Repairing performances of novel cement mortar modi ﬁ ed with graphene oxide and polyacrylate polymer

: Most cementitious repair materials have non ignorable drawbacks such as low strength, insu ﬃ cient bonding, and low anti - permeability. Although the bonding and anti - permeability of repair mortars modi ﬁ ed by polymer will be substantially improved, the compressive strength and component integrity will be decreased. Hereby six groups of cement mortars modi ﬁ ed by varied dosages of Graphene Oxide ( GO ) and PA copolymer ( GOPARMs ) were prepared. The ﬂ exural strength ( f t ) , compressive strength ( f c ) , f t / f c , bond strength ( f b ) , and chloride ion migration coef ﬁ cient ( λ c ) of GOPARMs were systematically studied by axial compressive, three - bending, pull - out, RCM method, along with microstructure analysis. When GO and PA dosages are ﬁ xed at 0.03 and 5 wt%, respectively, the f t , f c , f t / f c , f b , and λ c of GOPARMs reach the best comprehensive perfor -mances, which are 6.4, 46.5, 0.14, 6.73 MPa, and 1.179 × 10 − 12 m 2 /s. Compared with the control mortar, the f t , f c , f t /f c , and f b of GOPARMs are improved by 5.7, 12.3, 7, and 103%, respectively, and the corresponding λ c is dramatically reduced by 40.4%. Scanning electron microscope ( SEM ) shows that trace of GO can play a template nucleation e ﬀ ect on the hydra tion products ’ morphology and microstructure of GOPARMs. Meanwhile, cured PA polymer can form hydrophobic ﬁ lm and ﬁ ll the interfacial pores among hydration products, ﬁ nally superior repairing performances of GOPARMs with optimal mix can be achieved.


