Abstract
Ground phosphorous slag (PS) has not been widely used in construction due to its negative effects on the early-age performances of cementitious materials. The effects of calcium carbonate nanoparticles (NC) on strength development of mortar containing high content of PS were investigated at different curing ages. The NC was incorporated at 2% as partial mass replacements for binder. Hydration products and microstructure characterization was examined by X-ray diffraction (XRD), differential thermal gravity (DTG), thermogravimetric (TG), and scanning electron microscopy (SEM) analysis. Test results showed that NC improved both flexural and compressive strength of mortar containing high content of PS at 7, 28, 56, and 90 days. XRD, DTG-TG, and SEM analysis confirmed the filling effect of NC. Furthermore, the formation of the carboaluminate even at later age also improved the microstructure of mortar, which created a denser microstructure.
1. Introduction
Recently, sustainable development and environmental protection of concrete industry have attracted a growing attention [1, 2]. It is reported that partially replacing clinker by mineral admixtures, such as fly ash (FA), granulated blast furnace slag (GBFS), silica fume (SF), copper slag (CS), ground phosphorous slag (PS), and limestone powder (LS), is an effective way to alleviate the environmental burden [2–5]. These mineral admixtures also improve the mechanical performances and durability of cementitious materials [6–9]. In the past few decades, a large part of clinker is commonly replaced by FA due to its pozzolanic effect and filling effect [10]. However, the amount of FA is relatively small in Southwest of China, such as Guizhou province. Therefore, it is necessary to develop other industrial wastes as mineral admixtures for cementitious materials in these areas.
PS, a by-product of yellow phosphor production, could be used as a mineral admixture for cementitious materials due to its high content of CaO and SiO2 [11–13]. PS exhibits pozzolanic activity [14] due to its glassy microstructure, which is similar to that of GBFS [15, 16]. Previous findings demonstrate that PS can decrease hydration heat, increase later age strength, and improve durability of concrete [13, 15]. However, PS has not been widely used in construction due to its high content of phosphorous [16]. It is reported that phosphorous has negative effects on the early-age performances of cementitious materials [17–19]. Therefore, activation techniques, such as mechanical, thermal, and chemical treatment, have been studied extensively [20–22]. However, these activation techniques are relatively expensive.
Studies have shown that calcium carbonate has positive effects on the strength development of cementitious materials due to its accelerating effect on the rate of cement hydration [23–26]. Recently, the use of calcium carbonate nanoparticles (NC) in cementitious materials has received great interest among researchers [27–31]. The effects of NC on the performances of cementitious materials containing FA have been extensively reported [32–36]. NC is considered to be a good candidate for accelerating hydration of cement in FA mortar. However, very few studies have reported the effects of NC on cementitious materials with PS. This paper investigated the strength development of PS mortar containing NC. The hydration products and microstructure development of PS mortar containing NC was also examined by XRD, DTG, TG, and SEM analysis.
2. Materials and Methods
Portland cement (PC, P.I 42.5), PS, NC, silica sand, and deionized water were used in this experiment. The chemical compositions of PC, PS, and NC as given in Table 1 were determined by X-ray fluorescence. PS was provided by a ready mixed concrete company, Guizhou province, China. NC was purchased from Shanghai Yuanjiang Chemical Co., Ltd. of China. The NC came in dry powder form, with a rod like shape, as shown in Figure 1. Figure 2 shows that the main crystalline phase of NC is aragonite. Furthermore, it can be seen from Figure 2 that PS does not exhibit pronounced crystalline peaks indicating that PS is mainly amorphous in nature.


