Dual-functional utilization of phosphogypsum as cementitious binder and aggregate in concrete: Synergistic mechanisms and performance enhancement

PanChen1,2

ZhexinWang1

FengZhu1

ShujieWan1

MengyangHuang1

PengfeiLiu1

DongxuZhang1

CaiWu1,2

YaniLu1✉,2Emaillyn2016@hbeu.edu.cn

1School of Civil EngineeringHubei Engineering University432000XiaoganChina

2Hubei Province Engineering Research Center for Cement-Based Ultra-High Performance Concrete and Prefabricated Building Technology432000XiaoganChina

Pan Chen^{1, 2}, Zhexin Wang¹, Feng Zhu¹, Shujie Wan¹, Mengyang Huang¹, Pengfei Liu¹, Dongxu Zhang¹, Cai Wu^{1, 2}, Yani Lu^{1, 2,} *

1 School of Civil Engineering, Hubei Engineering University, Xiaogan, 432000, China

2 Hubei Province Engineering Research Center for Cement-Based Ultra-High Performance Concrete and Prefabricated Building Technology, Xiaogan, 432000, China

*Corresponding author: School of Civil Engineering, Hubei Engineering University, Xiaogan, 432000, China. E-mail address: lyn2016@hbeu.edu.cn

Abstract

Focusing on the environmental issues caused by phosphogypsum (PG) stockpiling and its potential resource utilization in construction materials, this study investigates the synergistic mechanisms of the dual-functional application of PG as both a cementitious binder and an aggregate in concrete systems. Through a comprehensive experimental program, the effects of PG utilization strategies on mechanical performance were systematically evaluated. The optimized concrete formulation achieved a PG incorporation rate of 38% by mass while maintaining superior mechanical properties, with a 28-day compressive strength reaching 39.3 MPa and a softening coefficient of 0.93, meeting the material requirements for building applications. Comparative tests revealed that the PG concrete system strengths ranked highest to lowest as follows: stone aggregate, pure cement paste, PG aggregates (PGAs), and cement-PG paste. Microstructural characterization revealed that the synergistic interaction between the PG-based binder and PGA promoted the formation of interwoven ettringite and C-S-H gel networks, effectively densifying the interfacial transition zone through pore refinement. The dual-functional utilization strategy not only resolves technical challenges in PG valorization but also establishes a scientific foundation for developing low-carbon building materials through industrial byproduct recycling.

Keywords:

Phosphogypsum concrete

Cementitious binder

Phosphogypsum aggregates

Mechanical properties

Microstructure

1 Introduction

Phosphogypsum (PG), the primary solid waste from the phosphorus chemical industry, has a global annual production exceeding 300 million tons. The open-air stockpiling of PG has led to heavy metal pollution and land resource occupation, becoming a global environmental issue^[1–4]. Concrete, as the most widely used construction material globally, consumes vast amounts of natural resources and generates significant carbon emissions. Promoting the substitution of industrial solid waste for natural aggregates and cementitious binder is a critical pathway for sustainable development in the construction industry^[5–7]. Although existing studies have confirmed that PG can be used as a cement retarder or roadbed filler, its application in structural concrete has long faced technical bottlenecks, such as low incorporation thresholds and significant mechanical performance degradation due to impurity components and low reactivity^[8–9]. In particular, the synergistic mechanism of the dual functions of PG (as both cementitious binder and aggregate) remains unclear, limiting breakthroughs in its resource utilization efficiency. Therefore, studying the composite efficacy of PG as both a cementitious binder and an aggregate in concrete systems provides theoretical support for constructing high-volume PG concrete material systems. This research holds significant practical importance for alleviating industrial solid waste disposal pressures and advancing low-carbon concrete technology innovation.

The application of PG in cementitious binders is an important approach to solid waste resource utilization. Studies have shown that untreated PG used alone as a cementitious binder suffers from defects such as low early strength and prolonged setting time. Islam et al.^[10] reported that when raw PG replaced 10% natural gypsum, the 28-day compressive strength of Portland cement reached 29.5 MPa, but when the substitution exceeded 15%, the strength decreased by 22% due to residual impurities. Hu et al.^[11] successfully prepared C30-grade PG concrete by optimizing the water-cement ratio and silica fume content, achieving a 28-day compressive strength of 45.1 MPa. Levickaya et al.^[12] compared the performance differences of PG from various sources and reported that PG exhibited greater compressive strength than natural gypsum at low water‒solid ratios because of its lower degree of particle agglomeration; however, water demand control still relies on grinding or the addition of a superplasticizer. Yang et al.^[13] developed nonautoclaved aerated concrete by increasing the sulfate reactivity of PG with calcium sulfoaluminate cement, but the drying shrinkage rate far exceeded the standard limits. These studies confirm that the standalone application of PG requires pretreatment or activators to balance mechanical performance and volumetric stability.