Introduction
With the large-scale construction of concrete infrastructure in China, cement production has increased year by year, where large amounts of carbon dioxide greenhouse gas are emitted. The concrete structure under harsh environment cannot reach the expected durability, which will shorten its service life. However, rebuilding and reconstruction will bring forth more carbon emissions. Actually, using high-performance repair materials to repair and strengthen the original damaged parts or cracks of concrete structure is more cost-effective and eco-environmental compared to the traditional rebuilding and reconstruction. The repaired structure can continuously hinder the invasion of external harmful substances and the service life can be effectively prolonged [1][2][3][4][5][6][7][8][9].
In order to achieve good compatibility with the original concrete structure, the cement-based repair mortars are often chosen as the repair materials [10][11][12][13][14][15]. But lots of weak interfaces, cracks and breaks, stain residues, low strength concrete, etc., exist in the old structure substrate, for which it is difficult to achieve structural interface repair and rapid strengthening effect with ordinary cement repair mortar. Many scholars studied the repair performance of Portland cement-based repair mortars on the substrate concrete. They found that its bond strength (f b ) on the substrate was not strong, and a non-negligible gap existed when the concrete dried and shrunk at different times, which jeopardized the wholeness between common repair mortars and old substrates [16,17]. Therefore, there is an urgent need to develop a special repair mortar with multi-functional performance such as fastsetting, high strength, low shrinkage, well-compatibility, and superior durability.
It is well known that viscous polymer latex can significantly improve the interfacial bond performance as frequently reported [18][19][20]. Compared to UHPC or ECC repair materials, the polymer-modified repair materials generally possess superior surface hydrophobic behavior, higher ductility, superior resistance to chloride ion permeability or corrosion, and lower shrinkage, whereas with relatively lower mechanical strength [21][22][23][24]. So, polymer-modified repair material is suitable to serve as high-performance cement-based repair materials [25,26]. In the 1980s, Ohama [27] and Konietzko [28] proposed the polymer netlike formation model and the dual web model in cement mortar, respectively. Ohama argued that the polymer would wrap a film around the hardened cement paste, while Grosskurth and Konietzko argued that the polymer and cement paste would interpenetrate and form a network structure. Today there are two main views on polymer-modified cement mortars (PCMs). One is based on the physical behavior of the polymer latex during the hydration of the cement [29,30]. During the hydration process, the polymer creates a network structure on the surfaces of the hydration products and cement particles or fills in the cracks of the cementitious material to improve its pore structure. The other view is that there is a physical interaction and chemical reaction between the polymer latex and the cement hydration, and the chemical reaction is a coordination reaction between Ca 2+ within the cementitious material and the polymer directly to improve the final performance through chemical bonding [31][32][33][34][35][36][37]. Kim [20] compared the bond strength of different materials modified repair mortars. The 28-day bond strength of styrene-butadiene rubber (SBR) modified repair mortar was 29% higher than the control group, which was the best result among all the materials. Yang et al. [22] evaluated the chloride permeability and microstructure of Portland cement mortar modified by SBR latex. The study showed that SBR can improve the resistance of cement mortars to chloride ion permeability and can bind hydration products. Jiang et al. [23] reported that the mortar mixed with polyacrylate latex (PA) and silica fume (SF) had high bond strength and excellent durability performance compared with the control mortar without PA and SF. It is worth noting that most studies have shown that the mechanical properties of PCMs dosed with some polymer latex in general reduced, especially dramatic decreases in compressive strength (f c ) [38][39][40][41]. Tian et al. [42] modified cementitious materials by using four organic polymers, namely silicone propylene latex, benzene propylene latex, waterborne epoxy latex, and acrylic latex, and revealed that most types of latex adversely affected the compressive properties of the cementitious materials. The f c and flexural strength (f t ) of the cementitious materials with 5 wt% dosage of polyacrylic latex decreased by about 28.0-31 MPa and 9.1-6 MPa, respectively. Pei et al. [43] tested the mechanical properties of PA-modified mortars with different monomer ratios and found that the f c of the modified mortar decreased by more than 50% when dosed with methyl methacrylate. Tian et al. [44] found that excessive PA addition significantly reduced the mechanical properties of the cement mortar, and the 28 days f c of mortar specimens mixed with 10 wt% PA was only about 36 MPa. Ma and Li [45] found that the f c of the mortar dosed with 10% PA decreased by 11.7% compared to the control mortar.
Obviously, the f c of PCM is still eager to substantial enhancement to achieve superior repairing property.
The dense microstructure of cement-based composites is the foundation for achieving high strength, toughness, and durability [46][47][48][49]. Therefore, the study of pursuing novel materials to regulate hydration products of cement-based composite has attracted considerable attentions. The employment of carbon nanotubes, graphene oxide (GO), and other materials can effectively toughen the cementitious materials [50][51][52][53][54][55]. Especially, GO possessing many oxygen-containing functional groups, high surface area, and amphiphilic hybrid sp 2 structure, can exert its active template and nucleation effect at very low dosage [56][57][58][59], and it can assume to have the role of template in the hydration products of the cementitious materials, accelerate the hydration rate, reshape the microstructure, and control the distribution of hydration products and form a regular and orderly microstructure [60][61][62][63][64]. Some scholars have found that suitable GO dose can significantly improve the tensile strength [65], f c and f t [66] of the cementitious materials, improve the microstructure [67] and durability [68], and reduce the expansion rate [52]. Meanwhile, the functional groups such as hydroxyl, carbonyl, and carboxyl groups of GO itself make it easy to react with organic polymers, and the reacted GO can also be effectively dispersed in cement matrix to embody its excellent properties in the corresponding composite [69,70].
Here combined with the merits of two kinds of materials, GO/acrylate polymer-modified repair mortar (GOPARMs) by hybridization of GO and PA was first prepared. The f t , f c , flexural-compressive strength ratio (f t /f c ), bond strength, resistance to chloride ion permeability, and microstructure of GOPARMs were also systematically investigated, to evaluate the practicability of GOPARMs as a novel repair material with excellent toughness, adhesion, and resistance to chloride ion permeability.
2 Materials and testing methods

