Effect of partial replacement of volcanic ash with slag on the performance of sustainable alkali-activated materials for lead-contaminated soil remediation

Abstract

Lead-contaminated soils pose critical environmental and ecological risks due to the persistence, mobility, and toxicity of Pb in terrestrial systems. This study investigates the use of alkali-activated binders derived from volcanic ash (VA) with partial replacement by ground granulated blast-furnace slag (GGBFS) for the stabilization/solidification of Pb-contaminated soil. The influence of slag content (0–40 wt%) under both oven and ambient curing regimes was systematically evaluated in terms of unconfined compressive strength (UCS), Pb immobilization efficiency, and environmental performance. A combination of leaching tests (TCLP), spectroscopic and microstructural analyses (XRD, FE-SEM, EDS, FTIR), and life cycle assessment (LCA) was employed. Results demonstrate that slag incorporation markedly enhances matrix densification through the synergistic formation of NASH and C(A)SH gels, leading to up to 75% higher UCS and more than 99% reduction in Pb leachability compared with VA-only binders. Microstructural evidence confirmed Pb incorporation into stable gel phases, while LCA revealed modest climate change benefits (∼5% reduction in CO₂ emissions) alongside trade-offs in ecotoxicity and resource depletion categories. Overall, the findings highlight the novelty and potential of VA–slag hybrid binders as eco-efficient and sustainable stabilizers for long-term remediation of heavy metal-contaminated soils.

Introduction

Soil contamination by heavy metals (HMs) is a critical environmental issue because of their persistence, bioaccumulative potential, and long-term risks to human and ecological health¹. Lead (Pb), in particular, is highly problematic due to its toxicity, non-biodegradability, and prolonged residence time in soil. Industrial and urban activities—including mining, smelting, battery recycling, and improper waste disposal—have led to elevated Pb concentrations in many soils^2,3, often exceeding regulatory thresholds and requiring effective remediation measures⁴.

Various remediation techniques have been applied, such as soil washing, vitrification, electrokinetic treatment, and phytoremediation^5–10. However, stabilization/solidification (S/S) is considered one of the most practical and cost-effective approaches, especially for in-situ applications^11,12. In this method, contaminants are immobilized by physical encapsulation and chemical stabilization within a hardened matrix, reducing their mobility and bioavailability^13–16.

Traditionally, ordinary Portland cement (OPC) has been the dominant binder used in S/S applications because of its availability and proven effectiveness^17–19. Nevertheless, its extensive use is increasingly questioned due to the high environmental footprint of cement production, which accounts for approximately 7–8% of global anthropogenic CO₂ emissions^20–24. This limitation has motivated the development of alternative low-carbon binders, particularly alkali-activated materials (AAMs)^25–28, which are synthesized by activating aluminosilicate-rich precursors with alkaline solutions^29–31. The reaction products, primarily sodium aluminosilicate hydrate (NASH) and calcium aluminosilicate hydrate (C(A)SH), form dense, durable matrices capable of immobilizing heavy metals more effectively than OPC-based hydrates^32,33.

A key advantage of AAMs is their strong ability to immobilize heavy metals. This is achieved through both chemical mechanisms—such as ion exchange and structural incorporation of metal ions into the aluminosilicate framework—and physical mechanisms including encapsulation within a dense gel matrix. Compared with conventional cement hydrates, the amorphous gels formed in AAMs generally exhibit higher sorption capacity and greater long-term stability^34–37.

The performance of AAMs strongly depends on the type and proportion of precursors used. Volcanic ash (VA), a natural pozzolan rich in reactive silica and alumina, is abundant and economically attractive in many regions³⁸. Slag, in contrast, is a calcium-rich industrial by-product from steel production that promotes the formation of calcium aluminosilicate hydrate (C(A)SH) gels, thereby improving early-age strength and chemical durability^39,40. When VA and slag are combined, a hybrid binder system is obtained that balances early mechanical performance with long-term immobilization capacity^41,42.

The synergy between VA and slag is highly influenced by their relative proportions and curing conditions. Increasing slag content typically accelerates alkali-activation reactions, enhances matrix densification, and improves early strength development^43,44. Higher VA proportions, however, may favor extended polymerization and the formation of aluminosilicate-rich matrices, contributing to long-term durability^38,45,46. In addition, curing temperature and humidity play a decisive role in reaction kinetics, gel morphology, and phase development. Elevated-temperature curing accelerates gel formation and strength gain, whereas ambient curing more closely represents field conditions but requires longer durations^47–50.

The novelty of this study stems from its focus on VA as the primary precursor, with partial replacement by ground granulated blast-furnace slag (GGBFS). While GGBFS has been extensively studied in alkali-activated binders, VA has received comparatively less attention despite its wide availability, low cost, and lower environmental footprint. By incorporating slag into VA-based systems, the calcium supplied by GGBFS modifies the alkali-activation reactions and favors the development of calcium-rich aluminosilicate hydrate phases (commonly described as C–(A)–S–H or N,C-(A)–S–H-type gels), together with sodium aluminosilicate hydrate (N–A–S–H) gel typical of low-calcium systems. The coexistence of these gels contributes to matrix densification, early strength development, and improved Pb immobilization. Such synergistic behavior distinguishes VA–slag hybrid binders from single-precursor systems and provides both engineering and environmental advantages. The present work is novel in integrating mechanical, leaching, microstructural, and life cycle assessment (LCA) analyses to establish a comprehensive basis for applying VA–slag binders in sustainable remediation of heavy metal-contaminated soils.