Composite cementitious systems enhance PG utilization efficiency through synergistic effects of multiple solid wastes. Gong et al.^[1] employed ultrasonic cyclic water washing to pretreat PG, where secondary hydration-generated ettringite increased the compressive strength of the composite cementitious binder by 12%. However, when the substitution exceeded 30%, unreacted gypsum crystals caused pore coarsening. Pratap^[14] reported that microsilica and PG synergistically promoted the formation of C-A-S-H gels, reducing the chloride ion permeability of geopolymer concrete to 1476 coulombs, but did not assess the degree of radioactive risk. Kong et al.^[15] constructed a ternary system of red sandstone-PG-cement, revealing that Ca²⁺ and SO₄²⁻ in PG could activate the Al³⁺ reactivity of red sandstone, but the material strength reached only 32.1 MPa, requiring the substitution ratio to be limited to ≤ 25%. Wang et al.^[16] developed autoclaved aerated concrete panels with 75% solid waste content, achieving a compressive strength of 5.2 MPa through multitemperature zone curing, but did not address the inhibitory effect of P₂O₅ in PG on hydration. Feng et al.^[17] introduced alkali modulus regulation in a ternary system of PG-fly ash-slag, increasing the axial compressive strength by 17%, but high substitution led to a 30% increase in porosity.

Pretreatment and admixtures can significantly improve the performance shortcomings of PG-based materials. Fu et al.^[18] enhanced the freeze‒thaw cycle strength retention rate of a PG-based cementitious binder by 40% through the combined addition of quicklime and silica fume, but excessive quicklime caused expansion cracks. Tian et al.^[19] used calcium sulfoaluminate cement to reduce the setting time to 102 minutes, but the drying shrinkage rate still exceeded the standard limit by sevenfold, necessitating optimized curing regimes. Huang et al.^[20] developed high-volume PG foamed concrete (53.1%), employing water glass to activate slag reactivity, which exhibited microexpansion characteristics in humid environments but had a strength of only 3.36 MPa, limiting its engineering application scope.

The preparation of PG into construction aggregates and its application in concrete systems not only alleviates the pressure of natural aggregate resource shortages but also provides an innovative pathway for the large-scale resource utilization of industrial solid waste. The preparation of PG-based aggregates is a prerequisite for their concrete application. Ding et al.^[21] used cold-bonding technology to prepare PG aggregates (PGAs) and reported that when the PG content was ≤ 80%, the 28-day cylindrical compressive strength of the aggregates reached 11.9 ~ 16.5 MPa, with a softening coefficient ≥ 0.82, meeting lightweight aggregate standards. This process generated ettringite and C-S-H gels through the hydration reaction of slag and cement, effectively filling pores and encapsulating unreacted PG particles, significantly reducing the leaching concentrations of phosphorus and heavy metals. However, when the PG content exceeded 80%, the porosity and water absorption of the aggregates sharply increased, leading to mechanical performance degradation, indicating the need to balance the solid waste content and aggregate performance.

In concrete applications, PGA must be compatible with cementitious systems to achieve synergistic effects. Sun et al.^[5] developed a high-volume PG concrete (HPGC) system that consumed 800 kg of PG per cubic meter by using PG as both a cold-bonded aggregate and a raw material for supersulfated cement. When the PG content in the aggregates was 80% and that in the cementitious binder was 60%, the 28-day compressive strength of the HPGC reached 18–40 MPa, with the phosphorus concentration and radioactivity indices in the leachate complying with Chinese standards. To address the challenge of fine aggregate substitution, Sun et al.^[22] further investigated the impact of PG-based cold-bonded fine aggregates on excess sulfate cement concrete (ESCC). They reported that when natural sand was fully replaced, the compressive strength of ESCC decreased by 17.2%, but optimization of the mix ratio could meet the 30 MPa strength requirement, with environmental safety significantly superior to that of conventional concrete.

Although significant progress has been made in the research of PG as both a cementitious binder and an aggregate, existing studies have focused predominantly on its single-function development and have not systematically explored its dual-function synergistic effects in concrete. Previous research has focused mainly on the impact of PG as a partial cement replacement on early hydration behavior and strength development or its role as an artificial aggregate in lightweight concrete and pore structure optimization. However, few studies have investigated the synergistic effects of PG when it is used as both a cementitious binder and an aggregate, particularly in terms of hydration reactions, ITZ modification, and mechanical performance enhancement.

This study systematically reveals the synergistic effects of PG in its dual roles through the design of comparative experiments, addressing the critical gap in understanding their combined effects on mechanical performance and microstructure evolution. The research framework encompasses material characterization, mix proportion design, mechanical testing, and microstructural analysis, establishing fundamental relationships between PG utilization strategies and concrete performance metrics.

2 Materials and methods

2.1 Experimental raw materials

PG was supplied by Jiuqi Building Materials Co., Ltd. (Zhucheng, China). Prior to experimentation, the organic matter and soluble impurities in the PG were removed through a water washing method, resulting in a P₂O₅ mass fraction less than 0.1% in the treated PG. The PG is a grayish-white powder with a specific surface area of approximately 500 m²/kg, and its primary physical characteristics are summarized in Table 1.