Materials and mix ratio design
The chemical composition of cement (C), 42.5 grade highbelite sulfur-aluminate cement (HBSAC), produced by Tangshan Polar Bear Cement Plant Co., is shown in Table 1. Sand (S), ordinary river sand, locally available, with fineness modulus of 2.64, medium sand, mud content of less than 1.0% after rinsing treatment, in accordance with the requirements of DL/T 5126-2001 [71] was used. Superplasticizer (Sup), polycarboxylic superplasticizer with 30% water reduction rate, produced by Shandong Academy of Construction Sciences was procured. Polyacrylate latex (PA), obtained from the polymerization of acrylic monomer, was purchased from Beijing Mengtai Weiye Building Materials Co., Ltd, its main characteristics are shown in Table 2. Polyether modified silicone defoamer, B-364 type, was purchased from Guangdong China-union Fine Chemical Co. GO suspension with 10 mg/mL concentration, with 49-56% carbon and 41-50% oxygen content, was purchased from Nanjing Xianfeng Nanomaterials Technology Company. Distilled water (W), homemade was available.
The PA latex and GO dosages were calculated according to the solid content of PA and GO, respectively. The mix ratio was referred to the previous experimental results of Luo et al. [72] and fixed at (C + PA):S:W = 1:1.7:0.4 after allowing for the water content in the PA and GO. Table 3 shows the experimental mix ratios and specimen grouping of GOPARMs.

Preparation of GO/PA suspension
The GO dispersion, Sup, and remaining water (total water required for the design ratio minus the water contained in PA and GO) were mixed in a beaker and sonicated for  20 min (sonication power 100 W and frequency 40 kHz); PA was added to the dispersed GO solution and magnetically stirred for 10 min, and the GO/PA suspension was accordingly prepared as reserved.

Preparation of substrate specimens for bond bending test
The bending bond strength (f b ) test was referred to JC/T 2381-2016 [73]. It is worth to note that in order to keep the moisture content and the roughness of bonding interface of the substrate as consistent as possible for different experimental groups of the substrates, we improved to the preparation process of the substrates. The specimen with the size of 40 × 40 × 160 mm was sawed into two splits of 40 × 40 × 80 ± 2 mm blocks using a cutting machine, and the cut surfaces of the test block were lightly polished using an angle grinder to make each cut surface to have the same interface roughness. The polished test blocks were ovendried at 105 ± 2℃ until the weight was unchanged, and then they were immersed in water to saturate, and their saturation rates were calculated. The substrate test blocks were oven-dried at 105 ± 2 ℃ until its water content was between 55 and 65%. The test blocks were film sealed and left to stand for 2 days.

Preparation of GOPARMs specimens
GOPARMs was prepared according to the pre-designed ratio with reference to the preparation method specified in GB/T17671-1999 [74]. The cement, water, and sand were first mixed at 40 rpm for 2 min at low speed; GO/ PA suspension was subsequently mixed into the above mixture for 2 min at low speed, then stirred rapidly at 80 rpm for 30 s, and the defoamer was dipped to eliminate the air bubble; after that, the well-mixed slurry was poured into the molds of varied sizes (40 mm × 40 mm × 160 mm, φ100 mm × 50 mm), vibrated about 60 times on a mortar jump table to compact and facilitate the possible air bubble escaping from the slurry. The mortar surface was smoothed and film covered, and the GOPARMs specimens were removed after 5 h setting and standard curing  till 7 days of age. The preparation process of GOPARMs is shown in detail in Figure 1.

