Super-hydrophobic magnesium oxychloride cement (MOC): from structural control to self-cleaning property evaluation

In this paper, magnesium oxychloride cement (MOC), which has a needle-like structure, is upgraded with super-hydrophobic surface using a facile method involving immersion in a FAS-ethanol solution. The influence of the molar ratios of the raw materials on the super-hydrophobic property was investigated. The phase compositions, microstructure, compressive strength, water resistance and wetting behaviour are studied in detail by X-Ray diffraction, scanning electron microscopy, a water contact angle measurement instrument, and mechanical testing. The water contact angle of as-prepared MOC reaches 152 ± 1° for the optimal mix design. The variation of the water contact angle of different mixes can be explained by the Cassie–Baxter model. The experiments using rolling off dust on the super hydrophobic surface present excellent self-cleaning ability. Moreover, proposed super hydrophobic surface exhibited excellent UV-durability, indicating a promising potential for the outdoor application.


Introduction
Magnesium oxychloride cement (MOC) has attracted extensive attention, as it is generally associated with ambient temperature curing and excellent materials properties such as fast setting, high mechanical strength, good resistance to abrasion and fire [1][2][3][4][5]. The behavior of MOC is mainly regulated by the composition and microstructure of the hydration products, produced through neutralization-hydrolysis-crystallization process in the ternary system of MgO-MgCl 2 -H 2 O [1,6]. At room temperature, the mechanism of formation of magnesium chloride hydrates in the MgO-MgCl 2 -H 2 O system can be described by the following chemical reactions (Eqs. 1-3) [6,7] Changing the proportion of MgO, MgCl 2 and H 2 O in this reaction system, such as molar ratios of MgO/ MgCl 2 and H 2 O/MgCl 2 , can lead to significant differences concerning chemical compositions and consequently different materials performance of MOC [2]. Specifically, Phase 5 (P5) has been commonly recognized as the most desirable reaction product in MOC-based composites, as it is believed that the Phase 5 crystals can provide the best mechanical properties [2,3,6]. Generally, a molar ratio of over five for MgO/MgCl 2 is often used for manufacturing the MOC-based product [2,6]. As a consequence, the hydration products usually consist of more P5 and some Mg(OH) 2 .
However, when exposed to water, Phase 5 will be easily decomposed to Mg(OH) 2 that leads to a poor water resistance, which seriously restricts its outdoor applications. An efficient and facile way to improve the water resistance of MOC is through controlling the molar ratios of MgO/MgCl 2 and H 2 O/MgCl 2 . Li et al. [2] investigated the influence of the molar ratios of raw materials on the properties of magnesium oxychloride cement. They found that there are three different structures of Phase 5 specified as plate-like, needlelike and gel-like crystals. Among them, needle like structure shows a better performance in terms of mechanical properties. However, as the packing of the needle-like structures resulting in more space for the water to penetrate, MOC rich in need-like Phase 5 always results in a poor water resistance and leads to a very low strength of MOC at ambient temperature after 28 days in water [8]. Another efficient method is to apply additives to improve the water resistance of MOC. Li and Yu [9] found that after the addition of active SiO 2 , the water resistance of MOC was increased, owing to the generation of new hydration products (S-1 gel) that lead to a denser matrix. Deng et al. [8] discovered that the addition of phosphate could increase the softening coefficient of 14 days' cured samples to above 0.8 after immersed in water for 15 days as more gel-like P5 is formed which increases the tortuosity of the MOC.
Preventing the water penetration by generating super-hydrophobic surfaces has already led to various new hybrid materials with astonishing properties [10][11][12]. A surface can be described as super hydrophobic if the contact angle (CA) of water on this surface is larger than 150° [13,14]. The best known example of super hydrophobic surface is the lotus leaf, on which water easily rolls off, leaving little or no residue and carrying away surface contamination, therefore showing excellent self-cleaning properties [10,15]. With the help of scanning electron microscopy (SEM), the mechanism behind this phenomenon has been resolved. The electron microscopy of the surface of lotus leaves shows protruding nubs of 20-40 lm apart each covered with a smaller scale rough of epicuticular wax crystalloids [12]. This wax only consists of -C-H and -C-O groups, which give the origin to the super hydrophobicity and selfcleaning properties of the lotus leaf. Since Tsujii et al. first fabricated biomimetic surfaces in the mid-1990s, numerous smart and efficient methods for attaining rough surfaces to prepare super hydrophobic surface have been reported [12]. In summary [16], there are two main kinds of surface microstructures to prepare super hydrophobic surface: one is the hierarchical micro-and nanostructure like the lotus leaf and the other one is the unitary micro-line structure like the ramee leaf.
Considering the complexity of the preparation of micro-, nano-, and lotus-type micro-/nanocomposite structures, unitary micro-line structure provides a facile solution to prepare super hydrophobic surfaces [16,17]. Ogawa et al. prepared super hydrophobic film-coated nano-fibrous membranes based on the inspiration of self-cleaning silver ragwort leaves. After the modification with fluoroalkylsilane (FAS), the surface of the nano-fibrous membrane presented super hydrophobicity with the highest water contact angle of 162°and the lowest water-roll angle of 2° [ 18]. Li et al. reported a simple solution-immersion method to fabricate a super hydrophobic surface on a cellulose-based material (i.e. cotton fabric or paper). The surface morphology of the fabric provided ideal roughness for trapping the air to build a super hydrophobic surface [19]. Jiang et al. [20] prepared Cu(OH) 2 nano-needle arrays presenting little contactangle hysteresis, even under certain hydrostatic pressures, by a facile chemical-base deposition method. The average length of the needles was 5 lm and the average diameter of the nano-needle is 300 nm. However, due to the restriction of the microstructure of the cement-based materials, fabrication of super hydrophobic surface with cementbased materials is still limited..
The present study aims to propose a novel method for the fabrication of super hydrophobic surface on MOC by applying the unique needle-like structure Phase 5 of MOC. The procedure is fairly facile to carry out and no special technique or equipment is required, so it is very cost-effective. The compressive strength and water resistance of the MOC samples were studied. The phase composition and microstructural properties are determined by using X-ray diffraction (XRD) and scanning electron microscopy (SEM). The water-repellency ability was evaluated by performing the water contact angle and water sliding angle tests using a Dataphysics OCA20 contact-angle system. As UV-durability property, which is crucial for practical outdoor use, mainly depends on the surface chemical composition and surface structures. As such UVdurability of the as prepared surface is also examined. The results show that the designed super MOC surface possesses a super hydrophobic property, together with an excellent water-repellency and self-cleaning ability.

