Skip to main content
NCBI home page
Scientific Reports logoLink to Scientific Reports
. 2026 Mar 30;16:10659. doi: 10.1038/s41598-026-46673-w

Optimization of precursor proportions and performance characteristics of iron ore tailings-based one-part geopolymers

Wenqi Kou 1, Mingxing Gao 1,✉, Ting Zhao 1, Huiping Wang 2
PMCID: PMC13040013  PMID: 41912766

Abstract

To promote high-value utilization, bulk iron ore tailings (IOT) were combined with fly ash (FA) and ground granulated blast furnace slag (GGBS) to prepare one-part geopolymers. A simplex lattice design was used to investigate the effects of precursor proportions on compressive strength and fluidity. The FA–GGBS system exhibited continuous strength development with curing age, whereas the IOT–GGBS system showed early-age strength stagnation. In the IOT–FA–GGBS ternary system, compressive strength is promoted by a relatively high GGBS content together with moderate proportions of IOT and FA. Polynomial regression models for 28-day compressive strength and fluidity were highly significant (P < 0.0001), with F-values of 52.57 and 22.01 and R² values of 0.9228 and 0.8334, respectively. Model optimization identified an optimal precursor composition of 35% IOT, 40% FA, and 25% GGBS. Response surface analysis showed that compressive strength was enhanced by synergistic gelation and the micro-aggregate effect of IOT when FA and IOT contents were both 30–50%, while fluidity improved at FA contents of 40–60% and IOT contents of 20–40%. Microscopically, calcium–aluminum–silicate–hydrate and sodium–aluminum–silicate–hydrate gels coexisted in the FA–GGBS and IOT–FA–GGBS systems, whereas the high Si/Al ratio (2.9) in the IOT–GGBS system limited gel development and strength.

Keywords: Iron ore tailings, One-part geopolymer, Solid Na2SiO3, Simple lattice design

Subject terms: Engineering, Materials science, Mathematics and computing

There are 1.4 billion tonnes of iron ore tailings generated in the world every year1. According to statistics, more than 5.0 × 109 t of iron ore tailings have been cumulatively stockpiled in China, and the amount continues to increase at a rate of about 3.0 × 108 t annually, while the comprehensive utilization rate remains below 10%2. The large-scale accumulation of iron ore tailings not only occupies arable land and wastes valuable resources, but also poses serious threats to the surrounding environment and public safety3. Therefore, promoting the resource utilization of iron ore tailings has become an urgent issue that must be addressed.

Studies on iron ore tailings have shown that their chemical composition is mainly composed of SiO2, Al2O3, Fe2O3, CaO, and MgO, and they have been applied to a certain extent in fields such as catalysis, adsorption, and fuel production4. Owing to their silica-alumina characteristics, recent studies have attempted to incorporate iron ore tailings into geopolymer precursor systems. Geopolymers are three-dimensional network inorganic cementitious materials formed through depolymerization and polycondensation reactions of aluminosilicate precursors activated by alkaline activators5. Compared with ordinary Portland cement, geopolymers not only exhibit excellent mechanical strength and durability, but also achieve a reduction in carbon emissions of 53%–72% during production, demonstrating significant low-carbon and environmental6–8. Therefore, geopolymers are considered one of the most promising inorganic cementitious materials for replacing ordinary Portland cement9.

However, due to the inherently low reactivity of iron ore tailings, measures such as mechanical grinding, thermal pretreatment, or the incorporation of reactive aluminosilicate materials are usually required to enhance their reactivity and cementitious performance in geopolymer systems13. Thermal treatment is highly energy-intensive. Therefore, incorporating reactive supplementary materials like fly ash and metakaolin is considered a more favorable approach to enhancing the performance of geopolymers based on iron ore tailings. These materials are rich in reactive silicate components, which can effectively promote the formation of cementitious products, thereby enabling the preparation of iron ore tailings-based geopolymers with relatively high compressive strength under alkaline conditions5,14.

At present, most existing studies on iron ore tailings–based geopolymers have focused on two-part systems activated by liquid alkaline solutions15–17. However, practical applications are hindered by difficulties associated with the transportation, storage, and handling of highly alkaline solutions. In contrast, one-part geopolymers are produced by pre-mixing precursor materials with solid alkaline activators, requiring only the addition of water during use, which offers advantages such as ease of construction, improved safety, and strong engineering adaptability, thereby exhibiting broader application prospects18,19. Current research on one-part geopolymers predominantly relies on conventional precursors such as fly ash (FA) and ground granulated blast furnace slag (GGBS), whereas studies employing iron ore tailings (IOT) as precursor materials remain limited20. In this context, compared with existing iron ore tailings–based geopolymer studies that mainly focus on two-part systems, systematically incorporating iron ore tailings into one-part geopolymer systems is of significant importance for promoting their engineering applications.

In one-part geopolymer systems, the types of precursor materials and their proportions directly influence the dissolution behavior, gel structure evolution, and the resulting mechanical and durability properties21. Based on this, this study explores a one-part geopolymer system incorporating IOT as the primary raw material, activated by solid Na₂SiO₃. A mix proportioning scheme incorporating IOT, FA, and GGBS was systematically established using a simplex lattice design. Response surface methodology was then applied to investigate the effects of precursor composition on the compressive strength and fluidity of the one-part geopolymers. Representative specimens were further selected for FTIR and SEM–EDS analyses to examine the reaction products and microstructural characteristics, thereby elucidating the microscopic mechanisms governing the development of macroscopic properties.

