Mechanical Properties and Water Stability of High Ductility Magnesium Phosphate Cement-Based Composites (HDMC)

In this study, the compressive test and four-point flexural test were carried out to explore the water stability as well as mechanical properties of high ductility magnesium phosphate cement-based composites (HDMC). The effects of ambient curing age (7 d and 28 d), water immersion age (7 d, 28 d, and 56 d), water/binder ratio (W/B), and magnesium oxide/potassium dihydrogen phosphate ratio (M/P) on the mechanical properties (compressive strength, first-crack strength, ultimate flexural strength, ductility index, and toughness index) and water stability of the HDMC were examined. The results showed that the 28-day ambient curing could lead to higher retention rates of strength, ductility, and toughness than 7-day ambient curing, indicating better water stability; however, it did not result in significant improvement in the mechanical properties of the HDMC. As the water immersion age increased, the mechanical properties of the HDMC with 7-day ambient curing showed an obvious downward trend; the mechanical properties of the HDMC with 28-day ambient curing did not show an obvious decrease and even could be increased in many cases, especially when the water immersion age was 56 days; and the change of water stability was consistent with that of the mechanical properties. If all indexes and their corresponding retention rates were considered comprehensively, the W/B ratio of 0.16 and the M/P ratio of 5 seemed to be the optimum values for the HDMC. The scanning electron microscopy analysis confirmed that the water immersion had a large adverse effect on the HDMC and thus reduced their mechanical properties.


Introduction
Magnesium phosphate cement (MPC) refers to an inorganic cementitious material that is produced through the acid-base neutralization reaction of MgO and phosphate [1][2][3]. In previous studies, dead-burned MgO and ammonium phosphate were used as raw materials to produce magnesium ammonium phosphate cement, but the release of ammonia during the hydration process hindered its wide application. Then researchers used potassium dihydrogen phosphate to replace ammonium phosphate to produce MPC. The MPC has excellent characteristics such as fast setting and hardening, high early strength [4], high bonding strength with old concrete [5], small drying shrinkage [5], solidification of liquid wastes containing heavy metals [6], excellent wear resistance [7], frost resistance performance [8], high-temperature resistance [9,10], etc. It has a promising engineering application prospect for highways, bridges, airport runways, and other places with rapid repair needs. It also has research prospects for 3D printing [11], alternative biomaterials for hard tissue applications [12], fiber-reinforced inorganic polymer composites [13], etc.
Similar to ordinary Portland cement-based materials, the MPC-based composites have the characteristics of brittleness and low strain capacity. The addition of fibers can increase the strength and toughness of the MPC-based composites and reduce the number and width of matrix cracks due to their bridging effect. It was reported that the addition of Two kinds of industrial-grade potassium dihydrogen phosphates (P) in the form of white crystalline powder were mixed with a mass ratio of 1:1 and then were used as an acidic compound. The particle size ranges of these two potassium dihydrogen phosphates were 180-315 µm and 425-600 µm, which were denoted as fine P and coarse P, respectively. The borax (B) was used as a cement retarder to delay the setting time. The physical properties of the M, the P, and the B are summarized in Table 2. Note: the missing content has not been tested and provided by the manufacturer.
The quartz sand with a particle size range of 120-180 µm was used as fine aggregates, with its technical parameters shown in Table 3. The first-grade fly ash (FA) was selected, with its chemical composition shown in Table 4. The PVA fibers used were produced by Kuraray Company of Japan, with their specific performance indicators shown in Table 5. The particle size distribution of MgO, quartz sand, and FA is shown in Figure 1. The particle size distribution of KH 2 PO 4 is shown in Figure 2.

Mix Proportion
Based on the experimental study in the authors' research group, when the mass ratio of sand to binder (MPC + FA), denoted as S/B, is 0.2 and the borax dosage is 6% by mass of MgO, the mechanical properties and ductility of specimens are relatively good. Therefore, in this study, for all specimens, these two numbers are selected. The PVA fiber

Mix Proportion
Based on the experimental study in the authors' research group, when t of sand to binder (MPC + FA), denoted as S/B, is 0.2 and the borax dosage i of MgO, the mechanical properties and ductility of specimens are rel

Mix Proportion
Based on the experimental study in the authors' research group, when the mass ratio of sand to binder (MPC + FA), denoted as S/B, is 0.2 and the borax dosage is 6% by mass of MgO, the mechanical properties and ductility of specimens are relatively good. Therefore, in this study, for all specimens, these two numbers are selected. The PVA fiber volume ratio is 1.6%, and the FA is used to replace 20% (by mass) MPC. A reference mix proportion is adopted, in which the mass ratio of water to binder (MPC + FA), denoted as W/B, is 0.16, and the molar ratio of MgO to KH 2 PO 4 , denoted as M/P, is 4. To explore the effects of ambient curing age (duration in the ambient curing room) and water immersion age (duration in water following the ambient curing) on the HDMC, two ambient curing ages (7 days and 28 days) and three water immersion ages (7 days, 28 days, and 56 days) are considered on the reference mix proportion (W/B = 0.16 and M/P = 4). To explore the effects of W/B and M/P on the HDMC, four W/B ratios (0.14, 0.16, 0.18, and 0.20) and four M/P ratios (3, 4, 5, and 6) are considered while both the ambient curing age and the water immersion age are 7 days. The mix proportions are shown in Table 6.