The consistency is a main factor for the material workability. The standard consistencies of PC, PC containing 40% PS (“binder” for short), and binder containing 2% of NC are 25.6%, 26.4%, and 27.1%, respectively. The binder containing 2% of NC demands more water than that of binder. The increased water demand is attributed to the high specific surface area of PS and NC.
Mix proportion of the specimens is given in Table 2. The amount of NC is 2% by weight of binder. A water-to-binder ratio is maintained at 0.4 for all the specimens, and the sand to binder ratio is 3. Mixing procedures were performed for all specimens in a rotary mixer. Firstly, PC, PS, and NC were added to the mixer and mixed at a low speed for 2 minutes. Then, added silica sand and mixed for another 1 minute. As a final stage, deionized water was added and mixed at high speed for more 3 minutes. Specimens were cast into their molds and demolded after 24 hours. Then, specimens were cured in saturated lime water at 20 ± 2°C until testing ages. Specimens used for XRD, DTG, TG, and SEM analysis did not include silica sand. The composition of these pastes was the same proportions of binders as in mortar.
The flexural and compressive strengths of mortar were conducted at 7, 28, 56, and 90 days, respectively. XRD test was done by D8 Advance X-ray Diffraction Instrument. Operating conditions were set a 40 kV and 40 mA using a Cukα X-ray source. 2θ range was 5° to 60°. DTG-TG was used to calculate the weight loss of hydration products and calcium hydroxide content in specimens. SEM (JSM-IT300) analysis was performed for microstructural investigation.
3. Results and Discussion
3.1. Strength Development
Flexural strength and compressive strength of all mortars are presented in Figures 3(a) and 3(b), respectively. It can be seen that the mortar containing 40% of PS reduced both the flexural strength and compressive strength at 7 and 28 days. However, the strength of PS40 developed rapidly after 28 days. At the age of 56 days, both the flexural strength and compressive strength of PS40 exceeded that of PC, which can be attributed to the formation of additional C-S-H gel through hydration reaction of PS with CH and densification of microstructure [37–39]. Then, the strength development of PS40 slowed down. The strength of PS40 at 90 days was equivalent to the strength at 56 days. Compared to PS40, the addition of 2% NC increased the flexural strength and compressive strength of mortar at all ages. The flexural strength and compressive strength of PS38NC2 at 7 days were higher than that of PS40, 10.4% and 8.2%, respectively. The increased strength is mainly due to the accelerating effects of NC. It is known that NC act as nucleation sites for the hydration products, such as C-S-H and CH, which will accelerate the hydration reaction of cement at early ages, thereby improving the flexural and compressive strength of mortar [40–43]. The flexural strength and compressive strength of PS38NC2 at 28 days were higher than those of PS40, 19.0% and 22.4%, respectively. The increased strength can be attributed to the filler effects [44]. After 56 days, both the flexural strength and compressive strength of PS38NC2 still exceeded that of PS40. The improved flexural strength and compressive strength are due to the densification of microstructure, which is discussed in XRD analysis section, DTG-TG analysis section, and SEM analysis section.

(a)

(b)
3.2. XRD Analysis
In order to identify different phases of specimens, XRD analysis was performed. Figures 4(a) and 4(b) show the XRD pattern of specimens at 28 and 90 days, respectively. The main phases detected were calcium hydroxide (CH), calcite, ettringite (AFt), alite, belite, carboaluminate, and quartz. Calcite in PS40 and part of calcite in PS38NC2 is caused by carbonation of CH during sample preparation. It can be assumed that the amount of CH carbonized in PS38NC2 is the same as that of in PS40 because of the same sample preparing method. The PS38NC2 showed an increment in CH peak intensity compared to that of PS40 at 28 days. The addition of 2% of NC increased the intensity peak of CH from 1755 to 2182 at 2θ = 18.08°. This indicates that NC acts as nucleation for the precipitation of CH and C-S-H at early ages, which accelerate the hydration of cement [45]. Furthermore, it is significant to note that the addition of 2% NC decreased the intensity peak of AFt but increased the intensity peak of carboaluminate at 28 days, indicating that the NC promotes the conversion of AFt to carboaluminate. The formation of carboaluminate helps in the improvement of cementitious materials [9, 46, 47].

(a)

(b)
It can be seen from Figures 4(a) and 4(b) that the intensity peak of alite, belite, and CH in all specimens decreases with the increasing hydration ages. The reduction of alite peaks and belite peaks indicates the formation of hydration products in specimens. It is recognized that the peak intensity of CH is related to C-S-H produced by pozzolanic reaction of mineral admixture at later ages [32]. The decrease of CH intensity indicates the increase of C-S-H formed by pozzolanic reaction at later ages. Thus, the reduction of CH peaks is due to the pozzolanic reaction of PS, which causes the consumption of CH. Therefore, the change of CH peaks could indicate the reaction degree of PS at later ages.
It is also noted that PS38NC2 showed a decrement in CH peak intensity compared to that of PS40 at 90 days. The intensity peak of CH decreased from 1210 to 1082 at 2θ = 18.08°. This phenomenon implies that NC could increase the reaction degree of PS, which modifies the performances of mortar containing PS. It can be seen from Figures 4(a) and 4(b) that the carboaluminate is formed. This may be attributed to the reaction between NC and aluminates. Kakali et al. [48] also reported that in pastes containing CaCO3, either as a chemical reagent or as a limestone constituent, the ettringite’s transformation to monosulfate is delayed, while calcium aluminate monocarbonate is preferably formed instead of monosulfate. Furthermore, NC decreased the intensity peak of AFt but increased the intensity peak of carboaluminate at 90 days, indicating that the NC promotes the conversion of AFt to carboaluminate even at late ages.
3.3. DTG-TG Analysis
The observations made through XRD analysis were further validated through DTG-TG analysis. The results of DTG/TG analysis of specimen’s hydration for 28 days and 90 days are shown in Figures 5(a) and 5(b), respectively. The TG curves indicate the changes in mass of specimens due to heating from room temperature to about 1000°C. The first weight loss, occurred between room temperature and 105°C, is attributed to the evaporation of free water from hydrated cement matrix. The second weight loss, occurred between 105 and 400°C, is the result of decomposition of hydration products (C-S-H, C-A-H, AFt, and AFm phases). The third weight loss, observed between 400 and 500°C, corresponds to the dehydroxylation of CH. The CH content can be calculated according to Taylor’s formula [32]. The results are listed in Table 3. The fourth weight loss appears at about 620°C, which corresponds to the decarbonation of well-crystalline calcium carbonate.