Materials and methodology

Raw materials

The precursor materials included IOT, FA, and GGBS. The IOT were obtained from Zhongxing Mining Co., Ltd. (Huhhot, China), and were subjected only to mechanical activation processes such as crushing and sieving, without any additional high energy-consuming pretreatment, to produce iron ore tailings powder with a particle size of 200 mesh. The FA (Class I) and GGBS (S95) were both purchased from Henan Wuhu Environmental Protection Technology Co., Ltd. (Zhengzhou, China). Solid Na2SiO3 was supplied by Qingdao Haiwan Chemical Co., Ltd. (Qingdao, China), with a modulus of 2.73 (58.31% SiO2 and 21.36% Na2O); solid NaOH with a purity of 99% was used to adjust the modulus.

The main chemical compositions of the precursor materials are presented in Table 1. As shown in Table 1, IOT mainly consists of SiO2, Al2O3, and CaO, with a Si/Al ratio of approximately 3.23 and a total silica-alumina content exceeding 60%, indicating that it can serve as a potential aluminosilicate source in geopolymer systems. The CaO content of FA is only 5.61%, classifying it as low-calcium fly ash, which limits its contribution to early strength development in geopolymers. In contrast, GGBS contains 34.23% SiO₂ and 34.22% CaO, and this high-calcium characteristic provides an effective complement to FA22,23.

Table 1.

Main chemical compositions of raw materials. Unit: wt %.

Precursor material SiO2 Al2O3 CaO Fe2O3 MgO SO3
IOT 46.96 14.52 10.17 12.71 6.06 0.03
FA 43.26 23.13 5.61 2.51 0.95 0.82
GGBS 34.23 17.61 34.22 1.01 6.21 1.62
Open in a new tab

The particle size distributions and particle characteristics of the precursor materials are shown in Fig. 1; Table 2, respectively. As indicated in Fig. 1; Table 2, the particle size distribution peaks of IOT and GGBS are sharper and more concentrated, whereas the distribution curve of FA is broader and more gradual. Although IOT exhibits a relatively large specific surface area, its particle size distribution is relatively wide, suggesting effective grinding activation but with the presence of some coarse particles. GGBS shows the largest specific surface area and the smallest median particle size (D50), indicating the finest particles, which is beneficial for promoting early reactions and optimizing the packing density of the paste. In contrast, FA presents the most dispersed particle size distribution, with the largest characteristic particle size and the smallest specific surface area.

Fig. 1.

Fig. 1

Open in a new tab

Particle size distribution of precursor materials.

Table 2.

Particle characteristics of raw materials.

Precursor materials Specific surface areas/ (m2/kg) Characteristics particle sizes/µm
D10 D50 D90
IOT 796.8 3.189 24.141 72.910
FA 642.5 4.355 27.964 108.847
GGBS 1073.9 2.511 11.863 29.602
Open in a new tab

The XRD patterns of the precursor materials are shown in Fig. 2. IOT is mainly composed of well-crystallized mineral phases, including strong albite diffraction peaks, as well as clearly identifiable characteristic minerals such as Fe-rich cordierite, phlogopite, and kaolinite. Notably, a pronounced broad “hump” is observed in the 2θ range of 20°–40°, confirming that mechanical grinding has successfully introduced a certain amount of amorphous (glassy) phases, thereby significantly enhancing its potential reactivity24,25. FA is dominated by crystalline phases, exhibiting strong diffraction peaks of quartz and mullite, which are typical indicators of high-temperature combustion, and its glassy phase content is relatively low. In contrast, the XRD pattern of GGBS is characterized by a broad diffuse hump without sharp crystalline diffraction peaks, indicating that it is composed almost entirely of highly chemically reactive glassy phases.

Fig. 2.

Fig. 2

Open in a new tab

XRD pattern of precursor materials.

The SEM images of the precursor materials are shown in Fig. 3. From the microstructural morphology, IOT particles are irregular in shape with layered surfaces, FA particles exhibit smooth surfaces with good sphericity, and GGBS particles are angular and irregular with smooth surfaces and sharp edges.

Fig. 3.

Fig. 3

Open in a new tab

SEM spectrum of precursor materials. (a) IOT. (b) FA. (c) GGBS.

Sample preparation

IOT, FA, GGBS, and solid Na₂SiO₃ were added to a mixer according to the designed proportions and dry-mixed for 3 min. Subsequently, water was slowly and uniformly added, followed by an additional 3 min of mixing. The homogeneous geopolymer paste was then poured into cubic molds with dimensions of 50 mm × 50 mm × 50 mm and cured under standard conditions (temperature 20 ± 2 °C, relative humidity ≥ 95%) for 24 h. After demolding, the specimens were removed and further cured in a standard curing chamber until the specified ages. Compressive strength and fluidity tests were conducted in accordance with GB/T 8077-201226 and GB/T 17,671–202127 to evaluate the mechanical properties and workability of the specimens. The compressive strength was measured using a universal electronic testing machine (YAW-300E, Zhongluchang, Jinan, China), with three parallel specimens tested for each sample. In addition, specimens cured for 28 days were crushed and immersed in anhydrous ethanol for 24 h to terminate hydration. The samples were then dried in a vacuum oven at 40 °C for 24 h until a constant mass was achieved. The dried samples were ground and sieved, and the powder passing through a 0.075 mm sieve was used for FTIR analysis, while the remaining material was used for SEM–EDS analysis. A Fourier transform infrared spectrometer (Spectrum Two, Perkin Elmer Instruments Inc. Boston, USA) was employed to analyze the characteristic chemical functional groups of the materials, with a wavenumber range of 500–4000 cm− 1 and a resolution of 2 cm− 1. The microstructural morphology and chemical composition were characterized by using SEM and EDS (Phenom PW-100-516, Thermo Fisher Scientific Inc. Eindhoven, Holland). EDS was performed by the face-scanning method.

