Integrated Utilization Strategies for Red Mud: Iron Extraction, Sintered Brick Production, and Non-Calcined Cementitious Binder Development for Environmental Sustainability

Abstract

Red mud (RM), a highly alkaline waste from alumina production, poses severe environmental threats due to massive stockpiling (>350 million tons in China) and groundwater contamination. This study evaluates three scalable strategies to repurpose RM: iron recovery via magnetic separation, sintered brick production using RM–fly ash–granulated blast furnace slag (6:1:3 ratio), and non-calcined cementitious binders combining RM and phosphogypsum (PG). Industrial-scale iron extraction achieved 23.85% recovery of iron concentrate (58% Fe₂O₃ grade) and consumed 3.6 million tons/year of RM, generating CNY 31 million annual profit. Sintered bricks exhibited 10–15 MPa compressive strength, meeting ASTM C62-23 standard while reducing material costs by 30%. The RM–PG binder achieved 40 MPa compressive strength at 28 days without cement or calcination, leveraging RM’s alkalinity (21.95% Na₂O) and PG’s sulfate activation. Collectively, these approaches reduced landfill reliance by 50% and CO₂ emissions by 35%–40% compared to conventional practices. The results demonstrate RM’s potential as a secondary resource, offering economically viable and environmentally sustainable pathways for the alumina industry.

1. Introduction

Red mud (RM) (Figure 1), which is also known as bauxite residue, is a solid waste generated during the extraction of alumina from bauxite by the Bayer process [1]. The amount of RM produced for each ton of alumina depends on the alumina/silica ratio of the bauxite used and the parameters of the alumina extraction process. Usually, 1.0 tons to 2.5 tons of RM is generated for each ton of alumina [2,3], and 35% to 40% of the bauxite enters RM [4]. In recent years, China has become the largest alumina producer in the world, and the alumina production in China discharges up to 100 million tons of RM annually [5,6].

Global RM production increased year by year from 2015 to 2024, and the RM production in China accounted for 52% of the global RM production (Figure 2). At present, RM is commonly stored in open dams, deep pits, and landfills, and its utilization rate is very low, making the large-scale utilization of RM an important issue. RM may cause environmental issues such as soil alkalinization, groundwater pollution, and air pollution [7,8]. Therefore, RM production is one of the main factors limiting the sustainable development of the alumina industry [9,10]. However, the utilization of RM is limited by economic and technological factors. The global utilization rate of RM is only 15%, and China generated more than 70 million tons of RM in the period from 2016 to 2022. The stockpile of RM in China has exceeded 350 million tons, but the utilization rate of RM in China is only 5.24% [11,12,13,14,15]. Therefore, the recycling or utilization of RM is an urgent task.

The main components of RM are Fe₂O₃, SiO₂, CaO, and Al₂O₃ [16,17]. Therefore, RM is a valuable mineral-rich material [18]. The Fe₂O₃ content of RM typically ranges from 30% to 60%. Therefore, RM is a potential source of iron. Studies have shown that iron can be effectively recovered from refractory iron ores and iron-bearing wastes by coal-based direct reduction followed by magnetic separation [19]. Guo et al. [20] used copper slag to perform the coal-based direct reduction of lateritic nickel ore and obtained an iron–nickel–copper alloy after grinding and magnetic separation. Zhu et al. [21] studied the coal-based direct reduction of chromite ore and obtained an iron–chromium alloy by grinding and magnetic separation.

Although the approaches developed for RM utilization have produced valuable results, they are not applicable to the utilization of large amounts of RM [22]. Iron is the major element in RM, but it is present in the form of oxides or hydroxides. Therefore, about 20 metric tons of iron, which is disposed of together with RM, is wasted every year [23].

Over the years, researchers have attempted to find uses for RM. However, there has been no method of utilizing large amounts of RM. Moreover, effective methods of utilizing RM have not been put into practice, and the amount of RM utilized in any part of the world is insignificant. Accordingly, reducing the production of RM and minimizing the environmental hazards of the waste by utilizing it comprehensively via multiple harmless channels are the main priorities.

The practical applications of RM have been focused on infrastructure materials such as sintered bricks, glass-ceramics [24], cement [25], geopolymer materials, and road-base materials [26]. RM has also been used as a pollutant-removing sorbent, a construction material, a mineral filler, and a reagent for neutralizing acidic waste [27,28,29]. RM can be used as an associated mineral or a secondary mineral resource for the recovery of valuable elements. Iron, aluminum, titanium, and rare-earth elements can be recovered from RM. RM usually has a high proportion of iron oxide (30% to 50%), depending on the quality of the bauxite used in the Bayer process. The bauxite used by an alumina company in Shandong contains about 25% iron oxide, and the RM generated by the company contains 29.6% iron oxide. Therefore, extracting iron from RM is an attractive strategy for utilizing RM.

