flexural strength presented in Table 4 suggest that the tested mortars present rather weak mechanical characteristics, but these results are consistent with the mechanical properties of the lime mortars generally used in restoration and conservation works. However, the values of the flexural strength from Figure 3 are higher compared to previously reported data for different restoration mortars [34]. Table 4. Flexural test results (following the Tukey test: groups of specimens marked with the same letter show no statistically significant differences between them; the groups marked with different letters show statistically significant differences between them). It is known that flexural strength values are affected by the size of the inner pores of the mortar samples, which may contain water that weakens the strength of these materials. In our study, the highest flexural strength was obtained by the Fly ash-containing mortar sample (see Table 4). From the results obtained, the ductility character of the mortar can be observed. The force required to deform the specimen begins to increase during the test, up to the maximum point of the stress-strain curve which corresponds to the load at maximum load value. The action of this value is the formation of the "neck" specimen, which is accentuated reasonably quickly because the deformation continues to occur only in the neck area with lower forces (because this section of the specimen decreases continuously). Thus, the specimen breaks at load at break value, which corresponds to the point where the material failed. Following the Anova test, statistically significant differences were found among all the tested mortars. However, no significant differences regarding the maximum bending stress were observed among the mortar samples containing GO (M-GO-1% and M-GO-5%) or a mixture of GO-Ag and GO-Fly ash (M-GO-Ag-GO-Fly ash).

Mortar Sample Load at Maximum Load (N) ± SD Maximum Bending Stress (MPa) ± SD Load at Break (N)± SD
According to [38], the mechanical properties of the construction materials depend on the thickness of the tested layer, i.e., the value of the compressive strength increases quickly when the thickness of the mortar layer falls below 25 mm, while the maximum strength is usually obtained for thicknesses corresponding to the normal thickness of the horizontal joints (about 10 ÷ 12 mm).
In our study, the thickness of the layer of the mortars was 40 mm, and the results of the mechanical tests lay within the middle of the accepted ranges for lime mortars (0.1-1.0 N/mm 2 ). The values obtained for the compressive and flexural strengths suggest that the mortars tested present rather poor mechanical characteristics, but they are able to ensure compatibility with the mechanical properties of the bricks to which these mortars are addressed.
In 2017, Faria [34] investigated for the first time the incorporation of GO (in percentages of 0.05%, 0.1%, 0.5%, and 1%) on natural hydraulic lime mortars, determined by the demand to improve the properties of traditional materials used in the rehabilitation of heritage buildings. If, for small percentages of GO additions, the results of flexural strength were maintained and, for compression strength, were slightly higher compared to the reference mortar for larger GO additions, the flexural strength is slightly reduced while the compressive strength increases.
In our study, minor variations were observed in the mechanical properties that decrease by increasing the amount of GO but which show a maximum supported force and a breaking strength greater than the reference mortar, a sign that their ductility increases. Thus, although the overall improvement is not remarkable, the fact that the mechanical strength does not change much allows these mortars to maintain compatibility with historical materials, for applications such as replacing plasters or preserving and repairing historic buildings where low mechanical properties are required, so these results can be seen to be positive. Figure 4 shows the results of the adhesive strength of the mortars determined with the pull-off method. Three types of fractures occurred within all tested mortars. The lowest values were obtained from the mortar-adhesive interface fracture ( Figure 4c); therefore, they were eliminated from the applied statistical analysis. The presence of the fracture at the mortar-brick interface (Figure 4a) correlates with a higher hardness of the tested mortar, compared to the fracture occurring inside the mortar layer (Figure 4b). This appears when the mortar presents a higher adhesion strength in the pores of the support material.
Also known as peak load, the load at maximum load is the maximum load produced during an impact test. Quite often, this point may also correspond to the onset of material damage or complete failure. As shown in Table 5, the addition of GO leads only to slight improvements in the adhesion strength of the brick substrate compared to the reference mortar. Instead, the calculated value of the adhesion strength for the mortar sample incorporating the mixture of GO-Ag and GO-Fly ash is two times higher compared to the reference mortar (Table 5). Table 5. Adhesive strength of the mortars (following the Tukey test: groups of specimens marked with the same letter show no statistically significant differences between them; the groups marked with different letters show statistically significant differences between them). Although some significant statistical differences among the five mortar samples were obtained with regard to their adhesion resistance, the Tukey test emphasized that these differences are not important in the case of the reference (M) and GO-containing mortars (M-GO 1% and M-GO 5%).