X-ray diffraction (XRD) analysis was conducted to investigate the crystalline phase composition of PG, as illustrated in Fig. 1. The XRD results demonstrate that the predominant mineral phases in PG are calcium sulfate hemihydrate (CaSO₄·0.5H₂O) and calcium sulfate dihydrate (CaSO₄·2H₂O). Distinct diffraction peaks for CaSO₄·0.5H₂O are observed at 2θ angles of 14°, 25°, and 29°, whereas characteristic peaks for CaSO₄·2H₂O appear at 2θ angles of 12°, 22°, and 29°. The sharp diffraction peaks in the pattern indicate well-crystallized phases in the sample.

Table 1
Physical properties of PG
Properties	Initial Setting Time	Final Setting Time	2-Hour Wet Compressive Strength	Soluble anhydrite	Calcium sulfate hemihydrate	Calcium sulfate dihydrate	Crystallization water
Results	4 min	13 min	6.89MPa	5.98%	77.65%	0.55%	4.70%

Fig. 1

XRD patterns of PG

Fig. 2

PGA Fig. 3 XRD patterns of PGA

The PGA was supplied by Hubei Jvhai Environmental Technology Co., Ltd., as shown in Fig. 2. PGA is produced by high-temperature calcination of raw PG in a rotary kiln to eliminate or neutralize harmful substances and impurities such as phosphorus and fluorine, followed by processing into aggregate particles of varying particle sizes. In this study, a PGA with a particle size range of 9–15 mm was employed, which exhibited a grayish-white spherical morphology with a smooth and rounded surface and a density of 1.7 ~ 1.8 g/cm³. Figure 3 displays the XRD pattern of PGA, revealing that CaSO₄·0.5H₂O was the predominant mineral phase. The sharp diffraction peaks observed in the pattern indicate a well-crystallized structure. Although it is derived from the same raw PG material shown in Fig. 1, PGA has a more homogeneous crystalline composition and higher purity of calcium sulfate hemihydrate after high-temperature calcination.

The cement used is 42.5R grade ordinary Portland cement, with a specific surface area of 385 m²/kg and a specific gravity of 3.15 g/cm³. According to the Chinese standard GB175-2023, the compressive strengths at 3, 7, and 28 days are 18.7, 29.8, and 44.5 MPa, respectively. Figure 4 shows the XRD pattern of the cement. As shown in the figure, the main mineral phases of the cement are dicalcium silicate (C₂S), tricalcium silicate (C₃S), calcium carbonate (CaCO₃), and a small amount of silicon dioxide (SiO₂). SiO₂ exhibited distinct diffraction peaks at diffraction angles (2θ) of approximately 26° and 30°. Silicon dioxide is one of the essential components of cement, and an appropriate amount of SiO₂ contributes to the formation of stable calcium silicate hydrate (C-S-H) gel. CaCO₃ shows relatively clear diffraction peaks at diffraction angles (2θ) of approximately 29° and 48°. The presence of CaCO₃ is due to the inclusion of limestone or other calcium carbonate-containing minerals in the raw materials, which were not fully decomposed during the cement production process. C₂S and C₃S are the primary mineral components of cement. C₂S exhibits distinct diffraction peaks at diffraction angles (2θ) of approximately 32° and 42°, whereas C₃S shows clear diffraction peaks at diffraction angles (2θ) of approximately 33° and 52°. These minerals primarily generate calcium silicate hydrate (C-S-H) and calcium hydroxide during the hydration process of cement, which are the main sources of cement strength development.

Fig. 4

XRD pattern of the cement

The other raw materials used in this experiment include stone and sand, with the sand having a fineness modulus of 2.38. The admixtures include retarders and superplasticizers. The retarder is primarily used to delay the hydration reaction when PG is employed as a cementitious binder, thereby extending the setting time and improving the workability. The superplasticizer mainly serves to reduce the water content in the concrete while ensuring its workability, thereby increasing the strength and durability of the concrete.

2.2 Sample Design

Four comparative experimental groups (labeled S1 to S4) were designed to investigate the influence of PG on the mechanical properties of concrete when used as both a cementitious binder and an aggregate. The mix proportions of the raw materials used in the experiments are shown in Table 2. The water-to-binder ratio for all the samples was set to 0.34. S1 is the reference sample, where PG serves as both a cementitious binder and an aggregate. In S1, PG accounts for 50% of the mass of the cementitious binder, and PGA constitutes 30% of the total volume. Compared with S1, S2 is identical in all aspects except that the PGA is replaced by an equal volume of stone. The two groups (S1 and S2) are compared to study the effects of replacing PGA with stone on the mechanical properties of the concrete. Compared with S1, S3 is identical in all aspects except that the PG is replaced by an equal mass of cement. S1 and S3 are compared to study the effects of replacing PG with cement as the cementitious binder on the mechanical properties of concrete. Compared with S3, S4 is identical in all aspects except that the PGA is replaced by an equal volume of stone. S3 and S4 are compared to study the effects of replacing PGA with stone when cement is used as the cementitious binder on the mechanical properties of concrete.

