MOC Doped with Graphene Nanoplatelets: The Influence of the Mixture Preparation Technology on Its Properties

The ongoing tendency to create environmentally friendly building materials is nowadays connected with the use of reactive magnesia-based composites. The aim of the presented research was to develop an ecologically sustainable composite material based on MOC (magnesium oxychloride cement) with excellent mechanical, chemical, and physical properties. The effect of the preparation procedure of MOC pastes doped with graphene nanoplatelets on their fresh and hardened properties was researched. One-step and two-step homogenization techniques were proposed as prospective tools for the production of MOC-based composites of advanced parameters. The conducted experiments and analyses covered X-ray fluorescence, scanning electron microscopy, energy-dispersive spectroscopy, high-resolution transmission electron microscopy, sorption analysis, X-ray diffraction, and optical microscopy. The viscosity of the fresh mixtures was monitored using a rotational viscometer. For the hardened composites, macro- and micro-structural parameters were measured together with the mechanical parameters. These tests were performed after 7 days and 14 days. The use of a carbon-based nanoadditive led to a significant drop in porosity, thus densifying the MOC matrix. Accordingly, the mechanical resistance was greatly improved by graphene nanoplatelets. The two-step homogenization procedure positively affected all researched functional parameters of the developed composites (e.g., the compressive strength increase of approximately 54% after 7 days, and 37% after 14 days, respectively) and can be recommended for the preparation of advanced functional materials reinforced with graphene.


Introduction
Graphene is a carbon-based, two-dimensional nanomaterial and consists of a single monolayer of graphite with a thickness of one atom (0.34 nm). Due to its structure, graphene exhibits extraordinary physical properties such as a high specific surface area, ultrahigh tensile strength and modulus of elasticity, and excellent electrical, optical, and thermal parameters. Graphene nanoplatelets (GNPs) are composed of a few layers of graphene and have a total thickness of less than 100 nm. GNPs inherit many advantages of graphene and, in contrast to it, GNPs can be produced on a large-scale and are cheaper than carbon-based nanotubes and nanofibers [1][2][3]. The price of GNPs is about 400 EUR/kg, which is not high with respect to its low content in the composition of graphene-doped composites.
In recent years, the addition of GNPs to Portland cement-based composites to enhance, in particular, their mechanical, electrical, and durability-related properties was studied intensively. Wang et al. [4] reported that the incorporation of 0.05 wt. % of GNPs increased no study has been found to investigate the effect of MOC-graphene mixture preparation technology on the parameters of resulting composites. Therefore, the main objective of this study is to evaluate comprehensively the influence of the MgCl 2 -GNPs aqueous suspension dispersion prior to the preparation of MOC-GNPs fresh mixture, and the additional homogenization of this fresh mixture after the ordinary mixing procedure on the basic structural, microstructural, and mechanical properties of the hardened composites. The obtained results will provide a basic framework for the design of MOC-GNP nanocomposites.

Chemicals
The composition of the investigated materials is given in Table 1 Table 1, this ratio was identical for both MOC-graphene mixtures. It is expected, due to the graphene doping, that the unit price of MOC-G materials will be around 0.75 EUR higher than that of pure MOC.

Synthetic Procedures
For the evaluation of homogenization influence on the fresh and hardened paste properties, MOC-graphene composites were prepared with mixing methods based on one-step and two-step homogenization, respectively. Both methods started with the first step of homogenization, where graphene nanoplatelets were dispersed in MgCl 2 solution using a mechanical rotor-stator homogenizer Ultra-Turrax T-18 (IKA, Staufen im Breisgau, Germany). This first mixing was conducted for 5 min at 10,000 rpm. After that, the ordinary mixing procedure based on the technical standard EN 14016-2 [23] was applied. Magnesia powder and MgCl 2 solution for MOC-REF, or graphene-MgCl 2 solution in the case of MOC-GNPs composites, respectively, were placed into a planetary type mortar mixer (ELE). The mixing started at a low speed regime (paddle 140 rpm, mixing head 62 rpm) for 1 min, followed by high-speed mixing (paddle 285 rpm, mixing head 125 rpm) for 0.5 min. Then, the mixing was stopped to clean the mixer head and the sides of the mixing bowl. After that, mixing continued for another 1 min at high speed followed by casting of MOC-REF and MOC-G samples. The second step of homogenization was applied for MOC-G-H samples only. This composite mix was additionally homogenized using a mechanical homogenizer for another 5 min at 10,000 rpm. The two-step preparation of composites is documented in Figure 1. head 62 rpm) for 1 min, followed by high-speed mixing (paddle 285 rpm, mixing head 125 rpm) for 0.5 min. Then, the mixing was stopped to clean the mixer head and the sides of the mixing bowl. After that, mixing continued for another 1 min at high speed followed by casting of MOC-REF and MOC-G samples. The second step of homogenization was applied for MOC-G-H samples only. This composite mix was additionally homogenized using a mechanical homogenizer for another 5 min at 10,000 rpm. The two-step preparation of composites is documented in Figure 1. All samples were cast into steel molds and were exposed to a laboratory temperature of 23 ± 2 °C and relative humidity of 30 ± 5% for 4 h, followed by demolding and air-curing under the same laboratory conditions for 7 and 14 days, respectively.