Specimen Preparation
Firstly, the weighed magnesium oxide, potassium dihydrogen phosphate, quartz sand, borax, and fly ash were added to the single-horizontal shaft concrete mixer and mixed for 120 s to make the mixture uniform. Secondly, water was added and mixed for 120 s, during which PVA fibers were added several times to ensure their uniform dispersion. Thirdly, the mixture continued to be mixed for another 120 s after the addition of PVA fibers. Lastly, the mixture was cast into 50 × 50 × 50 mm 3 plastic molds and 400 × 100 × 15 mm 3 steel molds to prepare cubic specimens for compressive test and thin plate specimens for four-point flexural test, respectively.
The cast specimens were stored in the curing room (temperature: 25 ± 2 • C, relative humidity: 45 ± 5%) for 2h, after which the molds were removed. After that, the specimens continued to be stored in the curing room to a certain age (7 days or 28 days) and then were immersed in a water tank for a certain age (7 days, 28 days, or 56 days).

Compressive Test
The compressive test was carried out on the cubic specimens (50 × 50 × 50 mm 3 ) according to ASTM C109 [44]. Three samples were tested for each group, and the average value was adopted. The test instrument is a pressure testing machine (200T), as shown in Figure 3, with a loading speed of 0.9 kN/s.

Four-Point Flexural Test
The four-point flexural test was conducted on the thin plate specimens, as shown in Figure 4. The specimen size was selected based on the study from Xu et al. [45]. The loading speed was adjusted to 0.3 mm/min for continuous loading. The collected midspan deflection values and the corresponding load values of the specimens were plotted into a load-deflection curve.

Scanning Electron Microscopy (SEM) Test
The scanning electron microscopy (SEM) test was conducted to examine the samples at the microscopic level. The samples were broken and cut into 1 mm 2 size using a precision gas cutting machine and then put into the field emitted scanning electron microscope (Zeiss, Oberkochen, Germany) for the SEM test. The effects of the water immersion on the matrix and its bonding with fibers were analyzed by observing the changes in the microstructure of the samples. Figure 5 shows a typical strain hardening curve of a fiber-reinforced cement-based composite material. The first cracking point is the position where the load-deflection curve changes from linear to nonlinear, which is called the limit of proportion (LOP) according to ASTM C1018 [46]. The load and deflection at the LOP are called cracking load (PLOP) and cracking deflection (δLOP), respectively. The modulus of rupture (MOR) is defined as the point at which the load-deflection curve begins to soften. The load and deflection at the MOR are called ultimate load (PMOR) and ultimate deflection (δMOR), respectively.

Four-Point Flexural Test
The four-point flexural test was conducted on the thin plate specimens, as shown in Figure 4. The specimen size was selected based on the study from Xu et al. [45]. The loading speed was adjusted to 0.3 mm/min for continuous loading. The collected mid-span deflection values and the corresponding load values of the specimens were plotted into a load-deflection curve.

Four-Point Flexural Test
The four-point flexural test was conducted on the thin plate specimens, as shown in Figure 4. The specimen size was selected based on the study from Xu et al. [45]. The loading speed was adjusted to 0.3 mm/min for continuous loading. The collected midspan deflection values and the corresponding load values of the specimens were plotted into a load-deflection curve.

Scanning Electron Microscopy (SEM) Test
The scanning electron microscopy (SEM) test was conducted to examine the samples at the microscopic level. The samples were broken and cut into 1 mm 2 size using a precision gas cutting machine and then put into the field emitted scanning electron microscope (Zeiss, Oberkochen, Germany) for the SEM test. The effects of the water immersion on the matrix and its bonding with fibers were analyzed by observing the changes in the microstructure of the samples. Figure 5 shows a typical strain hardening curve of a fiber-reinforced cement-based composite material. The first cracking point is the position where the load-deflection curve changes from linear to nonlinear, which is called the limit of proportion (LOP) according to ASTM C1018 [46]. The load and deflection at the LOP are called cracking load (PLOP) and cracking deflection (δLOP), respectively. The modulus of rupture (MOR) is defined as the point at which the load-deflection curve begins to soften. The load and deflection at the MOR are called ultimate load (PMOR) and ultimate deflection (δMOR), respectively.