Table 2
Mix proportion design
Samples	Cement/ kg·m^− 3	PG/ kg·m^− 3	Sand/ kg·m^− 3	Stone/ kg·m^− 3	PGA/ kg·m^− 3	Water/ kg·m^− 3	Superplasticizer/ kg·m^− 3	retarder/ kg·m^− 3
S1	304.2	304.2	810	-	510	206.9	12.16	1.2
S2	304.2	304.2	810	810	-	206.9	12.16	1.2
S3	608.4	-	810	-	510	206.9	12.16	1.2
S4	608.4	-	810	810	-	206.9	12.16	1.2

Each experimental group consisted of 18 samples, totaling 72 samples, with dimensions of 100 mm × 100 mm × 100 mm. After mixing according to the proportions in Table 2, the mixture was uniformly stirred in a mixer and poured into molds. After curing at room temperature for 1 day, the samples were placed in a standard curing chamber (temperature, 20°C; humidity, 95%) and cured for 3, 7, or 28 days. The mechanical properties and microstructures of the samples were subsequently measured.

2.3 Test methods

2.3.1 Microstructure analysis

The XRD patterns were obtained via a BRUKER D8 ADVANCE XRD diffractometer. Prior to testing, the samples were ground into powder and dried. As shown in Fig. 5. The diffraction measurement angle range was set from 10° to 70°.

Fig. 5

XRD data (from left to right: cement, PGA, PG and the diffractometer)

The surface microstructure of the concrete samples was analyzed via a JSM−6510 scanning electron microscope (SEM), with samples taken from the interior of the concrete. To ensure the clarity of the scanning images, the sample surfaces were coated with gold to increase their conductivity, as shown in Fig. 6. On the basis of these observations, magnification levels ranging from 2000–8000 were selected for scanning.

Fig. 6

SEM scanning samples and scanner

2.3.2 Mechanical properties

The flowability of the concrete samples was determined according to the Chinese standard GB/T 50080 − 2016. Since the flowability of the prepared samples in this experiment was relatively good, the inverted slump cone emptying test was used to measure the flowability of the concrete mixture. During the test, the concrete was poured into an inverted slump cone, sealed, and then inverted. The time taken for the concrete to completely flow out of the cone was recorded. The shorter the time is, the better the flowability of the concrete. The test was completed within 150 seconds, and the results were averaged from two tests, which were accurate to 0.1 seconds.

The water absorption of the concrete samples was determined according to the Chinese standard GB/T 50081 − 2019. First, the samples were immersed in water at 20 ± 2°C, and the saturated surface-dry mass (m_s) of each sample was measured. Then, the samples were placed in a forced-air drying oven with the temperature controlled at 105 ± 5°C, and the oven-dried mass (m_d) was measured. The water absorption (W_a) of the concrete samples was calculated via the following formula:

The compressive strength of the concrete samples was determined according to the Chinese standard GB/T 50081 − 2019. An electrohydraulic servo pressure tester was used for the test, with a loading rate of 0.5 MPa/s. Each sample was tested three times. Since the samples in this experiment were of nonstandard dimensions, the obtained strength values were multiplied by a size conversion factor of 0.95. The arithmetic mean of the three test results was taken as the strength value for that group of samples.

The stress‒strain curves of the concrete samples were tested via the RMT-301 Rock and Concrete Mechanical Testing System, with a device capacity of 1500 kN. During the test, the load was applied at a uniform rate, and the load and strain data were recorded simultaneously to plot the stress‒strain curve.

The softening coefficient was determined through the following testing procedure. First, the compressive strength of two samples was measured under both saturated (

) and dry (

) conditions. The softening coefficient (

) of the concrete was subsequently calculated via the following formula:

3 Results and Discussion

3.1 Workability and Appearance

Figure 7(a) shows the state of the test samples. The excellent fluidity is attributed primarily to the use of a high-efficiency superplasticizer and the smooth surface characteristics of the sand and PGA. The slump was measured via the inverted slump cone emptying test, and the emptying times for the concrete mixtures of S1 to S4 were 15 s, 16 s, 18 s, and 20 s, respectively. The results indicate that the fluidity of S1 to S4 decreases sequentially. Through comparative analysis, it is evident that incorporating an appropriate amount of PG and PGA can enhance the fluidity of the concrete.