In this context, the present study explores the efficacy of a sustainable S/S approach for Pb-contaminated soil using alkali-activated binders derived from volcanic ash and slag. The investigation focuses on the effect of gradually replacing VA with slag (ranging from 0 to 40% by weight) on unconfined compressive strength (UCS) and Pb immobilization efficiency, under both ambient and oven-curing regimes. Mechanical and leaching performance are monitored over multiple curing periods (1, 7, 28, and 90 days), while advanced microstructural analyses—including X-ray diffraction (XRD), field-emission scanning electron microscopy (FE-SEM), energy-dispersive X-ray spectroscopy (EDS), and Fourier-transform infrared spectroscopy (FTIR)—are employed to elucidate the underlying stabilization mechanisms at the mineralogical and microstructural levels.

Materials and methods

Materials

Soil

The soil used in this study was collected from the campus of Iran University of Science and Technology (35.755° N, 51.509° E). To ensure homogeneity and consistency across specimens, the soil was oven-dried at 105 °C and passed through a 2 mm sieve to remove debris and coarse particles. Its geotechnical properties were characterized following standard procedures. Atterberg limits, determined per ASTM D423 and D424, indicated a liquid limit (LL) of 26%, plastic limit (PL) of 18%, and plasticity index (PI) of 9%. Particle size distribution, based on ASTM D422 and D7928, identified the soil as clayey sand (SC) under the Unified Soil Classification System (ASTM D2487). The chemical composition of the soil was analyzed using X-ray fluorescence (XRF) spectroscopy, and the results are reported in Table 1. Additionally, standard Proctor compaction tests (ASTM D1557) were performed to determine the maximum dry unit weight and the corresponding optimum moisture content, which were found to be 1.74 g/cm³ and 14%, respectively.

Table 1.

Chemical composition of soil, slag, and volcanic ash.

Component	SiO2	Al2O3	CaO	Fe2O3	Na2O	MgO	K2O	MnO	TiO2
Soil (wt%)	54.06	18.46	16.81	2.42	1.76	2.53	2.09	0.88	0.99
Slag (wt%)	36.64	15.42	36.24	0.28	1.32	6.71	1.09	0.89	1.41
Volcanic ash (wt%)	55.86	21.87	10.13	3.31	5.82	1.71	0.88	0.12	0.30

Alkali-activated materials

In this study, ground granulated blast furnace slag (GGBFS) and volcanic ash were obtained from Esfahan Steel Company and the Taftan volcanic region in Iran, respectively. Prior to use, both materials were air-dried, sieved through a 2 mm mesh, and stored in sealed containers to prevent moisture ingress. Their particle size distributions, shown in Fig. 1.

Fig. 1.

Grain size distribution of the soil, slag, and volcanic ash.

The chemical compositions of VA and slag were determined by XRF analysis and are summarized in Table 1. To activate the aluminosilicate phases in the precursors, a 6 M sodium hydroxide (NaOH) solution was used. Analytical-grade NaOH (99% purity; density = 2.13 g/cm³) was procured from Neutron Pharma Chemical Company.

Methods

Sample preparation

Artificial contamination of the soil with lead was carried out to simulate field-relevant pollution levels. A lead nitrate solution was prepared by dissolving Pb(NO₃)₂ in deionized water to achieve a target concentration of 2500 mg Pb per kg of dry soil. The adopted contamination level of 2500 mg Pb/kg dry soil was selected to simulate severe pollution conditions that may occur in industrial and post-industrial sites such as mining, smelting, and battery recycling areas. For comparison, the USEPA residential soil screening level for Pb is 400 mg/kg, while the EU guideline typically considers 300–500 mg/kg as critical thresholds for intervention. The Brazilian regulation ABNT NBR 10.004 categorizes soils with Pb concentrations exceeding 1000 mg/kg as hazardous waste. Thus, the chosen level represents an upper-end contamination scenario that exceeds international thresholds by a factor of 5–8. This deliberate choice provides a conservative framework to evaluate the efficiency and robustness of volcanic ash–slag alkali-activated binders under extreme conditions, thereby ensuring that their performance would remain effective even in highly contaminated soils.

The contaminated solution, equivalent to 20% of the dry soil mass, was gradually added to the clean, oven-dried soil while mixing thoroughly for 10 min to ensure uniform distribution of the contaminant. The prepared mixture was sealed in polyethylene bags and stored at ambient laboratory conditions (~ 25 °C) for 10 days to allow sufficient interaction between Pb ions and soil particles. Thereafter, the contaminated soil was re-dried in an oven at 105 °C for 24 h before stabilization treatments were applied.

The total binder dosage was fixed at 10% by weight of the contaminated soil. This binder consisted of varying proportions of volcanic ash and slag, with slag content replacing VA at 0%, 10%, 20%, 30%, and 40% by weight. For simplicity, mixtures are denoted as V100G0, V90G10, V80G20, V70G30, and V60G40, respectively. In all mixtures, the total liquid content (activating solution) was kept constant at 14% of the combined mass of the soil and binder to ensure consistent workability.

The mix preparation involved a multi-stage blending protocol to ensure homogeneity. First, VA and slag were dry-blended for approximately 4 min. The NaOH solution was then gradually introduced to the dry binder, and mixing continued for another 5 min to initiate partial activation. The pre-contaminated soil was subsequently added to the slurry, and mixing was extended for an additional 5 min to ensure uniform dispersion of binders and activator throughout the soil matrix.