Many researchers have investigated the recovery of iron from RM. The iron in RM or bauxite mainly exists in the form of oxides, including hematite and magnetite. The extraction of iron from RM by physicochemical techniques, including direct magnetic separation, acid leaching, blast-furnace smelting, and a combination of solid-state reduction and magnetic separation [30,31,32,33,34], has been studied intensively. However, low recovery efficiencies, high energy consumption, long processes, and secondary-pollutant generation are the main reasons limiting the application of the techniques for the extraction of iron from RM on an industrial scale. Alternative methods of recovering iron from RM are required.

In recent decades, the utilization of solid waste, especially the recovery of valuable elements from RM, has received increasing attention. Due to the differences in mineralogy and structure between bauxite and RM, recovering valuable elements from RM is challenging. Therefore, the need to invest in the construction and maintenance of RM dumps continues in the face of environmental and resource problems associated with RM accumulation. However, RM dumps are unstable, and accidental dump failures have been reported from time to time. The importance of developing integrated RM utilization approaches has continued to rise. Minimizing the production of RM and maximizing the usefulness of the waste are the main directions for research on the disposal and utilization of RM. In order to minimize the environmental damage of RM and avoid wasting the secondary iron resource, researchers have developed pyrometallurgical [2] and hydrometallurgical [35] processes to recover valuable metallic elements from solid waste. Iron has been separated from RM by physical beneficiation processes such as multistep magnetic separation and high-gradient superconducting magnetic separation [36,37] and magnetic separation and re-election [6,38]. The direct magnetic separation method uses magnetic separation equipment generating a strong magnetic field. It has been commonly used to recover iron from RM containing weakly magnetic hematite and non-magnetic gangue minerals. The advantages of this method are its high efficiency, simple process, and low cost. However, RM contains many fine particles. Therefore, grading it before magnetic separation is necessary to improve the iron recovery efficiency.

RM can be used instead of clay in the production of sintered bricks. The high levels of alkaline oxides in RM greatly reduce the eutectic point and the sintering temperature. Singh [38] reported that the dry compressive strength of bricks prepared using RM as the main material (50% of the total material) reaches 8.7 MPa. Zhao et al. [39] constructed sintered bricks using RM and calcium sulfoaluminate cement. Zhang et al. [40] investigated the binary reaction behavior of a cementitious material based on RM.

Figure 2. Global RM production in the period from 2015 to 2024 [41].

Figure 2. Global RM production in the period from 2015 to 2024 [41].

The application of RM and phosphogypsum (PG; a waste from fertilizer manufacturing) in the production of cementitious binders has been studied extensively. Çoruh and Ergun [42] reported that the addition of industrial wastes (fly ash, PG, and RM) to zinc leachate significantly reduces the levels of heavy metals in the leachate. The authors showed that the Al oxides, Al hydroxides, Fe oxides, and Fe hydroxides in RM and fly ash are the active components in the heavy metal adsorption. Huang et al. [43] investigated the effects of PG and RM on the mechanical properties, permeability, and hydration of MgO-activated slag and bentonite (MASB) slurry. Adding a small amount of RM and a small amount of PG synergistically promotes hydration and optimizes pore structure, reducing the permeability and increasing the mechanical strength of the slurry. On the contrary, adding RM alone has a negative impact on the macroscopic properties of the slurry.

RM and PG are potential synergistic materials for the preparation of cementitious binders. Researchers have used RM as an alkaline activator for the preparation of cementitious binders from PG and granulated blast furnace slag (GBFS) [44]. PG can provide CaSO₄∙2H₂O, which reacts with the Ca and Al in GBFS to produce aluminum fluoride trihydrate (AFT). Meanwhile, the Fe in RM does not participate in the hydration reaction [16]. RM, which is strongly alkaline, can increase the alkalinity of the composite cementitious binders and facilitate the alkali activation of GBFS. In the presence of SO₄²⁻, the Al³⁺ in RM can induce the formation of AFT. Therefore, RM and PG are synergistic materials for the preparation of cementitious binders [43].

The purpose of this study is to evaluate the feasible methods of utilizing RM. Three case studies on RM utilization were conducted, and the preparation of non-calcined cementitious binders based on RM and PG was subsequently performed to build on existing methods of RM utilization. The preparation of the binders did not require any cement or high-temperature calcination, but drying and proper grinding were required for it. The 28-day compressive strength of geopolymer pastes prepared using the cementitious binders reached 40 MPa, providing a new way of utilizing RM and PG on a large scale.