Fig. 7

Test samples. (a) Mixture, (b) Sample appearance

Figure 7(b) shows the state of the concrete after molding. The sample has a grayish-white color with a relatively smooth surface. Both the S1 and S3 samples incorporate the PGA, but compared with S1, S3 is more prone to stratification of the PGA. The primary reason is the significant density difference between the aggregate and the cementitious binder in the S3 sample, which causes the PGA to float to the surface during the molding and vibration process. Stratification not only affects the appearance of the concrete but also tends to cause defects in its mechanical properties. In contrast, the S1 sample includes PG in the cementitious binder, which has a density similar to that of PGA, resulting in minimal stratification.

3.2 Failure mechanism analysis

Analyzing the failure mechanisms of concrete samples is crucial for understanding the mechanical performance of PG concrete under loading. Figure 8 shows typical failure images for each concrete sample under vertical pressure. Upon examining the failed samples, it was observed that the failure modes varied among the different groups.

For S1 (which used both PG and PGA), the primary failure mode was debonding between the aggregate and the paste. Additionally, significant breakage of large-sized PGAs was observed. The top and bottom surfaces of the sample exhibited obvious cracks, and the sides of the concrete experienced extensive spalling and fragmentation under pressure. This finding indicates that the failure of S1 primarily involved interfacial failure between the paste and aggregate, breakage of PGA, and mortar fragmentation.

In contrast, S2 (which replaced PGA with stone) maintained better integrity during failure. The stone aggregate remained unbroken, whereas the concrete surface in contact with the pressure plate experienced spalling. Mortar fragmentation occurred in areas without aggregates, indicating that the failure of S2 mainly involved interfacial failure between the paste and aggregate and mortar fragmentation.

For S3 (which replaced PG with cement but included PGA), the failure surface exhibited obvious stratification. PGA detached from the cement matrix and broke. The cracks on the failure surface were irregular and varied in width, indicating that the failure primarily involved interfacial failure between the paste and aggregate, breakage of the PGA, and minor paste failure.

For S4 (which uses traditional raw materials without PG), the failure involved separation between the aggregate and the paste. The top and sides of the sample cracked and spalled, but the stone aggregate remained relatively intact, with no observed breakage. This finding indicates that the failure of S4 mainly involved interfacial failure between the paste and the aggregate and paste failure.

Fig. 8

Typical failure modes for each sample. (a)S1, (b)S2, (c)S3, (d)S4

3.3 Density

As shown in Fig. 9, the densities of the concrete after 28 days of curing are presented. The measured densities for S1 to S4 were 2184.1, 2384.3, 2193.6 and 2521.7 kg/m³, respectively. These results indicate that the addition of PG or PGA can effectively reduce the density of concrete. Compared with S4, which uses traditional raw materials, the density of S1 decreased by 13.4%. The maximum relative error between the measured densities and the theoretical densities for S1 to S4 is 6.1%, demonstrating that the material mix design has high reliability compared with the theoretical values. Overall, the incorporation of PG and related products is highly important for the development of lightweight concrete.

Fig. 9

Sample density

3.4 Water absorption

The durability of concrete is closely related to its water absorption characteristics. Generally, the lower the water absorption of concrete is, the better its durability. The water absorption of concrete is primarily determined by its pore structure, and the incorporation of PG can influence this pore structure. The water absorption of the samples was measured via the dry‒wet method. The mass difference of the concrete samples in dry and wet states was measured, and the water absorption of the concrete was calculated. As shown in Fig. 10, the water absorption test results of the concrete samples indicate that the water absorption rates of the samples follow the order S2 > S1 > S3 > S4, with values of 15.6%, 8.4%, 6.8%, and 3.5%, respectively. The lower water absorption of S3 and S4 is attributed mainly to the higher cement content, indicating that cement as a cementitious binder can form a denser structure. The incorporation of PG in the cementitious binder slightly reduces the compactness of the concrete. Compared with S1 and S2, S1 uses a regular spherical PGA, whereas S2 uses irregular stone aggregates. The water absorption of S1 is lower than that of S2, suggesting that the use of PGA increases the compactness of the concrete compared with that of the stone aggregate.

Fig. 10

Water absorption rates

3.5 Microstructure analysis

Figure 11 shows the microstructure of the concrete samples at 28 days. From the micrographs of S1 to S4, the hydration products of the concrete can be observed, including a large amount of flocculent calcium silicate hydrate (C-S-H), some layered calcium hydroxide (Ca(OH)₂), and a small amount of needle-like or rod-like ettringite (AFt). The microstructure is relatively dense. C-S-H gel binds various solid-phase particles in the concrete together, providing the concrete with good strength and integrity. Calcium hydroxide is generated during the hydration of cement through the reaction of silicate minerals with water. It has high solubility and alkalinity and plays an important role in maintaining the high-alkali environment of the cement paste and influencing the formation of other hydration products. The formation of ettringite helps enhance the early strength of concrete, as it rapidly fills the pores in the cement paste, making the structure denser.