Characterization of cured GOPARMs specimens
Compressive and flexural strength tests for the specimens with size of 40 mm × 40 mm × 160 mm was referred to GB/T17671-1999 [74], as shown in Figure 2(a), and the average experimental data of three specimens in each group represent the final result. The resistance to chloride ion permeability test was referred to GB/T50082-2009 [75], as shown in Figure 2   , the corresponding f c reaches the maximum value of 46.5 MPa, which is 4.5% higher than that of ordinary mortar without GO or PA dosing (44.5 MPa). And it's noting, the f c of GOPARMs has somewhat decrease when GO dosage continues to increase.
It can be seen from Figure 3(b) that under the same condition of GO dosage, the f t of GOPARMs altered with PA dosage is different from that of f c . In the case of low PA dosage (5 wt%), the f t of GOPARMs increases, and its f t reaches the maximum 6.7 MPa when GO dosage is 0.03 wt%, which is almost 20% improvement with respect to the plain mortar (5.6 MPa) without any PA and GO; as the PA dosage continues to increase, the f t enhancement gradually decreases. And under the same condition of PA dosage, the effect of GO on the f t of GOPARMs has the same trend as its effect on its f c . Under the condition of low dosage of PA (lower than 0.03 wt%), the higher the GO dosage is, the higher is the f t of GOPARMs, whereas when the GO dosage climbs up to 0.05 wt%, the corresponding f t decreases significantly.
Obviously, at the optimal dosages of 5 wt% PA and 0.03 wt% GO, the corresponding f c and f t are 46.5 and 6.4 MPa, respectively.
As known, PA dosage has a negative effect on the f c of GOPARMs. Actually, with the hydration reaction of cement, PA will form polymer films staggered inside the mortar or wrap the hydration products. As shown in Figure 4(a) and the yellow circles in Figure 4(b), there exists obvious milky white films in the specimens with 5 wt% PA dosage, but not in the control specimen without PA dosage. Because the elastic modulus of the polymer film is much lower than that of the main product C-S-H gel, after the cement mortar is hardened, it leads to the reduction in the elastic modulus, and the f c of GOPARMs. In addition, the loading of viscous polymer will inevitably introduce air bubbles, as shown by the red rectangles in Figure 4(b). Although appropriate amount of defoamer has been added, still the air-entraining effect of the polymer cannot be avoided, and the introduction of larger pore size bubbles will increase the porosity of the mortar, so the f c of GOPARMs will have an unignorable drop with high polymer dosage.
It is worth noting that there exists an obvious increase (7%) in f t of GOPARMs at low PA dosage (0 wt% GO dosage). This is because as the hydration reaction of cement proceeds, PA can form a film inside the mortar to fill the small gaps between or wrap the hydration products, which are attached to the surface of the cementitious gel and improves the bulk flexural strength (f t ) and toughness (f t /f c ) of the final mortar. When the filling and wrapping effect of the polymer reaches saturation, the f t of cement mortar reaches the highest. However, if the polymer dosage continues to increase, the hydration products per unit area decreases, resulting in a decrease in the continuity between hydration products per unit area, the polymer will overwhelm the cement particles, thus affecting the hydration degree [45], and the corresponding macroscopic f c and f t of the cement mortar decrease. In addition, the air-entraining effect of polymer will also have a negative impact on the f c and f t of the cement mortar, as shown by Figure 4(a) and (b). Accordingly, the microstructure in group No. P5G3 specimen has much more air bubbles than that in No. P0G0 group.
Obviously, the increase in f c and f t at low GO dosage is owing to the nucleation effect and reshaped microstructure of GO on cement hydration products. It can render AFt, AFm, AH, CH, and C-S-H gel to form regular and orderly arrangements through the templating effect of GO itself, forming flower-shaped or multifaceted shape crystals or gels [34,62,63]. These crystals or gels mostly grow in the pores or loose positions in the mortar structure and are mutually connected with each other, with the tendency to form a network, which in turn increases the f c and f t of the mortar. With the continuous increase in GO dosage, the dispersion effect of GO is insufficient and agglomeration phenomenon may occur, which will lead to the accumulation and irregular arrangement of hydration products, and the bridging capacity between hydration decreases, its f c and f t accordingly falls down [64,69].

Fracture toughness
The fracture toughness of GOPARMs can be characterized by its ratio of f t /f c , with higher f t /f c indicating superior fracture toughness of the composite [76]. Figure 5 shows the changes in the f t /f c of GOPARMs with different dosage of PA and GO. It can be seen that the f t /f c of the composite gradually increases with the increase in PA dosage, and the variation in f t /f c is great in the range of 0-10 wt% dosage (f t /f c increased by 74.6, 63.1, 63.1, and 46.9% at 0, 0.01, 0.03, and 0.05 wt% GO dosage, respectively), while it is relatively low in the range of 10-20 wt% dosage (f t /f c increased by -6.6, 15.6, 0.5, and 2.1% at 0, 0.01, 0.03, and 0.05 wt% GO dosage, respectively). This indicates that the fracture toughness of GOPARMs can be effectively improved with the appropriate dosage of PA. Actually, PA itself has very good deformability, it can well fill the small pores inside the GOPARMs, suitable PA dosage renders GOPARMs to have superior fracture toughness [43]. Figure 5 also shows that there is no significant variation in the f t /f c of GOPARMs with GO dosage under the same PA dosage condition. It shows that the GO dosage has no significant effect on the f t /f c of GOPARMs. In terms of f t /f c , the optimal dosage of PA and GO are 20 and 0.01 wt%, respectively, the corresponding f t /f c is 0.245. It should be noted that at this dosage level, the corresponding f c of GOPARMs decreases significantly owing to low elastic modulus of PA at high dosage.