Scanning Electron Microscopy (SEM) Test
The scanning electron microscopy (SEM) test was conducted to examine the samples at the microscopic level. The samples were broken and cut into 1 mm 2 size using a precision gas cutting machine and then put into the field emitted scanning electron microscope (Zeiss, Oberkochen, Germany) for the SEM test. The effects of the water immersion on the matrix and its bonding with fibers were analyzed by observing the changes in the microstructure of the samples. Figure 5 shows a typical strain hardening curve of a fiber-reinforced cement-based composite material. The first cracking point is the position where the load-deflection curve changes from linear to nonlinear, which is called the limit of proportion (LOP) according to ASTM C1018 [46]. The load and deflection at the LOP are called cracking load (P LOP ) and cracking deflection (δ LOP ), respectively. The modulus of rupture (MOR) is defined as the point at which the load-deflection curve begins to soften. The load and deflection at the MOR are called ultimate load (P MOR ) and ultimate deflection (δ MOR ), respectively.  (1) Strength index
where PLOP is the cracking load (N), PMOR is the ultimate load (N), L is the span le the thin plate specimens (mm), and b and h are the width and height of the thi specimens (mm), respectively.
(2) Ductility index The ductility index (DI) is defined as the difference between the ultimate de (δMOR) and the cracking deflection (δLOP), as shown in Equation (3). The grea difference is, the better the ductility of specimens can be.
where δMOR is the ultimate deflection (mm), and δLOP is the cracking deflection (mm The toughness index (Tpeak) is defined as the difference between the area un load-deflection curve up to the ultimate deflection and the area up to the c deflection, as shown in Equation (4). The larger the difference is, the better the tou of specimens can be.
where SMOR is the area under the load-deflection curve up to the ultimate deflection (k and SLOP is the area under the load-deflection curve up to the cracking deflection (kN• (1) Strength index The first-crack strength, f LOP (MPa), and ultimate flexural strength, f MOR (MPa), are calculated using Equations (1) and (2), respectively.
where P LOP is the cracking load (N), P MOR is the ultimate load (N), L is the span length of the thin plate specimens (mm), and b and h are the width and height of the thin plate specimens (mm), respectively.
(2) Ductility index The ductility index (DI) is defined as the difference between the ultimate deflection (δ MOR ) and the cracking deflection (δ LOP ), as shown in Equation (3). The greater the difference is, the better the ductility of specimens can be.
where δ MOR is the ultimate deflection (mm), and δ LOP is the cracking deflection (mm). (

3) Toughness index
The toughness index (T peak ) is defined as the difference between the area under the load-deflection curve up to the ultimate deflection and the area up to the cracking deflection, as shown in Equation (4). The larger the difference is, the better the toughness of specimens can be.
where S MOR is the area under the load-deflection curve up to the ultimate deflection (kN·mm), and S LOP is the area under the load-deflection curve up to the cracking deflection (kN·mm).

Water Stability Evaluation
Water stability was evaluated by defining the strength retention rate, ductility retention rate, and toughness retention rate. The larger the value is, the better the water stability of specimens is.
The strength retention rate (k c ) is calculated using Equation (5).
where f c1 and f c2 are the strength (including compressive strength, first-crack strength, and ultimate flexural strength) (MPa) of the specimens without water immersion and with water immersion, respectively. The ductility retention rate (k DI ) is calculated using Equation (6).
where DI 1 and DI 2 are the ductility indexes of the thin plate specimens without water immersion and with water immersion, respectively. The toughness retention rate (k T ) is calculated using Equation (7).
where T 1 and T 2 are the toughness indexes of the specimens without water immersion and with water immersion, respectively. Figure 6a,b shows the compressive strength and compressive strength retention rate of the HDMC with 7-day and 28-day ambient curing, respectively. For non-immersed HDMC, the 7-day compressive strength is 44.8MPa, which is slightly higher than the 28-day compressive strength (40MPa). Also, for immersed HDMC, regardless of water immersion age, the 7-day ambient curing can result in higher compressive strength than the 28-day ambient curing. Figure 6 also shows that as the water immersion age increases from 0 to 56 days, the compressive strength of the HDMC with 7-day ambient curing shows an obvious downward trend; however, the compressive strength of the HDMC with 28-day ambient curing decreases to a small extent, especially when the water immersion age changes from 7 to 56 days. Figure 6 shows that the compressive strength retention rate of the HDMC with 28-day ambient curing is slightly higher than that of the HDMC with 7-day ambient curing, which indicates that the increase of ambient curing age can enhance the water stability of HDMC. This is because increasing the ambient curing age can make the hydration of MPC more complete, the structure more compact, the number of phosphates not involved in the reaction reduced, and the dissolution of phosphates after water immersion decreased significantly. In addition, Figure 6a shows that for the HDMC with 7-day ambient curing, the compressive strength retention rate shows a downward trend as the water immersion age increases, reaching 86% when the water immersion age is 56 days. Figure 6b shows that for the HDMC with 28-day ambient curing, the compressive strength retention rate does not change too much as the water immersion age increases from 7 to 56 days, remaining at about 93%. This indicates that the hydration of the HDMC with 28-day ambient curing is more completed, so the water immersion age has little effect on the compressive strength.