A comparison of the micrographs reveals that S1, which uses PG as both a cementitious binder and an aggregate, has relatively abundant hydration products. A significant amount of ettringite crystals can be observed, tightly filling the interior of the cement paste and forming a denser microstructure. In contrast, S2 uses stone as an aggregate. The SEM images reveal numerous microcracks and pores at the interface between the cement paste and the rough surface of the stone. The SEM image of S3 shows fewer hydration products in the cement paste, with a distinct layered surface and more pores. The SEM image of S4 displays a large amount of calcium hydroxide interwoven on the surface of the C-S-H gel, with agglomerated ettringite crystals in localized pore regions, indicating that the hydration products of the cement fill the voids and cracks and that the degree of cement hydration is relatively high.

Fig. 11

SEM image of the concrete. (a) S1, (b) S2, (c) S3, (d) S4

In PG concrete, phosphorus originates primarily from PG. During the hydration process, a portion of the phosphorus reacts with components such as aluminates in the cement and is incorporated into hydration products such as ettringite. This process involves the substitution of some sulfate ions, forming phosphorus-containing hydration products, which are then immobilized within the solid phase structure of the concrete. A smaller portion of the phosphorus may exist in the pore solution of the concrete in the form of soluble phosphorus. However, as the concrete hardens and the pore structure becomes more refined, the migration and leaching of this phosphorus are restricted. This reduces the potential environmental pollution risk posed by phosphorus to some extent and endows PG concrete with a certain degree of self-stabilizing characteristics in terms of phosphorus immobilization.

3.6 Compressive strength

Figure 12 shows the compressive strength of the concrete samples at 3 days, 7 days, and 28 days. The figure shows that the 3-day strengths of S1 to S4 were 26.0, 25.8, 31.6, and 46.2 MPa, respectively. The 7-day strengths were 29.6, 29.3, 34.8, and 58.8 MPa, respectively. The 28-day strengths were 39.3, 39.2, 41.1, and 68.3 MPa, respectively. The strength increase in S4 was the most significant, increasing by 22.1 MPa from 3 days to 28 days. S1 and S2 showed strength increases of 13.3 MPa, whereas S3 increased by 9.5 MPa. The test results for S1 indicate that when PG serves as both a cementitious binder and an aggregate, the 28-day strength still reaches 39.3 MPa. This meets the GB 50010 − 2010 requirement (minimum C20 strength for reinforced concrete structures).

S1 and S2 differ in aggregate type but exhibit similar strengths, indicating that replacing stone aggregate with PGA does not significantly reduce the compressive strength of concrete. The primary reason is that the failure of the S1 and S2 samples is determined by the strength of the cement-PG paste. As shown in Fig. 8(a), most of the PGA in S1 is broken, suggesting that the strength of the PGA is close to that of the cement-PG paste. Compared with that in S2, the bond between PGA and cement paste in S1 is tighter, with better interfacial bonding, resulting in a slightly greater strength than that in S2.

S1 and S3 both use PGA but differ in cementitious binder. The strength of S3 is slightly greater than that of S1, indicating that the strength is mainly determined by the cement-PG paste and PGA strength. The strength of pure cement paste is greater than that of cement-PG paste.

S3 and S4 both use pure cement as the cementitious binder but differ in aggregate. The 28-day compressive strength results show that S4 is 65.8% greater than S3 is, indicating that the sample strength is primarily determined by the aggregate. The strength of PGA is lower than that of pure cement paste, whereas the strength of stone aggregate is greater than that of pure cement paste.

S2 and S4 both use stone as an aggregate, but S2 incorporates PG in the cementitious binder. The 28-day compressive strength results show that S4 is 70.0% greater than S2 is, indicating that the sample strength is mainly determined by the paste strength. The strength of pure cement paste is significantly greater than that of cement-PG paste.

From the above comparative analysis, it can be concluded that in concrete incorporating PG and PGA, the strength decreases in the following order: stone aggregate, pure cement paste, PGA, and cement-PG paste, with the strength of the cement-PG paste being close to that of PGA. This strength ranking provides guidance for configuring high-strength concrete incorporating PG.

Fig. 12

Compressive strength

3.7 Stress‒strain curves

To investigate the influence of PG on the constitutive relationship of concrete, concrete samples cured for 28 days were selected. A gradual increase in load was applied via a compression testing machine until the samples failed, and the pressure and deformation data were recorded. On the basis of these data, stress‒strain curves were plotted, as shown in Fig. 13. The figure illustrates the stress‒strain characteristics of concrete samples S1 to S4. All curves exhibit a development pattern similar to that of ordinary concrete and can be divided into four stages: the elastic stage, the elastoplastic stage, the yield stage, and the failure stage.