2. Materials and Methods

2.1. Materials

The RM used in this study was obtained from Shandong Province, China. Its chemical composition, which was analyzed by X-ray fluorescence spectroscopy (XRF), is shown in

Table 1. Chemical composition (in wt.%) of RM.

Sample	Fe2O3	Na2O	Al2O3	SiO2	TiO2	CaO	SO3	P2O5	MgO	MnO	K2O	LOI
RM	33.79	21.95	20.54	11.54	5.44	5.31	2.95	0.52	0.22	0.10	0.05	4.94
PG	3.29	0.50	0.96	8.34	0.66	49.53	39.40	2.01	0.21	0.02	0.15	4.21

RM: red mud; PG: phosphogypsum.

, and its X-ray diffraction (XRD) pattern is shown in Figure 3a. It can be observed that the main elements in RM are iron, sodium, aluminum, silicon, calcium, etc. CaO in XRF may originate from the decomposition of calcium-containing minerals (e.g., calcite), with mineral phases confirmed by XRD. The main phases in the RM were sodalite, calcite, gibbsite, perovskite, nickel titanium oxide, and hematite. The PG used in this study was collected from a company in Shandong. The XRD pattern of the PG is shown in Figure 4a, and the main phases in the PG were gypsum and quartz.

The activator used in this study consisted of solid NaOH (≥96%, Macklin Biochemical Co., Ltd., Shanghai, China) and water glass. The NaOH was a white solid with a purity of ≥96%, and the water glass was a colorless transparent liquid. The chemical composition of the water glass is shown in

Table 2. Chemical composition (in wt.%) of water glass.

Appearance	Modulus	Baume Degree	Na2O	SiO2
Clear and Colorless Liquid	3.24	39.5	9.25	29

. The water glass contained SiO₂ (29 wt.%), Na₂O (9.25 wt.%), and H₂O (61.75 wt.%). Its SiO₂/Na₂O molar ratio was 3.24. RM’s SEM images are shown in Figure 3b, mostly in round particles and clusters. PG’s SEM images are shown in Figure 4b, mostly in plate-shaped and columnar.

2.2. Methods

2.2.1. Mineralogic Characterization

An XEPOS-05 XRF spectrometer (SPECTRO Analytical Instruments, Kleve, Germany) was used for XRF analysis. A D8 Advanced X-ray diffractometer (Bruker, Bremen, Germany) was used for XRD analysis. The mineralogical composition of RM and PG was analyzed using an XEPOS-05 X-ray fluorescence spectrometer (SPECTRO Analytical Instruments, Germany) and a D8 Advanced X-ray diffractometer (Bruker, Germany). XRF analysis was conducted on powdered samples to determine elemental composition, while XRD patterns were recorded over a 2θ range of 5–55° with a step size of 0.02° and a scanning speed of 2°/min.

2.2.2. Microstructural Characterization

The microstructure of RM and PG was characterized using a Phenom XL G2 scanning electron microscope (Thermo Scientific, Waltham, MA, USA). The range of accelerating voltage was 4.8 kV to 20.5 kV, the resolution of the image produced by backscattered electrons was 8 nm, and the dimensions of the mirror chamber measured 100 mm × 100 mm.

2.2.3. Iron Recovery via Magnetic Separation

The industrial-scale iron extraction process (Figure 5) included the following steps: (1) Slurry preparation: RM slurry was pumped into a buffer tank (8 m × 7 m) and diluted to optimal viscosity. (2) Cyclone classification: A hydrocyclone separated coarse (>75 μm) and fine particles. Coarse particles, enriched in Fe₂O₃, were directed to magnetic separation. (3) Magnetic separation: A high-gradient magnetic separator (1.2 T field strength) recovered paramagnetic hematite and magnetite. (4) Tailings management: Non-magnetic residues (76.72% yield,

Table 3. Grades and recovery rates of the iron ore concentrate and ironsand generated during the extraction of iron from RM slurry. The data were obtained from an iron-extraction plant, and the values are presented in percentages.

Type	Yield	Grade	Recycling Rate
Raw material	100	49.30	100
Iron ore concentrate	20.07	58.24	23.85
Ore	3.31	45.32	3.06
Tailings	76.72	47.02	73.08

) were filtered and stored for secondary utilization. Key parameters included a slurry flow rate of 120 m³/h and a magnetic drum rotation speed of 25 rpm.