Experimental design

Preliminary experiments were conducted to determine the following parameters. With respect to the water-to-solid ratio, values below 0.30 led to excessively high paste viscosity, while ratios above 0.38 resulted in overly fluid mixtures, both adversely affecting moldability; thus, a ratio of 0.34 was adopted. The dosage of the composite solid activator (a mixture of Na₂SiO₃ and NaOH) was optimized. Increasing the activator content from 16% to 18% (by mass of the precursor) improved the compressive strength, whereas a further increase to 20% resulted in a slight reduction; thus, 18% was selected. The modulus (SiO2/Na2O molar ratio) of the composite activator was set to 1.0, as it provided higher compressive strength than lower (0.8) or higher (1.2) moduli. Based on this formulation (18% total activator dosage with a modulus of 1.0), the Na2O equivalent content contributed by the activator was calculated to be 7.7% of the total precursor mass, quantitatively representing the overall alkalinity of the system. It was observed that a GGBS content below 10% resulted in slow strength development or insufficient setting, whereas contents above 30% led to excessively rapid setting, adversely affecting fresh-state workability. Similar trends have been reported in previous studies, where moderate GGBS contents effectively enhanced early-age strength while avoiding excessively rapid setting28,29. Therefore, the GGBS content was limited to the range of 10–30% in this study, and levels of 10%, 20%, and 30% were selected for the mixture design. Using these limits as constraints, a simple lattice design method was employed to systematically investigate the effects of precursor proportioning on the performance of IOT–FA–GGBS ternary one-part geopolymers. The specific precursor mix proportions are illustrated in Fig. 4, and the corresponding experimental results are presented in Table 3.

Fig. 4.

Fig. 4

Open in a new tab

Raw materials proportion diagram.

Table 3.

Results of the simple lattice design.

Sample 28-day compressive strength/ MPa Fluidity/ mm Sample 28-day compressive strength/ MPa Fluidity/ mm Sample 28-day compressive strength/ MPa Fluidity/ mm
BI1 5.8 211 T2 12.5 248 T8 26.8 216
BI2 16.3 179 T3 16.3 254 T9 26.4 235
BI3 31.5 134 T4 28.8 257 T10 28.5 229
BF1 32.4 222 T5 23.8 256 T11 39.1 178
BF2 34.8 182 T6 24.5 239 T12 42.5 157
BF3 38.4 140 T7 23.7 184 T13 30.4 145
T1 8.1 208
Open in a new tab

Results and discussion

Strength development of binary systems influenced by precursor composition

To systematically investigate the synergistic effects of precursors in the IOT–FA–GGBS ternary one-part geopolymer system, compressive strength tests were first performed on two binary systems, namely FA–GGBS and IOT–GGBS. The experimental mix proportions were selected from the boundary compositions shown in Fig. 4 (BI1–BI3 and BF1–BF3), resulting in a total of six mixtures. The evolution of compressive strength with curing age for the two binary systems is shown in Fig. 5.

Fig. 5.

Fig. 5

Open in a new tab

Compressive strength of the binary systems. (a) FA - GGBS. (b) IOT - GGBS.

As shown in Fig. 5, at the same GGBS content, the compressive strength of the FA–GGBS system at all curing ages is significantly higher than that of the IOT–GGBS system. The FA–GGBS system exhibits a continuously increasing strength development pattern. Taking specimen BF1 as an example, its compressive strength increases by 290% from 3 to 28 days, indicating that even at a low GGBS content, a high proportion of FA can continuously generate gel products through sustained pozzolanic reactions, thereby promoting long-term strength development. In contrast, the strength development of the IOT–GGBS system exhibits early-stage stagnation. When the GGBS content is lower than 30%, the compressive strength of the system enters a plateau after 7 days. This is mainly because the strength development of this system relies predominantly on the early hydration of the GGBS, while IOT mainly acts as a physical filler and contributes little to the mid- to late-stage cementitious reactions. Consequently, after the early reaction stage, further strength development is limited. Overall, these results demonstrate that GGBS plays a critical role in governing the strength development of the binary systems.

Strength development of ternary systems influenced by precursor composition

A three-dimensional column chart was plotted by organizing part of the ternary mix proportions in Fig. 4 with the IOT/(IOT + FA) mass ratio (0.3, 0.5, and 0.7) as the x-axis, curing age as the y-axis, and compressive strength as the z-axis, reflecting the relationship between precursor proportions and the strength development of the IOT–FA–GGBS one-part geopolymer, as shown in Fig. 6.

Fig. 6.

Fig. 6

Open in a new tab

Compressive strength of the ternary system. (a) GGBS content: 10%. (b) GGBS content: 20%. (c) GGBS content: 30%.