Fig. 13

Stress‒strain curves

Table 3
Mechanical test results
Samples	Modulus of elasticity /GPa	Maximum Stress /MPa	Strain Corresponding to Maximum Stress/ mm/mm ×10 ^− 3
S1	8.60	43.90	11.20
S2	9.55	43.36	9.71
S3	10.01	44.58	8.86
S4	18.66	73.69	8.35

In the elastic stage, the stress is directly proportional to the strain, allowing for the calculation of the elastic modulus of the concrete, as presented in Table 3. Specimen S4 has the highest elastic modulus at 18.66 GPa, whereas S1 has the lowest elastic modulus at 8.60 GPa. Compared with that of ordinary concrete, the elastic modulus of S1 is relatively low, primarily due to the weakened ITZ caused by the incorporation of PG.

In the elastoplastic stage, the relationship between stress and strain becomes nonlinear. As the load increases, the peak stress of the concrete increases. The peak stresses for S1 to S4 are 43.90 MPa, 43.36 MPa, 44.58 MPa, and 73.69 MPa, respectively, with corresponding peak strains of 11.20%, 9.71%, 8.86%, and 8.35%, respectively. The concrete samples subsequently enter the failure stage, where the stress sharply decreases, the strain significantly increases, and the samples exhibit noticeable cracking and fragmentation.

3.8 Softening coefficient

The softening coefficient reflects the extent to which the strength of concrete decreases after water absorption and is an important parameter for evaluating the durability of concrete. To investigate the effect of incorporating PG on the softening coefficient of concrete, the softening coefficients of four concrete samples were measured, as shown in Fig. 14.

Fig. 14

Softening coefficient

The softening coefficient of S1 is 0.93. Although PG was used as both the cementitious binder and aggregate in S1, its durability performance remained relatively good. The primary reason is that the hydration products, such as ettringite, generated during the hydration of PG fill the pores, increasing the compactness of the structure and thereby improving durability. The softening coefficient of S2 was 0.84, where PG was used as the cementitious binder and stone was used as the aggregate. The durability performance of this combination is moderate, mainly because the interfacial bonding between PG and the stone aggregate is not as tight as that between the PG aggregate and the PG cementitious binder in S1, leading to more pores and microcracks at the interface. The softening coefficient of S3 is the highest, reaching 0.97, indicating that the combination of cement as the cementitious binder and PG as the aggregate performs best in terms of durability. The main reason is that the hydration products of cement form strong bonds with PGA, reducing the porosity and microcracks at the interface, thereby improving its strength retention in wet environments. The softening coefficient of S4 is the lowest, at approximately 0.80. Although the hydration products of cement generally exhibit good bonding properties, the surface characteristics of stone aggregate may result in a lower interfacial bonding strength, affecting its strength retention in wet environments.

4 Conclusions

This study systematically explores the synergistic mechanism of PG as both a cementitious binder and an aggregate in concrete, revealing the impact of its dual-function application on material properties. Through comparative experiments, the effects of different PG incorporation methods on the microstructure, hydration products, and mechanical properties of concrete were analyzed. The main conclusions are as follows:

(1) By utilizing PG as both a cementitious binder and an aggregate, the utilization rate of PG was significantly improved, achieving a maximum incorporation of 38%. This effectively alleviates the environmental pressure caused by PG stockpiling.

(2) PG concrete exhibited excellent mechanical properties, successfully achieving a 28-day compressive strength of 39.3 MPa, which meets the strength standards for building structures. The softening coefficient of samples mixed with PG cementitious binder and PG aggregate reached 0.93, indicating that excellent strength stability can be maintained even in high-humidity environments.

(3) Comparative tests revealed that the strength of PG concrete, from highest to lowest, followed this order: the stone aggregate system, the pure cement paste system, the PGA system, and the cement‒PG composite paste system. These findings indicate that the compatibility between the strength of PG aggregates and the cementitious system is crucial for overall performance, providing a theoretical basis for the mix design of high-strength PG concrete.

(4) SEM analysis revealed that abundant hydration products, including ettringite and C-S-H gel, formed in the PG concrete. These products effectively filled the pores and increased the compactness of the ITZ, thereby improving the mechanical properties and durability of the concrete.

This study provides a scientific basis for the application of PG in concrete, demonstrating its significant potential as a sustainable building material. However, despite the broad prospects for PG application, there are still several limitations, such as sensitivity to environmental conditions and potential long-term durability issues, which require further research and resolution in future applications.

CRediT authorship contribution statement

Pan Chen: Resources, Writing—review & editing, Writing—original draft, Conceptualization. Zhexin Wang: Writing—review & editing, Methodology. Feng Zhu: Visualization, Data curation. Shujie Wan: Investigation, Data curation. Mengyang Huang: Methodology, Investigation, Data curation. Pengfei Liu: Data curation, Formal analysis. Dongxu Zhang: Visualization, Software. Cai Wu: Writing—review & editing. Yani Lu: Resources, Funding acquisition.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data Availability

The data will be made available upon request.

Funding

Declaration

The authors gratefully acknowledge the financial support provided by the Hubei Provincial Department of Education Scientific Research Project (Grant No. B2024149) and the National Natural Science Foundation of China (Grant No. 41907259) .