(a)

(b)
It can be observed from Figure 5(a) and Table 3 that the weight loss of hydration products and CH content in PS38NC2 is approximately 6.74% and 8.30%, respectively, which is higher than that of PS40. This could be due to the accelerating effect of NC on the hydration of cement at early ages. It also can be seen in Figure 5(b) and Table 3 that the weight loss of hydration products in PS38NC2 is equivalent to that of PS40 at 90 days and the weight loss of CH in PS38NC2 is lower than that of PS40 at 90 days, indicating the accelerating effect of NC on the pozzolanic reaction of PS at later ages. These behaviors are consistent with the test results of strength shown in Figure 3, which implies that NC increases the hydration products, therefore, improving the mechanical performances of mortar containing PS both at early and later ages.
3.4. SEM Analysis
To study the strength development of specimens, the morphology of hydration products and microstructure of specimens at 28 and 90 days was examined by SEM. Typical SEM images of PS40 and PS38NC2 at 28 days are presented in Figures 6(a) and 6(b). The PS40 has an apparently compact microstructure filled with hydration products, shown in Figure 6(a), where an abundance of C-S-H, CH, and AFt can be seen. Compared to the PS40, the PS38NC2 (Figure 6(b)) has a denser microstructure, as can be resolved from SEM imaging. This indicates that the microstructure of mortar containing PS is altered by 2% NC at 28 days.

(a)

(b)
Figures 7(a) and 7(b) present typical SEM images of PS40 and PS38NC2 at 90 days, respectively. It seems that more hydration products could be formed at 90 days than that of at 28 days. This is because of the pozzolanic reaction between PS particles and CH at later ages, which improve the density and thus the mechanical performance of mortar [13]. The microstructure of PS38NC2, as shown in Figure 7(b), is denser than that of PS40, which can be attributed to the filling effect of NC. Furthermore, from the results of XRD analysis and DTG-TG analysis, the formation of the carboaluminate and the seeding effect also improve the microstructure of mortar containing NC.

(a)

(b)
The backscatter scanning electron (BSE) microscope observations on PS40 and PS38NC2 have been carried out to observe the microstructure changes. Figures 8(a) and 8(b) show the typical BSE images of specimens at 28 and 90 days, respectively. The phases, such as unhydrated cement grains, CH, C-S-H, and pores, in these images can be identified through their brightness [35]. The unhydrated cement grains appear brightest, followed by PS grains, hydration products, and finally the pores. It seemed that many hydration products could be formed on the surface of PS particles. Due to the high alkalinity of hydration products, part of PS grains disappeared at 28 days because of the pozzalanic reaction of PS, which is very helpful for the accelerated hydration of PS and the improvement of mortar strength.

(a)

(b)
In Figure 8, it is clearly seen that the PS38NC2 has fewer porous areas and more dense areas than PS40 at 28 days, which indicates that the microstructure of PS38NC2 is denser than that of PS40. Similar phenomenon can also be seen in Figure 9.

(a)

(b)
4. Conclusions
This work investigated the effects of incorporating NC as partial replacement for binder on the hydration products and strength development of mortar containing high content of PS. From test results of the experiments conducted in this work, the following conclusions can be drawn:(1)Incorporating 40% of PS reduced both the flexural strength and compressive strength of mortar at early ages. However, the strength of mortar containing PS surpassed that of plain mortar after 56 days, which can be attributed to the pozzolanic reaction of PS with CH at later ages.(2)Incorporating 2% of NC resulted in higher flexural and compressive strength of mortar containing 40% of PS compared to that of the control mortar at both early and later ages.(3)Test results of XRD and DTG-TG indicate that NC could increase the reaction degree of PS at later ages. Furthermore, incorporating 2% of NC in mortar containing PS promotes the conversion of AFt to carboaluminate even at later ages.(4)SEM and BSE observation showed that incorporating 2% of NC resulted in a denser microstructure of mortar containing 40% of PS compared to that of the control mortar, which can be attributed to the filling effect of NC. Furthermore, the formation of the carboaluminate improved the microstructure of mortar containing PS.(5)The improved microstructure of mortar containing high content of PS due to the addition of NC through XRD, DTG-TG, and SEM analysis confirmed that the mechanical strength of mortar can be enhanced compared with that of control mortar.
Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare no conflicts of interest.
Acknowledgments
This research was funded by Natural Science Foundation of China (grant no. 51669004) and Science and Technology Program of Guizhou Province (grant no. [2017]7346).