Strength Index, Ductility Index, and Toughness Index and Corresponding Retention Rates
The strength index (first-crack strength and ultimate flexural strength), ductility index, and toughness index of the HDMC with different ambient curing ages and water immersion ages can be calculated by using Equations (1)- (7). The effect of ambient curing age and water immersion age on these parameters is shown in Figure 7. It is shown that compared to the HDMC with 7-day ambient curing, the HDMC with 28-day ambient curing have higher first-crack strength and ultimate flexural strength only when they are immersed in water for 28 days or 56 days, higher ductility index only when they are immersed in water for 7 days or 56 days, and higher toughness index only when they are immersed in water for 56 days. In general, the 28-day ambient curing does not lead to significantly improved strength index, ductility index, and toughness index compared to 7-day ambient curing. Figure 7 also shows that for the HDMC with 7-day ambient curing, as the water immersion age increases, the first-crack strength, ultimate flexural strength, and

Strength Index, Ductility Index, and Toughness Index and Corresponding Retention Rates
The strength index (first-crack strength and ultimate flexural strength), ductility index, and toughness index of the HDMC with different ambient curing ages and water immersion ages can be calculated by using Equations (1)- (7). The effect of ambient curing age and water immersion age on these parameters is shown in Figure 7. It is shown that compared to the HDMC with 7-day ambient curing, the HDMC with 28-day ambient curing have higher first-crack strength and ultimate flexural strength only when they are immersed in water for 28 days or 56 days, higher ductility index only when they are immersed in water for 7 days or 56 days, and higher toughness index only when they are immersed in water for 56 days. In general, the 28-day ambient curing does not lead to significantly improved strength index, ductility index, and toughness index compared to 7-day ambient curing. Figure 7 also shows that for the HDMC with 7-day ambient curing, as the water immersion age increases, the first-crack strength, ultimate flexural strength, and toughness index gradually decrease except for that the 56-day water immersion age leads to a slightly higher ultimate flexural strength than 28-day water immersion age; the ductility index increases first and then decrease, reaching the highest value of 22.07 mm when the water immersion age is 28 days. For the HDMC with 28-day ambient curing, as the water immersion age increases, the first-crack strength shows a gradual increase; the ultimate flexural strength decreases first and then increases, reaching the highest value of 9.85 MPa when the water immersion age is 56 days; the ductility index first increases and then decreases, reaching the highest value of 20.81 mm when the water immersion age is 7 days; and the toughness index shows an ascending trend on the whole, reaching the maximum value of 12.75 kN·mm when the water immersion age is 56 days. Therefore, generally, as the water immersion age increases, the strength index, ductility index, and toughness index of the HDMC with 7-day ambient curing become worse; however, those of the HDMC with 28-day ambient curing are not so bad and even can be improved in many cases, especially when the water immersion age is 56 days. toughness index gradually decrease except for that the 56-day water immersion age leads to a slightly higher ultimate flexural strength than 28-day water immersion age; the ductility index increases first and then decrease, reaching the highest value of 22.07 mm when the water immersion age is 28 days. For the HDMC with 28-day ambient curing, as the water immersion age increases, the first-crack strength shows a gradual increase; the ultimate flexural strength decreases first and then increases, reaching the highest value of 9.85 MPa when the water immersion age is 56 days; the ductility index first increases and then decreases, reaching the highest value of 20.81 mm when the water immersion age is 7 days; and the toughness index shows an ascending trend on the whole, reaching the maximum value of 12.75 kN•mm when the water immersion age is 56 days. Therefore, generally, as the water immersion age increases, the strength index, ductility index, and toughness index of the HDMC with 7-day ambient curing become worse; however, those of the HDMC with 28-day ambient curing are not so bad and even can be improved in many cases, especially when the water immersion age is 56 days.  Figure 7 shows that the retention rates of first-crack strength, ultimate flexural strength, ductility, and toughness of the HDMC with 28-day ambient curing are all higher than those of the HDMC with 7-day ambient curing, indicating better water stability. It is because, with the increase of ambient curing age, the number of unreacted phosphates in the HDMC matrix is gradually reduced; the number of hydration products is gradually increased; and the structure of the matrix is denser. Figure 7 also shows that for the HDMC with 7-day ambient curing, with the increase of water immersion age, the retention rates of first-crack strength, ultimate flexural strength, and toughness decrease; and the retention rate of ductility increases first and then decreases, reaching the highest value of 117.84% when the water immersion age is 28 days. For the HDMC with 28-day ambient curing, with the increase of water immersion age, the first-crack strength retention rate shows a gradual increase, reaching the highest value of 128.23% after 56 days of water immersion; the ultimate flexural strength retention  Figure 7 shows that the retention rates of first-crack strength, ultimate flexural strength, ductility, and toughness of the HDMC with 28-day ambient curing are all higher than those of the HDMC with 7-day ambient curing, indicating better water stability. It is because, with the increase of ambient curing age, the number of unreacted phosphates in the HDMC matrix is gradually reduced; the number of hydration products is gradually increased; and the structure of the matrix is denser. Figure 7 also shows that for the HDMC with 7-day ambient curing, with the increase of water immersion age, the retention rates of first-crack strength, ultimate flexural strength, and toughness decrease; and the retention rate of ductility increases first and then decreases, reaching the highest value of 117.84% when the water immersion age is 28 days. For the HDMC with 28-day ambient curing, with the increase of water immersion age, the firstcrack strength retention rate shows a gradual increase, reaching the highest value of 128.23% after 56 days of water immersion; the ultimate flexural strength retention rate slightly decreases first and then increases, reaching the highest value of 117.87% after 56 days of water immersion; the ductility retention rate first increases and then decreases, reaching the highest value of 126.17% after 7 days of water immersion; and the toughness retention rate shows an ascending trend generally, reaching the highest value of 145.95% after 56 days of water immersion. It can be seen that the changes of strength retention rate, ductility retention rate, and toughness retention rate are consistent with those of strength index, ductility index, and toughness index. Therefore, generally, as the water immersion age increases, the water stability of the HDMC with 7-day ambient curing becomes worse; however, the water stability of the HDMC with 28-day ambient curing can be improved in many cases, especially when the water immersion age is 56 days. Figure 8 shows the compressive strength and its retention rate of non-immersed (denoted as "7d") and immersed HDMC (denoted as "7 + 7d") with different W/B ratios. It can be seen that with the increase of the W/B ratio, the compressive strength of the non-immersed HDMC decreases from 49.5 to 43.8 MPa, and the compressive strength of the immersed HDMC decreases from 47.1 to 41.4 MPa. This is because when the W/B ratio is increased, the pores in the matrix can be increased due to the evaporation of excess water, thereby reducing the compressive strength of the HDMC [36,47].