Interfacial bond strength 3.3.1 Effect of PA dosage on the f b of GOPARMs
As shown in Figure 6, the f b of GOPARMs with the substrate is significantly enhanced by low PA dosage, and the most obvious improvement effect is achieved by 5 wt % PA dosage. Without any GO dosage, the f b of GOPARMs increases by about 39% with respect to the baseline; whereas with the addition of 0.03 wt% GO, the f b reaches the maximum value (6.73 MPa), which is about 103% enhancement compared with the baseline group. It is worth to note that with higher PA dosage, the increment amplitude of f b of the GOPARMs also decreases. When the PA dosage is 10 wt%, the f b of the specimens without GO dosage increases by about 36% compared with the baseline; when the GO dosage is 0.03 wt%, it accordingly increases by about 45% compared with the baseline. While the f b of the specimens is even lower than that of the baseline when the PA dosage reaches 20 wt%.  The f b of mortar with low PA dosage increases because the polymer in the mortar can penetrate into the pores of the mortar, rendering the contact surface of the old and new test blocks more tightly connected, so the f b of the mortar will be elevated. When PA dosage keeps increasing, the content of the cementitious material in the mortar decreases and the polymer content increases, the bonding performance of the cement mortar decreases when the f b of the cementitious material decreases to a greater extent than the f b provided by the polymer.

Effect of GO on the f b of GOPARMs
From Figure 6, the f b of GOPARMs and the substrate will be enhanced with a small amount of GO, and the enhancement effect on the f b of GOPARMs is related to PA dosage. The best strengthening effect of GO on the f b of the specimen can be achieved when the PA dosage is 5 wt%, and the f b gradually increases up to 6.73 MPa when the GO dosage is between 0 and 0.03 wt%. While GO dosage continues to increase, the corresponding f b steadily decreases. Under the condition of 10 wt% PA dosage, the f b gradually increases up to 4.82 MPa when GO dosage is between 0 and 0.03 wt%, and when the GO dosage continues to increase, the f b starts to decrease, and sharply drops down to 4.00 MPa. Under the condition of 20 wt% PA dosage, the GO dosage has no obvious effect on the improvement of the bonding performance, and even causes a significant decrease in the f b .
The mechanism of GO enhancing the f b of GOPARMs is similar to that of enhancing the f c of GOPARMs, which can induce the generation of flower-shaped and polyhedral crystals at the pores and cracks in the cementitious composites, thereby reduce the pores and cracks, and significantly improve the strength, toughness, and interface adhesion of the mortar.
In terms of f b , 5 wt% PA and 0.03 wt% GO are the optimal dosage, for the f b of GOPARMs to reach up to 6.73 MPa.

Chlorine ion permeation resistance
performance of GOPARMs Figure 7 shows the rapid chloride ion migration coefficient method (RCM) test operation and the color rendering diagram of chloride ion permeation of specimens. Seven groups of specimens were selected for this test, namely No. P0G0, P5G0, P5G1, P5G3, P5G5, P10G0, and P20G0, and the mix ratios of each group are shown in Table 3. The effect of GO and PA dosages on the resistance to chloride ion permeability of mortar was investigated by analyzing the chloride ion migration coefficients (λ c ) of No. P5G0, P5G1, P5G3, P5G5, P0G0, P10G0, and P20G0, respectively. The λ c is calculated according to equation (1): where, D RCMnon-stationary λ c of concrete, accurate to 0.1 × 10 −12 m 2 /s; Uabsolute value of the voltage used (V); Taverage of the initial and ending temperatures of the anode solution (°C); Lspecimen thickness (mm), accurate to 0.1 mm; X daverage value of chloride ion penetration depth (mm), accurate to 0.1 mm; ttest duration (h).