As shown in Fig. 6, the strength development pattern of the ternary system differs from that of the binary systems and exhibits a pronounced dependence on mix proportions. When the GGBS content is 10%, increasing the IOT content leads to a consistent decrease in compressive strength at all curing ages, and the strength growth with age becomes progressively less pronounced, indicating that strength development is severely inhibited. In contrast, when the GGBS content is increased to 30%, the system with a moderate IOT content (IOT/(IOT + FA) = 0.5) exhibits the largest strength increase from 7 to 28 days, reaching 10.3 MPa, and achieves the highest 28-day compressive strength. These results highlight that both higher GGBS content and a moderate IOT fraction are key factors controlling the compressive strength, providing guidance for selecting mixture proportions in this ternary system.

Response surface modeling and significance analysis of compressive strength and fluidity

Regression models describing the relationships between the 28-day compressive strength, fluidity and precursor proportions were fitted to the data in Table 3 using Design-Expert software (Version 13.0), as presented below.

graphic file with name d33e850.gif 1
graphic file with name d33e854.gif 2

where YS represents the 28-day compressive strength, YL denotes the fluidity, A is the FA content, B is the IOT content, and C is the GGBS content.

The analysis of variance (ANOVA) and model adequacy evaluation results for the fitted models are presented in Tables 4 and 5. The results indicate that the F-values of the regression models for the 28-day compressive strength and fluidity are 52.57 and 22.01, respectively, with both p-values being less than 0.0001, demonstrating that the two models are highly significant. In addition, the signal-to-noise ratios (Adeq Precision) of the two models are 25.0192 and 14.2066, respectively, both far exceeding the recommended value of 4, indicating an adequate signal and reliable experimental data. The adjusted coefficients of determination (R²adj) and predicted coefficients of determination (R²pred) for the 28-day compressive strength model are 0.9052 and 0.8821, respectively, while those for the fluidity model are 0.7955 and 0.7396. The differences between each pair of values are less than 0.2, suggesting good model fitting and reliable predictive capability.

Table 4.

Analysis of variance and reliability analysis for the 28-day compressive strength regression model.

Source Sum of Squares df Mean Square F-value P-value
Model 2588.27 5 517.65 52.57 < 0.0001 Highly significant
Linear Mixture 2333.99 2 1167 118.52 < 0.0001 Highly significant
AC 93.61 1 93.61 9.51 0.0054 Highly significant
BC 42.22 1 42.22 4.29 0.0503 significant
AB²C 69.13 1 69.13 7.02 0.0146 significant
Residual 216.62 22 9.85
Lack of Fit 164.09 13 12.62 2.16 0.1248 not significant
Pure Error 52.53 9 5.84
Cor Total 2804.89 27
Open in a new tab

R²=0.9228 R²adj=0.9052 R²pred=0.8821 Adeq Precision = 25.0192.

Table 5.

Analysis of variance and reliability analysis for the fluidity regression model.

Source Sum of Squares df Mean Square F-value P-value
Model 37621.91 5 7524.38 22.01 < 0.0001 Highly significant
Linear Mixture 32146.33 2 16073.17 47.02 < 0.0001 Highly significant
AB 4345.42 1 4345.42 12.71 0.0017 Highly significant
AC 894.62 1 894.62 2.62 0.12
BC 1131.83 1 1131.83 3.31 0.0825
Residual 7520.77 22 341.85
Lack of Fit 6015.35 13 462.72 2.77 0.0658 not significant
Pure Error 1505.42 9 167.27
Cor Total 45142.68 27
Open in a new tab

R²=0.8334 R²adj=0.7955 R²pred=0.7396 Adeq Precision = 14.2066.

Model validation and determination of optimal mixture proportions

A comparison between the experimental values and the regression model–predicted values of the 28-day compressive strength and fluidity is shown in Fig. 7.

Fig. 7.

Fig. 7

Open in a new tab

Comparison of model predicted values with experimental values for geopolymer. (a) 28-day compressive strength. (b) Fluidity.

As shown in Fig. 7; Tables 4 and 5, the model predictions exhibit good agreement with the experimental data, with the coefficients of determination (R²) for compressive strength and fluidity being 0.9228 and 0.8334, respectively. This confirms that the regression models can accurately predict the effects of precursor proportions on the compressive strength and fluidity of the one-part geopolymer.

Using the Optimization module in Design-Expert (Version 13.0), the precursor proportions were optimized, yielding an optimal composition of 35% IOT, 40% FA, and 25% GGBS. Verification tests showed that the measured compressive strength at this composition was 33.5 MPa, representing a 5.07% deviation from the predicted value of 31.8 MPa. Similarly, the measured fluidity was 187 mm, deviating by 4.28% from the predicted value of 195 mm. Both deviations are below 6%, indicating that the response surface models fitted to the simple lattice design possess good predictive accuracy and that the optimization results are reliable. Therefore, the models are suitable for accurate performance prediction and mix proportion optimization of one-part geopolymers.

Response surface-based analysis of precursor effects on the performance of ternary systems

Based on the regression models established above, three-dimensional response surfaces were generated, as shown in Fig. 8. The color gradient of the surfaces, ranging from blue to red, corresponds to increasing 28-day compressive strength and fluidity of the system, and the surface elevation follows the same trend. As illustrated in Fig. 8, both the 28-day compressive strength and fluidity response surfaces of the ternary system clearly reveal the synergistic regulatory effects arising from the interactions among IOT, FA, and GGBS.

Fig. 8.

Fig. 8

Open in a new tab

3D response surface graph. (a) 28-day compressive strength. (b) Fluidity.