Compressive Strength and Its Retention Rate
From Figure 8, it can be seen that the compressive strength retention rate of HDMC is less than 1. This is because, under the water immersion condition, some hydration products are dissolved, and some phosphates dissolve to reduce the hydration degree of MgO grains and thus the formation of hydration products. Figure 8 also shows that the compressive strength retention rate of HDMC decreases slightly with the increase of the W/B ratio. When the W/B ratio is large, although the formation of hydration products decreases and some of the hydration products dissolve, increasing the matrix defects, the content of unreacted MgO grains increased, and the superimposition effect of the center particles is stronger than that of the unreacted HDMC matrix, which may make the compressive strength retention rate change little [48].
Materials 2021, 14, x FOR PEER REVIEW 12 of 23 rate slightly decreases first and then increases, reaching the highest value of 117.87% after 56 days of water immersion; the ductility retention rate first increases and then decreases, reaching the highest value of 126.17% after 7 days of water immersion; and the toughness retention rate shows an ascending trend generally, reaching the highest value of 145.95% after 56 days of water immersion. It can be seen that the changes of strength retention rate, ductility retention rate, and toughness retention rate are consistent with those of strength index, ductility index, and toughness index. Therefore, generally, as the water immersion age increases, the water stability of the HDMC with 7-day ambient curing becomes worse; however, the water stability of the HDMC with 28-day ambient curing can be improved in many cases, especially when the water immersion age is 56 days. Figure 8 shows the compressive strength and its retention rate of non-immersed (denoted as "7d") and immersed HDMC (denoted as "7 + 7d") with different W/B ratios. It can be seen that with the increase of the W/B ratio, the compressive strength of the nonimmersed HDMC decreases from 49.5 to 43.8 MPa, and the compressive strength of the immersed HDMC decreases from 47.1 to 41.4 MPa. This is because when the W/B ratio is increased, the pores in the matrix can be increased due to the evaporation of excess water, thereby reducing the compressive strength of the HDMC [36,47].

Compressive Strength and Its Retention Rate
From Figure 8, it can be seen that the compressive strength retention rate of HDMC is less than 1. This is because, under the water immersion condition, some hydration products are dissolved, and some phosphates dissolve to reduce the hydration degree of MgO grains and thus the formation of hydration products. Figure 8 also shows that the compressive strength retention rate of HDMC decreases slightly with the increase of the W/B ratio. When the W/B ratio is large, although the formation of hydration products decreases and some of the hydration products dissolve, increasing the matrix defects, the content of unreacted MgO grains increased, and the superimposition effect of the center particles is stronger than that of the unreacted HDMC matrix, which may make the compressive strength retention rate change little [48].

Strength Index, Ductility Index, and Toughness Index and Corresponding Retention Rates
The strength index (first-crack strength and ultimate flexural strength), ductility index, and toughness index, together with their retention rates, can be calculated using Equations (1)- (7). The effect of the W/B ratio on these parameters is shown in Figure 9. It can be seen that for the non-immersed HDMC, with the increase of W/B ratio from 0.14 to 0.20, the first-crack strength increases slightly first and then decreases, reaching the maximum value of 7.74 MPa when the W/B ratio is 0.18; the ultimate flexural strength decreases gradually; the ductility index first increases and then decreases, reaching the highest value of 18.73 mm when the W/B ratio is 0.16; and the toughness index first increases and then decreases, reaching the maximum value of 13.23 kN·mm when the W/B ratio is 0.16. For the immersed HDMC, with the increase of W/B ratio from 0.14 to 0.20 the first-crack strength and ultimate flexural strength show a decreasing trend; however, the ductility index and toughness index show an upward trend on the whole except when the W/B ratio is 0.18. Therefore, for non-immersed HDMC, when the W/B ratio is 0.16, the strength index, ductility index, and toughness index all can be satisfactory. For immersed HDMC, the strength index is the best when the W/B ratio is 0.14, and the ductility index and toughness index are the best when the W/B ratio is 0.20; these three indexes are moderate when the W/B ratio is 0.16. The strength index (first-crack strength and ultimate flexural strength), du index, and toughness index, together with their retention rates, can be calculated Equations (1)- (7). The effect of the W/B ratio on these parameters is shown in Figur can be seen that for the non-immersed HDMC, with the increase of W/B ratio from 0 0.20, the first-crack strength increases slightly first and then decreases, reachin maximum value of 7.74 MPa when the W/B ratio is 0.18; the ultimate flexural str decreases gradually; the ductility index first increases and then decreases, reachin highest value of 18.73 mm when the W/B ratio is 0.16; and the toughness index increases and then decreases, reaching the maximum value of 13.23 kN•mm when th ratio is 0.16. For the immersed HDMC, with the increase of W/B ratio from 0.14 to 0. first-crack strength and ultimate flexural strength show a decreasing trend; howeve ductility index and toughness index show an upward trend on the whole except whe W/B ratio is 0.18. Therefore, for non-immersed HDMC, when the W/B ratio is 0.1 strength index, ductility index, and toughness index all can be satisfactory. For imm HDMC, the strength index is the best when the W/B ratio is 0.14, and the ductility and toughness index are the best when the W/B ratio is 0.20; these three indexe moderate when the W/B ratio is 0.16.
(a)  Figure 9a,b shows that the retention rates of first-crack strength and ultimate f strength first decrease and then increase with the increase of the W/B ratio. When th ratio is 0.14, the retention rate of first-crack strength is the highest, and when the W/ is 0.20, the retention rate of ultimate flexural strength reaches the maximum va 115.53%. Figure 9c,d shows that the retention rates of ductility and toughness sh overall upward trend as the W/B ratio increases, reaching the maximum values of 1 and 163.7%, respectively, when the W/B ratio is 0.20.
Generally, the strength index, ductility index, and toughness index are satis when the W/B ratio is 0.16; the retention rates of strength, ductility, toughness seem the highest when the W/B ratio is 0.20, indicating the best water stability, followed W/B ratio of 0.16. Therefore, the W/B ratio of 0.16 can be the best choice for the H considering both the mechanical properties and the water stability. Figure 10 shows the compressive strength and its retention rate of non-immers immersed HDMC with different M/P ratios. It can be seen that, regardless of immersion, the compressive strength of the HDMC increases with the increase of th ratio. As the M/P ratio increases from 3 to 6, the compressive strength of the non-imm HDMC increases from 33.7 to 49.2 MPa, and the compressive strength of the imm HDMC increases from 30.1 to 46.3 MPa.

Compressive Strength and Its Retention Rate
By comparison, the compressive strength of the immersed HDMC is lower than the non-immersed HDMC. Therefore, the compressive strength retention rate is le 100%, as shown in Figure 10. Besides, Figure 10 shows that the compressive strength re rate of HDMC first increases and then decreases as the M/P ratio increases. When th  Figure 9a,b shows that the retention rates of first-crack strength and ultimate flexural strength first decrease and then increase with the increase of the W/B ratio. When the W/B ratio is 0.14, the retention rate of first-crack strength is the highest, and when the W/B ratio is 0.20, the retention rate of ultimate flexural strength reaches the maximum value of 115.53%. Figure 9c,d shows that the retention rates of ductility and toughness show an overall upward trend as the W/B ratio increases, reaching the maximum values of 137.48% and 163.7%, respectively, when the W/B ratio is 0.20.
Generally, the strength index, ductility index, and toughness index are satisfactory when the W/B ratio is 0.16; the retention rates of strength, ductility, toughness seem to be the highest when the W/B ratio is 0.20, indicating the best water stability, followed by the W/B ratio of 0.16. Therefore, the W/B ratio of 0.16 can be the best choice for the HDMC considering both the mechanical properties and the water stability. Figure 10 shows the compressive strength and its retention rate of non-immersed and immersed HDMC with different M/P ratios. It can be seen that, regardless of water immersion, the compressive strength of the HDMC increases with the increase of the M/P ratio. As the M/P ratio increases from 3 to 6, the compressive strength of the non-immersed HDMC increases from 33.7 to 49.2 MPa, and the compressive strength of the immersed HDMC increases from 30.1 to 46.3 MPa.