Effect of PA dosage on the resistance to chloride ion permeability of GOPARMs
As seen from Figure 8, with the increase in PA dosage, the λ c first gradually decreases, and when the PA dosage increases to 5 wt% (GO dosage 0), the λ c is the smallest, only 1.495 × 10 −12 m 2 /s, which is 24.3% less than that of the specimens without PA dosage (P0G0). Continuing to increase the PA dosage, the λ c of GOPARMs gradually increases, and when the PA dosage reaches 20 wt% (GO dosage 0.03 wt%), the λ c reaches 29.763 × 10 −12 m 2 /s, which is much larger than that of the blank group. Therefore, an adequate amount of PA can improve the resistance to chloride ion permeability of mortar to a certain extent, but too large PA dosage will have a negative effect on the impermeability of the mortar. The reason why fewer PA can increase the resistance to chloride ion permeability of mortar is that the polymer can penetrate into the pores of the mortar, and cement hydration absorbs the water from outside and the water in PA. The organic polymer film can be formed after the PA dries, and this film has a good filling effect on the small pores and improves the material resistance to chloride ion penetration. It is worth noting that the addition of too much PA will make the mortar very viscous, and many larger air bubbles will be introduced during the preparation of mortar specimens, which leads to a reduction in the compactness and resistance to chloride ion permeability.

Effect of GO dosage on the resistance to chloride ion permeability of GOPARMs
From Figure 8, we can clearly observe that with the increase in GO dosage, the λ c decreased first, and the smallest λ c of 1.179 × 10 −12 m 2 /s was obtained when the GO dosage was 0.03 wt% (PA dosage 5 wt%), which was about 13.6% lower than that of the P5G0 group without GO dosage. However, when the GO dosage was increased to 0.05 wt%, the λ c increased again to 1.934 × 10 −12 m 2 /s, which was 41.8% higher than that of P5G0. The above experimental phenomenon indicates that the addition of a small amount of GO can reduce the λ c of the specimens, i.e., the resistance to chloride ion permeability is improved, but the excessive amount of GO will adversely affect the resistance to chloride ion permeability of the specimens. The effect of change in GO on the microstructure of mortar can explain the above situation well. Low amount of GO has good active template and nucleation effect, which can effectively improve the arrangement structure of hydration products, reduce the pore structure inside GOPARMs, and improve the impermeability of the mortar. Excessive amount of GO is not easy to disperse in the cementitious materials, and agglomerated GO cannot play its excellent performance, but will lead to the agglomeration and accumulation of hydration products, reduce the connection between hydration crystals, and the internal compactness of GOPARMs to be reduced, which is not conducive to its impermeability performance.
In terms of the resistance to chloride ion permeability, 5 wt% PA and 0.03 wt% GO are the optimal dosage, respectively, and the corresponding λ c of GOPARMs is 1.179 × 10 −12 m 2 /s.

Micromorphological analysis of GOPARMs
In order to observe the microscopic morphology of GOPARMs and further verify the microscopic mechanism of the changes in its mechanical properties, bonding properties, and resistance to chloride ion penetration, in this study, the crosssections were observed by field-emitting scanning electron microscopy.
3.5.1 Micromechanical analysis of PA to improve the performance of GOPARMs Figure 9 shows the microscopic morphology of hydration products of GOPARMs with different PA admixtures. The blank group in Figure 9(a) shows typical cement hydration products, including needle-rod AFt, amorphous AFm, CH, AH, and bulk C-S-H gels with a relatively loose structure and more voids and cracks. In Figure 9(b), with PA dosage of 5 wt%, PA has obviously played a role in filling the pores and connecting the hydration products, the gel structure is obviously denser compared to the blank group, and the polymer starts to form a small amount of encapsulation on the hydration products. In Figure 9(c), the presence of continuous rod-like gels can be clearly observed with PA dosage of 10 wt%, and these gels are mixed gels of polymer and hydration products, i.e., the polymer has obvious encapsulation connection to the needle and rodlike hydration products. However, some of the gels intertwined with the cement hydration products appeared to be agglomerated into clumps, which produced fine cracks compared to GOPARMs with 5 wt% PA dosage. The results of the different elemental contents detected in the orange marked points of PA-modified mortars in Figure 9(c) are given in Table 4. It can be found that the C/O is about 0.13, which is much higher than the C/O of ordinary cementitious materials [34]. Moreover, Ca, Si, S, and Al in the mortars, which are the key elements that constitute the hydration products of SAC, can also be detected by EDS. The junction and the EDS and SEM test results can prove that the hydration products of the cementitious materials and PA mixed with each other. And PA can fill the voids between the hydration products or encapsulate the hydration products.
After microscopic tests, it is difficult to observe a very obvious polymer film formation regardless of the PA dosage, and it is speculated that the improving mechanism of the PA on the performance is not the one proposed by most scholars to enhance the macroscopic performance of mortar by forming a good integral polymer film between hydration products [77][78][79]. Tian et al. [34] demonstrated that PA can complexation react with Ca 2+ within the cementitious materials to generate calcium carboxylate functional groups, which can enhance the properties of the cementitious materials through chemical bonding. In summary, a small amount of PA can not only improve the compactness of mortar material by filling the internal pores of GOPARMs   but also enhance the material properties by forming ligand bonds through chemical reactions, the specific complexation reactions are shown in Figure 10. However, the large amount of PA will lead to the increase in large internal pores due to the air-entraining effect of PA, which will reduce the degree of material compactness and adversely affect the performance of GOPARMs.