The homogenized mixture was then cast into cylindrical molds measuring 38 mm in diameter and 76 mm in height. A smooth steel rod was used to compact the mixture manually in the molds to minimize entrapped air and ensure dense packing. Once molded, the specimens were demolded and subjected to two distinct curing regimes: oven curing (OC) and ambient curing (AC). In OC, specimens were cured at a constant temperature of 50 °C, whereas AC was conducted in a controlled environment chamber set at 25 ± 2 °C and 80 ± 2% relative humidity. To evaluate the effect of curing time, specimens were tested after 1, 7, 28, and 90 days of curing. Each sample label indicates its binder composition and curing method (e.g., V80G20-OC or V60G40-AC). Curing durations of 1, 7, 28, and 90 days were considered to evaluate the time-dependent behavior of the stabilized soil.

Unconfined compressive strength (UCS)

The unconfined compressive strength (UCS) of both untreated and treated specimens was evaluated following the procedure outlined in ASTM D2166. The tests were conducted by applying axial loading at a constant strain rate of 1 mm/min until failure occurred. To ensure the reliability and consistency of the results, three replicate specimens were tested for each mix design, and the average UCS value was calculated and reported accordingly.

Leachability test

To evaluate the environmental performance of the stabilized systems, lead leaching potential was examined using the Toxicity Characteristic Leaching Procedure (TCLP), as specified in Method 1311 of the U.S. Environmental Protection Agency (USEPA). Prior to testing, specimens were crushed to ensure that no particle exceeded 9.5 mm in diameter. An acetic acid-based extraction fluid with a pH of 2.88 ± 0.05 was prepared and mixed with the solid samples at a liquid-to-solid mass ratio of 20:1.

The slurry was transferred to sealed polyethylene bottles and agitated for 18 h at 30 rpm using a rotary extractor. Following agitation, the suspension was filtered through Whatman No. 42 filter paper to separate the leachate. The lead concentrations in the filtrates were quantified using inductively coupled plasma optical emission spectroscopy (ICP-OES). All TCLP tests were conducted in triplicate, and average values were reported.

Microstructural analyses

Mineralogical phases present in the contaminated soil, as well as the reaction products resulting from the alkali activation of aluminosilicate precursors with NaOH, were characterized using X-ray diffraction (XRD) analysis. Quantification of crystalline and amorphous phases from XRD patterns was carried out using the Hall–Petch method, which estimates the relative proportion of amorphous content based on the integrated intensities of broad humps compared with crystalline peak areas. To ensure reliability, duplicate measurements were performed, and the variation in calculated amorphous fractions remained within ± 2%, confirming acceptable repeatability of the results.

To further investigate the elemental composition of the soil specimens, energy-dispersive X-ray spectroscopy (EDS) was employed. EDS analysis was conducted by surface mapping on multiple regions of each specimen, with at least five spectra collected per sample. Reported values are presented as mean ± standard deviation to account for local heterogeneity. Moreover, the microstructural features of the samples were examined using field emission scanning electron microscopy (FE-SEM), which enabled detailed observation of the particle morphology in the untreated contaminated soil and the amorphous gel structures developed in the stabilized specimens. In addition, Fourier-transform infrared spectroscopy (FTIR) was performed to analyze the functional groups and chemical bonding in the geopolymer matrix.

Life cycle assessment (LCA)

To assess the environmental impacts of the stabilization and solidification methods, a comprehensive life cycle assessment (LCA) was performed. This study compared two treatment scenarios for lead-contaminated soil: one using solely volcanic ash and another employing a binary blend of volcanic ash and slag. The LCA was conducted following ISO 14040 and 14044 standards, adopting a cradle-to-gate scope that encompassed all processes from raw material extraction to the preparation of stabilizers ready for field application. Eleven midpoint impact categories from the ReCiPe 2016 methodology were selected, focusing on key environmental concerns relevant to cementitious material production—such as climate change potential (GWP100), human and ecological toxicity, and resource depletion (including metals and water). Characterization factors were applied according to ReCiPe’s guidelines.

System boundaries included raw material extraction, processing, manufacturing, and transportation, assuming a uniform transport distance of 100 km for all components to maintain consistency in comparison. Processes downstream of binder production—such as site preparation, soil mixing, compaction, and long-term performance—were excluded due to their similarity across treatments and negligible influence on comparative results.

Results

Compressive Strength

Figure 2 presents the evolution of unconfined compressive strength (UCS) in both uncontaminated and Pb-contaminated soils treated with varying slag replacement levels (0% to 40% by weight of volcanic ash), under two curing conditions: oven-dried (OC) and ambient (AC) conditions. Overall, UCS values increased with curing time, slag content, and in uncontaminated specimens.

Fig. 2.

Effect of different slag replacement levels on the UCS.

In uncontaminated soils (Fig. 2a,b), a notable rise in UCS was observed with increasing slag content at all curing ages. Under OC, the UCS of specimens with 40% slag reached approximately 15.3 MPa after 90 days—more than eightfold that of untreated soil. Ambient-cured specimens also displayed a similar trend, albeit with slightly lower absolute strengths.

Pb contamination led to a general reduction in UCS, especially in early curing stages. However, the incorporation of slag substantially mitigated this adverse effect. As shown in Fig. 2c,d, the UCS of contaminated samples improved significantly with higher slag contents. For instance, the 40% slag specimen under OC exhibited a UCS of 14.31 MPa at 90 days—nearly eight times that of untreated contaminated soil. Under AC, the same mix achieved around 8 MPa.