Compressive Strength and Its Retention Rate
By comparison, the compressive strength of the immersed HDMC is lower than that of the non-immersed HDMC. Therefore, the compressive strength retention rate is less than 100%, as shown in Figure 10. Besides, Figure 10 shows that the compressive strength retention rate of HDMC first increases and then decreases as the M/P ratio increases. When the M/P ratio is 5, the compressive strength retention rate reaches the maximum value of 98.5%, indicating the best water stability. This may be because a large amount of MgO makes the microstructure of the HDMC matrix relatively compact [4]. However, when the M/P ratio is 6, the compressive strength retention rate decreases slightly, which may be because that the excessive MgO content makes the early hydration reaction more fully complete, and the growth of compressive strength is more obvious after 7 days of ambient curing.   Figure 11 shows the strength index (first-crack strength and ultimate flexural strength), ductility index, and toughness index of non-immersed and immersed HDMC with different M/P ratios. For the non-immersed HDMC, with the increase of M/P ratio from 3 to 6, there is no obvious change in the first-crack strength; the ultimate flexural strength first increases and then decreases, reaching the maximum value of 12.16 MPa when the M/P ratio is 5; and the ductility index and toughness index also increase first and then decrease, reaching the maximum values of 18.73 mm and 14.07 kN·mm when the M/P ratio is 4 and 5, respectively. For the immersed HDMC, as the M/P ratio increases from 3 to 6, the first-crack strength decreases gradually; the ultimate flexural strength, ductility index, and toughness index first increase and then decrease, reaching the maximum values of 10.28 MPa, 29.36 mm, and 18.11 kN·mm, respectively, when the M/P ratio is 5. It can be seen that, for both non-immersed and immersed HDMC, the strength index, ductility index, and toughness index are satisfactory when the M/P ratio is 5.
With the increase of M/P ratio from 3 to 6, the first-crack strength retention rate and ultimate flexural strength retention rate of HDMC show a decreasing trend generally, reaching the maximum values of 98.51% and 134.58%, respectively, when the M/P ratio is 3; the ductility retention rate and toughness retention rate first decrease and then increase, reaching the maximum values of 295.6% and 195.91%, respectively, when the M/P ratio is 6.
In general, when the M/P ratio is 5, the strength index, ductility index, and toughness index are satisfactory. However, the retention rate analysis indicates that the HDMC with the M/P ratio of 5 shows moderate water stability, and the best water stability is reached when the M/P ratio is 3. With the increase of M/P ratio from 3 to 6, the first-crack strength retention ra ultimate flexural strength retention rate of HDMC show a decreasing trend gen reaching the maximum values of 98.51% and 134.58%, respectively, when the M/P ra the ductility retention rate and toughness retention rate first decrease and then in reaching the maximum values of 295.6% and 195.91%, respectively, when the M/P ra In general, when the M/P ratio is 5, the strength index, ductility index, and tou

Microstructure of HDMC
The microstructure of HDMC is examined from SEM images. Figure 12a shows the matrix of the non-immersed HDMC with 7-day ambient curing. It can be seen that the struvite-K crystals in the matrix are mainly wedge-shaped and arranged closely. Figure 12b shows the bonding between fibers and the matrix of the non-immersed HDMC with 7-day ambient curing. It can be seen from the fibers that the cross-section of fibers does not change obviously, and there are some scratches, grooves, and hydration products on the surface of fibers, which indicates that the bonding between fibers and the matrix is relatively good, and the fibers may not be fractured but be pulled out from the matrix.

Microstructure of HDMC
The microstructure of HDMC is examined from SEM images. Figure 12a shows the matrix of the non-immersed HDMC with 7-day ambient curing. It can be seen that the struvite-K crystals in the matrix are mainly wedge-shaped and arranged closely. Figure 12b shows the bonding between fibers and the matrix of the non-immersed HDMC with 7-day ambient curing. It can be seen from the fibers that the cross-section of fibers does not change obviously, and there are some scratches, grooves, and hydration products on the surface of fibers, which indicates that the bonding between fibers and the matrix is relatively good, and the fibers may not be fractured but be pulled out from the matrix.  Figure 13a shows the matrix of the HDMC with 7-day water immersion following 7day ambient curing. It can be seen that there are many small substances attached to struvite-K crystals, which are believed to be unreacted MgO or MgO obtained from struvite-K decomposition. Therefore, the struvite-K generated by the hydration reaction is relatively  Figure 13a shows the matrix of the HDMC with 7-day water immersion following 7-day ambient curing. It can be seen that there are many small substances attached to struvite-K crystals, which are believed to be unreacted MgO or MgO obtained from struvite-K decomposition. Therefore, the struvite-K generated by the hydration reaction is relatively less, and the gap between the crystals is relatively larger, which shows that the 7-day water immersion has a large adverse effect on the matrix, and the mechanical properties of the HDMC can be decreased. This is consistent with that the compressive strength (42.5 MPa) of the HDMC with 7-day water immersion is lower than that of the non-immersed HDMC by 5.1%. Figure 13b shows the bonding between fibers and the matrix of the HDMC with 7-day water immersion. It can be seen that the gap between the fibers and the matrix becomes larger after water immersion, and the bonding is not good enough. The surface of fibers is smoother than that of the fibers in the non-immersed HDMC, and there are not large amounts of hydration products attached. The possible reason is that the unreacted potassium dihydrogen phosphate dissolves after water immersion and makes the solution weakly acidic. Then the hydration reaction proceeds in reverse, which leads to the decomposition of the hydration product, struvite-K. This is consistent with that after 7 days of water immersion, the ultimate flexural strength of the HDMC decreases slightly. less, and the gap between the crystals is relatively larger, which shows that the 7-day wa immersion has a large adverse effect on the matrix, and the mechanical properties of HDMC can be decreased. This is consistent with that the compressive strength (42.5 M of the HDMC with 7-day water immersion is lower than that of the non-immersed HDM by 5.1%. Figure 13b shows the bonding between fibers and the matrix of the HDMC with day water immersion. It can be seen that the gap between the fibers and the matrix becom larger after water immersion, and the bonding is not good enough. The surface of fiber smoother than that of the fibers in the non-immersed HDMC, and there are not la amounts of hydration products attached. The possible reason is that the unreac potassium dihydrogen phosphate dissolves after water immersion and makes the solut weakly acidic. Then the hydration reaction proceeds in reverse, which leads to decomposition of the hydration product, struvite-K. This is consistent with that after 7 d of water immersion, the ultimate flexural strength of the HDMC decreases slightly.