Author Contribution

P.C. conceptualized the study and wrote the original draft. P.C. and Z.W. reviewed and edited the manuscript. Z.W. and M.H. developed the methodology. F.Z. and D.Z. performed visualization. F.Z., S.W., M.H., and P.L. curated data. S.W. and M.H. conducted the investigation. P.L. performed formal analysis. D.Z. developed the software. C.W. reviewed and edited the manuscript. Y.L. provided resources and acquired funding. All authors reviewed the manuscript.

References

Gong, Y., Dong, S., Liu, L. & Wu, F. A sustainable composite cementitious material manufactured by phosphogypsum waste. Appl. Sci. 12, 12718 (2022).

Smadi, M. M., Haddad, R. H. & Akour, A. M. Potential use of phosphogypsum in concrete. Cem. Concrete Res. 29, 1419–1425 (1999).

Murali, G. & Azab, M. Recent research in utilization of phosphogypsum as building materials. J. Mater. Res. Technol. 25, 960–987 (2023).

Meng, W., Dong, Y., Zang, M., Zou, N. & Lu, H. Mechanical properties and chemical microscopic characteristics of all-solid-waste composite cementitious materials prepared from phosphogypsum and granulated blast furnace slag. Mater. Today Commun. 41, 110809 (2024).

Sun, T. et al. A new eco-friendly concrete made of high content phosphogypsum based aggregates and binder: Mechanical properties and environmental benefits. J. Clean. Prod. 400, 136555 (2023).

Hamada, H. M., Shi, J., Abed, F., Humada, A. M. & Majdi, A. Recycling solid waste to produce eco-friendly foamed concrete: A comprehensive review of approaches. J. Environ. Chem. Eng. 11, 111353 (2023).

Long, H., Gu, Z., Qin, Y., Xing, A. & Cheng, R. Synergistic preparation of cemented backfill from multiple coal-based solid wastes and investigation of its properties. Mater Today Commun 113025 (2025).

Yang, X. et al. Cementation characteristics and hydration mechanism of magnesium slag-phosphogypsum composite cementitious materials. Mater. Today Commun. 41, 110547 (2024).

Luo, Y. et al. Use of untreated phosphogypsum as a raw material for autoclaved aerated concrete preparation. J. Build. Eng. 64, 105607 (2023).

10.

Islam, G. S., Chowdhury, F. H., Raihan, M. T., Amit, S. K. S. & Islam, M. R. Effect of phosphogypsum on the properties of Portland cement. Procedia Eng. 171, 744–751 (2017).

11.

Hu, N. et al. Experimental design, mechanical performance and mechanism analysis of C30 phosphogypsum-based concrete. Constr. Build. Mater. 443, 137673 (2024).

12.

Levickaya, K., Alfimova, N., Nikulin, I., Kozhukhova, N. & Buryanov, A. The Use of Phosphogypsum as a Source of Raw Materials for Gypsum-Based Materials. Resources 13, 69 (2024).

13.

Yang, L., Yan, Y. & Hu, Z. Utilization of phosphogypsum for the preparation of non-autoclaved aerated concrete. Constr. Build. Mater. 44, 600–606 (2013).

14.

Pratap, B. Analysis of the synergistic effects of micro silica on the mechanical and durability properties of geopolymer concrete incorporating phosphogypsum. J. Build. Eng. 93, 109831 (2024).

15.

Kong, C. et al. Preparation and micromechanics of red sandstone–phosphogypsum–cement composite cementitious materials. Materials 16, 4549 (2023).

16.

Wang, C. Q. et al. Preparation and properties of high blending phosphogypsum-desulfurization ash-waste soil based functional prefabricated autoclaved aerated concrete slabs. Constr. Build. Mater. 423, 135879 (2024).

17.

Feng, Y. et al. Improving geopolymer concrete performance with hazardous solid waste phosphogypsum. J. Build. Eng. 95, 110333 (2024).

18.

Fu, R. et al. The effects of admixtures on the durability properties of phosphogypsum-based cementitious materials. Constr. Build. Mater. 411, 134379 (2024).

19.

Tian, T. et al. Utilization of original phosphogypsum for the preparation of foam concrete. Constr. Build. Mater. 115, 143–152 (2016).

20.

Huang, S. et al. Preparation of sustainable foamed concrete using high-volume phosphogypsum and its hydration mechanism investigation. J. Mater. Civil Eng. 36, 04024034 (2024).

21.

Ding, C., Sun, T., Shui, Z., Xie, Y. & Ye, Z. Physical properties, strength, and impurities stability of phosphogypsum-based cold-bonded aggregates. Constr. Build. Mater. 331, 127307 (2022).

22.

Sun, T., Xu, D., Ouyang, G., Wang, Z. & Ye, Z. Mechanical properties and environmental implications of excess-sulfate cement concrete with phosphogypsum-based cold-bonded fine aggregates. J. Build. Eng. 95, 110008 (2024).

Yes