Micromechanical analysis of GO to improve the performance of GOPARMs
Some scholars have found that GO can be modified by Sup or PA through non-covalent or covalent bonding modifications during physical mixing [69,80,81]. It can effectively improve the dispersion of GO, which is the basis for GO to exert its excellent performance [82]. Figure 11 shows the morphology of hydration products of GOPARMs doped with 0.03 wt% GO. It can be clearly seen that the internal hydration products of mortar are denser and more orderly after GO dosage. This will lead to a significant increase in the denseness of the cement mortar material, which is beneficial to the improvement of the macroscopic properties of the material. The dense and orderly distribution of hydration products is due to the small amount of GO in GOPARMs which exerts active template and nucleation effect. The high specific surface area of GO provides a good nucleation structure for the orderly growth of hydration products, and GO itself can also play a good filling and linking effect between hydration products and cement particles, optimizing the threedimensional network structure inside GOPARMs. Therefore, GO can play a good toughening and densification effect on GOPARMs, which can effectively improve the mechanical properties and impermeability of the material [83][84][85][86].
It is concluded that all the properties of GOPARMs meet the specification requirements, and its f b and λ c are significantly superior to those of the ordinary repair mortar, which is suitable as a cost-effective repair and reinforcement material application in concrete structure.

Conclusion
In this article, GO dispersion and PA were physically blended and the GO/PA mixture was synergistically modified cement mortar, and the mortar was characterized with mechanical and durability behavior (f c , f t , f t /f c , f b , and λ c ). Some results are concluded: 1) With the increase in PA dosage, the f c of mortar steadily decreases, and the f t /f c steadily increases, but the f t , f b , and λ c of the mortar first increases and then decreases. 2) With the increase in GO dosage, the f c , f t , and f b of mortar are all up first and then down, yet the trend of f t /f c is not obvious, the optimal GO dosage is at 0.03 wt%.
3) The doped GO can play an active template and nucleation effect among the hydration products of GOPARMs, resulting in improved texture and microstructure; while the doped PA can encapsulate the hydration products to a certain extent, playing the role of filling the internal pores and connections, and forming coordination bonds between the hydration products, thereby improving the mechanical toughness, bonding, and resistance to chloride ion permeability of GOPARMs. 4) After comprehensive consideration, the f c , f t , f t /f c , f b , and λ c of GOPARMs with the dosage of 5 wt% PA and 0.03 wt% GO reach 46.5 MPa, 6.4 MPa, 0.14, 6.73 MPa, and 1.179 × 10 −12 m 2 /s, respectively, which are enhanced by 5.7, 12.3, 7, and 103%, and the corresponding λ c is reduced by 40.4%. Obviously, a cost-effective and competitive repair mortar with superior bonding and resistance to chloride ion permeability can be predicted.
Acknowledgments: The authors are obliged to thank for the reinforcement mechanism discussion with Prof. Qiuyi Li (Qingdao Agriculture University) and Prof. Jing Zhong (Harbin Institute of Technology (Shenzhen)).