As shown in Fig. 8(a), the 28-day compressive strength surface of the system exhibits a pronounced ridge of increasing strength extending from the high GGBS content region toward the FA vertex, while transitioning into a steep gradient zone of sharply decreasing strength toward the IOT vertex. When the GGBS content is 30% and both FA and IOT fall within specific proportion ranges (approximately 30–50% FA and IOT), the system exhibits relatively high compressive strength. This behavior is mainly attributed to the synergistic interaction among FA, IOT, and GGBS in the ternary system. Under this synergy, FA not only directly participates in the reaction but also improves the interfacial transition zone and mitigates the adverse effects of IOT, enabling it to function as an effective reinforcing component within the system. Compared with either FA–GGBS or IOT–GGBS binary systems, the incorporation of a third precursor allows the ternary system to establish a more balanced reaction environment, characterized by a relatively sufficient Ca²⁺ supply and a moderate Si/Al ratio. Such chemical conditions are generally considered favorable for the cooperative development of calcium–aluminum–silicate–hydrate (C–A–S–H) and sodium–aluminum–silicate–hydrate (N–A–S–H) gels, promoting the formation of a denser gel network and ultimately enhancing the compressive strength of the system30–33.

As shown in Fig. 8(b), the fluidity response surface shows a convex morphology, increasing from high- to low-GGBS regions. Within FA and IOT contents of 40–60% and 20–40%, respectively, the adverse effect of GGBS on fluidity is significantly reduced. Notably, the ternary blend with 20% GGBS, 50% FA, and 30% IOT achieves a fluidity of 235 mm, exceeding that of a binary system with only 10% GGBS (Table 3). This enhancement stems from an optimized FA-IOT particle packing system, where the spherical FA particles provide physical lubrication (ball-bearing effect) and IOT fills voids34,35. Concurrently, the high-fluidity region may also be associated with a relatively higher Si/Al ratio and moderated early-age Ca2+ driven reactions, which are generally considered to limit excessive gel precipitation and help maintain particle mobility31,36. Therefore, the fluidity of the ternary system is governed not solely by the GGBS content, but by the combined proportions and interactions of among FA, IOT, and GGBS.

Microanalysis

Three representative samples with the highest compressive strength, namely T12 from the IOT–FA–GGBS system, BI3 from the IOT–GGBS system, and BF3 from the FA–GGBS system, were selected for FTIR and SEM–EDS analyses. By comparing their functional group structures, microstructural morphologies, and elemental distributions, the microscopic mechanisms by which precursor materials influence the performance of one-part geopolymers were systematically elucidated.

FTIR

The FTIR spectra of samples T12, BF3, and BI3 after 28 days of curing are presented in Fig. 9. All samples exhibit essentially identical main peak positions, indicating the formation of a similar primary gel structure. The absorption bands at approximately 3430 cm− 1 and 1640 cm− 1 are assigned to the asymmetric stretching and bending vibrations of O–H bonds, respectively, associated with chemically bound water in the geopolymers23,37. The absorption peak near 1400 cm− 1 is attributed to the stretching vibration of O–C–O, resulting from the formation of carbonates due to the reaction of atmospheric CO2 with alkalis38. The absorption band in the range of 540–580 cm⁻¹ corresponds to the bending vibration of Si–O–Al bonds39.

Fig. 9.

Fig. 9

Open in a new tab

FTIR spectrum of representative samples.

A broadened absorption band near 1000 cm⁻¹ is assigned to the asymmetric stretching vibration of Si–O–T (T = Al or Si) in silicate tetrahedra, indicating the reorganization and polycondensation of silicate and aluminate species and the formation of amorphous aluminosilicate gels, namely N–A–S–H and C–A–S–H gels40. Within this wavenumber range, both samples T12 and BF3 exhibit a single peak, with the peak position following the order BF3 > T12. Because a shift toward lower wavenumbers in this region is associated with a higher degree of polymerization, while the response intensity mainly reflects the extent of gel development41, the lower peak position observed for T12 indicates a more condensed and highly cross-linked gel network. This superior polymerization degree is the key factor explaining its higher 28-day compressive strength (42.5 MPa for T12 vs. 38.4 MPa for BF3). In addition, the vibration peak observed at 870 cm⁻¹ for sample T12 originates from the stretching vibration of Al–O–Si bonds, further confirming the formation of calcium-containing aluminosilicate hydrates, such as ettringite or C–A–S–H gel42,43. In contrast, sample BI3 exhibits two distinct peaks at 1049 cm⁻¹ and 980.3 cm⁻¹ within this region, indicating an incomplete reaction state, in which unreacted siliceous phases coexist with the formed geopolymer gel. This disordered and less polymerized microstructure directly correlates with its lower compressive strength of 31.5 MPa, as the unreacted phases act as defects and disrupt the formation of a continuous, load-bearing gel network.

SEM–EDS

The SEM images and EDS analysis results of samples T12, BF3, and BI3 are presented in Fig. 10; Table 6, respectively.

Fig. 10.

Fig. 10

Open in a new tab

SEM of representative samples. (a) T12. (b) BF3. (c) BI3.

Table 6.

The proportion of each element content by EDS test.