Mortar Sample
The results of the adhesive strength tests showed that the mortars containing 5 wt.% GO and the combination of GO-Ag and GO-Fly ash present better adhesion strengths (0.23 MPa) compared to the reference mortar sample (0.10 MPa) ( Figure 5). Moreover, in the case of the M-GO-Ag-GO-Fly sample, the pull-off test revealed that more fractures occur in the mortar layer than at the mortar-brick interface.

Morphological and Structural Characterization
The SEM images obtained on the mortar surfaces are shown in Figure 6. Despite their homogeneous composition, the surface texture of the five mortars showed visible differences, as illustrated in Figure 6. Thus, the reference mortar (Figure 6a-c) presented a high porosity within the carbonate matrix, in which two types of crystalline growth emerged, i.e., with large and small particles. Instead, good internal cohesion between the components of the mortars containing GO and various additives was observed in Figure 6d-o.
Taking into account that the only difference between the five samples is related to the addition of some minor components, it could be asserted that the adhesion between the matrix and the aggregate is influenced not only by the texture of the aggregate particles, the presence of the GO-based added combinations and the nature and texture of these nanoparticles, but also by the viscosity of the mixture and the superficial tension [39].

Surface Deterioration of the Mortars Exposed to Chemical Attack
In an attempt to evaluate the behavior of the mortars in various corrosive environments, the reference mortar (M) and the mortar specimen presenting the best mechanical properties (M-GO-Ag-GO-Fly ash) were further subjected to chemical attacks. The surfaces of the mortars after exposure to the chemicals were examined using SEM, and the results are shown in Figure 7.
The attack of the nitric acid on the mortar matrix leads to the formation of calcium nitrate, which is a very soluble salt [40]. As a result, insoluble precipitates surrounding the areas of the unaffected material are formed. Consequently, variations of the pH in the degraded surface layer are expected, from very low values due to the acidic solution being in contact with the mortar, to high alkaline values that are characteristic of the unaffected material. The formed precipitates are soft, very porous, and present cracks on the surface caused by contractions during decalcification [40]. The corrosive substances can create cracks in the surface layer of the mortar, and cause material failure over time. Hence, for this type of construction material, the most important issue is the observation of a 'neutral' behavior to chemical attacks on the surface. As seen in Figure 7, a greater particle release from the surface layer of the tested mortar takes place, as the pH of the corrosive solutions decreases. The most degraded surface areas were noticed for the mortars in contact with the mixture of sulfuric acid and nitric acid (Figure 7c,f) and with the acetic acid solution (Figure 7b,e).
The results of the adhesive strength tests showed that the mortars containing 5 wt.% GO and the combination of GO-Ag and GO-Fly ash present better adhesion strengths (0.23 MPa) compared to the reference mortar sample (0.10 MPa). Moreover, in the case of the M-GO-Ag-GO-Fly sample, the pull-off test revealed that more fractures occur in the mortar layer than at the mortar-brick interface. It was reported [34] that the addition of well-dispersed GO (0.01-0.06% by weight) and other various nanoparticles could improve the strength and durability of cements and mortars due to their filler effect, because smaller particles can easily penetrate spaces created by the joining of larger particles [34]. Particularly, GO possesses hydroxyl (-OH), epoxy (-CH(O)CH-), carboxyl (-COOH), and carbonyl (-C=O) groups that are attached to the surface and to the edges of its sheets, which assist their effective dispersion in the water, due to electrostatic stabilization. Likewise, the addition of GO in the cement matrix reduces the need of using surfactants, which are commonly added to prevent the nanomaterials agglomeration caused by van der Waals attractive forces. Nonetheless, the carboxyl groups of GO are able to interact with the cement hydrates, creating a 3D network structure in the cementitious matrix [41], which could minimize the acid attack on the material surface.
The lifespan of mortar structures depends mainly on the durability of the substrate to which it is attached, and their durability is directly related to the presence of water due to their increased porosity.

Conclusions
In the present paper, several mortars incorporating graphene oxide without and with different mixtures of functionalized additives (Ag, Fly ash, ZnO and TiO 2 ) were prepared. Various mechanical, physico-chemical, and microscopical tests were carried out to evaluate the durability and mechanical properties of the natural lime mortar composites that were intended to be further used in cultural heritage restoration processes.
The results obtained showed that the compressive and flexural strengths of the mortars incorporating nanomaterials are reduced compared to the reference sample, while their adhesion strength to the brick substrate increased. The best results were obtained by incorporating a mixture of GO-Ag and GO-Fly ash into the reference mortar sample. The M-GO-Ag and GO-Fly ash also showed good resistance to chemical attacks.