Preparation of Red Mud-Electrolytic Manganese Residue Paste: Properties and Environmental Impact

Highlights

What are the main findings?

• When the RM-EMR mass ratio is 2:3 and the activator content reaches 60%, the optimal 28-day unconfined compressive strength reaches 35 MPa, and strength development follows a “rapid growth–gradual stabilization” pattern.
• EMR showed high NH₄⁺-N leaching performance (302 mg/L), and under the alkaline conditions induced by the activator (pH > 11), NH₄⁺ were converted to gaseous NH₃, and more than 50% of the composite test block leaching values met the regulatory limits.
• The RM-EMR system can effectively fix Mn and Cd, while Al and Se show high leaching behavior, especially the extremely high leaching concentrations of Al under water immersion conditions.

What are the implications of the main findings?

• These findings indicate that proper raw material ratio and activator dosage control can synergistically improve early/long-term strength, supplying key design parameters for solid waste-based cementitious material engineering use.
• The results show alkalinity regulation has a dual function: it may raise pH but reduces ammonia pollution via chemical transformation, offering new insights for environmental safety control.
• The results demonstrated that RM-EMR is effective in immobilizing heavy metals, and elevated Al/Se leaching requires raw material pretreatment (e.g., removing soluble Al), thus providing guidance for process refinement.

Abstract

Red mud (RM) and electrolytic manganese residue (EMR) possess inherently complementary acid–base characteristics, and their synergistic utilization offers a promising approach to simultaneously enhance mechanical performance and mitigate environmental risks. However, the environmental behavior of RM-EMR composites, particularly in terms of pH buffering, ammonium nitrogen (NH₄⁺-N) control, and heavy metal mobility, remains inadequately understood. In this study, a composite cementitious paste was developed using RM and EMR, and the effects of the RM-to-EMR ratio (1: 3, 2: 3, 1: 1, and 3: 2), alkali activator dosage (30%, 40%, 50% and 60% by weight), and curing time (3 day, 7 day, 14 day, and 28 day) under standard curing conditions on unconfined compressive strength (UCS) were systematically evaluated. Leaching tests were conducted to assess pH evolution, NH₄⁺-N release, and heavy metal migration. The results showed that the optimal 28-day UCS of 35 MPa was achieved with an RM-to-EMR mass ratio of 2:3 and an activator dosage of 60%. EMR contributed to NH₄⁺-N leaching concentrations as high as 302 mg/L; however, under alkaline conditions (pH > 11), over 50% of the block samples met regulatory limits due to the transformation of NH₄⁺ into gaseous NH₃. Furthermore, Mn and Cd were effectively immobilized. In contrast, Al and Se exhibited elevated leaching, with Al showing particularly high concentrations under water leaching conditions. These results underscore the importance of raw material pretreatment and system optimization. Overall, this study provides new insights into the environmental behavior and safe resource utilization of RM and EMR in cementitious systems.

1. Introduction

Cement, as the most widely used binder in the construction industry, has long been a major contributor to global carbon emissions [1]. It is estimated that the production of one ton of cement releases approximately 0.85 tons of CO₂ into the atmosphere [2], posing a significant obstacle to achieving global carbon neutrality goals. In response, the development of green, low-carbon alternatives to traditional cementitious materials has become both a pressing research priority and an engineering necessity. Among various options, geopolymer-based binders synthesized from industrial solid wastes have garnered increasing attention in recent years. Compared with ordinary Portland cement, geopolymers can reduce carbon emissions by 80–90% and exhibit superior chemical stability and resistance to corrosion [3]. These materials are typically composed of aluminosilicate-rich precursors that form a stable three-dimensional network structure under alkali activation, resulting in excellent mechanical performance and environmental durability.

Red mud (RM), a highly alkaline industrial by-product generated during alumina extraction via the Bayer process, represents a critical challenge for solid waste management. For every ton of alumina produced, approximately 1–2 tons of RM are generated [4]. With expanding bauxite mining activities and declining ore grades, global RM production continues to rise, leading to a cumulative stockpile of over 4.5 billion tons by 2023 [5]. RM is predominantly disposed of through landfilling, which not only occupies vast areas of land but also poses severe environmental hazards due to its high alkalinity (pH 10–13), including alkali leaching, heavy metal migration, and long-term contamination of surrounding soil and water systems [6]. In some cases, RM stockpiles have reached heights exceeding 40 m, elevating the risk of geological disasters such as dam failure and slope instability. To mitigate the environmental risks and promote the resource utilization of RM, researchers worldwide have explored diverse reuse strategies, such as recovery of valuable metals, preparation of construction materials, pollutant remediation, and carbon capture and sequestration [7]. For instance, Agrawal et al. extracted titanium and scandium from RM using a microwave acid-baking and water-leaching method, achieving leaching efficiencies of 73.3% and 88.4%, respectively [8]. Zhao et al. developed non-fired bricks using 70% RM and sulfoaluminate cement, whose mechanical and environmental properties met relevant standards [9]. Wang et al. formulated RM-based cementitious backfill material that, under optimal conditions (60% slurry concentration, RM: fly ash (FA) = 3:2, 5% cement addition), achieved 3-day and 28-day compressive strengths of 1.25 MPa and 2.37 MPa, respectively, with a low permeability coefficient of 6.005 × 10⁻⁶ cm/s [10]. Similarly, Nikbin et al. partially replaced cement with 25% RM to produce lightweight concrete (density: 1685–1789 kg/m³) with satisfactory compressive strength and significantly reduced carbon emissions. Compared with the control (0% RM), the RM addition reduced CO, NO_x, Pb, and SO₂ emissions by approximately 32.5%, 17.1%, 31.8%, and 22.4%, respectively [11]. A life cycle assessment conducted by Morsali et al. further demonstrated that RM-based concrete could reduce 18 environmental impact indicators, including human carcinogenic toxicity [12]. In the context of cementitious materials, Li et al. reported that optimizing key parameters, such as precursor composition, alkali type, and curing regime, can significantly enhance the mechanical and environmental performance of RM-based geopolymeric binders, enabling their high-volume utilization [13]. RM has also demonstrated potential in carbon capture and sequestration (CCUS) applications. Liu et al. found that carbonation of RM can fix between 5 and 175 g of CO₂ per kilogram of material, primarily through the formation of stable carbonates from alkaline components such as CaO and Na₂O [14]. Furthermore, increasing attention is being paid to the synergistic utilization of RM with other solid wastes. For example, Ma et al. showed that the high alkalinity of RM can be effectively moderated via acid neutralization, CO₂ treatment, or salt replacement, while residual free alkalis may serve as catalytic agents in cement or geopolymer reactions [15]. Synergistic use strategies are also being clarified through studies such as that of Zhang et al., who combined RM with coal gangue to produce a novel binder [16]. They identified key hydration products-calcium silicate aluminate hydrate (C-A-S-H) gels with amorphous-nanocrystalline composite structures and ettringite (AFt) phases-and proposed a four-stage hydration model, offering a theoretical basis for multi-waste synergy. Despite these promising developments, the overall resource utilization rate of RM remains below 10% [17]. Persistent challenges such as alkali metal migration, long-term environmental stability, and product adaptability continue to hinder its large-scale application [18]. Therefore, the synergistic preparation of building materials using RM in combination with other industrial solid wastes, such as EMR, is expected not only to improve resource utilization efficiency but also to offer viable pathways for the co-disposal and high-value reuse of multiple hazardous wastes.

EMR is a bulk acidic solid waste generated during the electrolytic production of manganese metal. Its primary components are silica (SiO₂), alumina (Al₂O₃), and gypsum (CaSO₄·2H₂O). EMR is typically characterized by high moisture content, high viscosity, and low reactivity. In addition, it contains harmful substances such as ammonia nitrogen (NH₃-N), soluble manganese, and various heavy metals. The massive accumulation of EMR not only consumes valuable land resources but also poses significant environmental risks to soil and water systems. As such, EMR has become a critical bottleneck impeding the green transformation and high-quality development of the electrolytic manganese industry. In response to these challenges, increasing attention has been directed towards the resource utilization of EMR. In recent years, a variety of studies have explored its incorporation into building materials and functional materials. In the field of building materials, EMR has shown promise due to its potential pozzolanic activity and capacity for heavy metal immobilization. For example, Wu et al. developed geopolymer binders (GOCM and ESGM) using EMR, achieving 28-day compressive strengths of 22.8–23.0 MPa [3,19]. Mi et al. optimized the EMR-coal gangue-fly ash ratio (0.43:0.34:0.23) using the Design-Expert software (V8.0.6.1). Under a liquid-solid ratio of 0.9 and curing at 60 °C for 24 h, the geopolymer reached a 28-day strength of 12.0 MPa, with manganese leaching concentration as low as 0.123 mg/L. Further efforts have been made to prepare cementitious materials based on EMR [20]. He et al. synergistically used EMR with barium slag, limestone, and bauxite to synthesize belite-calcium sulfoaluminate cement (B-CSA). By utilizing low-melting-point components in EMR, they reduced the clinker firing temperature to 1300–1350 °C while achieving a 28-day strength of up to 60 MPa [21,22,23]. Wang et al. produced belite-ye’elimite-ferrite (BYF) cement by calcining a mixture of 45.5% EMR with CaO and Al₂O₃ at 1200 °C, achieving compressive strengths of 45.9 MPa and 107 MPa at 3 and 60 days, respectively, while significantly reducing CO₂ emissions (0.58 kg/kg) and production costs [24]. Beyond construction materials, EMR has demonstrated promising applications in the field of functional materials. For instance, Tang et al. fabricated a Mn₃O₄-Fe₃O₄ magnetic catalyst from manganese slag, capable of removing 97% of enrofloxacin from wastewater within 30 min [25]. Tang et al. further proposed a selective leaching-solvent extraction-adsorption process, achieving recovery rates of 95% for Mn and 98% for As [26]. Ma et al. synthesized a dual-network hydrogel from NaOH-activated EMR and polyacrylic acid, which exhibited high adsorption capacities for Pb²+ (153.85 mg/g), Cd²+ (113.63 mg/g), and Cu²⁺ (54.35 mg/g) [27]. In another study, Shu et al. applied iron-rich EMR in combination with an electric field to treat phosphogypsum leachate, achieving over 99.8% removal of Mn²⁺, PO₄³⁻, and F-, and 58.4% removal of NH₄⁺-N [28]. Zhu et al. prepared monolithic catalysts by coating Cu-Ce-doped EMR with a silica-alumina sol, which achieved 99% removal of toluene at 240 °C and complete removal at 270 °C [29]. EMR has also been utilized in producing aggregates [30], ceramics [31], and no-burn bricks [32]. This application provides a new idea to achieve simultaneous solidification of manganese and ammonia nitrogen. Wang et al. (2020) investigated the migration and transformation mechanisms of Mn and NH₃-N in EMR through pH-dependent leaching experiments and geochemical modeling [33]. The results indicated that Mn primarily exists in the acid-extractable state, with its solubility governed by the formation of rhodochrosite and pyrolusite, while NH₃-N is released as NH₃ gas under alkaline conditions. Despite these advances, the overall level of EMR resource utilization remains low [33]. Current approaches are often limited by inefficiencies, high costs, or lack of synergy among multiple objectives. Therefore, there is an urgent need to establish a more efficient, synergistic, and low-carbon pathway for EMR resource utilization.

Despite extensive studies on RM- and EMR-based construction materials, most research has treated these two industrial residues as independent precursors, focusing primarily on mechanical performance enhancement. From an environmental chemistry perspective, however, RM and EMR exhibit inherently complementary acid–base characteristics: RM is strongly alkaline (pH 10–13), whereas EMR is typically acidic due to residual sulfate and ammonium-bearing species. This complementarity provides a unique opportunity to regulate the pH evolution of composite systems, which is a key factor controlling ammonium nitrogen transformation, heavy metal solubility, and long-term leaching behavior. Nevertheless, the deliberate coupling of RM and EMR through acid–base regulation and its environmental implications remain poorly understood, and research on their synergistic utilization is still in its early stage. Most existing studies focus on the isolated use of individual solid wastes, which often cannot simultaneously satisfy mechanical performance and environmental safety requirements. Recently, however, increasing attention has been paid to exploiting the complementary characteristics of RM and EMR for synergistic performance enhancement. For example, Zhou et al. prepared a composite mortar using RM, EMR, and steel slag (SS), and found that when the mixture contained 12% RM, 3% EMR, and 15% SS, the 28-day flexural and compressive strengths reached 7.2 MPa and 41.4 MPa, representing improvements of 18.0% and 25.5%, respectively, compared with mortar containing SS alone. Mechanistic analysis indicated that the free alkali in RM and sulfate in EMR synergistically activated the SS, promoting the formation of ettringite (AFt) and calcium silicate hydrate (C-S-H) gels. This led to significant microstructural optimization, including a 26.35% reduction in harmful pore volume and a 43.57% increase in harmless pores [17]. Tian et al. investigated the synergistic preparation of low-carbon limestone calcined clay cement (LC³) using a 1:1 RM/EMR blend thermally activated at 600 °C, combined with cement clinker, limestone powder, and gypsum. The resulting binder exhibited a 28-day compressive strength of 43.4 MPa (at a water-cement ratio of 0.35), while CO₂ emissions and energy consumption were reduced by 26.2% and 14.9%, respectively. Microstructural analysis revealed that reactive SiO₂, Al₂O₃, and CaSO₄ from the solid waste blend jointly promoted the formation of monocarboaluminate and secondary hydration products, thereby enhancing the mechanical performance and durability of the cementitious system [34]. Hu et al. developed an RM-EMR composite modifier and demonstrated that asphalt modified with a 1:1.5 RM/EMR ratio at a 15% dosage exhibited optimal penetration, softening point, and ductility, indicating favorable road performance. Molecular dynamics simulations further confirmed that the RM-EMR composite improved asphalt-filler interfacial bonding through filling and adsorption mechanisms, significantly enhancing interfacial interaction [35]. Peng et al. developed a multi-solid-waste-based cementitious material by combining RM, EMR, and ground granulated blast furnace slag (GGBS). Under optimal conditions (20% EMR, 15% RM, 52% GGBS, 13% Ca(OH)₂, and a water-cement ratio of 0.5), the 28-day compressive and flexural strengths reached 27.9 MPa and 7.5 MPa, respectively. The main hydration products were AFt and C-S-H. And the composite gelling system can effectively solidify and stabilize the harmful substances in EMR and RM, reducing the harm to the environment, and is thus an environmentally friendly cementitious material [36]. Previous studies have demonstrated the potential of multi-solid-waste integration for both mechanical performance and environmental impact reduction. For instance, a low-carbon mortar incorporating RM, ground coal bottom ash, marble dust, and recycled turbine blade fibers achieved approximately 20% improvements in compressive, tensile, and flexural strengths at moderate cement replacement levels, while reducing water absorption and minimizing environmental impacts such as acidification and climate change [37]. Similarly, red mud–slag composites optimized via response surface methodology showed enhanced 28-day flexural and compressive strengths (12.26 MPa and 69.83 MPa) and improved microstructural properties through synergistic formation of C–S–H and C–A–S–H gels at the fiber–matrix interface [38].

These studies on RM- and EMR-based construction materials have mainly focused on mechanical performance, microstructural optimization, or life cycle benefits. However, they often target only a limited range of contaminants, and the stabilization behavior of multiple harmful species under varying leaching conditions remains insufficiently understood. Moreover, RM and EMR are typically treated as independent precursors, overlooking their complementary acid–base characteristics, which could be leveraged to regulate pH evolution and thereby influence both mechanical performance and environmental risk. As a result, the coupled effects of pH regulation, alkali activation, and multi-contaminant stabilization have not yet been systematically clarified in RM–EMR composite systems, limiting their environmentally safe large-scale application. Against this backdrop, the present study employs RM and EMR as primary raw materials to develop a composite cement-neat paste system. The effects of the RM/EMR ratio, alkali activator dosage, and curing time on UCS are systematically investigated. In parallel, acid and water leaching tests are conducted to evaluate the pH behavior of raw materials and composites, the release of ammonium nitrogen (NH4+-N), and the leaching characteristics of heavy metals. This study aims to advance the synergistic resource utilization of RM and EMR, mitigate environmental hazards such as alkaline leachate and heavy metal contamination from stockpiled wastes, and provide both theoretical insights and practical strategies for the co-disposal and high-value utilization of industrial solid wastes.

2. Materials and Methods

2.1. Materials

The raw materials used in this study included RM sourced from the Pingguo Aluminum Plant, EMR obtained from the Guangxi Da Manganese Slag Yard in Jingxi City, and a self-developed alkali activator. The mineralogical compositions of RM and EMR were identified by X-ray diffraction (XRD), as shown in Figure 1. The XRD pattern of RM indicates that aragonite and calcite are the dominant crystalline phases, together accounting for approximately 60–65% of the mineral assemblage. In addition, minor phases such as opal (amorphous SiO₂), gibbsite (Al(OH)₃), and goethite (FeOOH) were also detected. The coexistence of carbonate minerals and reactive aluminosilicate phases suggests that RM can simultaneously contribute alkalinity buffering and potential reactivity in alkali-activated systems. For EMR, the XRD results reveal a more complex and less crystalline mineralogical structure. The primary phases include quartz (SiO₂) and gypsum (CaSO₄·2H₂O), accompanied by minor amounts of iron-bearing phases and amorphous components. This indicates that its silica-rich framework and sulfate-bearing phases play an important role in regulating alkalinity and participating in hydration or secondary reaction processes.

The BET surface area of RM typically ranges from approximately 60 to over 180 m²/g, depending on its processing history and mineralogical composition, which can significantly affect dissolution kinetics, alkali consumption, and subsequent reaction product formation. In contrast, EMR exhibits a relatively low to moderate BET surface area of 11.8–35.6 m²/g, primarily due to the prevalence of crystalline phases such as quartz and gypsum. The specific gravities of RM and EMR are 2.8 and 2.1, respectively. Prior to use, both RM and EMR were subjected to mechanical grinding for 30 min to ensure adequate fineness. After grinding, the residue on an 80 μm sieve was controlled to be ≤10.0 wt%, thereby minimizing particle-size-related variability and enhancing the uniformity of reaction and leaching behavior. The chemical compositions of the raw materials are presented in

Table 1. Chemical composition of RM and EMR (wt%).

Title 1	RM	EMR
Calcium oxide (CaO)	16.2	9.3
Quartz (SiO2)	13.7	35.04
Aluminium oxide/alumina (Al2O3)	17.6	5.97
Ferric oxide (Fe2O3)	26.7	12.9
Sulphur oxide (SO3)	0.2	13.67
Anatase (TiO2)	5.3	-
Manganese dioxide (MnO2)	-	7.6
Sodium oxide (Na2O)	5.9	-
Magnesium oxide (MgO)	0.7	-
Potassium oxide (K2O)	0.1	-
Chromium(III) oxide (Cr2O3)	0.2	-
LOI	11.6	15.52

. The RM was characterized by a combined SiO₂ and Al₂O₃ content of 31.3 wt%, indicating favorable pozzolanic and cementitious properties. EMR, as a typical acidic industrial solid waste, is enriched in SiO₂, Fe₂O₃, and SO₃, with their total content reaching approximately 61.61 wt%, which provides both silica sources and sulfate species relevant to alkalinity regulation and hydration product evolution.

2.2. Mix Proportions

The RM-EMR paste (REP) was prepared by incorporating a self-developed alkali activator into the raw material blend. The activator consisted of a blended powder system composed of industrial solid wastes—including ground granulated blast-furnace slag, fly ash, steel slag powder, silica fume, gypsum, and limestone powder—and supplementary alkaline and mineral components from the construction and chemical industries, namely quicklime, cement clinker, sodium sulfate, sodium silicate, and calcium chloride. In this activator system, alkaline species (primarily supplied by sodium silicate and quicklime) were responsible for establishing a high-pH environment and promoting aluminosilicate dissolution, while calcium-bearing phases (e.g., quicklime, clinker, steel slag, and limestone powder) facilitated the formation of calcium silicate hydrate (C–S–H) and calcium aluminosilicate hydrate (C–A–S–H) gels. Sulfate sources (gypsum and sodium sulfate) further contributed to ettringite (AFt) formation and pH buffering, thereby influencing both strength development and contaminant immobilization. The detailed mix proportions of REP samples are listed in

Table 2. Constituent proportions of REP mixtures.

Mixture	RM (wt%)	EMR (wt%)	Activator (wt%)	RM: EMR	Water-to-Binder Ratio	Remarks
P1	35%	35%	30%	1: 1	0.4
P2	15%	45%	40%	1: 3	0.4
P3	24%	36%	40%	2: 3	0.4
P4	30%	30%	40%	1: 1	0.4	Environmental analysis
P5	36%	24%	40%	3: 2	0.4
P6	12.5%	37.5%	50%	1: 3	0.4
P7	20%	30%	50%	2: 3	0.4
P8	25%	25%	50%	1: 1	0.4
P9	30%	20%	50%	3: 2	0.4
P10	10%	30%	60%	1: 3	0.4
P11	16%	24%	60%	2: 3	0.4
P12	20%	20%	60%	1: 1	0.4
P13	24%	16%	60%	3: 2	0.4

. Twelve test groups were designed to investigate the effects of activator content (30 wt%, 40 wt%, 50 wt%, and 60 wt%), RM-to-EMR mass ratios (1: 3, 2: 3, 1: 1, and 3: 2), and curing time (3 day, 7 day, 14 day, and 28 day) on UCS of the REP system. The water-to-binder ratio (defined as the mass of water relative to the total mass of RM, EMR, and activator) was 0.4 for all mixtures. To further assess the environmental performance of the system, samples with the P4 mix proportion (CMJ-08 to CMJ-13) were prepared for leaching and pH analysis. In addition, the individual raw materials, including RM, EMR, and the activator components CJ-G, CJ-M, CJ-F, and CJ-L, were also subjected to environmental testing. Notably, sample CMJ-08 represents a paste formulation with a pH value of 8, and the remaining samples follow a similar naming convention based on their designed pH levels.

2.3. Specimen Preparation

Based on the mix proportions detailed in Table 2, RM, EMR, and the self-developed activator were first dry-mixed to achieve uniform blending. Subsequently, a predetermined amount of water was added, and the mixture was thoroughly stirred until a homogeneous and consistent paste was obtained. During the mixing process, particular attention was paid to ensuring uniform dispersion and eliminating visible agglomerates.

The freshly prepared paste was subjected to a flowability test and then immediately cast into standard cubic molds with dimensions of 100 mm × 100 mm × 100 mm using vibration compaction, as illustrated in Figure 2. After casting, the surface of the specimens was leveled to achieve a smooth and flat finish. The molded specimens were then placed in a standard curing chamber under controlled conditions for subsequent curing.

2.4. Test Methods

The UCS of the REP specimens was measured using a 20 kN microcomputer-controlled electronic universal testing machine (Model LD24.204, Lishi Scientific Instruments Co., Ltd., Shanghai, China). Each specimen was subjected to axial loading until failure, and the peak load recorded at failure was used to calculate the UCS. To ensure data accuracy and reliability, a minimum of three replicate specimens were tested for each mix group, and the average value was reported as the final UCS result.

The pH values of the leachates from the REP samples, as well as from the individual raw materials (RM, EMR, and activator components), were determined in accordance with the Chinese national standard GB/T 15555.12-1995 Solid waste-Glass electrode test-Method of corrosivity [40]. Each sample was tested in triplicate using a calibrated glass electrode pH meter. The absolute standard deviation of parallel measurements was controlled within 0.15 pH units, and the arithmetic mean of the three measurements was adopted as the final result. The pH values serve as a key indicator of the leachate’s acidity or alkalinity and provide a basis for subsequent heavy metal leaching assessments.

The leaching concentrations of metal elements were determined in accordance with the standard HJ 766-2015 Solid Waste-Determination of metals-Inductively coupled plasma mass spectrometry (ICP-MS) [41]. For the test, REP samples were evaluated in their intact solid block form, while RM, EMR, and activator components were tested in powdered form. Each sample was fully immersed in both neutral (pH 7.0 ± 0.2) and acidic (pH 4.0 ± 0.2) leaching solutions and subjected to a continuous toxicity leaching procedure to simulate environmental exposure conditions. The tests were conducted under ambient laboratory conditions, with temperatures ranging from 22.2 °C to 25.4 °C and relative humidity maintained between 51% and 63%. Following leaching, the supernatants were filtered and analyzed using inductively coupled plasma mass spectrometry (ICP-MS) to determine the concentrations of dissolved metal elements. The results provide critical information on the release behavior of potentially hazardous metals from REP and its constituent materials, offering insight into the environmental safety of the composite system.

3. Results and Discussion

3.1. Unconfined Compressive Strength

3.1.1. The Influence of Activator Proportion

Figure 3 illustrates the effect of activator proportions on UCS of the neat paste at an RM:EMR ratio of 1:1. As shown in Figure 3, the UCS values at 3 d, 7 d, and 14 d increased progressively with increasing activator proportion. This trend can be attributed to the combined action of alkaline species and calcium-bearing phases in the self-developed activator system. Specifically, alkaline components, mainly sodium silicate and quicklime, establish a high-pH environment that promotes the dissolution of aluminosilicate phases from RM and EMR, while calcium-rich constituents, such as quicklime, cement clinker, steel slag, and limestone powder, facilitate the precipitation of C–S–H and C–A–S–H gels, thereby accelerating early strength development [17].

However, the 28 d UCS decreased when the activator proportion reached 50%. This phenomenon can be explained by an imbalance between alkali concentration and the availability of reactive Si and Al species in the system. Under such conditions, excessive alkalinity may induce rapid precipitation of low-polymerization gels, leading to a less homogeneous and relatively porous microstructure, which limits the contribution to long-term strength. Similar behavior has been reported in RM–EMR-based alkali-activated systems, where excessive alkali input without sufficient reactive precursors resulted in restrained later-age strength development. This is consistent with reference [17].

The highest 28 day strength (35 MPa) was observed at an activator proportion of 60%, whereas the lowest value (20 MPa) occurred at 30%. This strength enhancement suggests that a sufficiently high dosage of the composite activator not only sustained the alkaline environment but also supplied additional calcium- and sulfate-bearing phases. The presence of gypsum and sodium sulfate promoted AFt formation and contributed to pH buffering, while the increased calcium availability favored the continuous formation and densification of C–(A)–S–H gels. As a result, the synergistic effects among alkaline species, calcium sources, and sulfate phases led to a denser gel network and improved long-term load-bearing capacity, which aligns with the mechanisms of hydration promotion and microstructure optimization reported in RM–EMR composite cementitious materials [17].

3.1.2. The Influence of Curing Time

Figure 4 presents the evolution of UCS of the neat paste with curing age at an RM:EMR ratio of 1:1. As shown in Figure 4, the UCS increased rapidly during the early curing period, followed by a gradual slowdown as the curing age progressed. The 7 d to 28 d strength ratios of the neat paste were 73.5%, 55.6%, 64.8%, and 75.4% at activator percentages of 30%, 40%, 50%, and 60%, respectively, indicating that a large proportion of the final strength was developed during the early curing period. This strength evolution pattern is consistent with previous studies on RM-containing cementitious materials, which reported that the active SiO₂ and Al₂O₃ in RM mainly participate in secondary hydration reactions at early curing stages, contributing significantly to early strength gain [34]. As curing time increases, the contribution of RM to further strength development becomes limited due to its relatively low intrinsic reactivity and the gradual depletion of early-reacting phases, resulting in a reduced strength growth rate at later ages [34]. Moreover, RM–EMR–dominated systems contain a considerable fraction of inert mineral phases, such as Fe₂O₃ (Table 1), which do not continuously participate in hydration reactions. With increasing curing time, the accumulation of these inert components limits the formation of additional cementitious products, thereby slowing down the strength development at later ages [36].

3.1.3. The Influence of RM to EMR Ratio

Figure 5 illustrates the variation in UCS of the neat paste with the different RM:EMR ratios at a fixed activator percentage of 50%. As shown in Figure 5, the UCS values at 3 d, 7 d, and 14 d increased with increasing RM:EMR ratio, indicating that RM plays a dominant role in early-age strength development. This behavior can be primarily attributed to the alkaline nature of RM. When the RM content is within an appropriate range, the increase in RM dosage raises the alkali concentration of the binder system, thereby promoting geopolymerization or hydration reactions and generating more cementitious products [36]. However, the 28 d UCS decreased at RM:EMR ratios of 1:1 and 3:2. The highest 28 d UCS (32.5 MPa) was achieved at an RM:EMR ratio of 2:3, while the lowest value (22.3 MPa) was recorded at a ratio of 1:3. This non-monotonic evolution suggests that when the RM content becomes excessive, the proportion of inert components in the system increases, which suppresses the participation of active SiO₂ and Al₂O₃ in hydration reactions and reduces the formation of effective binding gels. This dilution effect explains the observed reduction in long-term strength at high RM:EMR ratios. RM is characterized by relatively low intrinsic activity compared with highly reactive precursors such as GGBS, and its primary mineral components are dominated by hematite (Fe₂O₃), with limited reactive silica–alumina phases [36]. In addition, RM exhibits limited pozzolanic activity at later curing stages. Although active SiO₂ and Al₂O₃ in RM can participate in secondary hydration reactions at early ages, their contribution diminishes with prolonged curing time, leading to a decrease in strength development rate at later ages [34]. This characteristic further explains why increasing RM content does not continuously enhance the 28 d UCS.

By contrast, EMR exhibits markedly lower pozzolanic activity, with the strength activity index (SAI) consistently remaining below 65%, indicating that only a limited fraction of its mineral phases actively participate in hydration reactions, while quartz remains largely inert [34]. Nevertheless, when RM and EMR are combined at suitable proportions, their complementary chemical characteristics can be beneficial. Specifically, the free alkali released from RM and the sulfate-containing phases in EMR can jointly regulate the reaction environment and stimulate the hydration activity of other mineral components, contributing to matrix densification and strength development [17]. As a result, an RM:EMR ratio of 2:3 provides a more favorable balance between alkalinity enhancement and inert phase dilution. This balanced composition mitigates the adverse effect of excessive RM while maintaining sufficient alkalinity and particle-filling effects, leading to sustained strength development and the highest 28 day UCS of 32.5 MPa. Similar synergistic effects between RM and EMR at appropriate proportions have also been reported in composite binder systems, where optimal RM–EMR contents resulted in superior mechanical performance [17].

3.2. Environmental Impact

3.2.1. Leachate pH

Figure 6 presents the pH values of leachates from raw materials and the RM-EMR synergistic neat paste after water leaching. As shown in Figure 6a, the leachates of the activator raw materials (CJ-G, CJ-M, and CJ-L) exhibited strongly alkaline characteristics, with pH values above 11. In contrast, the leachates from RM and EMR were alkaline (pH 10.77) and mildly acidic (pH 6.95), respectively. As shown in Figure 6b, all neat paste specimens exhibited alkaline leachates, which is consistent with the intentionally controlled alkaline conditions (initial pH 8–13) adopted during paste preparation. The maintenance of an alkaline leaching environment indicates the continuous presence of alkaline and calcium-bearing hydration products within the hardened matrix. Previous studies on RM–EMR-based composite cementitious materials have demonstrated that the formation of hydration products such as C–S–H and AFt contributes to matrix densification and provides a stable alkaline buffering capacity, which is beneficial for the stabilization of hazardous constituents [36]. Moreover, a positive correlation was observed between the initial alkalinity and the leachate pH, with powdered samples yielding higher pH values than block-shaped specimens. This behavior can be attributed to the larger specific surface area of powdered samples, which facilitates the dissolution of alkaline phases and accelerates ion exchange between the solid matrix and the leaching solution.

3.2.2. NH₄⁺-N Concentration

Figure 7 illustrates the NH₄⁺-N leaching concentrations of raw materials and the RM-EMR synergistic neat paste after water leaching. As shown in Figure 7a, EMR exhibited a significantly high NH₄⁺-N leaching concentration, reaching up to 302 mg/L, which is associated with the presence of ammonium-bearing residues generated during the electrolytic manganese production process. In contrast, the NH₄⁺-N concentrations of the other raw materials were all below the 0.5 mg/L threshold specified for Class III groundwater in the GB/T 14848-2017 Standard for Groundwater Quality [42]. Figure 7b shows that in powdered specimens, all neat paste samples exhibited NH₄⁺-N leaching concentrations exceeding the standard, ranging from 26.4 to 104 mg/L. In contrast, for block specimens, only CMJ-08, CMJ-09, CMJ-10, and CMJ-11 exceeded the standard, with concentrations between 2.12 and 7.09 mg/L. This phenomenon is attributed to the presence of alkaline compounds in RM (e.g., sodium hydroxide and calcium hydroxide), which release hydroxide ions (OH^-) upon dissolution. These hydroxide ions react with ammonium ions (NH₄⁺), which are relatively stable in aqueous environments, to form ammonium hydroxide (NH₃·H₂O), an unstable compound. Upon heating or agitation, ammonium hydroxide decomposes into ammonia gas (NH₃) and water (H₂O). When the system pH exceeds 9, the concentration of NH₃–N approaches a minimum due to volatilization of NH₃, as described in Equation (1) [33]. Consequently, systems with higher hydroxide ion (OH⁻) availability generally exhibit lower residual concentrations of NH₄⁺–N following leaching.

NH₄⁺ + OH^- → NH₃·H₂O → NH₃ + H₂O

(1)

Although EMR exhibited a high initial NH₄⁺-N concentration (302 mg/L), the alkalinity of RM is insufficient to completely eliminate ammonia nitrogen, resulting in residual NH₄⁺-N. Notably, the NH₄⁺–N leaching concentrations of specimens CMJ-12 and CMJ-13 remained within the regulatory limits, indicating that enhanced system alkalinity, together with improved matrix densification, plays a critical role in suppressing ammonia nitrogen release. Similar trends of reduced pollutant leaching associated with increasing curing age and matrix densification have also been reported for RM–EMR-based composite cementitious materials [36].

3.2.3. Leaching Toxicity

Figure 8 shows the leaching concentrations of F, Hg, As, Se, Al, and Cr⁶⁺ from the raw materials under both water leaching and acid leaching conditions. All measured concentrations were evaluated by direct comparison with the threshold values specified for Class III groundwater in the GB/T 14848–2017 Standard for Groundwater Quality, which is widely adopted in China for assessing environmental safety under potential groundwater exposure scenarios. As illustrated in Figure 8, the leaching concentrations of all six elements (F, Hg, As, Se, Al, and Cr⁶⁺) in RM exceeded the corresponding regulatory limits. Similarly, EMR exhibited excessive leaching concentrations of F, Hg, Se, and Al. Among the activator raw materials, CJ-G showed elevated levels of Hg, Se, and Cr⁶⁺; CJ-M exceeded the limits for Hg, Se, and Al; CJ-F showed high concentrations of F, Hg, and Se; while CJ-L exceeded the standard only for Hg. These results indicate that, when assessed as individual raw materials, both RM, EMR, and certain activators pose potential environmental risks if directly exposed to leaching environments, thereby highlighting the necessity of stabilization and solidification prior to engineering application.

Figure 9 and

Table 3. Related materials (test block as a whole) water leaching results (mg/L).

Number	Corrosive	Cr6+	NH4+-N	Mn	As	Cd	Pb	Hg
Groundwater Class III standard		0.05	0.5	0.1	0.01	0.005	0.01	0.001
1# (pH8)	10.03	ND	6.81	ND	0.0012	ND	ND	ND
2# (pH9)	10.1	ND	7.09	0.0052	0.0019	ND	ND	ND
3# (pH10)	10.32	ND	5.26	ND	0.001	ND	ND	ND
4# (pH11)	10.7	ND	2.12	ND	0.0012	ND	ND	ND
5# (pH12)	10.75	ND	0.421	ND	0.0012	ND	ND	ND
6# (pH13)	11.04	ND	0.404	ND	ND	ND	ND	0.00004

illustrate the leaching concentrations of F, Hg, As, Se, Al, and Cr⁶⁺ (ND, not plotted) from RM/EMR synergistic neat paste test blocks under water leaching and acid leaching conditions. In comparison with the raw materials, the leaching concentrations of most hazardous elements were markedly reduced after solidification, indicating the effective immobilization capacity of the RM–EMR synergistic system. This behavior is consistent with previous findings, which have demonstrated that composite cementitious materials derived from RM and EMR can efficiently stabilize hazardous elements through a combination of adsorption, physical encapsulation, and chemical incorporation into hydration products. Moreover, the immobilization efficiency has been reported to increase with curing age as the cementitious matrix becomes progressively denser [36].

As shown in Figure 9, the Al content in all test blocks (CMJ-08 to CMJ-13) exceeded the regulatory limits in the powdered water leaching tests. This can be attributed to the high Al contents of the RM and CJ-M raw materials. Specifically, as shown in Figure 8e, the Al contents in RM and CJ-M reached 9430 μg/L and 2930 μg/L, respectively, indicating that the elevated Al levels in the leachates primarily originated from these components. Moreover, the Al content in the water-leached samples were consistently higher than those observed in the acid-leached counterparts. This difference can be explained by the chemical forms of Al present in RM, which predominantly exist as diaspore (β-AlOOH) and alumina (Al₂O₃). Under weakly alkaline conditions, these phases remain largely insoluble. However, in strongly alkaline environments, diaspore and alumina can partially dissolve to form sodium meta-aluminate (NaAlO₂) or NaAl(OH)₄ [43], as described by Equations (2) and (3). The formation of soluble aluminate species facilitates the release of Al into the aqueous phase, resulting in elevated Al concentrations during water leaching.

AlO(OH) +NaOH → NaAlO₂+ H₂O

(2)

Al₂O₃ + 2NaOH → 2NaAlO₂ + H₂O

(3)

Furthermore, during the acid leaching process, NaAlO₂ undergoes a neutralization reaction with sulfuric acid (H₂SO₄), producing sodium sulfate (Na₂SO₄), aluminum sulfate Al₂(SO₄)₃, and water (H₂O), as shown in Equation (4). However, when the amount of sulfuric acid is insufficient, the reaction may instead yield aluminum hydroxide (Al(OH)₃) precipitate and sodium sulfate, as illustrated in Equation (5) [15,44]. Considering the low dosage of acid leaching agent—applied at a solid-to-liquid ratio of 10:1 (L/kg)—along with the high pH (ranging from 10.4 to 12.08) of the neat paste test blocks (Figure 6), the formation of Al(OH)₃ precipitates is likely during acid leaching. This precipitation reduces the concentration of dissolved aluminum ions, thereby resulting in lower aluminum leaching concentrations in acid leaching tests compared to those observed under water leaching conditions.

2NaAlO₂ + 3H₂SO₄ + 2H₂O → Na₂SO₄ + Al₂(SO₄)₃ + 4H₂O

(4)

2NaAlO₂ + H₂SO₄ + 2H₂O → Na₂SO₄ + 2Al(OH)₃↓

(5)

In the water leaching experiments, Se leaching concentrations in neat paste test blocks CMJ-8 to CMJ-12 (prepared under alkaline conditions with pH ranging from 8 to 12) exceeded the regulatory limit. In contrast, all test blocks exhibited excessive Se leaching under acid leaching conditions, as shown in Figure 9d. This can be attributed to the high Se content in the raw materials RM and EMR, which exhibited Se concentrations of 75.2 μg/L and 124 μg/L in water leaching, and 106 μg/L and 234 μg/L in acid leaching, respectively. Among the tested samples, CMJ-13 showed the lowest Se leaching concentrations after both water and acid leaching, with values of 7.9 μg/L and 12.9 μg/L, respectively, indicating effective immobilization in this formulation. For As, only test block CMJ-09 slightly exceeded the standard in the acid leaching test, with a concentration of 10.9 μg/L. Although the As concentration in the RM raw material was relatively high (299 μg/L), all other neat paste test blocks remained within the permissible range. This suggests that the exceedance in CMJ-09 may be due to experimental error or local heterogeneity in material composition.

Furthermore, the EMR raw material contained manganese (Mn) and cadmium (Cd) concentrations significantly above the standard limits, with water leaching results reaching 3.80 × 10⁵ μg/L for Mn and 13.8 μg/L for Cd, and acid leaching concentrations of 1.1 × 10⁶ μg/L and 35 μg/L, respectively. However, Table 3 indicates that the leaching concentrations of Mn and Cd in all RM-EMR synergistic neat paste specimens were below the detection limits. This suggests that under alkaline conditions, soluble Mn²⁺ and Cd²⁺ ions from EMR were effectively immobilized. Specifically, Mn²⁺ can react with OH^- to form water-insoluble compounds such as Mn(OH)₂ [28], Mn₃O₄, MnO₂, and MnOOH. These results confirm that the synergistic utilization of RM and EMR leverages the alkalinity of RM to immobilize heavy metals such as manganese and cadmium, thereby reducing their environmental risk.

From an engineering perspective, it should be emphasized that the exceedance of regulatory limits for elements such as Al and Se was observed exclusively in powdered samples, whereas all block-shaped specimens satisfied the relevant requirements under both water and acid leaching conditions. Block-shaped specimens more closely represent the actual service state of cementitious materials in engineering applications, where structural integrity and low permeability significantly restrict the migration of Al-, Se-, and other potentially mobile species. Therefore, the leaching behavior of block-shaped specimens is considered more representative of field performance. Nevertheless, potential accidental damage scenarios, such as mechanical crushing during demolition or extreme loading conditions, warrant consideration. In such cases, mitigation strategies may include surface encapsulation, secondary stabilization through carbonation or pozzolanic reactions, controlled disposal in alkaline-buffered environments, or reuse of crushed materials as encapsulated fillers rather than directly exposed powders. These measures can effectively reduce reactive surface area and suppress the release of Al^- and Se-bearing species, thereby ensuring environmental safety even under unfavorable conditions.

4. Conclusions

In this study, RM and EMR were used as primary raw materials to develop a composite cementitious neat paste system. The interrelationships among mix proportion optimization, activator dosage, mechanical performance, and environmental safety were systematically investigated. The main conclusions are as follows:

(1)

Mix Proportion Optimization and Mechanical Properties

When the mass ratio of RM to EMR was 2:3 and the activator dosage was 60%, the 28 d UCS of the neat paste system reached a maximum of 35 MPa. At a 1:1 RM-to-EMR ratio, the effect of activator dosage on strength development exhibited a stage-dependent trend, with a noticeable strength reduction at 50% dosage. Strength development followed a “rapid growth-gradual stabilization” pattern, with 7 d UCS reaching 55.6–75.4% of the 28 d value. These results indicate that appropriate control of raw material ratios and activator dosage can synergistically enhance both early and long-term strength, providing critical design parameters for engineering applications.

(2)

Alkaline Environment Regulation and Ammonia Nitrogen Removal Mechanism

The neat paste system exhibited strong alkalinity (pH > 11), with powdered samples showing more pronounced alkaline leaching due to their larger specific surface area. EMR was identified as the main contributor to ammonium (NH₄⁺) pollution, with NH₄⁺-N leaching concentrations reaching up to 302 mg/L, far exceeding regulatory thresholds. Under alkaline conditions, NH₄⁺ transformed into gaseous NH₃, significantly reducing leachate concentration. Only 50% of block samples exceeded the standard, while highly alkaline systems (e.g., CMJ-12 and CMJ-13) met compliance. These findings highlight that alkalinity regulation plays a dual role, simultaneously facilitating ammonia mitigation while potentially influencing the mobility of amphoteric elements.

(3)

Environmental Behavior of Heavy Metals and Harmful Elements

Raw materials exhibited exceedances of Al, Se, F, Hg, As, Cr⁶⁺, and other elements, with Se and Cd in EMR being particularly notable. Under water leaching conditions, the alkaline environment promoted the dissolution of Al, forming NaAlO₂ and resulting in significantly higher leaching concentrations compared to acid leaching. In contrast, acid leaching facilitated the formation of Al(OH)₃ precipitates, thereby reducing Al mobility. Selenium exhibited bimodal leaching behavior and generally exceeded limits under acid leaching. Mn and Cd were undetected after forming stable phases in the alkaline system, verifying RM’s stabilizing effect on heavy metals. Overall, the system can effectively immobilize certain heavy metals; however, the leaching risks of Al and Se remain, highlighting the need for targeted mitigation strategies in practical applications, such as surface encapsulation, secondary stabilization via carbonation or pozzolanic reactions, and controlled reuse of crushed materials to minimize contaminant release.

(4)

Limitations and future research directions

Despite the promising results, several limitations of the present study should be acknowledged. First, the environmental assessment was primarily based on short-term batch leaching tests, which may not fully capture long-term leaching behavior under complex field conditions such as carbonation, wet–dry cycling, or fluctuating pH environments. Second, the elevated alkalinity required for ammonia nitrogen removal may exacerbate the leaching of amphoteric elements such as Al and influence Se mobility, posing potential environmental risks over extended service periods. Third, mechanistic understanding of mechanical performance and environmental behavior was derived from bulk tests and literature, but direct micro- and nanoscale evidence remains limited. In addition, the activator composition and dosage were optimized at the laboratory scale, and their performance variability under industrial production conditions remains uncertain.

Future research should therefore focus on (i) developing targeted pretreatment or modification strategies to reduce Al and Se mobility, such as mineral phase regulation or selective adsorption; (ii) strengthening mechanistic investigations through advanced micro- and nanoscale characterization techniques (e.g., SEM–EDS, XPS, and FTIR) to elucidate the immobilization pathways of ammonia nitrogen and heavy metals; (iii) evaluating long-term environmental performance through accelerated aging tests and field-simulated exposure scenarios; (iv) optimizing activator systems to achieve a balance between mechanical performance, ammonia mitigation, and heavy metal immobilization; and (v) conducting life-cycle and risk-based assessments to support the safe and sustainable engineering application of RM–EMR-based cementitious materials.