Conclusions
In this study, magnesium phosphate cement (MPC), PVA fibers, and fly ash w combined to prepare high ductility magnesium phosphate cement-based compos (HDMC). The effects of ambient curing age, water immersion age, magnesi oxide/potassium dihydrogen phosphate (M/P) ratio, and water/binder (W/B) ratio on water stability and mechanical properties of the HDMC were examined. Seve important conclusions can be drawn.
For the HDMC with the same water immersion age, the 28-day ambient curing lead to higher retention rates of strength, ductility, and toughness than 7-day ambi curing, indicating that the extension of ambient curing age can enhance the water stabi

Conclusions
In this study, magnesium phosphate cement (MPC), PVA fibers, and fly ash were combined to prepare high ductility magnesium phosphate cement-based composites (HDMC). The effects of ambient curing age, water immersion age, magnesium oxide/potassium dihydrogen phosphate (M/P) ratio, and water/binder (W/B) ratio on the water stability and mechanical properties of the HDMC were examined. Several important conclusions can be drawn.
For the HDMC with the same water immersion age, the 28-day ambient curing can lead to higher retention rates of strength, ductility, and toughness than 7-day ambient curing, indicating that the extension of ambient curing age can enhance the water stability of HDMC; however, a longer ambient curing age does not result in significant improvement in the mechanical properties of HDMC.
As the water immersion age increases from 0 to 56 days, the mechanical properties of the HDMC with 7-day ambient curing shows an obvious downward trend; however, the mechanical properties of the HDMC with 28-day ambient curing does not show an obvious decrease and even can be increased in many cases, especially when the water immersion age is 56 days. In terms of water stability, its change is consistent with that of the mechanical properties.
With the increase of the W/B ratio, the changes of strength index, ductility index, and toughness index are complicated. For non-immersed HDMC, the strength index, ductility index, and toughness index all can be satisfactory when the W/B ratio is 0.16. For immersed HDMC, the strength index is the best when the W/B ratio is 0.14, and the ductility index and toughness index are the best when the W/B ratio is 0.20; these three indexes are moderate when the W/B ratio is 0.16. The retention rates of strength, ductility, toughness seem to be the highest when the W/B ratio is 0.20, indicating the best water stability, followed by the W/B ratio of 0.16. Therefore, if all indexes and their corresponding retention rates are considered comprehensively, the W/B ratio of 0.16 seems to be the optimum value for HDMC.
With the increase of the M/P ratio, although the changes of the strength index, ductility index, and toughness index are not always consistent, it can be seen that for both nonimmersed and immersed HDMC, the strength index, ductility index, and toughness index are satisfactory when the M/P ratio is 5. However, the retention rates of these indexes indicate that the HDMC with the M/P ratio of 5 shows moderate water stability, and the best water stability can be reached when the M/P ratio is 3.
The microstructure analysis indicates that for non-immersed HDMC, the struvite-K crystals in the matrix are mainly wedge-shaped and arranged closely; the bonding between fibers and the matrix is relatively good; and the fibers may not be fractured but be pulled out from the matrix. However, for immersed HDMC, the gap between the struvite-K crystals is relatively large, and the surface of fibers is smooth, with no large amounts of hydration products attached, which shows that the water immersion has a large adverse effect on the HDMC and thus reduce their mechanical properties.
Author Contributions: Conceptualization, methodology, supervision, project administration and funding acquisition, H.F.; formal analysis, investigation and writing-original draft preparation, Y.W.; formal analysis and writing-review and editing, A.G.; data curation, X.Z. All authors have read and agreed to the published version of the manuscript.

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 some privacy.