Materials
The starting raw materials for preparing the MOC cement paste included light-burnt magnesia powder (MgO), magnesium chloride hexahydrate (MgCl 2-6H 2 O), and tap water. Both the MgO and MgCl 2 Á6H 2 O were provided by Sinopharm Chemical Reagent, China.

Preparation of the super hydrophobic MOC samples
The magnesium chloride hexahydrate was dissolved in water before use. The MOC samples were prepared by mixing MgO and the MgCl 2 Á6H 2 O solution. To find out an appropriate formulation of the MOC to provide sufficient surface roughness, MOC cement pastes using MgO and MgCl 2 Á6H 2 O with different molar ratios were designed, as shown in Table 1. The paste is cast into the steel moulds applying a normal cement paste mixing step following EN 196-1 [21]. Based on the mixed recipe, the samples are designated as MOC-510, MOC-68 and MOC-610. To prepare hydrophobic MOC samples, after 48 h' curing in the air, MOC products were placed in the FAS ethanol 5% solution at room temperature for 24 h. Subsequently they were dried at room temperature for 1 h.

Characterization
The XRD analysis was performed by using a Bruker D8 advance powder X-ray diffractometer with a Cu tube (20 kV, 10 mA) with a scanning range from 5°to 65°(2h), applying a step 0.02°and 0.2 s/step measuring time. The powder sample was prepared by crushing and grinding the solid MOC cement paste with a ball milling equipment, and then passing it through a sieve with a screen aperture of 75 lm. A contact angle meter DataPhysics SCA20 (DataPhysics Germany) was used to measure the static water contact angle and sliding angle with 10 ll water drop at the ambient temperature. The morphologies of the MOC were observed using a field emission scanning electron microscopy (FESEM) of Quanta 450 (FEI USA) with the operation voltage of 5 kV and spot size was set to 5.0.

Self-cleaning behavior evaluation
The wetting behavior of the samples was assessed in a simple qualitative manner by placing droplets of water

UV durability evaluation
The UV durability test under different environmental conditions was performed in a set-up as shown in Fig. 1. In the present work, the experimental setup is composed of a light source, transport gas (air) supply, flow rate valves, relative humidity meter valve, and parameter measurement apparatus like temperature and relative humidity. The applied light source consists of three ultraviolet lamps of 25 W each, emitting an ultraviolet radiation (UVA) in the range of 300-400 nm. The irradiance can be adjusted by a light intensity controller. The light intensity is measured with UVA radiometer. The light intensity E is kept constant at 14 W/m 2 to simulate outdoor conditions. The summary of the test conditions is shown in Table 2.
3 Results and discussion