Analytical Techniques
Morphology of the samples was studied by scanning electron microscopy (SEM), optical microscopy (OM), and high-resolution transmission electron microscopy (HR-TEM). The sample surface was determined by Brunauer, Emmett, and Teller (BET). The phase composition was studied via X-ray diffraction (XRD) and Fourier-transform infrared spectroscopy (FTIR). Rotation viscometry was used to measure the change in viscosity of fresh composite mixtures.

Results and Discussion
The microstructure of the typical graphene nanoplatelet provided by HR-TEM is shown in Figure S1 (SI). BET surface of graphene is introduced in Figure S2 (SI).
MOC-graphene composites and MOC reference without graphene were prepared. The samples were termed as MOC-REF (a reference sample with pure Magnesium Oxychloride Cement), MOC-G (Magnesium Oxychloride Cement with Graphene) and MOC-G-H (Magnesium Oxychloride Cement with Graphene Homogenized by Ultra-Turrax). The photograph of the prepared samples is shown in Figure 2. All samples were cast into steel molds and were exposed to a laboratory temperature of 23 ± 2 • C and relative humidity of 30 ± 5% for 4 h, followed by demolding and air-curing under the same laboratory conditions for 7 and 14 days, respectively.

Analytical Techniques
Morphology of the samples was studied by scanning electron microscopy (SEM), optical microscopy (OM), and high-resolution transmission electron microscopy (HR-TEM). The sample surface was determined by Brunauer, Emmett, and Teller (BET). The phase composition was studied via X-ray diffraction (XRD) and Fourier-transform infrared spectroscopy (FTIR). Rotation viscometry was used to measure the change in viscosity of fresh composite mixtures.
Among macrostructural parameters of the hardened composites, bulk density ρ b (kg·m −3 ), specific density ρ s (kg·m −3 ), and total open porosity Ψ (%) were measured. The microstructure of the formed composites was characterized by mercury porosimetry. Lastly, compressive strength f c (MPa) and flexural strength f f (MPa) were tested. The experimental details can be found in Supporting Information (SI). The expanded combined uncertainty of the applied test methods was as follows: ρ b 1.4%, ρ s 1.2%, Ψ 2.0%, f c 1.4%, f f 1.4%.

Results and Discussion
The microstructure of the typical graphene nanoplatelet provided by HR-TEM is shown in Figure S1 (SI). BET surface of graphene is introduced in Figure S2   The fracture surface of the composite samples was studied using optical microscopy. The micrographs (see Figure 3) show the difference between the composites containing graphene and the reference sample, where the color of the sample changes not only with the content of graphene but also with the rate of homogenization. The overall structure of the fracture surface is compact with no visible cracks or air bubbles.  Figure 4. The dependence of shear stress and velocity gradient conforms to the Bingham model of ideal plastic fluid [33] and for higher-velocity gradients to rheopectic fluid. For velocity gradient <50 s −1 , the effect of the addition of graphene on viscosity and thus rheological performance was negligible, which is promising for the design and development of multi-component MOC composites enriched with nanoadditives. It can be expected that the admixing of fillers and mineral admixtures will decrease the viscosity of the The fracture surface of the composite samples was studied using optical microscopy. The micrographs (see Figure 3) show the difference between the composites containing graphene and the reference sample, where the color of the sample changes not only with the content of graphene but also with the rate of homogenization. The overall structure of the fracture surface is compact with no visible cracks or air bubbles. The fracture surface of the composite samples was studied using optical microscopy. The micrographs (see Figure 3) show the difference between the composites containing graphene and the reference sample, where the color of the sample changes not only with the content of graphene but also with the rate of homogenization. The overall structure of the fracture surface is compact with no visible cracks or air bubbles.  Figure 4. The dependence of shear stress and velocity gradient conforms to the Bingham model of ideal plastic fluid [33] and for higher-velocity gradients to rheopectic fluid. For velocity gradient <50 s −1 , the effect of the addition of graphene on viscosity and thus rheological performance was negligible, which is promising for the design and development of multi-component MOC composites enriched with nanoadditives. It can be expected that the admixing of fillers and mineral admixtures will decrease the viscosity of the  Figure 4. The dependence of shear stress and velocity gradient conforms to the Bingham model of ideal plastic fluid [33] and for higher-velocity gradients to rheopectic fluid. For velocity gradient <50 s −1 , the effect of the addition of graphene on viscosity and thus rheological performance was negligible, which is promising for the design and development of multicomponent MOC composites enriched with nanoadditives. It can be expected that the admixing of fillers and mineral admixtures will decrease the viscosity of the composite mixtures. Therefore, the minimal influence of graphene nanoplatelets on the viscosity of examined composites is very beneficial. For lower-velocity gradients, the differences between the viscosity and shear stress of the tested fresh mixtures were well apparent. They can be assigned to the continuous precipitation of MOC constituents. In this case, the lowest shear stress, and thus the dynamic viscosity, exhibited the well homogenized mixture MOC-G-H. This is a positive finding that proves the applicability of the proposed homogenization technique in the production of MOC composites doped with graphene.
Materials 2021, 14, 1450 6 of 14 composite mixtures. Therefore, the minimal influence of graphene nanoplatelets on the viscosity of examined composites is very beneficial. For lower-velocity gradients, the differences between the viscosity and shear stress of the tested fresh mixtures were well apparent. They can be assigned to the continuous precipitation of MOC constituents. In this case, the lowest shear stress, and thus the dynamic viscosity, exhibited the well homogenized mixture MOC-G-H. This is a positive finding that proves the applicability of the proposed homogenization technique in the production of MOC composites doped with graphene. The phase composition of the 14-day samples was analyzed using XRD. The obtained diffraction patterns are shown in Figure 5. The only crystalline phase present in the pattern of all the samples was Mg3(OH)5Cl·4H2O (ICDD 04-014-8836), which corresponds with the stoichiometry of MOC Phase 5. Graphene was not visible in the diffraction patterns due to its very low content in the composites. Due to that, the diffraction patterns of all three samples were quite similar.
The microstructure of MOC-REF and prepared MOC composites combined with graphene was analyzed using SEM. The SEM micrographs of the prepared composites are shown in Figure 6. The prepared samples showed the presence of needle-shaped crystals typical for MOC-based materials. The typical dimensions of the needles were 1-6 µm in length and ~0.5 µm in thickness. Graphene was not visible because micrographs were obtained from the fracture surface. It was previously confirmed that graphene significantly improves the mechanical properties (flexural strength and compressive strength) [32]; therefore, the fracture occurred in areas with low graphene concentration.
The EDS elemental maps were measured to obtain the chemical composition for comparison of reference and prepared composites. EDS detected the presence of Mg, O, Cl, and C and confirmed the high purity of the prepared samples. Obtained composition was similar for all samples corresponding to the stoichiometry of pure Phase 5. All elemental maps are displayed in Figure 7. Relatively higher carbon content was detected, most probably due to the surface reaction with carbon dioxide forming chlorartinite [34].  The microstructure of MOC-REF and prepared MOC composites combined with graphene was analyzed using SEM. The SEM micrographs of the prepared composites are shown in Figure 6. The prepared samples showed the presence of needle-shaped crystals typical for MOC-based materials. The typical dimensions of the needles were 1-6 µm in length and~0.5 µm in thickness. Graphene was not visible because micrographs were obtained from the fracture surface. It was previously confirmed that graphene significantly improves the mechanical properties (flexural strength and compressive strength) [32]; therefore, the fracture occurred in areas with low graphene concentration.  The EDS elemental maps were measured to obtain the chemical composition for comparison of reference and prepared composites. EDS detected the presence of Mg, O, Cl, and C and confirmed the high purity of the prepared samples. Obtained composition was similar for all samples corresponding to the stoichiometry of pure Phase 5. All elemental maps are displayed in Figure 7. Relatively higher carbon content was detected, most probably due to the surface reaction with carbon dioxide forming chlorartinite [34].  The collected FT-IR spectra are presented in the form of spectral lines in Figure 8 and as a summarization of the major absorption band assignments in Table 2 The collected FT-IR spectra are presented in the form of spectral lines in Figure 8 and as a summarization of the major absorption band assignments in Table 2  The analyzed spectrum of graphene evidences only one weak band appearing at 1586 cm −1 , which can be attributed to the stretching vibration of the C=C bond [37][38][39]. In the case of MOC composites this band is shifted to higher wave numbers, specifically at 1611 cm −1 . This band also partially overlaps with the intensive band of the stretching vibration of Mg-O in MgCl2·8H2O. Based on this phenomenon the peak resolution was found to be at 1611 cm −1 in the case of MOC-G and MOC-G-H samples.   The analyzed spectrum of graphene evidences only one weak band appearing at 1586 cm −1 , which can be attributed to the stretching vibration of the C=C bond [37][38][39]. In the case of MOC composites this band is shifted to higher wave numbers, specifically at 1611 cm −1 . This band also partially overlaps with the intensive band of the stretching vibration of Mg-O in MgCl 2 ·8H 2 O. Based on this phenomenon the peak resolution was found to be at 1611 cm −1 in the case of MOC-G and MOC-G-H samples.
The macrostructural parameters of the hardened composites are shown in Table 3. In both curing ages, the influence of the graphene addition on the drop in porosity was visible. Moreover, the advanced mixing method enabled a further decrease in the porosity accessible for helium molecules [40]. The drop in porosity measured for 14-day cured samples was 21% for MOC-G and 33% for MOC-G-H, respectively, which clearly characterized the effectiveness of the proposed homogenization procedure. The differences in the bulk density and specific density were small, but in comparison with the reference material, the bulk density was typically higher for graphene-doped composites; contrary to that, the specific density was reduced by the incorporation of the nanoadditive into the composite mixture. Table 3. Basic structural parameters of the hardened composites. The pore size distribution curves obtained by mercury intrusion porosimetry are given in Figures 9 and 10. Apparently, the microstructural data correspond with the total porosity data presented in Table 3. The differences between the total porosity data obtained by MIP and calculated on the basis of the bulk density and specific density values are small, especially considering the samples' size and principals of the applied techniques. The macrostructural parameters of the hardened composites are shown in Table 3. In both curing ages, the influence of the graphene addition on the drop in porosity was visible. Moreover, the advanced mixing method enabled a further decrease in the porosity accessible for helium molecules [40]. The drop in porosity measured for 14-day cured samples was 21% for MOC-G and 33% for MOC-G-H, respectively, which clearly characterized the effectiveness of the proposed homogenization procedure. The differences in the bulk density and specific density were small, but in comparison with the reference material, the bulk density was typically higher for graphene-doped composites; contrary to that, the specific density was reduced by the incorporation of the nanoadditive into the composite mixture.  The pore size distribution curves obtained by mercury intrusion porosimetry are given in Figures 9 and 10. Apparently, the microstructural data correspond with the total porosity data presented in Table 3. The differences between the total porosity data obtained by MIP and calculated on the basis of the bulk density and specific density values are small, especially considering the samples' size and principals of the applied techniques.   The microstructural parameters of 14-day cured samples are summarized in Table 4. The pore volume visibly decreased with the addition of graphene, but the average pore size was slightly higher for graphene-enriched materials, compared to the reference composite MOC-REF. This was due to the increase in the volume of pores in the diameter range 100-10 µm, and the decrease in the volume of nanometer-sized pores (5-35 nm).  The investigated mechanical parameters are presented in Figure 11. As a potential construction material, MOC composites exhibited high compressive strength and flexural strength as similarly presented, e.g., by Misra et al. [41], Góchez et al. [42], and most recently Lauermannová et al. [43]. The effect of graphene addition on the mechanical resistance of the developed composites is well apparent. The use of the nanoadditive in the form of graphene nanoplatelets greatly improved both the compressive and flexural strength, which corresponded well with the total porosity data and pore size distribution. The two-set homogenization procedure led to the increased growth of the compressive strength. On the other hand, the improvement of the flexural strength due to the modified homogenization procedure was not great, but still sufficient for the practical use of graphene-reinforced MOC materials in construction practice. Obviously, the two-step homogenization procedure improved the distribution of graphene particles in The microstructural parameters of 14-day cured samples are summarized in Table 4. The pore volume visibly decreased with the addition of graphene, but the average pore size was slightly higher for graphene-enriched materials, compared to the reference composite MOC-REF. This was due to the increase in the volume of pores in the diameter range 100-10 µm, and the decrease in the volume of nanometer-sized pores (5-35 nm). Table 4. Microstructural parameters of the 14-day composites.

MOC-REF MOC-G MOC-G-H
Total pore volume (cm 3 ·g −1 ) 0.0844 0.0644 0.0622 Average pore diameter (µm) 0.0295 0.0317 0.0320 The investigated mechanical parameters are presented in Figure 11. As a potential construction material, MOC composites exhibited high compressive strength and flexural strength as similarly presented, e.g., by Misra et al. [41], Góchez et al. [42], and most recently Lauermannová et al. [43]. The effect of graphene addition on the mechanical resistance of the developed composites is well apparent. The use of the nanoadditive in the form of graphene nanoplatelets greatly improved both the compressive and flexural strength, which corresponded well with the total porosity data and pore size distribution. In comparison with the reference composite MOC-REF, the increase in the compressive strength was remarkable in both curing ages. For the 7-day samples, the increase in the compressive strength was 34.6% for MOC-G, and 53.6% for MOC-G-H, respectively. In the case of 14-day matured samples, the increase in the compressive strength values was 28. authors, who reported the strength development of MOC mixtures with curing age [44]. This was due to the solidification of the precipitated MOC phases, i.e., the free intergrowth of crystal phases and, thus, densifying of their microstructure [45]. Nevertheless, in the estimation of the mass transport in the researched materials, and thus their durability, one must consider the fact that the porosity essentially affects the mechanical strength, whereas the transport processes depend on the structure and the size distribution of the pores [46].

Conclusions
The effect of the mixture homogenization procedure in the preparation of magnesium oxychloride cement paste doped with graphene nanoplatelets on its fresh and hardened properties was studied. The addition of graphene and the process of two-step homogenization resulted in the following aspects: 1. The addition of graphene did not increase the viscosity of the fresh mixtures, which is positive for their presumed application in the design and development of the advanced types of multi-component MOC composites for construction applications. 2. Both presented mixing procedures designed for MOC mixtures with graphene enabled the production of materials with decreased porosity and greatly improved mechanical resistance, in comparison with the ordinary MOC paste. 3. The two-step mixing technique even improved the structural and mechanical parameters acquired for MOC-graphene composite homogenized using the one-step technology.
This represents very promising results for the future research and prospective application of carbon-based nanoadditives in the production of building materials. The two-set homogenization procedure led to the increased growth of the compressive strength. On the other hand, the improvement of the flexural strength due to the modified homogenization procedure was not great, but still sufficient for the practical use of graphenereinforced MOC materials in construction practice. Obviously, the two-step homogenization procedure improved the distribution of graphene particles in MOC fresh mixture, which consequently resulted in the upgraded parameters of the hardened composites.
For all researched composites, the prolonged curing time significantly increased both observed mechanical parameters. This is in agreement with the findings of other authors, who reported the strength development of MOC mixtures with curing age [44]. This was due to the solidification of the precipitated MOC phases, i.e., the free intergrowth of crystal phases and, thus, densifying of their microstructure [45]. Nevertheless, in the estimation of the mass transport in the researched materials, and thus their durability, one must consider the fact that the porosity essentially affects the mechanical strength, whereas the transport processes depend on the structure and the size distribution of the pores [46].

Conclusions
The effect of the mixture homogenization procedure in the preparation of magnesium oxychloride cement paste doped with graphene nanoplatelets on its fresh and hardened properties was studied. The addition of graphene and the process of two-step homogenization resulted in the following aspects: 1.
The addition of graphene did not increase the viscosity of the fresh mixtures, which is positive for their presumed application in the design and development of the advanced types of multi-component MOC composites for construction applications.

2.
Both presented mixing procedures designed for MOC mixtures with graphene enabled the production of materials with decreased porosity and greatly improved mechanical resistance, in comparison with the ordinary MOC paste.

3.
The two-step mixing technique even improved the structural and mechanical parameters acquired for MOC-graphene composite homogenized using the one-step technology.
This represents very promising results for the future research and prospective application of carbon-based nanoadditives in the production of building materials.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to privacy.