These findings are reinforced by the 90-day UCS data, which further underscore the beneficial role of slag incorporation. Under oven-curing conditions, replacing 40% of volcanic ash with slag resulted in strength increases of approximately 12% in uncontaminated and 20% in contaminated soils compared to the VA-only system. More notably, under ambient curing—where geopolymer development is typically limited—the inclusion of slag led to strength gains of over 50% in uncontaminated and nearly 75% in contaminated samples.

Leaching Behavior

Figure 3 illustrates the TCLP-extracted lead concentrations from stabilized contaminated soils cured under oven-dried (OC) and ambient (AC) conditions after 28 and 90 days. In all cases, the incorporation of slag into the alkali-activated binder markedly reduced Pb leachability, with increasing slag content leading to progressively lower leachate concentrations.

Fig. 3.

Effect of different slag replacement levels on lead concentration.

In oven-cured specimens (Fig. 3a), the untreated sample exhibited an extremely high Pb concentration (27.81 ppm), far exceeding the USEPA regulatory limit of 5 ppm. Upon treatment with a binder containing 100% volcanic ash (0% slag), Pb concentration dropped significantly to 3.2 ppm at 28 days, reflecting an 89% reduction relative to the untreated soil. This effect became more pronounced with increasing slag content. For instance, at 40% slag replacement, the Pb concentration declined sharply to approximately 0.13 ppm at 28 days—a reduction of over 99% compared to the untreated soil. By 90 days, the Pb concentration in this mix further decreased to below 0.1 ppm.

A similar trend was observed under ambient curing (Fig. 3b), though the absolute values of Pb leaching were generally higher than their oven-cured counterparts. The untreated sample showed Pb concentrations 37.3 ppm. At 0% slag, leachate Pb concentration was approximately 5.1 ppm after 28 days, only modestly below the regulatory limit. However, increasing the slag content to 40% reduced the Pb concentration to around 0.36 ppm at 28 days (∼99% reduction), which further decreased to 0.33 ppm by day 90. Comparatively, oven curing consistently enhanced the stabilization efficiency due to accelerated alkali-activation reactions and gel densification. However, even under ambient conditions, the hybrid binder system demonstrated excellent performance. At 40% slag, both curing regimes successfully reduced Pb leaching to less than 1% of the initial untreated level by day 90.

After 90 days, the inclusion of 40% slag resulted in a dramatic improvement in Pb immobilization efficiency under both curing regimes. Under oven curing, the addition of slag reduced Pb leachability by approximately 97% compared to the 100% volcanic ash binder. Under ambient conditions, this improvement was slightly lower but still substantial, achieving a more than 92% reduction.

Importantly, even under suboptimal ambient conditions, the VA–slag system achieved leachate concentrations well below the USEPA threshold, underscoring its robustness and long-term environmental reliability.

X-ray diffraction (XRD)

Figure 4 presents the mineralogical profiles of untreated and stabilized Pb-contaminated soils, as determined by X-ray diffraction (XRD) analysis. Figure 4a shows the diffraction patterns, while Fig. 4b quantifies the relative contributions of crystalline and amorphous phases.

Fig. 4.

XRD analysis results: (a) XRD patterns, and (b) quantified crystalline and amorphous phase.

The XRD pattern of the untreated soil reveals a predominantly crystalline structure, with sharp and intense peaks corresponding mainly to quartz (SiO₂) and calcite (CaCO₃). These minerals reflect the inert, non-reactive nature of the original soil matrix, which lacks significant binding potential. As indicated in Fig. 4b, the untreated soil exhibits an 83% crystalline phase content and only 17% amorphous content.

In contrast, the treated samples display substantial mineralogical transformation. In all stabilized mixes, additional crystalline phases—specifically albite (NaAlSi₃O₈) and anorthite (CaAl₂Si₂O₈)—are detected alongside residual quartz and calcite. These new phases suggest the incorporation of alkali and calcium into the aluminosilicate network during alkali-activation reactions.

Compared to the untreated soil, the amorphous content increases substantially in all stabilized samples, ranging from 28% in V60G40–AC to 32% in V60G40–OC. This enhancement reflects the successful dissolution of aluminosilicate precursors and subsequent polycondensation reactions, leading to the development of a disordered yet highly stable binding matrix.

Notably, the incorporation of slag (i.e., V60G40 mixes) leads to slightly higher amorphous content than VA-only mixes (V100G0), with oven-cured samples achieving the maximum amorphous proportion (32% in V60G40–OC). This improvement can be attributed to the calcium-rich nature of slag, which promotes the co-formation of both NASH and C(A)SH gel phases—key contributors to microstructural densification and heavy metal immobilization. In contrast, VA-only mixes primarily generate NASH gels with comparatively slower kinetics.

While the curing condition had a modest impact on overall phase composition, oven curing slightly favored amorphous gel development due to accelerated geopolymerization. For example, V100G0–OC showed a 1% higher amorphous fraction than its AC counterpart. However, the influence of slag content on mineralogy was more pronounced than that of curing regime alone.

Energy-dispersive X-ray spectroscopy (EDX)

Figure 5 shows the EDS elemental compositions of untreated soil, V100G0–OC, V100G0–AC, V60G40–OC, and V60G40–AC specimens. Reported values represent the mean of five spectra obtained from surface mapping across multiple regions of each sample. While Fig. 5 provides a visual overview of average elemental contributions, Table 2 complements these results by presenting the mean ± standard deviation values.

Fig. 5.

Results of EDS analysis.

Table 2.

Average elemental compositions from EDS mapping (mean ± SD, n = 5 spectra).

Chemical element	Untreated	V100G0–OC	V100G0–AC	V60G40–OC	V60G40–AC
Si (wt%)	34.7 ± 1.1	32.7 ± 1.3	32.9 ± 1.0	26.0 ± 1.2	27.6 ± 1.1
Al (wt%)	11.7 ± 0.7	11.0 ± 0.6	8.6 ± 0.5	8.8 ± 0.6	8.5 ± 0.5
Na (wt%)	0.9 ± 0.3	6.6 ± 0.4	7.3 ± 0.3	8.5 ± 0.3	6.1 ± 0.3
Ca (wt%)	5.9 ± 0.2	10.0 ± 0.3	7.1 ± 0.2	15.0 ± 0.5	19.9 ± 0.4
O (wt%)	42.9 ± 1.8	36.4 ± 2.0	42.3 ± 1.9	40.2 ± 1.7	36.2 ± 1.6
Pb (wt%)	3.9 ± 0.6	3.3 ± 0.2	1.8 ± 0.2	1.5 ± 0.1	1.7 ± 0.1

The Na content in the untreated soil was found to be very low, which is attributed to the absence of geopolymer gel formation in these samples. In contrast, the treated samples (V100G0−OC, V100G0−AC, V60G40−OC, and V60G40−AC) exhibited higher Na concentrations. Along with the appropriate presence of silicon, aluminum, and calcium ions, these elements provided favorable conditions for the formation of NASH and C(A)SH gels.

Moreover, the Si/Al and Na/Al atomic ratios—known indicators of geopolymerization potential—were within the optimal range in all treated samples. An increase in the Si/Al ratio was observed in V100G0−OC and V100G0−AC compared to the untreated soil, indicating the development of a geopolymer network. Furthermore, higher Ca/Si ratios were detected in V60G40−OC and V60G40 − AC samples, suggesting a greater tendency for the formation of C(A)SH gels.

Field emission scanning electron microscopy (FE-SEM)

Figure 6 illustrates the FE-SEM micrographs of untreated and stabilized Pb-contaminated soil specimens after 90 days of curing.

Fig. 6.

FE-SEM analysis results: (a, b) non-stabilized soil; (c) soil stabilized with 100% VA (V100G0) under oven-dried condition; (d) soil stabilized with 100% VA (V100G0) under ambient condition; (e) soil stabilized with 60% VA and 40% slag (VA60G40) under oven-dried condition; and (f) soil stabilized with 60% VA and 40% slag (VA60G40) under ambient condition.

The microstructure of the untreated soil (Fig. 6a,b) exhibits a rough and granular texture with high porosity and no evidence of geopolymeric gel formation. This observation is consistent with the predominantly crystalline phases and low sodium content revealed by XRD and EDS analyses.

For the VA-only system (V100G0, Fig. 6c,d), both oven- and ambient-cured samples reveal porous regions, microcracks, and partially reacted volcanic ash particles. Amorphous NASH gel is present along with residual calcite and hydroxy-sodalite crystals. However, the gel coating around particles remains incomplete, particularly under ambient curing, leading to a relatively porous matrix. These features agree with the moderate amorphous content (up to 31% in V100G0–OC) and the Na/Al and Si/Al ratios obtained from EDS.

The hybrid VA–slag system (V60G40, Fig. 6e,f) demonstrates a denser and more cohesive microstructure compared with VA-only mixtures. Porosity and microcracking are significantly reduced, especially under oven curing. Both NASH and C(A)SH gels are clearly visible, with C(A)SH appearing as a compact, fibrous phase filling the intergranular voids. This morphology corresponds to the higher amorphous content (up to 32% in V60G40–OC) and elevated Ca/Si ratios identified by EDS, highlighting the beneficial role of slag in enhancing matrix densification.

Quantitative image analysis further confirmed these differences. The untreated soil exhibited porosity above 35% with a wide pore size distribution (0.5–4.0 μm). VA-only specimens showed partial gel coverage (~ 40–45% of particle surfaces), porosity of 25–28%, and pore sizes ranging from 0.8–3.0 μm. In contrast, VA–slag specimens exhibited markedly lower porosity (15–18%), narrower pore size distribution (0.3–1.5 μm), and higher gel coverage (~ 65–70%). These findings provide quantitative evidence of slag’s positive influence on matrix densification, supporting the improved mechanical strength and Pb immobilization observed in V60G40 systems.

Fourier-transform infrared spectroscopy (FTIR)

Figure 7 presents the FTIR spectra of untreated and stabilized samples after 90 days of curing. The untreated soil shows broad absorption bands around 3400 cm^-1 and 1630 cm^-1, which are attributed to the O–H stretching and H–O–H bending vibrations of adsorbed water, respectively. These features persist in the stabilized samples but exhibit reduced intensity, indicating a lower content of free or loosely bound water due to the formation of binding phases.

Fig. 7.

FTIR spectra of non-stabilized and stabilized soil samples.

A significant shift and broadening of the band in the region of 950–1000 cm^-1 is observed in all stabilized samples compared to the untreated soil. This band is associated with the asymmetric stretching vibrations of Si–O–T (T = Si or Al) in the aluminosilicate framework. The emergence and enhancement of this band point to the formation of geopolymeric gels—mainly NASH in the V100G0 mixes and a combination of NASH and CASH in the V60G40 mixes. The presence of slag in the V60G40 formulations leads to a more intense and slightly shifted Si–O–T band, consistent with the incorporation of Ca into the aluminosilicate network.

Moreover, the samples exhibit absorption bands in the region of 450–550 cm^-1, corresponding to bending vibrations of Si–O–Si and Si–O–Al bonds, further confirming the formation of amorphous aluminosilicate structures. The AC samples generally show sharper and more intense peaks across the spectrum compared to their OC counterparts, reflecting a more advanced reaction process and greater polymerization under ambient curing conditions.

The band in the 3000–3500 cm^-1 range in AC samples also overlaps with the characteristic stretching vibration of –OH groups associated with CSH gels, particularly in V60G40–AC, reinforcing the coexistence of calcium-rich hydration products in slag-containing mixes.

Table 3 summarizes the main absorption bands observed in the FTIR spectra, their chemical assignments, and comparison with values reported in the literature. The characteristic bands between 950–1000 cm^-1 correspond to asymmetric stretching of Si–O–T (T = Si or Al), confirming the formation of geopolymeric NASH and C(A)SH gels. The bands at 1400–1440 cm^-1 indicate the presence of carbonates, while the broad absorption between 3000–3500 cm^-1 reflects O–H stretching in both water and hydration products. The peaks around 1630 cm^-1 are associated with H–O–H bending of absorbed water. No peak deconvolution was performed, as the analysis was intended to capture the main functional group transformations; however, the major shifts and broadening trends are consistent with reported geopolymer spectra in the literature^51–55.

Table 3.

Summary of FTIR absorption bands and assignments.

Band position (cm-1)	Assignment	Observed trend in this study	Literature references
3000–3500	O–H stretching (water, CSH gels)	Broad band, reduced intensity after stabilization	50
1630	H–O–H bending of absorbed water	Present in all samples, weaker in stabilized soils	51
1400–1440	CO32- stretching (carbonates)	Detected in VA-only mixes, reduced in VA–slag	52
950–1000	Si–O–T asymmetric stretching (NASH, C(A)SH gels)	Shifted and broadened in VA–slag systems	53
450–550	Si–O–Si, Si–O–Al bending vibrations	More intense in slag-containing binders	54

Life cycle assessment (LCA)

Life cycle inventory data for VA, slag and NaOH were obtained from previous studies that used material sourced from the same supplier as in the present study^15,35,56. Supplementary inventory data were extracted from the Ecoinvent v3.8 database, which provides background data for industrial materials based on regional production systems. For the LCA, the functional unit was defined as 1 m³ of stabilized lead-contaminated soil exhibiting comparable UCS after 28 days of oven curing. Based on compaction tests, the maximum dry mass of contaminated soil per cubic meter was assumed to be 1740 kg.

For the reference binder system (100% volcanic ash), the required materials per cubic meter of soil included 261 kg of VA (15 wt% of dry soil), 243.6 kg of alkaline solution (14 wt% of dry soil). The alkaline solution was formulated as a 6 molar NaOH solution, comprising 47.88 kg of solid NaOH dissolved in 195.72 kg of water, yielding a NaOH concentration of approximately 6 M and a total solution mass equivalent to 14% of the soil’s dry weight. In the hybrid VA–slag system (V60G40), the blended binder (10 wt% of dry soil) consisted of 104.4 kg of VA and 69.6 kg of slag (corresponding to 60% and 40% of the total binder, respectively), with the same 243.6 kg of 6M NaOH solution used as the activator. All other mix design parameters—such as liquid-to-binder ratio, activator concentration, and curing method—were kept constant to isolate the environmental impact of precursor substitution. This standardized approach ensured a fair comparison of the environmental performance between different binder formulations.

The life cycle impact assessment was carried out using SimaPro software (version 9.5, https://simapro.com). All modeling assumptions, input flows, and boundary definitions were applied consistently across the two binder formulations to ensure that the comparative results reflect only differences in material composition and associated production processes. Figure 8 presents their percentage contributions and logarithmic-scale emission amounts across 11 impact categories.

Fig. 8.

Contribution and emission amount of environmental impact categories.

The LCA results do not indicate a significant environmental advantage for the V60G40 mixture over the V100G0 formulation. In fact, across most impact categories, the environmental profiles of the two binders are broadly comparable. For instance, the climate change potential of the V60G40 system is slightly lower, decreasing from 68.3 to 65.4 kg CO₂ equivalent—an improvement of just around 5%—despite the incorporation of slag.

Marine eutrophication was also marginally higher in the slag-free system, while human toxicity potential showed a notable decline—from 18.1 to 14.7 kg 1,4-DB eq—indicating reduced emissions of hazardous substances during production and use. For other impact categories such as photochemical oxidant formation and particulate matter formation, both mixtures showed nearly identical results, implying that binder composition had negligible influence on these aspects.

A closer examination of the percentage contribution analysis reveals that the V60G40 system tends to impose a marginally higher environmental burden in specific impact categories, including terrestrial ecotoxicity, marine ecotoxicity, freshwater ecotoxicity, and metal depletion. In the remaining categories, the environmental contributions of both mixtures are largely comparable, indicating that replacing volcanic ash with granulated blast furnace slag in this context does not consistently result in increased environmental impacts.

Discussion

The findings of this study provide a comprehensive understanding of the performance and environmental implications of volcanic ash-based alkali-activated binders with and without slag as a partial replacement. The multi-scale analyses—ranging from mechanical strength to microstructure, leaching behavior, and life cycle impacts—reveal the critical influence of slag incorporation on both engineering properties and sustainability metrics of stabilized Pb-contaminated soils.

The progressive increase in UCS with slag content and curing time demonstrates the synergy between calcium-rich slag and aluminosilicate geopolymer networks. While VA-only mixes showed reasonable strength gain due to the formation of NASH gels, the inclusion of slag (up to 40%) facilitated the co-formation of CASH and hybrid N,C(A)SH gels, especially under oven conditions. These dual gel systems significantly densify the matrix and accelerate early-age strength development, as evidenced by the 15.3 MPa UCS at 90 days in the V60G40–OC mix—more than 8 times that of untreated soil and up to 75% improvement compared to VA-only formulations.

The adverse influence of Pb on early alkali-activation reactions was evident in contaminated samples, particularly during the first 7 days. However, the slag-containing systems demonstrated an enhanced ability to overcome heavy metal interference, eventually achieving UCS values comparable to uncontaminated counterparts. This resilience is attributed to the higher availability of reactive Ca²⁺ ions from slag, which not only participate in early hydration reactions but may also aid in heavy metal immobilization through ion exchange, precipitation, or incorporation into amorphous gels^57,58.

Leaching test results provided further validation of slag’s critical role in enhancing Pb stabilization. Although VA-only binders succeeded in reducing leachate Pb concentrations below regulatory thresholds, the inclusion of slag led to an order-of-magnitude improvement in immobilization efficiency. This effect was particularly pronounced under oven-cured conditions, where the Pb concentration in the V60G40–OC specimen declined to below 0.1 ppm by day 90—translating to over 99.6% reduction in leachability. Such substantial immobilization is indicative of multiple concurrent mechanisms operating within slag-containing systems.

The immobilization of Pb in VA–slag alkali-activated matrices occurs through multiple concurrent mechanisms^13,35. Microstructural and spectroscopic analyses confirmed the formation of NASH and C(A)SH gels, which provide both chemical and physical pathways for Pb stabilization. Chemically, Pb²⁺ ions may substitute for Ca²⁺ or Na⁺ within the gel framework or be incorporated into secondary crystalline phases such as anorthite and albite, as indicated by XRD patterns. The FTIR spectra revealed shifts in the Si–O–T (T=Si, Al) bands, suggesting structural incorporation of Pb into the aluminosilicate network. EDS results further demonstrated localized Pb enrichment within Ca-rich regions, supporting the hypothesis of ion exchange and incorporation into C(A)SH gels. Physically, FE-SEM images showed a denser microstructure with reduced porosity in slag-containing systems, indicating that Pb was also encapsulated within a compact gel matrix. Similar findings have been reported in previous studies, where heavy metals were immobilized through a combination of ion exchange and gel incorporation^57,58.

Overall, the Pb immobilization pathway in VA–slag binders can be summarized as: (i) sorption and ion-exchange of Pb²⁺ on the surfaces of N–A–S–H and C–(A–S–H / N,C–(A)–S–H-type gels and newly formed aluminosilicate phases, (ii) partial structural incorporation of Pb into these gels and into secondary Pb-bearing aluminosilicate or carbonate phases, and (iii) physical encapsulation of Pb within the densified amorphous matrix as gel formation refines the pore structure.

It is worth noting that Pb is amphoteric and can form soluble hydroxo-complexes at very high alkalinity; therefore, in alkali-activated matrices its immobilization is more accurately described by a combination of sorption, incorporation into gel phases, and microstructural encapsulation rather than by a simple model of complete precipitation as an insoluble hydroxide.

The synergistic effect of VA and slag enhances these mechanisms by providing a balance between aluminosilicate polymerization (VA) and Ca²⁺ availability (slag), resulting in more efficient and stable long-term Pb stabilization.

Microstructural and spectroscopic analyses corroborate these mechanical and environmental performance trends. XRD patterns revealed that untreated soil was predominantly crystalline and inert, while all treated specimens exhibited elevated amorphous content—a hallmark of active geopolymerization. The amorphous fraction in the V60G40–OC sample increased from 17% to over 32%, reflecting the formation of complex hybrid gels. Furthermore, the presence of secondary crystalline phases such as anorthite and albite in treated systems suggests successful incorporation of Ca and Na into the evolving aluminosilicate network.

FE-SEM micrographs visually confirmed the microstructural transformations induced by slag incorporation. VA-only matrices remained relatively porous and poorly bonded, particularly under ambient curing, whereas slag-rich systems exhibited well-developed gel coatings, reduced microcracking, and enhanced matrix cohesion. The fibrous morphology associated with C(A)SH gels was clearly observed in V60G40 samples, highlighting their contribution to mechanical strength and Pb retention.

Quantitative FE-SEM data provide further evidence of slag’s beneficial role in refining the microstructure of alkali-activated binders. Porosity decreased from ~ 28% in VA-only systems to ~ 15% in VA–slag mixtures, accompanied by narrower pore size distribution and higher gel coverage (~ 70%). These improvements explain the substantial enhancement in UCS and Pb immobilization efficiency. A denser matrix enriched with both C(A)SH and NASH phases not only increases load-bearing capacity but also enhances the physical encapsulation of Pb ions. Overall, these findings demonstrate that slag incorporation promotes pore refinement and gel continuity, thereby establishing a strong link between microstructural parameters and the observed macroscopic performance.

FTIR spectra provided further molecular evidence of gel formation and interaction with Pb species. All stabilized samples showed shifts in the Si–O–T (T = Si or Al) bands toward lower wavenumbers and increased broadness, indicating the formation of disordered geopolymeric gels. The more intense and shifted bands in V60G40 samples suggest a higher degree of cross-linking and possible Ca²⁺ incorporation into the gel network. Interestingly, ambient-cured samples showed sharper infrared absorption bands in the 450–1000 cm^-1 region, which could point to slower yet more complete polymerization in the absence of thermal activation. The FTIR findings also support the hypothesis that multiple immobilization mechanisms—including chemical bonding and physical encapsulation—are simultaneously active in slag-enhanced systems.

The life cycle assessment reveals nuanced insights that challenge the assumption that slag always enhances environmental performance. While the incorporation of slag marginally reduced climate change potential, it resulted in higher impacts in categories such as marine ecotoxicity, terrestrial ecotoxicity, and metal depletion. Moreover, heavy metals present in slag itself may contribute to ecotoxicity potentials, offsetting its environmental advantages in other areas. From a broader perspective, this study advances the field of sustainable soil remediation by demonstrating the feasibility of using industrial by-products in tandem—VA and slag—to stabilize contaminated soils.

While the present work demonstrates the laboratory-scale efficiency of VA–slag alkali-activated binders, several factors must be considered for field-scale application. First, natural soils are inherently heterogeneous in terms of mineralogy, moisture content, and contaminant distribution, which may influence binder dispersion and reaction kinetics. Moreover, actual contamination levels often vary widely, and while the selected Pb concentration (2500 mg/kg) represents a severe scenario, field cases may involve lower or mixed heavy-metal loads, requiring adjustment of binder dosage and composition. Second, long-term durability under cyclic wetting–drying, freeze–thaw, and carbonation conditions remains a critical issue that warrants extended performance evaluations. Third, economic and logistical aspects—including the sourcing, quality control, and transportation of slag—must be assessed, particularly when industrial by-products are used in large volumes. These practical considerations highlight that although VA–slag binders exhibit strong potential as sustainable stabilizers, future research should focus on validating their performance under diverse field conditions, optimizing mix designs for variable contamination, and conducting cost–benefit analyses to ensure large-scale feasibility.

Conclusion

This study comprehensively investigated the effect of partially replacing volcanic ash with slag in alkali-activated material systems for the stabilization/solidification of lead-contaminated soils. By integrating mechanical testing, leaching analysis, microstructural and spectroscopic characterization (XRD, FE-SEM, EDS, FTIR), and LCA, the research provides a multi-dimensional evaluation of both performance efficacy and environmental sustainability.

The incorporation of slag significantly enhanced the mechanical properties of the treated soils. After 90 days of curing, specimens with 40% slag exhibited UCS up to 75% higher than those with VA alone, and more than 13-fold greater than untreated soil. This improvement is attributed to the synergistic formation of C(A)SH and NASH gels, resulting in denser, more cohesive matrices with superior load-bearing capacity. Lead immobilization was also markedly improved by slag addition. After 90 days, the inclusion of slag reduced Pb leachability by approximately 97% under oven curing and over 92% under ambient conditions, relative to VA-only binders. These reductions correspond to over 99% lower leachate concentrations compared to untreated contaminated soil. The enhanced performance is linked to the calcium-rich nature of slag, which facilitates the chemical incorporation of Pb into stable gel and crystalline structures, as well as its physical encapsulation within the hardened matrix.

Nonetheless, the environmental impacts of slag use warrant consideration. While LCA results showed a modest 5% reduction in greenhouse gas emissions with slag-containing binders, they also revealed increased burdens in categories such as marine and freshwater ecotoxicity and metal depletion. This trade-off underscores the importance of optimizing binder formulations not only for technical performance but also for life-cycle environmental outcomes.

Although this study provides strong evidence of the efficiency of VA–slag alkali-activated binders for Pb-contaminated soils, certain limitations should be acknowledged. First, the experiments were performed under controlled laboratory conditions using artificially contaminated soils, which do not fully capture the heterogeneity of real field sites. Second, the study focused exclusively on Pb, whereas other heavy metals such as Cd, Zn, and As may interact differently with the geopolymer matrix and should be systematically investigated. Third, the long-term durability of the stabilized soils under environmental exposure (e.g., wetting–drying, freeze–thaw, carbonation) requires further validation. Future research should therefore aim to (i) evaluate binder performance in naturally contaminated and heterogeneous soils, (ii) expand the scope to multiple contaminants, and (iii) explore pilot-scale or field-scale implementations, including cost–benefit assessments. Addressing these aspects will help bridge the gap between laboratory research and real-world applications, ensuring that VA–slag binders can be deployed as practical and sustainable remediation technologies.

In conclusion, the partial replacement of volcanic ash with slag in alkali-activated systems offers a powerful strategy to improve the stabilization/solidification of heavy metal-contaminated soils. These hybrid binders provide a compelling combination of mechanical strength, contaminant immobilization, and acceptable environmental performance, making them promising candidates for eco-efficient remediation technologies.

Author contributions

**Alireza Komaei:** Conceptualization, Methodology, Supervision, Formal analysis, Data Curation; Writing—Original Draft, Writing—Review & Editing. **Mohammad Amin Molaei:** Investigation, Methodology, Conceptualization, Project administration.

Funding

The authors received no specific funding for this work.

Data availability

All data generated or analyzed during this study are available from the corresponding author upon reasonable request. *Corresponding author* : *Alireza Komaei; email:* [*komaei@aut.ac.ir*](mailto:komaei@aut.ac.ir) .

Competing interests

The authors declare no competing interests.