Crystalline composition
The XRD patterns of the MOCs are presented in Fig. 2. It should be noted that for all the MOC specimens, the mineralogical phases consist of major Phase 5, minor Mg(OH) 2 and unreacted MgO and MgCO 3 . This is in agreement with the previous studies [2,4,6]. As the Phase 5 is the most favourable crystalline phase [2,6], the molar ratio of MgO/MgCl 2 is desired to be higher than 5 and the molar ratio of H 2 O/MgCl 2 is in the range of 6-15. However, comparing the XRD patterns shown in    Figure 3 shows the SEM micrographs of the MOC samples. It is clear that all the samples present a composite structure consisting of needle-like Phase 5 and gel-like Phase 5. Figure 3a shows the micrographs of the MOC-510, which demonstrates that the MOC surface has an uneven structure. The morphology consists of gel-like Phase 5 and needle-like P5 crystal structures. Because of the differences between the recipes of the raw materials, the proportions of the Phase 5 with different structures are different in these three mixes. It can be seen that MOC-510 presents more gel-like Phase 5 as compared to the other two samples. This can be attributed to the smaller amount of MgO in the reaction system. As stated in [2,6], MgO particles could act as a role of reaction seeds, which in turn increase the reaction sites in the MOC hydrated process. So less MgO seeds would increase the possibility of group growth of the MOC, which leads to more gel-like Phase 5. It can be observed that MOC-610 exhibits more needle-like Phase 5 than MOC-68 as shown in Fig. 3b, c, which is attributed to the higher amount of the free water in MOC-610. It has been reported in [2,6,22] that the free water will react with the MgO during the hardening process. Hence, the more water the system contains, the more interspaces would be produced in the matrix after hardening of the MOC is completed.

Compressive strength
The compressive strengths of different mixtures after air curing for 3 days, 7 days and 14 days were determined, as shown in Fig. 4. Furthermore, the compressive strength changes after the immersion in water for 48 h are compared and shown in Fig. 5 [2,6]. It should be noted that MOC-510 shown the weakest mechanical property. This can be attributed to the higher free water amount in its reaction system. As reported by [2,23], the strength of MOC increases with the decrease of molar ratio of H 2 O/MgCl 2 with a fixed molar ratio of MgO/MgCl 2 , while a higher molar ratio of MgO/MgCl 2 results in a higher strength with a fixed molar ratio of H 2 O/MgCl 2 . However, it should be noticed that the compressive strength of MOC-610 is almost the same as MOC-68 whose molar ratio of H 2 O/MgCl 2 is lower. This is because the mechanical performance of MOC is highly related to the microstructure of the hydration product in MOC. It can be seen from the SEM images of Fig. 3, the formed needle crystals of MOC-610 are long and thin and highly intergrown. Li [24] has reported that this unique structure is beneficial to disperse the strength and lead to superior mechanical strengths.
The water resistance improvement of MOC pastes after super hydrophobic modification was performed can be seen from the compressive strength changes after soaking samples for 48 h, as shown in Fig. 5a, b. It can be seen that the super hydrophobic modification significantly improves the water resistance of MOC. Without the modification, the compressive strengths of MOC-68, MOC-610 and MOC-510 is only 9.3 MPa and 9 MPa and 9.5 MPa, respectively after soaking for 48 h. The compressive strength retentions of the MOC-68, MOC-610 and MOC-510 are only 17.1%, 17.6% and 21.4%, respectively. This is due to the decomposition of the hydration products into Mg(OH) 2 in water, as reported by [6,25]. It should be noted that the compressive strength retention of MOC-510 is higher than the other two samples. This is due to more gel-like Phase 5 in MOC-510, which has been reported to efficiently improve the water resistance of MOC [2]. Owing to the super hydrophobic treatment, all of the MOC samples have a much higher compressive strength retention than unmodified samples. The compressive strength retentions of MOC-68, MOC-610 and MOC-510 after soaking 48 h were 91.3%, 99% and 91.2%, respectively, indicating the hydrophobic treatment is highly effective in terms of improving water resistance of MOC.