Sample Element content (by mass)/% Ca/Si Si/Al Ca/(Si + Al) (Ca + Na)/(Si + Al)
C O Na Al Si Ca
T12 14.85 41.50 6.37 8.50 15.74 7.59 0.5 1.9 0.3 0.6
BF3 14.36 42.13 4.92 11.14 16.91 5.80 0.3 1.5 0.2 0.4
BI3 13.20 40.88 8.61 4.77 13.80 9.97 0.7 2.9 0.5 1.0
Open in a new tab

As shown in Fig. 10, sample BI3 exhibits a large number of unreacted IOT and GGBS particles, accompanied by wide cracks. This indicates that the siliceous and aluminous phases in the iron ore tailings powder were not effectively activated and therefore could not sufficiently participate in polycondensation reactions, resulting in inadequate gel formation and a loose microcracked structure, which directly accounts for its lowest compressive strength (31.5 MPa) among the compared samples. Sample BF3 contains a considerable amount of incompletely reacted FA particles with an agglomerated distribution, suggesting insufficient depolymerization of the FA glassy phase. As a result, the overall reaction degree is limited, leading to the formation of weak zones and relatively poor structural compactness. In contrast to the two binary systems, the ternary system, represented by sample T12, exhibits a more homogeneous microstructure. Only a small number of partially reacted FA particles with smooth surfaces can be observed in sample T12, and its overall structure is the most uniform and compact among the three samples. This optimal microstructure is the fundamental reason why T12 achieved the highest 28-day compressive strength of 42.5 MPa.

As shown in Table 6, the Ca/(Si + Al) ratios of samples T12 and BF3 are both lower than 1.5, while their Ca/Si ratios fall within the range of 0–0.6, indicating the coexistence of C–A–S–H and N–A–S–H gels in these two samples41. With increasing Ca/Si and Ca/(Si + Al) ratios, the amount of geopolymer gel formed increases, leading to a progressively denser and stronger microstructure44. The EDS-derived ratios provide a chemical basis for this strength ranking. T12, with its moderate and balanced Ca/Si and Ca/(Si + Al) ratios—conducive to effective C-A-S-H/N-A-S-H gel coexistence—developed the strongest microstructure, culminating in the highest strength (42.5 MPa). BF3, with lower ratios, formed a slightly less robust gel network, consistent with its lower strength (38.4 MPa). In contrast, the relatively high Si/Al ratio of sample BI3 (2.9) mainly arises from the inherently high-silicon and low-aluminum chemical characteristics of this system. On the one hand, both IOT and GGBS are rich in SiO₂ but relatively deficient in Al₂O₃, resulting in an insufficient aluminum supply for constructing a continuous geopolymer network. On the other hand, the high-calcium environment introduced by GGBS preferentially promotes the formation of C–A–S–H gel with a Ca/Si ratio greater than 0.6, further consuming the limited reactive aluminum species. Consequently, a large fraction of the inert siliceous phase in the IOT cannot be effectively activated or dissolved and remains in the system as unreacted particles. This disrupts the continuity and homogeneity of the gel matrix and ultimately manifests as deteriorated mechanical performance at the macroscopic scale44.

Conclusions

The FA–GGBS geopolymer exhibits continuous strength development with age, whereas the IOT–GGBS system shows early-age stagnation. For the IOT–FA–GGBS ternary system, strength is strongly dependent on the mix proportion, with optimal performance at 30% GGBS and a moderate IOT/FA ratio (IOT/(IOT + FA) = 0.5), achieving the highest 28-day compressive strength and 7–28 day gain (10.3 MPa) .

Polynomial regression models based on a simplex lattice design were developed to predict 28-day compressive strength and fluidity as a function of precursor proportions. Both models were highly significant (P < 0.0001), with F-values of 52.57 and 22.01, and good predictive capability (R² = 0.9228 and 0.8334). Model optimization yielded an optimal precursor proportion of 35% IOT, 40% FA, and 25% GGBS.

Response surface analysis indicated that optimizing the proportions of IOT, FA, and GGBS can significantly enhance the compressive strength and fluidity of one-part geopolymers. When the GGBS content is 30% and the FA and IOT contents are approximately 30–50%, the reaction products optimize the interfacial structure and maintain a moderate Ca²⁺ concentration and Si/Al ratio, allowing IOT to fully exert its micro-aggregate effect and improve macroscopic mechanical performance. Fluidity is significantly improved when FA and IOT contents are approximately 40–60% and 20–40%, respectively, as FA provides physical lubrication (ball-bearing effect) and IOT fills particle voids, preventing aggregation of high water-demanding GGBS and reducing paste resistance.

Microscopic analysis showed that C–A–S–H and N–A–S–H gels coexisted in the IOT–FA–GGBS and FA–GGBS systems, whereas the IOT–GGBS system produced only C–A–S–H gel. FTIR peaks at 1049 and 980.3 cm⁻¹ in the IOT–GGBS system indicate incomplete reaction, with the high Si/Al ratio (2.9) leaving residual silicate that disrupted the gel network and lowered strength.

This study proposes a ternary system synergistically utilizing iron ore tailings, fly ash, and GGBS, offering a new approach for the high-value use of iron ore tailings in one-part geopolymers. To support its application in construction, future studies should assess long-term durability, including leaching of potentially harmful elements and performance under sulfate attack and carbonation. Furthermore, to accurately evaluate its environmental benefits, a life cycle assessment accounting for the carbon footprint of the solid alkali activator is essential. Such an assessment will provide the scientific basis needed to advance this material toward low-carbon construction.

Acknowledgements

The authors would like to express our sincere gratitude and appreciation to each other for our combined efforts and contributions throughout the course of this research paper.

Author contributions

Conceptualization, M.G. and T.Z.; methodology, W.K.; software, W.K.; validation, W.K., H.W. and M.G.; formal analysis, W.K.; investigation, H.W.; resources, T.Z.; data curation, W.K.; writing—original draft preparation, W.K.; writing—review and editing, M.G.; visualization, H.W.; supervision, M.G.; project administration, M.G.; funding acquisition, M.G. and T.Z.

Funding

This research was funded by the Central Government Guidance Funds for Local Science and Technology Development (No. 2024ZY0111). Natural Science Foundation of Inner Mongolia Autonomous Region (No. 2025QN04029).

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  • 1.Diaz-Piloneta, M., Terrados-Cristos, M., Ortega-Fernandez, F. & Martinez-Huerta, G. Alvarez-Cabal, V. Experimental characterization of iron mining tailings as sustainable material for thermal energy storage. Sci. Rep.15, 40218 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Lv, X. et al. Environmental impact, durability performance, and interfacial transition zone of iron ore tailings utilized as dam concrete aggregates. J. Clean. Prod.292, 126068 (2021). [Google Scholar]
  • 3.Lu, H. Y. et al. Effects of retarders on the rheological properties of coal fly ash/superfine iron tailings based 3D printing geopolymer: insight into the early retarding mechanism. Constr. Build. Mater.411, 134445 (2024). [Google Scholar]
  • 4.Lorenzini, L., Prates, C. D., Teixeira, A. P. C. & Pinto, P. S. Reuse of iron ore tailing to potassium silicate synthesis and to the production of geopolymers. New. J. Chem.47 (42), 19737–19745 (2023). [Google Scholar]
  • 5.Zhan, X., Tang, L., Yue, Z., Lu, H. & Wang, J. Enhanced mechanical properties and mechanism of iron ore tailings-based geopolymers modified by municipal solid waste incineration fly ash. J. Build. Eng.98, 111456 (2024). [Google Scholar]
  • 6.Aziz, A. et al. A state-of-the-art review of the structure and properties of laterite-based sustainable geopolymer cement. Environ. Sci. Pollut Res.30 (19), 54333–54350 (2023). [DOI] [PubMed] [Google Scholar]
  • 7.Liu, Q. et al. Alkali-hydrothermal activation of mine tailings to prepare one-part geopolymer: Activation mechanism, workability, strength, and hydration reaction. Ceram. Int.48 (20), 30407–30417 (2022). [Google Scholar]
  • 8.Bompa, D., Xu, B. & Corbu, O. C. Evaluation of one-part slag fly-ash alkali-activated mortars incorporating waste glass powder. J. Mater. Civ. Eng.34 (12), 05022001 (2022). [Google Scholar]
  • 9.Brandvold, A. S., Trindade, A. C. C. & Kriven, W. M. Rheological assessment of metakaolin-based geopolymer composites through squeeze flow. J. Am. Ceram. Soc.106 (7), 4038–4051 (2023). [Google Scholar]
  • 10.De Carvalho, A. R., da Silva Calderón-Morales, B. R., Borba, J. C. Jr. & Gleize, P. J. P. Proposition of geopolymers obtained through the acid activation of iron ore tailings with phosphoric acid. Constr. Build. Mater.403, 133078 (2023). [Google Scholar]
  • 11.Wang, A., Zheng, Y., Zhang, Z., Wang, Y. & Li, Y. The durability of alkali-activated materials in comparison with ordinary Portland cements and concretes: A review. Engineering6 (6), 695–706 (2020). [Google Scholar]
  • 12.Zhang, X., Zhang, S., Liu, H., Chen, B. & Wang, X. Disposal of mine tailings via geopolymerization. J. Clean. Prod.284, 124756 (2021). [Google Scholar]
  • 13.Tome, S. et al. Structural and physico-mechanical investigations of mine tailing-calcined kaolinite based phosphate geopolymer binder. Silicon14 (7), 3563–3570 (2022). [Google Scholar]
  • 14.Ferreira, I. C. et al. Reuse of iron ore tailings for production of metakaolin-based geopolymers. J. Mater. Res. Technol.18, 4194–4200 (2022). [Google Scholar]
  • 15.Wang, T., Ning, X., Yang, L., Rao, F. & Jiang, K. The synthesis of an iron tailings-based geopolymer with synergistic electromagnetic wave consumption property. Environ. Res.268, 120744 (2025). [DOI] [PubMed] [Google Scholar]
  • 16.Zhang, N. et al. Iron tailings-carbon fiber synergy in geopolymer composites: multi-objective optimization of self-sensing mortar. Mater. Struct.58, 286 (2025). [Google Scholar]
  • 17.Ning, X. et al. A novel electromagnetic wave absorption geopolymer originated from iron tailings and blast furnace slag. Mater. Struct.58, 18 (2025). [Google Scholar]
  • 18.Ooi, W. E. et al. Comparative mechanical and microstructural properties of high calcium fly ash one-part geopolymers activated with Na₂SiO₃ anhydrous and NaAlO₂. J. Mater. Res. Technol.15, 3850–3866 (2021). [Google Scholar]
  • 19.Chen, Q., Hu, G. & Wu, J. Comparative study on the prediction of the unconfined compressive strength of the one-part geopolymer stabilized soil by using different hybrid machine learning models. Case Stud. Constr. Mater.21, e03439 (2024). [Google Scholar]
  • 20.Mahendra, K. & Narasimhan, M. C. One part alkali-activated materials for construction – A review. Mater. Today: Proc.93, 182–188 (2023). [Google Scholar]
  • 21.Kumar, S. G. K. M., Kinuthia, J. M., Ott, J. & Adeleke, B. O. Geopolymer Chemistry and Composition: A Comprehensive Review of Synthesis, Reaction Mechanisms, and Material Properties-Oriented with Sustainable Construction. Materials18 (16), 3823 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Nath, P. & Sarker, P. K. Effect of GGBFS on setting, workability and early strength properties of fly ash geopolymer concrete cured in ambient condition. Constr. Build. Mater.66, 163–171 (2014). [Google Scholar]
  • 23.Jose, A., Nivitha, M. R., Krishnan, J. M. & Robinson, R. G. Characterization of cement stabilized pond ash using FTIR spectroscopy. Constr. Build. Mater.263, 120136 (2020). [Google Scholar]
  • 24.Cheng, Y., Tong, B., Sun, X. & Zhang, J. On the grinding effects of high-silicon iron tailings. Environ. Sci. Pollut Res.29 (49), 73326–73340 (2022). [DOI] [PubMed] [Google Scholar]
  • 25.Li, X. et al. Preparation of iron tailings-based cementitious materials based on mechanochemical reinforcement. Front. Built Environ.11, 1672953 (2025). [Google Scholar]
  • 26.GB/T 8077 – 2012. Test Methods for Uniformity of Concrete Admixture (Standards Press of China, 2012).
  • 27.GB/T 17671 – 2021. Test Method of Cement Mortar Strength (ISO Method) (Standards Press of China, 2021).
  • 28.Lee, N. K. & Lee, H. K. Setting and mechanical properties of alkali-activated fly ash/slag concrete manufactured at room temperature. Constr. Build. Mater.47, 1201–1209 (2013). [Google Scholar]
  • 29.Nasir, N. A. M. et al. Performance of high-content waste ash geopolymer-based composites for sustainable construction. Arch. Civ. Mech. Eng.25, 228 (2025). [Google Scholar]
  • 30.Ansari, G. A. & Kumar, S. S. Geopolymerization of fly ash and GGBS for sustainable industrial waste utilization. Sci. Rep.15, 34318 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Yi, C., Boluk, Y. & Bindiganavile, V. Experimental characterization and multi-factor modelling to achieve desired flow, set and strength of N-A-S-H geopolymers. Materials15, 5634 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Geetha, S. & Ramamurthy, K. Properties of geopolymerised low-calcium bottom ash aggregate cured at ambient temperature. Cem. Concr Compos.43, 20–30 (2013). [Google Scholar]
  • 33.Temuujin, J. & Van Riessen, A. Effect of fly ash preliminary calcination on the properties of geopolymer. J. Hazard. Mater.164, 634–639 (2010). [DOI] [PubMed] [Google Scholar]
  • 34.Yang, T. et al. Effect of fly ash microsphere on the rheology and microstructure of alkali-activated fly ash/slag pastes. Cem. Concr Res.109, 198–207 (2018). [Google Scholar]
  • 35.Ma, J., Wang, D., Zhao, S., Duan, P. & Yang, S. Influence of Particle Morphology of Ground Fly Ash on the Fluidity and Strength of Cement Paste. Materials14 (2), 283 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Hanjitsuwan, S. et al. Effects of NaOH concentrations on physical and electrical properties of high calcium fly ash geopolymer paste. Cem. Concr Compos.45, 9–14 (2014). [Google Scholar]
  • 37.Yu, P., Kirkpatrick, R. J., Poe, B., McMillan, P. F. & Cong, X. Structure of calcium silicate hydrate (C-S-H): near-, mid-, and far-infrared spectroscopy. J. Am. Ceram. Soc.82 (3), 742–748 (1999). [Google Scholar]
  • 38.Zhao, S. et al. Solidification/stabilization of municipal solid waste incineration fly ash using uncalcined coal gangue–based alkali-activated cementitious materials. Environ. Sci. Pollut Res.26 (25), 25609–25620 (2019). [DOI] [PubMed] [Google Scholar]
  • 39.Heikal, M. Y. M., Nassar, G. & El-Sayed, S. M. I. Physico-chemical, mechanical, microstructure and durability characteristics of alkali activated Egyptian slag. Constr. Build. Mater.69, 60–72 (2014). [Google Scholar]
  • 40.Perez-Cortes, P. & Escalante-Garcia, J. I. Gel composition and molecular structure of alkali-activated metakaolin-limestone cements. Cem. Concr Res.137, 106211 (2020). [Google Scholar]
  • 41.Wang, M. L., Zhang, Y., Shi, Z. P. & Li, Z. Properties and mechanisms of geopolymers constructed from polymerized CGP-OSA-GGBS ternary-based solid wastes. Constr. Build. Mater.450, 138710 (2024). [Google Scholar]
  • 42.Guo, L. et al. Preparation of coal gangue-slag-fly ash geopolymer grouting materials. Constr. Build. Mater.328, 126997 (2022). [Google Scholar]
  • 43.Thai, T. D. et al. Evaluation of Compressive Strength of Uncalcined Dredged Sediment and Fly Ash-based Geopolymer Using Response Surface Method. Int. J. Eng. Trans. B: Appl.38 (3), 670–679 (2025). [Google Scholar]
  • 44.Ma, H. et al. Study on compressive strength and durability of alkali-activated coal gangue-slag concrete and its mechanism. Powder Technol.368, 112–124 (2020). [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Articles from Scientific Reports are provided here courtesy of Nature Publishing Group

RESOURCES