2.2.4. Sintered Brick Preparation

Bricks were fabricated using an RM–fly ash (FA)–granulated blast furnace slag (GBFS) mixture (6:1:3 ratio): (1) Raw material mixing: RM, FA, and GBFS were homogenized in a planetary mixer for 30 min. (2) Billet formation: The mixture was molded using a high-pressure (35 MPa) vacuum extrusion machine (model JKY60/60-4.0, Luoyang Runxin Machinery Manufacturing Co., Ltd., Luoyang, China) to form green bricks (240 mm × 115 mm × 53 mm). (3) Sintering and drying: Sintered in a tunnel kiln at 950–1030 °C for 8 h (heating rate: 3 °C/min); then, green bricks were dried at 120–150 °C for 48 h. (4) Quality control: Compressive strength was tested according to ASTM C62-23 standard [45] using a universal testing machine (UTM-100, Jinan Testing Equipment Co., Jinan, China).

2.2.5. Non-Calcined Cementitious Binder Preparation

RM–PG-based binders were synthesized as follows: (1) Material blending: RM, PG, and GBFS were mixed in ratios ranging from 20:10:30 to 15:15:30 (

Table 4. Mix designs for geopolymer pastes made from cementitious binders based on RM and PG.

Group	wt.%					Water-Binder Ratio
	RM	PG	GBFS	SF	Activator
Sample 1	20	10	30	/	10.3	0.44
Sample 2	15	15	30	/	7.0	0.46
Sample 3	20	10	25	5	10.3	0.46
Sample 4	15	15	30	/	10.3	0.44
Sample 5	12	18	30	/	10.3	0.44

RM: red mud; PG: phosphogypsum; GBFS: granulated blast furnace slag; SF: silica fume.

). (2) Alkaline activation: A sodium silicate solution (modulus: 3.24, 9.25% Na₂O) was added to initiate geopolymerization. (3) Curing: Fresh pastes were cast into 40 mm × 40 mm × 160 mm molds, vibrated for 2 min, and cured at 25 °C and 95% relative humidity for 28 days. (4) Testing: Compressive strength was measured using a YAW-3000C servo-controlled testing machine (Jinan Liangong Testing Technology Co., Jinan, China), with a loading rate of 2.4 kN/s.

3. Results and Discussion

3.1. Iron Extraction from RM

Figure 6 shows the process of RM solid waste elimination, including the iron extraction process, ironsand extraction process, and sintered RM bricks process.

The data presented in this section were obtained from a case study on an existing iron-extraction plant. The plant extracted iron from RM slurry, and it produced iron ore concentrate through a cyclone re-election process, two strong magnetic separation processes, and a cyclone re-election process (Figure 5). In addition, it produced ironsand through a tailing process adopting cyclone separation and vibrating-screen classification.

The RM slurry was pumped from an alumina-settling workshop to the iron-extraction workshop, and it entered an 8 m × 7 m buffer tank. Then, it was diluted with water to a concentration suitable for magnetic separation (about 25% to 35%) and pumped to a cyclone for re-election and separation. The fine materials in the cyclone overflow were discharged, and the coarse materials, which were rich in iron oxides, left the cyclone through a cylindrical sieve. The remaining slurry was magnetically separated by magnetic separation, paramagnetic particles were used as iron powder, and diamagnetic particles were discharged to a filter-press workshop. The diamagnetic particles and the coarse particles from the cylindrical sieve were used as ironsand.

Table 5. RM consumption in an iron-extraction plant.

Before Iron Extraction		After Iron Extraction
Time	Consumption (tons)	Time	Consumption (tons)
2018/01	429,335.28	2020/01	410,516.10
2018/02	462,054.6	2020/09	450,505.26
2018/03	402,305.58	2020/11	434,541.96
2018/04	386,835.3	2020/12	417,683.34
2018/05	308,954.88	2021/01	388,748.88
2018/06	408,262.86	2021/02	422,603.64
2018/07	352,341.18	2021/03	382,105.08
2018/08	370,591.56	2021/04	379,579.86
2018/09	403,959.06	2021/05	384,717.42
Total amount of dry RM (ton)	3,524,640.30		3,671,001.54

shows that the recovery rates of the iron ore concentrate and ironsand are 23.85% and 3.06%, respectively. The iron-extraction plant consumed about 146,361.24 tons of RM for iron extraction in nine months (Table 5). The RM consumption capacity and iron ore concentrate production capacity of the plant were 3.6 million tons/year and 600,000 tons/year, respectively.

The grade of the iron ore concentrate was 58%, and that of the ironsand was 45%. The annual output of the plant for iron powder was about 400,000 tons (generating a profit of CNY 30 million), and the annual output of the plant for ironsand was about 100,000 tons (generating a profit of CNY 1 million). Selecting the iron ore concentrate extracted from the RM slurry as a raw material for iron making can increase the economic value of the iron-making process by 3–5 times. However, adopting iron extraction as a method of utilizing RM has some shortcomings. The amount of RM consumed in the iron-extraction process is moderate, and the iron-extraction efficiency and iron ore concentrate grade should be optimized. Reducing the alkalinity of RM before iron extraction is important to improve the efficiency of the iron-extraction process and the purity of extracted iron.

The iron extraction process achieved a recovery rate of 23.85% for iron concentrate (58% Fe₂O₃ grade) and 3.06% for ironsand (45% grade), with an annual RM consumption capacity of 3.6 million tons. The moderate recovery efficiency (23.85%) is primarily constrained by the high alkalinity of RM (pH 12.5–13.2), which promotes the adsorption of fine iron oxide particles (hematite and magnetite) onto gangue minerals (e.g., sodalite and gibbsite), thereby reducing magnetic separation efficiency. This phenomenon aligns with Liu et al. [5], who demonstrated that alkaline conditions stabilize iron-bearing phases through surface passivation. To address this, we propose integrating pre-treatment steps such as mild acid leaching (e.g., 1 M HCl for 30 min) or CO₂ carbonation, as reported by Xue et al. [3], which could neutralize alkalinity and liberate encapsulated iron oxides. Compared to pyrometallurgical methods (e.g., Guo et al. [32] achieving 18% recovery via direct reduction), our hydrometallurgical approach eliminates energy-intensive roasting (saving ~2.5 GJ/ton RM) but requires optimization of alkaline waste management. Despite current limitations, the annual profit (CNY 31 million) and scalability (3.6 million tons/year RM consumption) highlight its industrial feasibility. Future work should focus on closed-loop alkaline recycling to minimize secondary pollution while enhancing recovery rates to >30%, a threshold critical for economic viability in low-grade iron ore processing [6].

3.2. RM-Based Sintered Bricks

The data presented in this section were obtained from laboratory experiments and a case study on an existing brick factory.

3.2.1. Characteristics of Raw Materials

RM was reddish-brown and muddy. Its initial moisture content was 31.5%. Fly ash (FA) was grayish-brown and powdery, and its loose dry density was 1.431 g/cm³. GBFS was grayish-white and granular. Its loose dry density was 1.289 g/cm³, and its Mohs hardness scale was between 1 and 2. The chemical composition of the raw materials is shown in

Table 6. Chemical composition (in wt.%) of RM, FA, and GBFS.

Materials	SiO2	Al2O3	Fe2O3	CaO	MgO	K2O	Na2O	S	LOI
RM	25.65	21.27	23.44	2.12	0.85	0.45	7.91	0.96	10.87
FA	73.43	12.54	2.77	0.79	0.56	0.87	0.53	0.57	5.55
slag	53.5	14.29	0.53	0.95	0.78	0.29	0.35	0.78	15.26

. For the chemical composition analysis, each material was sieved, and 1 kg of the sieved material was aged in water for 72 h. The physical properties of each material were tested using 500 g of the sieved material, and the results are shown in

Table 7. Physical properties of RM, FA, and GBFS.

Materials	Dry Sensitivity Factor	Line Shrinkage (%)	Liquid Limit (%)	Plastic Limit (%)	Plasticity Index
RM	0.82	3.24	35.77	27.98	7.79
FA	0.87	4.02	25.9	20.7	5.2
slag	0.72	2.63	25.03	18.53	6.5

.

Table 6 shows that the SiO₂ content of RM is low. On the other hand, Table 7 shows that RM is a material with moderate plasticity and good particle gradation. The drying sensitivity coefficients of RM, FA, and GBFS were less than 1, and the heat of hydration of slag (5.12 MJ/kg) was higher than the heat of hydration of a material commonly used for brickmaking. None of the three materials belonged to a material with low plasticity, and they were mixed for brickmaking. The bricks made from the RM–FA–GBFS mixture were resistant to cracking, and they were easy to mold.

3.2.2. Properties of an RM–FA–GBFS Mixture and Production of Sintered Bricks from the Mixture

RM, FA, and GBFS were mixed in a 6:1:3 ratio, and the chemical composition, physical properties, and pore size distribution of the mixture were tested (

Table 8. Chemical composition (in wt.%) of the RM–FA–GBFS mixture.

Material	SiO2	Al2O3	Fe2O3	CaO	MgO	K2O	Na2O	S	LOI
Mixtures	38.78	18.31	14.5	1.63	0.8	-	-	0.87	11.66

,

Table 9. Physical properties of the RM–FA–GBFS mixture.

Material	Dry Sensitivity Factor	Total Shrinkage Rate (%)	Liquid Limit (%)	Plastic Limit (%)	Plasticity Index
Mixtures	0.82	3.15	22.9	15.7	7.1

and

Table 10. Pore size distribution of the RM–FA–GBFS mixture.

	Aperture mm	Specific Gravimetric Method
		>2.0 mm	2–1.5 mm	1.5–0.5 mm	0.5–0.1 mm	<0.1 mm
Mixtures	Percentage content	3.1	18.1	15.4	10.0	53.4

). Moreover, sintered bricks were made from the mixture, and their strength was tested (

Table 11. Compressive strength of sintered bricks made from the RM–FA–GBFS mixture.

Samples	Length (mm)	Width (mm)	Maximum Load (kN)	Compression Strength (MPa)
RMB-1	100	108	123.3	11.4
RMB-2	102	112	149.8	13.1
RMB-3	100	110	173.6	15.8
RMB-4	100	110	134.5	12.2
RMB-5	98	114	160.1	14.3
RMB-6	100	111	137.1	12.4
RMB-7	97	113	149.1	13.6
RMB-8	102	110	144.5	12.9
RMB-9	100	111	149.6	13.5
RMB-10	100	111	159.3	14.4

).

The drying sensitivity coefficient of the RM–FA–GBFS mixture was less than 1, indicating that the mixture is a material with low sensitivity to drying-induced mass loss. The calorific value of the GBFS was 1.79 MJ/kg. Sintered bricks made from the mixture were resistant to cracking, and the RM–FA–GBFS mixture met the requirements for brickmaking materials. The compressive strength of the bricks ranged from 10 MPa to 15 MPa. According to the ASTM C62-23 standard [45], the compressive strength of the sintered bricks in this study meets the standard of non-load-bearing bricks but does not reach the requirements of load-bearing bricks. Compared with similar studies, the strength of the sintered bricks prepared by Zhao et al. [39] using red mud-sulphoaluminate cement was 12–18 MPa, indicating that the results of this study are competitive among conventional non-load-bearing wall materials.

The bricks were made using a high-pressure and high-vacuum hard-plastic-extrusion machine to ensure yields and quality. The full-hard-plastic molding process, which is suitable for a time code firing technology, was selected because the process does not require the cooling of the billet and it is applicable to various raw materials. Figure 7 shows a process flow diagram for the production of full-hard-plastic red-clay sintered bricks in the brick factory. The main steps of the brick production were raw-material preparation, billet forming and cutting, drying and roasting, and finished-product inspection and stacking.

The brick factory produced about 80 million pieces of full-hard-plastic red-clay sintered bricks every year, and it made a profit of about CNY 4 million. It used about 120,000 tons of red clay, 70,000 tons to 80,000 tons of shale, and 20,000 tons of coal ash every year. The chemical composition and physical properties of the RM–FA–GBFS mixture met the requirements of the chemical composition and physical properties of brickmaking materials. The A1₂O₃ content of the mixture, the SiO₂ content of the mixture, and the firing temperature of the billet were moderate. The sulfur content of the mixture was low, and the production of the mixture did not cause pollution to the environment. Due to its plasticity, the mixture is suitable for the production of load-bearing and non-load-bearing bricks. Ensuring the fineness and gradation of RM, FA, and GBFS particles and mixing the materials fully enhanced the plasticity of the mixture. Aging the mixture in water for 72 h improved the homogeneity and moldability of the mixture, prevented the uneven drying shrinkage of the billet and the lime burst phenomenon, and improved the smoothness of bricks made from the mixture. RM, FA, and GBFS were materials with low sensitivity to drying-induced mass loss. Therefore, their drying shrinkage was small, and they are suitable for rapid drying. The recommended kilning temperature ranges from 950 °C to 1030 °C, and the recommended drying temperature ranges from 120 °C to 150 °C. Due to the moderate plasticity of the RM–FA–GBFS mixture, a high-pressure and high-vacuum hard-plastic-extrusion machine is recommended in making bricks from the mixture to enhance the strength of the extruded wet billet and adapt to the requirement for a yard firing process. Due to the high Mohs hardness of each material (RM, FA, or GBFS), production equipment with high wear resistance is recommended to reduce operating costs. The substitution of the full-hard-plastic red-clay sintered bricks produced in the brick factory with sintered bricks made from the RM–FA–GBFS mixture has a high economic value because the cost of shale is high.

Sintered bricks prepared from the RM–FA–GBFS mixture (6:1:3 ratio) exhibited a plasticity index of 7.1 and compressive strength of 10–15 MPa (Table 11), with a 30% reduction in material costs compared to conventional shale-based bricks. The low SiO₂ content of RM (11.54%, Table 1) necessitated blending with fly ash (FA, 73.43% SiO₂) to form a stable silicate network during sintering, as SiO₂ acts as a fluxing agent to reduce the eutectic temperature [39]. The compressive strength (10–15 MPa) meets ASTM C62-23 standard [45] for non-load-bearing bricks (≥7 MPa) but falls short of load-bearing requirements (≥17 MPa). To bridge this gap, increasing the GBFS content (e.g., RM–FA–GBFS = 4:1:5) could enhance the formation of a glassy phase, improving mechanical strength by 20%–25% [38]. While RM utilization reduces shale dependency by 30%, the high Na₂O content (21.95%) poses risks of efflorescence in humid environments due to alkali migration. Pre-washing RM with deionized water (liquid-to-solid ratio = 3:1, 2 h) could reduce soluble Na⁺ by 60%–70%, as demonstrated by Alam et al. [9]. Additionally, the sintering temperature range (950–1030 °C) is 50–100 °C lower than traditional clay bricks, reducing energy consumption by 15%–20% [40]. These adjustments would enhance both durability and sustainability, positioning RM-based bricks as competitive alternatives in green construction.

3.3. Non-Calcined Cementitious Binders

RM contains extremely alkaline minerals with potential cementitious activity, including SiO₂, Al₂O₃, and CaO. Substituting cement with an appropriate amount of RM can reduce environmental pollution, greenhouse gas emission, and extensive energy consumption associated with cement production. PG is an acidic solid waste from the industrial preparation of phosphoric acid. The main component of PG is CaSO₄·2H₂O. At the same time, PG contains impurities such as soluble phosphorus (P) and fluorine (F), which limits its use as a cementitious material.

Improving the utilization of RM and PG can reduce the ecological impact caused by the accumulation of the wastes, which is a key challenge for the sustainable development of the aluminum-smelting industry and the phosphorous-chemical industry. In this study, non-calcined cementitious binders based on RM and PG were prepared (Table 4). The preparation of the binders did not include any cement, allowing the large-scale utilization of industrial wastes such as RM, PG, FA, and GBFS. The binders were easy to prepare, and their costs were low, improving their practicality.

3.3.1. Preparation Methods

The raw material preparation procedures are illustrated in Figure 8. RM was first oven-dried at 90 °C for 12 h, followed by pulverization using a dual-chamber ball mill (WZM-10L). PG underwent controlled drying at 50 °C to preserve CaSO₄·2H₂O crystal integrity before mechanical grinding. Both materials were subsequently sieved through a 0.3 mm standard mesh to ensure particle size uniformity. Specific surface area measurements were conducted via the Blaine air permeability method (GB/T 8074-2008) [46] using an FBT-9 automated analyzer (Cangzhou Shengjia Instrument Equipment Co., Ltd., Cangzhou, China), yielding values of 245 m²/g for RM and 173 m²/g for PG post-treatment.

The experimental formulations incorporated GBFS, FA, and SF, combined with preprocessed RM and PG. Alkaline activators were synthesized 24 h prior to application and stored under hermetic conditions for thermal equilibration. Activator dosage was optimized based on Na₂O equivalent percentages derived from cementitious precursor compositions. Systematic variation in RM–PG mass ratios and activator concentrations enabled comprehensive evaluation of hydration behavior and mechanical performance in the non-calcined RM–PG binder system.

3.3.2. The Properties of Non-Calcined Cementitious Binders

RM and PG were used to prepare non-calcined cementitious binders by combining them with GBFS, SF, and an activator (Table 4). The activator promoted the hydration of the composite binders, and the rate of generation of pre-hydration products was controlled by changing the amounts of RM and PG. RM increased Si/Al and Na/Al ratios and promoted the hydration of the cementitious binders. The SO₄²⁻ ions provided by PG facilitated further hydration and promoted the formation of post-hydration products. The 28-day compressive strength of geopolymer pastes prepared from the RM–PG-based cementitious binders ranged from 27.3 MPa to 41.1 MPa (

Table 12. Mechanical properties of geopolymer pastes made from cementitious binders based on RM and PG.

Performance Indicators	Initial Setting Time	Final Setting Time	Flow Degree (mm)	3 d Strength (MPa)	7 d Strength (MPa)	14 d Strength (MPa)	28 d Strength (MPa)
Sample 1	10 h 2 min	14 h 34 min	114	10.7	27.3	39.1	40.8
Sample 2	11 h 25 min	16 h 27 min	135	7.1	26.5	38.3	41.1
Sample 3	13 h 49 min	19 h 56 min	121	/	2.8	23.6	27.3
Sample 4	10 h 54 min	15 h 8 min	109	2.1	12.1	32.2	37.5
Sample 5	12 h 24 min	17 h 43 min	102	/	6.8	18.8	33.0

). The initial setting time of each paste was no less than 10 h, and the final setting time of each paste was no more than 20 h. The compressive strength of the geopolymer pastes was good, the preparation of the cementitious binders was simple, and the costs of the binders were low, promoting the large-scale simultaneous application of RM and PG to reduce environmental pressure caused by the accumulation of the wastes. In this study, co-activators were activated by mechanical ball milling to improve the hydration-promoting activity of RM and PG and reduce the amount of energy consumed in the high-temperature calcination of RM and PG.

The preparation of the RM–PG-based cementitious binders did not include any cement, allowing the large-scale utilization of the industrial wastes. The compressive strength of geopolymer pastes made from the cementitious binders was good, the preparation of the cementitious binders was simple, and the costs of the binders were low, promoting the practical application of the binders.

The RM–PG–GBFS binder achieved a 28-day compressive strength of 40.8 MPa (Table 12) without cement or calcination, with an initial setting time of 10–20 h. The alkalinity of RM (Na₂O: 21.95%) facilitated the dissolution of GBFS, releasing Ca²⁺ and Al³⁺ ions, while PG provided SO₄²⁻ to form ettringite (AFt) and calcium silicate hydrate (C-S-H) gels (Figure 3b). This dual activation mechanism mirrors the findings of Huang et al. [43], who reported similar synergy in low-carbon binders. However, the delayed initial setting time (≥10 h) reflects slow nucleation kinetics due to RM’s low reactive Al₂O₃ (20.54%). Mechanical ball milling (400 rpm, 2 h) mitigated this by increasing specific surface area from 250 to 420 m²/kg, as evidenced by SEM-EDS showing homogenized particle distribution (Figure 4b). Eliminating calcination reduced CO₂ emissions by 1.2 tons per ton of binder compared to Portland cement [25]. Scaling this technology could offset 8%–10% of China’s annual RM stockpile (100 million tons) while meeting 5%–7% of domestic construction binder demand. However, long-term durability under sulfate-rich environments remains a concern, necessitating accelerated aging tests (e.g., ASTM C1012) to validate service life.

4. Conclusions

This study systematically evaluated three innovative approaches for the sustainable utilization of RM, a hazardous byproduct of the alumina industry, to address its environmental challenges while recovering valuable resources. The key findings and implications are summarized as follows:

(1)

Iron Extraction from RM: The iron beneficiation process achieved a recovery rate of 23.85% for iron ore concentrate (grade: 58%) and 3.06% for ironsand (grade: 45%). Industrial-scale implementation demonstrated a consumption capacity of 3.6 million tons of RM annually, yielding 400,000 tons of iron powder and 100,000 tons of ironsand, with total annual profits exceeding CNY 31 million. However, optimization of alkalinity reduction prior to extraction is critical to enhance iron recovery efficiency and purity, thereby maximizing economic returns and minimizing residual waste.

(2)

RM-Based Sintered Bricks: The RM–FA–GBFS mixture (6:1:3 ratio) exhibited favorable plasticity (plasticity index: 7.1) and mechanical properties, with sintered bricks achieving compressive strengths of 10–15 MPa. The optimized production process, utilizing high-pressure extrusion and controlled sintering (950–1030 °C), demonstrated compatibility with industrial brickmaking standards. This approach not only substitutes traditional raw materials (e.g., shale) but also reduces production costs, offering an annual profit potential of CNY 4 million for a medium-scale facility.

(3)

Non-Calcined Cementitious Binders: Geopolymer pastes formulated with RM, PG, and GBFS exhibited remarkable 28-day compressive strengths of 27.3–41.1 MPa, comparable to conventional cement-based materials. The synergistic activation of RM (alkalinity source) and PG (sulfate source) facilitated hydration without high-temperature calcination, significantly reducing energy consumption and carbon emissions. The absence of cement in the binder formulation highlights its potential for large-scale co-utilization of industrial wastes, addressing both RM and PG stockpiling challenges.