Hydrophobic performance
The water contact angle (CA) was used to evaluate the surface wettability of the designed MOC mixtures and the measured CAs are shown in Fig. 6. It can be seen that contact angles measured on the surfaces of different MOC pastes are different. As shown in Fig. 6a, the MOC-510 has a water contact angle about 130°, indicating a hydrophobic property [12,13]. After the increase of the MgO content, the CA of the MOC-610 increased to about 152°as shown in Fig. 6c, indicating that the MOC surface becomes super hydrophobic. On the contrary, the water repellent capability of MOC-68 decreases and the CA decreases to about 143°after lowering the amount of MgO.
The change of the surface microstructures of different recipes (Fig. 3a-c) is believed to be responsible for the increase of CA. According to the ideal Young equation: where c l-g , c s-g , and c l-s represent the liquid-gas, solidgas, and liquid-solid interfacial tensions, respectively, and h y is the contact angle, h cb is the apparent contact angle on the rough surface. Equation (5) describes the apparent contact angle in the heterogeneous regime [26,27]. The roughness ratio (r f ) is defined as the ratio of the true area of the surface to its projection area. The variable f is defined as the fraction of the project area of the surface that is wetted by water. For super hydrophobic MOC, air bubbles are enclosed in the pores among the hierarchical crosslinking need-like Phase 5 structures. Without the surface modification, water can easily spread on MOC as the P5 is very hydrophilic. Hence, the value of r l-s is very small in ideal Young equation. The value of h y ranges from 0 to 90°and the value of cos h y ranges from 0 to 1. In the Cassie-Baxter equation, both r f and f range from 0 to 1 [12,26]. The rougher the surface is, the smaller the values of r f and f are [26]. As a result, on a rough surface, the value of cosh cb vary between -1 and 0 and the value of h cb between 90°and 180°. Hence, a rougher surface will lead to a larger contact angle and more hydrophobic behaviour. When gel-like Phase 5 is generated among the needle-like structure, it will fill in the interspace that in turn decreases the air amount for air trapping. This can be used to explain the decrease of super hydrophobic ability from MOC-610 to MOC-510 as the amount gel-like Phase 5 increases. Nano-needle structures have been reported beneficial to fabricate the super hydrophobic surface [20]. Such a unique structure offers two advantages: (1)  needle structures are likely to minimize the contact area between the solid and liquid interface, and their remarkable length prevents contact area between the MOC surface and water, (2) the crosslinking structure can effectively trap air and vigorously support the enwrapped liquid-air interface. It is concluded that MOC products with needle-like Phase 5 are ideal raw materials to fabricate super hydrophobic surfaces. Furthermore, it has been reported that needle-like Phase 5 presents a better mechanical performance than gel-like or plate-like Phase 5 [2,6]. This design thus favours both super hydrophobicity and the mechanical strength of MOC products.
3.5 Self-cleaning and UV-resistance ability Barthlott et al. reported that water droplets can roll off from the lotus leaves and remove the dust particles [14,28,29]. As MOC products are often used for a decorative purpose (e.g. decoration boards), it will be very meaningful for MOC possessing the self-cleaning ability. Figure 7a shows the results of rolling water on the super hydrophobic MOC surface (MOC-610).
UV-resistivity which is crucial for practical outdoor applications, mainly depends on the surface chemical composition and surface structures [30][31][32][33]. According to [34,35], low-energy coating modification with a long chain organic molecule, such as FAS, is one possible solution to achieve the goal of fabricating UV-durable super hydrophobic surface. In order to assess the UV-durability of the as-prepared super hydrophobic surface, the samples were exposed to UV light for the period of 7 days at different temperatures and humidity levels, and the wetting behaviours were measured every 30 min. The static contact angle and sliding angle as a function of UV-irradiation time are presented in Fig. 8. It is shown that after 7 days' UV irradiation, the surface still exhibits a contact angle of 152.1°and a sliding angle of 8°at 25°C and 50% humidity and a contact angle of 152.8°and a sliding angle of 9°at higher temperature. Even under a higher humidity, the MOC surface presents a contact angle of 152.3°and a sliding angle of 8°, suggesting a superior UV-durability.
The results here confirm that the as-prepared MOC super hydrophobic surface is very stable under strong UV irradiation. This can be attributed to both the properties of the MOC and the low surface energy of FAS. It has been widely reported that MOC possesses high strength and stability used in dry conditions [2,6]. The rough needle-like structure of the MOC matrix is therefore stable. Secondly, FAS, which has been applied on different kinds of substrate such as metal, ceramic and polymers to fabricate UV-durable super hydrophobic surface [11], is very stable. This can be attributed to the long chain of the FAS coated on the surface, which can provide a large amount of C-F bonds. It should be noted that the applied chemicals as well as the modification methods are rather economically effective and simple, which make the large scale industrial production relatively  feasible. The bond energy is 485 kJ/mol that cannot be broken by the UV light ranging from 314 to 419 kJ/mol [30]. Hence, it is necessary to apply low surface energy chemicals with long C-F chain to prepare UV durable super hydrophobic surface with cement-based materials.

Conclusions
A super hydrophobic MOC surface with excellent selfcleaning ability was fabricated applying a facile solution immersion method with FAS. The influences of the molar ratio on the microstructure, compressive strength, water resistance and hydrophobic performance were investigated. The self-cleaning ability and UV-durability were characterized. Based the presented results, the following conclusions can be reached: