Sustainable Mortars Incorporating Industrial Rolling Mill Residues: Microstructural, Physical, and Chemical Characteristics

Abstract

New alternatives in the construction industry are essential for economic, sustainable, and environmental progress. In this context, this work investigated three sets of sustainable mortars incorporating industrial lamination waste, assessing their chemical, physical, microstructural, and mechanical properties to inform their development. Cylindrical and prismatic specimens were produced using the following incorporation methods: a reference mortar, mortars with mill scale addition, partial cement replacement with mill scale, and partial sand replacement with mill scale, at proportions of 10%, 20%, 30%, 40%, and 50%. Additionally, analyses including X-ray fluorescence (XRF), X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR), and scanning electron microscopy with energy-dispersive spectroscopy (SEM/EDS) were performed. Physical and mechanical tests, including bulk density, consistency index, capillary water absorption, axial compressive strength, and flexural tensile strength, were also conducted. XRF results indicated an increase in iron oxide content and a decrease in calcium oxide with the addition of mill scale. XRD confirmed the presence of compounds, such as alite and portlandite, which are common in cementitious mortars. FTIR spectra exhibited characteristic functional groups through absorption bands related to Si–O stretching. SEM micrographs revealed slight morphological changes in the composites as the quantity of industrial lamination waste increased, and EDS data supported the XRF findings. The addition of industrial lamination waste affected the spread index and density of the mixtures, while capillary water absorption decreased in some formulations with mill scale. The strength of the mortars increased with the incorporation of industrial lamination waste. In conclusion, using industrial lamination waste in mortars is a technically and environmentally feasible alternative that aligns with the principles of sustainable development and the circular economy in the construction industry.

1. Introduction

The transformation of waste into valuable resources for the construction industry aligns with sustainability principles because improving industrial practices is crucial for conserving the planet’s natural resources [1,2]. When considering waste generation, it is essential to recognize that rapid and global industrialization has increased production capacity, making goods and services more readily available. This process has helped improve people’s quality of life and driven economic growth. However, it also presents a significant challenge: how to manage all this waste effectively [3]. A key example is the steel industry, which plays a vital role in supplying materials to sectors such as shipbuilding, healthcare, transportation, automotive, and civil engineering. Despite its importance to the economy, it is also a significant source of polluting waste. If steel waste is not disposed of or stored correctly, it can cause severe environmental damage [2,4].

Waste generation is not the only sustainability concern. The construction industry accounts for approximately half of all natural resources. As a result, this sector has adopted more sustainable practices, such as the use of industrial waste to produce construction materials, including mortar and concrete. This reduces the need for new raw materials. Utilizing waste to produce construction materials is a practical way to reduce environmental impacts and can also lower costs per square meter. Overall, this approach helps reduce pollution, conserve natural resources, save energy, and decrease waste sent to industrial landfills [5,6].

The metallurgical industry also generates substantial amounts of waste and faces significant challenges in managing it. Even with specific laws in place, many companies ignore these rules, often due to a lack of technical expertise or difficulties in meeting legal requirements [7,8]. Due to its characteristics and properties, steel mill waste has been used to produce recycled construction materials. Managing this waste is one of the most complex and challenging tasks worldwide, as it has a significant environmental impact. For example, scale—a metallurgical waste—can be used in cement production to partially replace clinker, serve as a base for unpaved roads, or act as an aggregate in concrete [2].

While the use of mill scale in concrete production and as a partial replacement for clinker is well documented, recent studies have explored its potential as a fine aggregate in mortars [8,9,10,11]. However, most of these studies focus on lower replacement levels or specific properties such as electromagnetic shielding. A comprehensive comparative analysis that evaluates both the chemical effects (as a binder substitute) and the physical packing effects (as an acceptable aggregate substitute) within the same experimental matrix—particularly at high replacement levels (up to 50%)—remains limited. This study bridges this gap by systematically comparing these two incorporation routes, linking mechanical performance directly to microstructural changes (ITZ analysis and densification) to define the optimal role of mill scale in sustainable mortars. However, most of these studies focus on lower replacement levels or specific properties such as electromagnetic shielding [11]. A comprehensive comparative analysis that evaluates both the chemical effects (as a binder substitute) and the physical packing effects (as an acceptable aggregate substitute) within the same experimental matrix—particularly at high replacement levels (up to 50%)—remains limited, with most research focusing on sand replacement [8,9,12]. This study bridges this gap by systematically comparing these two incorporation routes, linking mechanical performance directly to microstructural changes (ITZ analysis and densification) to define the optimal role of mill scale in sustainable mortars. The positioning of this research in relation to the current state of the art is summarized in

Table 1. Benchmarking of recent studies on mill-scale incorporation in cementitious matrices.

Study	Material Replaced	Max. Replacement	Main Research Focus
Pereira et al. [8]	Fine Aggregate	Not specified	Mechanical behavior of concrete using mill scale as aggregate.
Ozturk et al. [11]	Fine Aggregate	15%	Development of mortars for electromagnetic wave shielding.
Parvathikumar et al. [12]	Fine Aggregate	20%	Durability characteristics and river sand replacement.
Siriwattanakarn et al. [9]	Fine Aggregate	100%	Physical, mechanical, and post-fire properties.
Present Study	Cement and Sand	50%	Comparison of chemical vs. physical mechanisms and ITZ analysis.

.

Furthermore, understanding the upper limits of incorporation is critical for maximizing the environmental benefits of waste valorization. While conservative replacement levels are commonly investigated to ensure safety, they limit the total volume of waste diverted from landfills. By extending the replacement range to 50% and employing advanced microstructural techniques (such as EDS mapping of the Interfacial Transition Zone), this research moves beyond simple feasibility testing. It aims to provide a distinct, mechanism-based explanation for the contrasting behaviors of mill scale as a binder versus an aggregate, thereby offering a scientifically grounded roadmap for the steel and construction industries to implement high-volume recycling strategies effectively [9,10,12].

From a chemical perspective, mill scale is composed of iron oxides, with this composition varying across layers, as depicted in the iron–oxygen (Fe-O) phase diagram. According to Ahmed et al. [6,13], mill scale can be considered a valuable raw material due to its high iron content, low impurity levels, and stable chemical composition. Additionally, the literature indicates that in many cases, mill scale is stored or improperly disposed of in landfills without proper reuse or segregation [6,11,14]. Further research is needed on the potential for reusing mill scale in the construction industry, particularly for mortar production. Thus, incorporating it into cementitious materials presents a promising strategy to reduce environmental impacts, decrease the extraction of natural raw materials, and support practices aligned with the principles of the circular economy.

In this work, sustainable mortars were prepared by replacing cement and sand (at 10%, 20%, 30%, 40%, and 50%) with industrial rolling waste. XRF, FTIR, XRD, SEM, EDS, and both compressive and tensile strength tests were performed.

2. Materials and Methods

2.1. Materials

The raw materials employed to prepare the studied mortars included Portland cement (CPII-Z-32), fine aggregate (fine sand), hydrated lime (CH-1), mill scale from the metallurgical industry (Figure 1), and distilled water. The chemical properties of cement, sand, and lime are detailed in

Table 2. Chemical compositions of cement, hydrated lime, and sand (unit in % by mass).

	MgO	Al2O3	SiO2	P2O5	SO3	Cl	K2O	CaO	TiO2	MnO	Fe2O3	SrO	ZrO3
Cement	1.01	5.48	18.54	0.57	4.25	0.04	1.66	65.18	-	-	3.18	0.09	-
Hydrated lime	0.62	0.37	0.84	0.10	0.14	0.28	0.13	97.77	-	-	-	-	-
Sand	-	1.87	92.17	-	0.14	0.05	0.16	0.63	3.09	-	0.41	-	1.47

, while the grain size distribution (granulometry) of both sand and mill scale is presented in

Table 3. Particle size distribution of fine aggregate and sand.

Particle Size Distribution
Sieve Opening (mm)	Test Method	Sand		Mill Scale
		Percentage Retained (%)	Cumulative Percentage (%)	Percentage Retained (%)	Cumulative Percentage (%)
4.75	NBR 17054	0	0	0.05	0.05
2.36		0.65	0.65	0.10	0.15
1.18		4.21	4.87	4.66	4.81
0.6		24.74	29.60	12.07	16.88
0.3		42.70	72.30	50.58	67.45
0.15		17.66	89.96	17.93	85.38
Background		10.04	100	14.62	100
Total		100	-	100	-

.

The mill scale used is a byproduct of the steel rolling process and consists predominantly of inorganic iron oxides (

Table 4. Chemical composition of mill scale (unit in % by mass).

	Al2O3	SiO2	SO3	Cl	K2O	CaO	MnO	Fe2O3
Mill scale	0.63	2.26	0.20	0.02	0.09	0.33	0.57	95.90

). It was obtained free of organic binders, resins, and hazardous volatile compounds, such as phenols or formaldehyde, ensuring that the study focuses on the interaction between the mineral phases and the cementitious matrix.

The mill scale used in this study was obtained from Gerdau S.A., located in Pernambuco, Brazil. To ensure standardized properties, the residue was oven-dried at 105 ± 5 °C for 24 h to remove moisture, then ground in a ball mill for 30 min at 40 cycles per minute. After processing, a particle size analysis was performed on two 300 g samples in accordance with NBR 17054 [15]. The sieving process utilized mesh openings of 4.75 mm, 2.36 mm, 1.18 mm, 0.6 mm, 0.3 mm, and 0.15 mm, following the same procedure adopted for the fine aggregate. This controlled particle-size distribution enables the scale to function effectively as a microfiller, thereby increasing mass density and improving packing in the composites.

To evaluate the microstructure, morphology, and chemical element mapping of the precursor material, scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDS) were performed on the raw mill scale (Figure 2). This characterization is essential for establishing baseline physical and chemical properties of the residue, thereby improving understanding of its interaction with the cementitious matrix. The morphological analysis confirmed the particles’ surface roughness and angularity, which are critical factors for mechanical interlocking at the interfacial transition zone (ITZ).

The morphology and chemical composition of the mill scale are presented in Figure 2. At a 500 µm scale (Figure 2a), the residue consists of predominantly misshapen fragments with markedly rough surfaces. Higher magnification (2 µm) reveals smaller particles adhered to the larger matrix, exhibiting highly irregular surfaces. The EDS spectrum (Figure 2) confirms that Iron (Fe) and Oxygen (O) are the primary constituents, with minor occurrences of Carbon (C), Copper (Cu), Aluminum (Al), Calcium (Ca), and traces of Silicon (Si) and Sulfur (S). This combination of high surface roughness and metallic composition is pivotal for the mechanical interlocking observed in the mortar’s interfacial transition zone.

2.2. Proportions of the Mortars

The mortar mix proportions, by volume, were 1:1:6, consisting of Portland cement, CH-I hydrated lime, and sand. Partial replacement was performed under two conditions: (a) partial replacement of cement with scale residue; and (b) partial replacement of sand with scale residue in proportions of 10%, 20%, 30%, 40%, and 50%, in both cement and fine aggregate. The water/cement ratio was set at 1.5. While this value may appear high for structural concrete or pure cement mortars, it is characteristic of mixed mortars containing hydrated lime (1:1:6 ratio). Hydrated lime has a high specific surface area, thereby increasing the water demand required to achieve the necessary plasticity for masonry applications. When considering the total binder content (cement + lime), the water-to-binder ratio is approximately 0.70, which falls within the standard range for rendering and masonry mortars, ensuring adequate workability and adhesion. The water-to-cement (w/c) ratio was kept constant at 1.5 across all mixtures to isolate the effect of mill-scale incorporation on mortar properties. This decision ensured that any observed variations in workability, density, and mechanical performance could be directly attributed to the waste’s physical and chemical characteristics rather than to fluctuations in water content. This standardization is essential for the reproducibility of the experimental matrix and for a scientifically grounded comparison between the SUB-C and SUB-A series. CP II–Z–32 cement, sand as fine aggregate, CH–I hydrated lime, and scale were used to make the mortars, allowing the impact of adding this residue into the mix to be evaluated. For each formulation, 12 cylindrical and six prismatic specimens will be molded, enabling a detailed analysis of the mortars’ mechanical and rheological properties. In this study, mixes with cement substitution are coded as SUB-C, whereas those with substitution in the fine aggregate (sand) are coded as SUB-A. The numerical values that follow each code represent the percentage of cement or sand replaced by scale residue. These abbreviations are consistently used in

Table 5. Configuration of the studied mortars.

Code	Mortar Configuration
REF	Reference mortar in a 1:1:6 cement:lime:sand ratio.
SUB-C-10%	Mortar with 10% replacement in a 1:1:6 cement:lime:sand ratio (replacing cement with scale).
SUB-C-20%	Mortar with a 20% replacement in a 1:1:6 cement:lime:sand ratio (replacing cement with scale).
SUB-C-30%	Mortar with a 30% replacement in a 1:1:6 cement:lime:sand ratio (replacing cement with scale).
SUB-C-40%	Mortar with a 40% replacement in a 1:1:6 cement:lime:sand ratio (replacing cement with scale).
SUB-C-50%	Mortar with a 50% replacement in a 1:1:6 cement:lime:sand ratio (replacing cement with scale).
SUB-A-10%	Mortar with a 10% replacement in a 1:1:6 cement:lime:sand ratio (replacing sand with scale).
SUB-A-20%	Mortar with a 20% replacement in a 1:1:6 cement:lime:sand ratio (replacing sand with scale). Mortar with a 30% replacement in a 1:1:6 ratio of cement:lime:sand (replacing sand with scale).
SUB-A-30%	Mortar with a 40% replacement in a 1:1:6 ratio of cement:lime:sand (replacing sand with scale).
SUB-A-40%	Mortar with a 50% replacement in a 1:1:6 ratio of cement:lime:sand (replacing sand with scale).
SUB-A-50%	Mortar with a 20% replacement in a 1:1:6 cement:lime:sand ratio (replacing cement with scale).

.

2.3. Processing of Mortars

To produce the test specimens (PCs), the following mortar mixes were prepared: a reference mix, a mix with partial replacement of cement, and a blend with partial replacement of sand.

Table 6. Materials quantities used in producing reference test specimens.

Reference
Code	Proportion	Test Specimen	Cement (kg)	Lime (kg)	Sand (kg)	Residue (kg)	Water (kg)
REF	1:1:6	12 cylindrical	0.7312	0.7312	4.387	-	1.023
		6 prismatic	0.6142	0.6142	3.685		0.8599

,

Table 7. Materials quantities used in preparing cement mass replacement test specimens.

Replacing the Cement Mass
Code	Proportion	Test Specimen	Cement (kg)	Lime (kg)	Sand (kg)	Residue (kg)	Water (kg)
SUB-C-10%	1:1:6	12 cylindrical	0.6581	0.7312	4.387	0.0731	1.023
SUB-C-20%		6 prismatic	0.5528	0.6142	3.685	0.0614	0.8599
SUB-C-30%	1:1:6	12 cylindrical	0.5850	0.7312	4.387	0.1462	1.023
SUB-C-40%		6 prismatic	0.4914	0.6142	3.685	0.1228	0.8599
SUB-C-50%	1:1:6	12 cylindrical	0.5118	0.7312	4.387	0.2193	1.023
SUB-C-10%		6 prismatic	0.299	0.6142	3.685	0.1842	0.8599
SUB-C-20%	1:1:6	12 cylindrical	0.4387	0.7312	4.387	0.2925	1.023
SUB-C-30%		6 prismatic	0.3685	0.6142	3.685	0.2457	0.8599
SUB-C-40%	1:1:6	12 cylindrical	0.3656	0.7312	4.387	0.3656	1.023
SUB-C-50%		6 prismatic	0.3071	0.6142	3.685	0.3071	0.8599

and

Table 8. Materials quantities used in preparing sand mass replacement test specimens.

Replacing the Cement Mass
Code	Proportion	Test Specimen	Cement (kg)	Lime (kg)	Sand (kg)	Residue (kg)	Water (kg)
SUB-A-10%	1:1:6	12 cylindrical	0.7312	0.7312	3.948	0.4387	1.023
SUB-A-20%		6 prismatic	0.6142	0.6142	3.317	0.3685	0.8599
SUB-A-30%	1:1:6	12 cylindrical	0.7312	0.7312	3.510	0.8775	1.023
SUB-A-40%		6 prismatic	0.6142	0.6142	2.948	0.7371	0.8599
SUB-A-50%	1:1:6	12 cylindrical	0.7312	0.7312	3.071	1.316	1.023
SUB-A-10%		6 prismatic	0.6142	0.6142	2.579	1.105	0.8599
SUB-A-20%	1:1:6	12 cylindrical	0.7312	0.7312	2.632	1.755	1.023
SUB-A-30%		6 prismatic	0.6142	0.6142	2.211	1.474	1.023
SUB-A-40%	1:1:6	12 cylindrical	0.7312	0.7312	2.193	2.193	1.023
SUB-A-50%		6 prismatic	0.6142	0.6142	1.842	1.842	0.8599

present the quantities of materials used in the cement-replacement mixes, expressed as percentages of the total. Additionally, a dedicated table will also be provided for the mixes with sand replacement. It is important to note that the water/cement ratio is 1.5 in all proportions.

The specimens were prepared in accordance with the parameters specified in NBR 7215 [16]. The initial step involved separating the cylindrical molds, measuring 5 cm in diameter and 10 cm in height, and weighing out the materials: cement, lime, sand, scale, and water. To facilitate removal from the molds, a liquid release agent was applied to all molds before casting, thereby reducing their release after the material had hydrated. Additionally, the sand was oven-dried to eliminate moisture. Once the materials had been accurately weighed, mortar preparation began. First, cement, lime, and scale (except for the reference mix) were added to the tank, which had been pre-mixed with a spatula. The tank was then attached to the mortar mixer (Figure 3a), and water was added. The mixture was stirred at a low speed for 30 s. When the 30 s had elapsed, the mortar mixer beeped, signaling it was time to add the sand. After adding the sand, the mixture was stirred at high speed for an additional 30 s. The mortar was then allowed to rest for 90 s, during which any material adhering to the tank walls and spatula was removed. Subsequently, the equipment was operated at high speed for an additional 60 s, completing the preparation of the cement paste. Once the mortar was ready, specimen molding commenced immediately. Using a spatula, the mortar was distributed in four roughly equal layers, each subjected to 30 uniform blows with a tamper, and the final layer was leveled (Figure 3b). The specimens were then kept in the mold for 24 h; afterward, they were demolded, labeled, and wet-cured until testing.

In turn, prismatic specimens were produced in accordance with the parameters specified in NBR 13279 [17]. The first step involved separating the prismatic molds, which measured 4 cm in width, 4 cm in height, and 16 cm in length. A liquid release agent was applied to the molds, and the sand was oven-dried. Additionally, the process performed in the mortar machine was conducted in the same manner.

To begin shaping, the mold was placed on the compaction table, and the initial layer of mortar was evenly added to each compartment of the mold (Figure 4a). Then, 30 drops were applied using the compaction table. Subsequently, the second layer of mortar was applied, repeating the process with an additional 30 drops (Figure 4b). Once these steps were finished, the surface was smoothed with a metal spatula. The specimens remained in molds for 48 h, after which they were demolded, labeled, and wet-cured until the failure date.

2.4. Mechanical Tests

2.4.1. Mass Density Determination

Mass density was determined in accordance with NBR 13278 [18]. To do this, the cylindrical mold, already calibrated and weighed, was filled in three layers of roughly equal height, with 20 blows applied to each layer, for a total of 60 blows per mold. After each layer was compacted, the container was dropped three times from an approximate height of 3 cm to further compact the mortar. Excess material was then smoothed with a metal spatula. Finally, the assembly (mold and mortar) was weighed, and its mass was recorded to calculate the mass density using Equation (1).

$$D={\displaystyle \frac{Mc-Mv}{Vr}}1000$$

(1)

where D = mass density (g/cm³); Mc = mass of the container containing the test mortar (g); Mv = mass of the empty container (g); Vr = volume of the container (cm³).

2.4.2. Water Absorption by Capillarity

The capillary water absorption test complied with the recommendations of NBR 9779 [19]. Its purpose was to measure capillary rise in the produced mortar samples (CPs) to assess mortar permeability, as increased porosity correlates with a higher void ratio. After a 28-day wet-curing process, the specimens were placed in an oven at 105 ± 5 °C (Figure 5a) and weighed continuously until a stable weight was reached. The CPs were then removed from the oven, air-cooled to 23 ± 2 °C, and weighed to record their initial mass (g). Markings were made on the lower base of each CP at a height of 5 mm. For testing, the CPs were positioned on supports, and the test container was filled with water to maintain a water level 5 ± 1 mm above the bottom face, thereby avoiding contact with other surfaces. During the test, three specimens were used, and their mass was measured at 3, 6, 24, 48, and 72 h of water exposure (Figure 5b). After each measurement, the CPs were promptly returned to the test container.

After the final weighing, the specimens underwent diametrical compression in accordance with NBR 7222 [20], thereby allowing observation of water rising within them. Absorption was calculated using Equation (2).

$$C={\displaystyle \frac{A-B}{S}}$$

(2)

where C = is the water absorption by capillarity expressed in (g/cm²); A = is the saturated mass of the test specimen that remains with one of its faces in contact with the water during the specified period of time, expressed in grams (g); B = is the mass of the dry test specimen, as soon as it reaches a temperature of 23 ± 2 °C, in (g); S = is the cross-sectional area, expressed in square centimeters (cm²).

2.4.3. Axial Compression Strength

To verify the material’s mechanical behavior, a compression test was conducted. A 20-ton (t) Engetotus manual hydraulic press was used, meeting the requirements established by NBR 7215 [16]. The test was performed on 9 CPs: 3 at 7 days and six at 28 days.

2.4.4. Flexural Tensile Strength

Following the mechanical tests, flexural tensile strength testing commenced, in accordance with the guidelines outlined in NBR 13279 [17]. An Engetotus electric press with a maximum capacity of 100 tons was used for this purpose. Three 7-day-old and three 28-day-old CPs were tested. After flexural rupture, each half was subjected to compression testing to allow for further analysis of the material’s strength.

2.4.5. Determining the Consistency Index (Flow Table)

The flow table test evaluates the mortar’s spreadability (fluidity) and helps analyze its workability in accordance with NBR 13276 [21]. For each proportion tested, a test was conducted to assess the effect of adding or replacing the scale in the mortar.

After preparing the mortar on a clean table, the truncated cone mold was filled and positioned centrally on the table (Figure 6b) to determine the consistency index. Before this step, the flow table apparatus used in the test is shown in Figure 6a. The mold should be held firmly and filled in three successive layers of about equal height. Each layer was compacted with the socket using the specified number of blows: 15 for the first layer, 10 for the second, and 5 for the third. The blows were evenly distributed to ensure proper compaction. Once the mortar was compacted, the mold was removed, and the mortar before spreading is shown in Figure 6c. The crank was then immediately rotated, delivering 30 strokes in 30 s. Finally, after the spreading was complete, the spread is shown in Figure 6d, and three measurements were taken at different points to calculate the average spread diameter.

2.5. Microstructural Characterization

2.5.1. X-Ray Fluorescence (XRF)

X-ray fluorescence (XRF) was employed to determine the composition of the precursor materials and the resulting cementitious composites. This analysis was performed using a Rigaku Primini (Rigaku, Kyoto, Japan) instrument operating at 40 kV and 1.25 mA. The samples were prepared as pellets in a press at a force of 10 tons (tf). For scale, boric acid (H₃BO₃) was used as a binding agent to prepare the pellets.

2.5.2. X-Ray Diffraction (XRD)

To analyze the microstructure of the scale and mortar, samples from the residue and each proportion were separated and subjected to X-ray Diffraction (XRD) testing. The samples were dried in an oven at 105 ± 5 °C for 24 h to remove moisture, then ground in a porcelain mortar and pestle. After this process, a portion of the sample was set aside for XRD testing. The samples were analyzed using a Rigaku SmartLab X-ray diffractometer (Rigaku, Kyoto, Japan) at an operating voltage of 40 kV, a current of 30 mA, a 2θ range of 5–80°, a scanning speed of 2°/min, and a sampling pitch of 0.02°.

2.5.3. Fourier Transform Infrared Spectroscopy (FTIR)

Fourier transform infrared spectroscopy (FTIR) was employed to identify the chemical bonds and molecular interactions present in the samples. The analysis was performed with a Bruker Alpha II spectrometer (Bruker, Billerica, MA, USA) using the universal transmission module. Sample preparation involved forming pellets, which were then placed in the sample holder. To do this, the sample was mixed with potassium bromide (KBr) and compressed in a Specac press (Bruker, Massachusetts, USA) under a 10-tonne load (10 tf) to form the pellet. Spectra were recorded in the 4000–600 cm⁻¹ range in transmittance mode, with a resolution of 4 cm⁻¹, for a total of 64 measurements.

2.5.4. Scanning Electron Microscopy (SEM/EDS)

Microstructural analysis was conducted using Scanning Electron Microscopy (SEM) (Tescan, Brno, Czech Republic). Samples of each material—such as scale, reference mortar, and other proportions of addition and replacement—were prepared for this purpose. The samples were metallized before SEM analysis. A Tescan MIRA 3 microscope, equipped with an energy-dispersive X-ray analyzer (EDS) (Tescan, Brno, Czech Republic), an FEG filament, and secondary electron detectors, was used for imaging. Images were captured at an accelerating voltage of 5 kV. For chemical analysis, an energy-dispersive X-ray spectrometer (EDS) (Ultim Max, Oxford Instruments, Oxford, UK) was employed in high-vacuum chamber mode at an accelerating voltage of 15 kV.

2.6. Statistical Analysis

The experimental results were subjected to a statistical analysis to verify the significance of the differences observed between the reference mortar and the modified mixtures. A One-Way Analysis of Variance (ANOVA) was performed, followed by Tukey’s post hoc test, with a significance level set at 5% (p < 0.05). The error bars presented in the graphs represent the standard deviation of the mean values obtained from the tested specimens (n = 12 for cylindrical and n = 6 for prismatic specimens).

3. Results and Discussions

3.1. Physical Aspects

Initially, the mortar consistency index was evaluated in accordance with NBR 13276 [21] (flow table test) (Figure 7), indicating that partial replacement of cement with mill scale maintained or slightly increased the mortar’s workability. The highest spread rates were recorded at 30% and 50% replacements, with values of 255 mm and 257 mm, corresponding to increases of 4.51% and 5.33%, respectively, compared to the reference sample. Unlike cement, which is highly hygroscopic and reacts rapidly with water, mill scale is composed of iron oxides that do not react quickly with water. This slow reactivity of iron oxides helps keep the mortar’s fluidity [22].

As shown in Figure 8, the SUB A 30% formulation exhibited a localized deviation from the overall downward trend, with a higher standard deviation. This behavior is likely linked to the irregular and angular morphology of the mill scale fragments (as shown in SEM, Figure 2), which increases internal friction. At the 30% threshold, the balance between the non-absorbent nature of the scale—which increases the available free water—and the physical interlocking of the metallic particles may lead to greater experimental variability in the flow table test. At higher concentrations (40% and 50%), the physical packing and the high density of the residue (5.5 g/cm³) clearly override the effect of free water, resulting in a more consistent reduction in workability.

From Pereira, Verney, and Lenz [8] and the results obtained, the scale requires an increase in water content to maintain workability, as observed. However, all incorporation and proportion experiments in this study were conducted at a w/c ratio of 1.5.

Maintaining a constant w/c ratio of 1.5 directly influenced these rheological results. Since mill scale particles are predominantly metallic and non-absorbent, replacing natural sand with this residue theoretically increases the volume of free water available in the paste. However, the drastic reduction in the consistency index observed for the SUB-A group suggests that the physical properties of the residue—such as its high mass density (5.5 g/cm³), irregular morphology, and surface roughness—exerted a greater influence on internal friction than the presence of additional free water. This mechanism also explains why mechanical strength was maintained in the SUB-A series: effective physical interlocking and dense packing of the residue particles compensated for the high water content of the mortar matrix.

Figure 9 shows the density of the mortars, illustrating a gradual, consistent increase in mass density with increasing scale content. The initial density of the reference mortar was 2250 kg/m³, and substitutions of 10%, 20%, and 30% increased the density by 0.89%, 1.78%, and 2.22%, respectively. Beyond 40%, density decreased by 1.33% relative to the reference value; at 50%, it stabilized and returned to its initial value.

As shown in Figure 10, a steady and consistent increase in mass density was observed with increasing scale content. The initial density of the reference mortar was 2250 kg/m³, establishing a baseline for comparison with the REF (2250 kg/m³). The increases were approximately 2.22%, 4.44%, 5.33%, 8.44%, and 10.67%, respectively.

The behavior observed in sustainable mortars shows distinct effects on the compaction and internal structure of the cement matrix. According to Gagliotti [6], the scale has a density of 5.5 g/cm³, typical of dense metallic and mineral materials. This characteristic largely accounts for the increase in mortar mass density, particularly when sand is partially replaced by scale.

When replacing sand, which has a density of about 2.6–2.7 g/cm³, adding scale doubles the density. This can improve mechanical properties, such as increased strength and lower permeability, as the matrix becomes less porous. Omrane and Rabeni [23] suggest a direct link between density, porosity, and strength. In the specific case of the mortars studied here, the increase in bulk density (Figure 9, Figure 10 and Figure 11) is primarily a result of the high specific gravity of the mill scale (5.5 g/cm³) compared to the sand it replaces. While in conventional materials a higher density may directly indicate lower porosity, in these composites the density increase is a physical consequence of the heavier metallic inclusions. Therefore, the actual refinement of the pore network and the reduction in effective porosity are better evidenced by the capillary water absorption results (discussed in Section 3.2), rather than the bulk density alone.

The proportions with 10% and 50% replacement showed water absorption values that were lower than or very close to the REF at all time points analyzed. However, the formulations with 20%, 30%, and 40% replacement exhibited absorption values higher than the REF, suggesting increased porosity or greater heterogeneity of the cement matrix, possibly due to an imbalance in the cement-to-aggregate ratio that compromised proper compaction. This increased absorption could adversely affect durability by facilitating penetration by aggressive agents.

Based on Figure 12, all mortars with sand replacement had lower absorption values than REF, which showed the highest absorption throughout the test. Among the different proportions, the formulations with 20% and 30% replacement exhibit the lowest water absorption values.

Figure 12 suggests that replacing sand with scale may have reduced capillary porosity in the mortars, likely due to the filling effect of the fine residue particles, which promotes a denser matrix that is less permeable to moisture. Additionally, the 10%, 40%, and 50% formulations outperformed REF, although their absorption values were slightly higher than those previously reported. This behavior could be linked to excess residual material, which at high levels can promote the formation of unwanted pores, even if it does not exceed the REF values.

In summary, each incorporation led to distinct behaviors; however, some similarities emerged across the various compositions. In all proportions, water absorption increased over time, as expected in cementitious materials. Additionally, across all compositions, some proportions outperformed the reference, exhibiting lower water absorption values. It is essential to note that replacing cement, a crucial component in the formation of C–S–H and portlandite, can adversely affect microstructural development, particularly at intermediate levels. According to Mehta and Monteiro [24] and Taylor [25], the proper formation of hydration products (C–S–H) relies on the presence of reactive cement, and excessive replacements can dilute the matrix and increase permeability.

3.2. Mechanical Behavior

Figure 13 presents the compressive strength results expressed as mean values ± standard deviation. Statistical analysis (ANOVA) followed by Tukey’s post hoc test confirmed a significant reduction (p < 0.05) in compressive strength for all replacement levels compared to the reference (REF). As indicated by the distinct grouping letters (a–e) in Figure 13, the decrease was progressive and statistically significant with each mill-scale increment, reaching a minimum at 50% replacement. Portland cement primarily forms hydration products, such as C–S–H, portlandite, and ettringite, which together provide mechanical strength [24]. Therefore, it can be concluded that scale lacks the same binding capacity or hydration potential as Portland cement, thereby compromising the final mechanical strength.

Although a reduction in compressive strength was observed, this behavior is consistent with the literature on mortars incorporating other waste materials, such as expanded polystyrene (EPS) or recycled plastics, in which strength losses are accepted in exchange for environmental benefits [10,26]. Despite the reduction, the mortars with intermediate cement replacement levels still exhibit sufficient strength for non-structural applications, such as masonry coating or partition blocks.

For mortars with partial sand replacement (Figure 14), the statistical analysis (ANOVA/Tukey) indicated no significant difference (p > 0.05) between the REF sample and the remaining formulations at 28 days, as indicated by the shared grouping letter ‘a’ in the chart. This statistical analysis supports the observation that replacing sand with mill scale, even at higher levels, preserves the mechanical integrity of the matrix. Although minor fluctuations in mean values were observed (such as the slight decrease at 30% followed by recovery at 50%), they fall within the margin of error (standard deviation), confirming that the mill scale acts effectively as a filler without disrupting matrix cohesion. Unlike cement replacement, sand does not affect hydration processes and primarily acts as a filler [24]. Therefore, replacing sand with an inert material, such as scale, does not cause such notable interference. Ozturk et al. [8] also examined the effects of using scale as a partial replacement for fine aggregate; in this study, they found that at levels up to 15%, both compressive and flexural tensile strengths increased. However, levels above 15% reduced mechanical properties due to the formation of irregular structures and agglomerates that impaired the compactness and adhesion of the cement paste.

Based on the axial compressive strength test, the scale’s performance varies with the method of incorporation. When scale is used as a partial cement substitute, mechanical strength gradually decreases, indicating that scale neither promotes agglomeration nor contributes to the hydration products required for matrix cohesion. However, when the residue is used as a partial sand substitute or even added directly, scale is technically feasible (depending on the proportion added), maintaining or even slightly enhancing compressive strength at moderate levels without disrupting chemical hardening processes [27].

The flexural tensile strength results for cement replacement (Figure 15) align with the axial compression findings. Statistical analysis (ANOVA/Tukey) confirmed that the reduction in tensile strength is significant (p < 0.05) for replacement levels above 10%, as shown by the distinct grouping letters (a–d) in Figure 15. This indicates that the loss of hydration products, such as C–S–H, directly impacts the matrix’s capacity to resist tensile stresses. Despite decreases in strength values, all proportions showed gains in strength between 7 and 28 days, indicating that hydration and strength development continue, albeit less efficiently due to the absence of the material responsible for the hydration reactions. Additionally, Figure 15 displays the flexural tensile strength of the reference mortar and the mortar with cement replaced by scale.

For the sand-replacement group (Figure 16), the ANOVA test followed by Tukey’s post hoc test showed no statistically significant loss of strength relative to the reference (p > 0.05) at all tested replacement levels (up to 50%), as indicated by the shared grouping letter ‘a’. In fact, formulations like SUB A 20% and SUB A 40% showed mean values numerically higher than the REF. However, these increases fall within the statistical margin of equivalence, confirming that mechanical performance is maintained. Replacing sand with scale did not directly impact the hydration products, allowing the matrix to develop correctly.

Furthermore, the flexural tensile strength also supports the results of the axial compression test, in which mortars with up to 50% sand replacement either maintained or exceeded the reference values. Therefore, the results show that using scale as a partial substitute for sand in mortars is feasible, effectively preserving mechanical properties throughout the tested range.

Subsequently, the compressive strength test of the prismatic CPs was performed (Figure 17 and Figure 18).

The results obtained from prismatic specimens (Figure 17 and Figure 18) provide further validation. Statistical testing on the SUB C group (Figure 17) confirms a significant and progressive reduction (p < 0.05) in strength corresponding to the cement dilution, as evidenced by the distinct grouping letters (a–f). Conversely, for the SUB A group (Figure 18), the analysis shows that compressive strengths for 10% and 20% replacements are statistically equivalent to the reference (letter ‘a’, p > 0.05). However, for levels from 30% to 50%, a significant decrease was observed (letter ‘b’), although these values remained within safe limits for non-structural applications. This supports the proposal’s feasibility from a sustainability perspective, without compromising the safety of the materials used in civil construction.

3.3. Chemical Aspects of the Mortars

In the context of mortars with partial replacement of cement by scale (

Table 9. Chemical composition of mortars with partial cement replacement by scale (wt%).

	CaO	SiO2	Al2O3	Fe2O3	MgO	P2O5	SO3	Cl	K2O	TiO2	ZrO2
REF	61.85	29.07	3.50	1.60	0.90	0.40	1.62	0.05	0.43	-	058
SUB C 10%	57.81	29.40	3.14	4.30	0.86	0.35	1.61	0.05	0.34	1.63	0.50
SUB C 20%	55.79	27.59	2.92	7.97	0.81	0.36	1.45	0.09	0.80	1.69	0.52
SUB C 30%	53.21	29.14	2.77	11.19	0.67	0.33	1.29	0.07	0.77	-	0.54
SUB C 40%	50.31	29.24	2.64	12.91	0.60	0.35	1.19	0.08	0.57	1.58	0.54
SUB C 50%	49.41	30.67	2.55	14.51	0.67	0.26	1.09	0.09	0.21	-	0.52

), the results showed changes in the principal oxides, such as CaO, Fe₂O₃, and SiO₂. As the replacement percentage increased, the Fe₂O₃ content rose significantly, from 1.60% to 14.51%, reflecting the high iron oxide content in the scale. Conversely, the CaO content steadily decreased from 61.85% in the reference sample to 49.41%. Additionally, there was a slight increase in SiO₂, likely due to the presence of silicate phases in the scale.

Table 10. Chemical composition of mortars with partial sand replacement by scale (wt%).

	CaO	SiO2	Al2O3	Fe2O3	MgO	P2O5	SO3	Cl	K2O	MnO	ZrO2
REF	61.85	29.07	3.50	1.60	0.90	0.40	1.62	0.05	0.43	-	0.58
SUB A 10%	50.39	25.39	3.10	16.97	0.88	0.36	1.51	0.08	0.95	-	0.35
SUB A 20%	44.91	22.94	3.09	25.76	0.85	0.34	1.39	0.04	0.37	-	0.29
SUB A 30%	39.58	20.52	3.08	33.47	0.94	0.32	1.34	0.03	0.30	0.21	0.19
SUB A 40%	35.55	18.71	2.86	39.77	0.82	0.32	1.23	0.03	0.28	0.26	0.15
SUB A 50%	32.61	16.05	2.56	45.72	0.87	0.26	1.15	0.05	0.32	0.27	0.12

shows the results for mortars with partial sand replacement by scale, in which a similar trend to that observed with the addition of the residue can be observed, although much more expressive, with a gradual increase in relation to Fe₂O₃ and a reduction in CaO. Another relevant point is the behavior of SiO₂, which shows a gradual, though more discreet, reduction, from 29.07% to 16.05%.

3.4. Structural Characteristics of the Sustainable Mortars

Figure 19 displays the diffractograms of the reference mortars and those with partial cement replacement by scale after 28 days of curing. The crystalline phases identified were identical to those previously reported and exhibited approximately the same scattering angles.

The intensity of the Portlandite (P) peaks gradually decreases as cement scale replacement increases, particularly at 40% and 50% replacement levels. This decline indicates reduced formation of hydration products because less cement is available for reaction. Similarly, the alite (C₃S) peak shows a decreasing intensity with increasing scale content, supporting the idea of reduced reactive potential in the cementitious system.

Figure 20 presents the diffractograms of reference mortars and formulations with varying levels of partial replacement of sand by scale after 28 days of curing. Overall, it is evident that replacing sand with scale yields diffractograms that are similar to those obtained with other substitutions, showing the same crystalline phases: portlandite (P), alite (C₃S), calcite (C), and quartz (Q). The peaks remain in consistent positions, with only the intensities of the quartz (Q) peaks changing with the scale proportion used.

Figure 21 presents the spectral results for mortars with partial cement replacement at different scales. The same absorption bands observed in the previous spectrum are visible, with slight differences in peak intensities. The peaks at 1055 cm⁻¹ and 773 cm⁻¹, characteristic of the Si–O bond, decrease in strength as the cement replacement percentage increases. This may indicate a reduction in silicate product formation, which affects the mechanical performance of the mortars. This is evident in the mechanical strength results, which is discussed later [9,28,29].

Finally, Figure 22 presents the results for the partial replacement of sand with scale, showing behavior similar to that of cement replacement, with a noticeable reduction in the intensities of the bands at 1055 cm⁻¹ and 773 cm⁻¹, both of which are associated with silica in the system. This decrease is due to the lower content of silicate compounds in the mixture, specifically sand, which can directly influence the microstructure and, consequently, the physical and mechanical properties of the mortars [12,30].

3.5. Microstructural Properties of the Mortars

Figure 23, Figure 24, Figure 25, Figure 26, Figure 27, Figure 28, Figure 29, Figure 30, Figure 31 and Figure 32 show SEM images of sustainable mortars. When cement is partially replaced with scale, the formation of portlandite, ettringite, and C–S–H remains consistent across all tested proportions. The appearance of portlandite, a compound with low mechanical strength, is notable. When sand is partially replaced by scale, portlandite, ettringite, and C–S–H are also observed in all proportions.

Figure 23 (10% Cement Replacement). The micrograph of the mortar with 10% cement replacement demonstrates that the presence of the residue did not inhibit the formation of essential hydrated phases. It is possible to clearly identify the formation of Calcium Silicate Hydrate (C–S–H), along with acicular Ettringite (AFt) crystals and Portlandite (CH) plates. The microstructure suggests that, at this replacement level, the matrix retains reasonable cohesion, as cement dilution is not yet severe enough to prevent the formation of connections among hydration products. However, a reduction in mechanical strength becomes statistically significant.

Figure 24 (20% Cement Replacement). In this image, the persistence of hydration products (C–S–H, CH, and AFt) distributed throughout the matrix is observed. However, the data indicate that intermediate cement replacements (such as 20%) tend to increase capillary porosity and water absorption. The micrograph shows a matrix in which the amount of binder paste decreases proportionally as the inert scale is introduced. The visible presence of Portlandite (CH) is critical, as it is a low-mechanical-strength, high-solubility crystal that contributes to the observed performance reduction relative to the reference.

Figure 25 (30% Cement Replacement). The micrograph for 30% replacement indicates that cement mineral phases still form, but matrix heterogeneity increases. The image highlights C–S–H clusters and Ettringite (AFt) crystals filling spaces. However, since the scale lacks pozzolanic activity (it does not react with Portlandite to form additional C–S–H), it acts merely as an inert filler. The increased proportion of residue “dilutes” the binder, resulting in a less-dense, mechanically weaker microstructure, which aligns with the progressive drop in compressive strength reported in the study.

Figure 26 (40% Cement Replacement). At 40% replacement, the microstructure exhibits a pronounced presence of inert particles within the hydration products. Although C–S–H and Ettringite are still identified, the continuity of the binder phase is interrupted by the high dosage of non-reactive material. Chemical analysis (discussed in the text via XRF and EDS) corroborates that there is a significant reduction in Calcium Oxide (CaO) and an increase in Iron Oxide (Fe₂O₃) in this phase. The image illustrates the “dilution effect”: there is less “glue” (hydrated cement) to bind the particles, compromising structural integrity.

Figure 27 (50% Cement Replacement). The image corresponding to the maximum replacement (50%) best illustrates the inert nature of the residue. It is possible to clearly visualize scale particles, identified as Iron Oxides (FeO), adhered to the matrix. The sharp boundary between the FeO and the cement paste, without a diffuse reaction zone, confirms the absence of pozzolanic activity. Although the scale is physically incorporated, the drastic reduction in cement (the reactive material) results in a C–S–H-deficient microstructure, which explains why this proportion exhibited the poorest mechanical performance in the study.

In Figure 23 and Figure 27, scale particles adhered to the mortar, marked by iron oxides (FeO). This suggests that the scale acted as an inert material, with no significant pozzolanic activity, but that it may have improved the density and possibly the durability of the mortars [12]. In conclusion, incorporating various types of scale into the essentially inert cementitious matrix can serve as a barrier, thereby promoting pore and microcrack formation.

Furthermore, the spectra obtained indicate that iron (Fe) is more prominently distributed in formulations with higher scale percentages. On the other hand, the silicon (Si) content associated with C–S–H remains relatively constant with addition, with slight reductions when replacements are incorporated. This behavior supports the hypothesis that, in this context, scale does not exert a significant pozzolanic action, serving only as a filler [31,32].

Figure 28 (10% Sand Replacement). The SEM micrograph of the mortar with 10% sand replacement reveals a cohesive microstructural integrity, characterized by the consistent formation of essential hydration products such as Calcium Silicate Hydrate (C–S–H), Portlandite (CH), and Ettringite (AFt). The imagery confirms that at this substitution level, the incorporation of mill scale does not inhibit the hydraulic reactivity of the cement binder, enabling the development of a dense matrix comparable to that of standard formulations. The observed distribution of these phases suggests that the residue acts as a compatible, inert aggregate, maintaining the necessary microstructural cohesion to account for the observed mechanical performance.

Figure 29 (20% Sand Replacement). At the 20% replacement level, the microstructural analysis indicates a denser cementitious matrix, attributed to the higher specific gravity of the mill-scale particles, which promotes tighter solid packing. The micrograph identifies the coexistence of mature hydration phases, including C–S–H microlayers and Calcium Carbonate (Calcite/CaCO₃), indicating that the inert filler integrates well without inducing deleterious cracking or void formation. This compact morphology is consistent with physical test results indicating reduced capillary porosity, suggesting that the scale effectively occupies void spaces and enhances the matrix’s impermeability.

Figure 30 (30% Sand Replacement). Figure 30 highlights the effectiveness of the Interfacial Transition Zone (ITZ) between the mill scale and the cement paste, demonstrating a “tight physical interface” with no significant detachment voids. The rough surface texture of the scale particles facilitates robust mechanical interlocking with the surrounding C–S–H and Ettringite (AFt) network. This physical adhesion mechanism explains the retention of compressive and tensile strengths observed in the mechanical assays, as the scale functions as a high-stiffness inclusion that, when properly encapsulated, reinforces the composite structure rather than acting as a defect.

Figure 31 (40% Sand Replacement). At 40% substitution, the micrograph clearly delineates the distinct boundary between the inert phase and the binder, accentuating scale particles rich in Iron Oxides (FeO) embedded within the C–S–H matrix. This observation confirms the non-pozzolanic nature of the residue; it does not chemically react to form new binding phases but instead serves as a dense, impermeable physical barrier. The presence of these laminar, metallic-mineral particles disrupts the continuity of the capillary pore network, thereby increasing tortuosity and impeding fluid transport, which directly correlates with enhanced durability in water absorption.

Figure 32 (50% Sand Replacement). The analysis of the mortar with 50% replacement reveals a microstructure dominated by a high volume of dense mill-scale particles (FeO) that appear as compact blocks. Despite the substantial load of inert material, the image demonstrates that the cement paste maintains adequate coverage of the iron particles, with C–S–H and Ettringite (AFt) detectable at the interfaces. While the high fines content may present rheological challenges, the micrograph indicates that the material’s high density (5.5 g/cm³) yields a low-permeability composite, in which the residue acts as a high-density filler that physically blocks the ingress of aggressive agents.

The SEM/EDS data for mortars with cement replacement (Figure 32, Figure 33, Figure 34, Figure 35, Figure 36, Figure 37, Figure 38 and Figure 39) show that the percentages of calcium (Ca), oxygen (O), carbon (C), and silicon (Si) display expected patterns that repeat across different proportions. Regarding sand replacement, due to operational limitations, EDS was obtained only for 10%, 20%, and 40% sand replacement. These results indicated a significant increase in oxygen (O) and iron (Fe) concentrations, since the sand was replaced in substantial quantities [33,34,35].

Figure 33 (EDS of 10% Cement Replacement). The EDS spectrum for the 10% cement replacement exhibits a chemical profile dominated by Calcium (Ca, 33.0%) and Oxygen (O, 45.8%), which is characteristic of a standard cementitious matrix rich in Calcium Silicate Hydrate (C–S–H) and Portlandite1. The Iron (Fe) content remains low at 0.6%, indicating that at this low replacement level, the mill scale is well dispersed and does not significantly alter the elemental balance of the hydrated paste. The strong Calcium peak indicates that the binder’s chemistry is preserved.

Figure 34 (EDS of 20% Cement Replacement). At 20% replacement, the spectrum remains chemically similar to the 10% formulation and the reference, with Calcium (Ca) rising to 38.5% and Silicon (Si) at 3.0%. The Iron (Fe) content remains stable at 0.6%, suggesting that the scale particles are effectively encapsulated within the matrix. The persistence of high Calcium peaks corroborates the SEM findings that hydration products are still the primary constituents, despite the initial dilution of the cement.

Figure 35 (EDS of 30% Cement Replacement). The analysis of the 30% replacement sample indicates a slight shift in elemental composition, with Iron (Fe) increasing marginally to 0.8%. While the Calcium (Ca, 31.7%) and Oxygen (O, 47.0%) levels continue to indicate the presence of hydration phases, the slight rise in iron signals the increasing presence of the residue6. This gradual chemical change is supported by the mechanical data, in which binder dilution begins to affect the matrix’s overall cohesion.

Figure 36 (EDS of 40% Cement Replacement). In this spectrum, the Iron (Fe) content doubles to 1.2% compared to previous levels, while Calcium (Ca) remains dominant at 41.2%. This increase reflects the higher volume of mill scale replacing the cement. The detection of distinct iron signals provides evidence for the physical presence of iron oxide particles, which disrupt the homogeneity of the calcium–silicate binder and are consistent with the “dilution effect” discussed in the text regarding mechanical strength loss.

Figure 37 (EDS of 50% Cement Replacement). The spectrum for 50% replacement indicates a marked change in the chemical profile, with the Iron (Fe) content increasing to 11.7%. This sharp increase confirms the massive presence of iron oxides (Fe₂O₃, FeO) substituting the hydraulic binder. The spectral shape changes to reflect this high impurity load, indicating that the matrix is now predominantly composed of inert filler rather than reactive cement, which directly explains the observed minimum compressive strength.

Figure 38 (EDS of 10% Sand Replacement). Switching to the sand replacement route, the 10% sample shows a higher Iron (Fe) content (1.9%) than the equivalent cement replacement. This is expected because the residue is replacing silica (SiO₂) rather than the binder. The spectrum shows a balanced distribution of Calcium (Ca, 41.2%) from the cement paste and Silicon (Si, 5.4%) from the remaining sand, indicating a healthy matrix in which the scale acts as a supplementary fine aggregate.

Figure 39 (EDS of 20% Sand Replacement). The 20% sand replacement spectrum reveals a substantial increase in Iron (Fe) to 8.9%. The inset mapping is particularly revealing, showing a distinct iron-rich region (marked as FeO) adjacent to the calcium-rich paste. This confirms the formation of a tight physical interface without chemical diffusion, validating the “inert filler” mechanism. The Calcium (Ca) level (34.2%) indicates that the cement paste remains chemically intact around the scale particles, thereby maintaining its binding capacity.

Figure 40 (EDS of 40% Sand Replacement). The final spectrum for 40% sand replacement shows a large Iron (Fe) peak at 31.0%, consistent with the significant substitution of natural sand by metallic waste. The Oxygen (O) level (32.4%), combined with the high iron content, confirms the predominance of iron oxides (Fe₂O₃, Fe₃O₄). Crucially, despite this high iron load, the Calcium (Ca) signal (15.0%) indicates the continued presence of C–S–H gel binding the aggregates. This supports the conclusion that sand replacement is the technically superior approach, as it permits high waste incorporation without chemically depleting the binder.

Energy-dispersive spectroscopy (EDS) of sand showed a steady increase in oxygen (O) and iron (Fe) levels in samples with 10%, 20%, and 40% scale replacement. This reflects the increasing incorporation of iron oxides (Fe₂O₃, Fe₂O₄, and Fe₂O₅), which are the primary components of this steelmaking process, as widely documented in the literature. Several authors describe the scale as predominantly ferric and high in iron, which explains the Fe levels of 1.9%, 8.9%, and 31.0% at the respective replacement amounts. These findings confirm theoretical predictions and support the observed link between scale content and iron enrichment in the cementitious matrix, since these particles remain as inert oxides and do not significantly participate in hydration reactions [36,37].

From a microstructural perspective, incorporating scale tends to modify the interfacial transition zone (ITZ) and matrix density due to the higher specific gravity of the replacement material and its rough surface, which can promote paste-aggregate adhesion. Studies also report that moderate substitutions can favor porosity reduction and the production of denser, less permeable matrices, thereby enhancing durability mechanisms. In contrast, high levels may necessitate rheological and consolidation adjustments, potentially resulting in a loss of mechanical performance. Thus, the EDS results are consistent with the behavior described in the literature, reinforcing that scale acts primarily as a semi-inert mineral phase, altering the physical and microstructural properties of the matrix without modifying the essential cement hydration mechanisms [9,33].

Thus, the results suggest that the technical feasibility of scaling depends on its use as a partial replacement for fine aggregate, particularly in the 10–40% range, where microstructural improvements yield better performance or, at a minimum, results comparable to those of reference [26]. At the same time, the cement replacement route should be avoided for structural applications or for those requiring higher mechanical stresses, unless combined with active additives that mitigate reactivity loss, such as fine pozzolans or silica fume. From a sustainability standpoint, sand replacement is the most effective option among the evaluated routes due to its enhanced performance, reduced natural resource extraction, and alignment with circular economy principles [10,38,39].

A closer examination of the EDS elemental mapping reveals a distinct interaction at the Interfacial Transition Zone (ITZ). The sharp contrast between the iron-rich regions (red areas), representing mill-scale particles, and the calcium-rich regions (pink areas), corresponding to the hydrated cement paste, confirms the predominantly inert nature of the waste [15]. The absence of a diffuse transition zone suggests limited chemical reaction between the scale and the cement matrix. However, the microstructure indicates a tight physical interface without significant detachment voids. This observation supports the mechanical behavior results: while the lack of chemical reactivity explains the strength reduction in cement replacement (where binding capacity is lost), the effective physical interlocking observed at the ITZ explains why sand replacement formulations maintained their mechanical performance [9,39].

The correlation between the microstructural analysis and the physical–mechanical tests provides a clear explanation for the divergent behaviors observed in the two substitution routes:

• The Mechanism of Strength Retention in Sand Replacement (SUB-A): As evidenced by the SEM micrographs (Figure 28, Figure 29, Figure 30, Figure 31 and Figure 32) and EDS mapping (Figure 38, Figure 39 and Figure 40), the mill scale particles exhibit a rough texture and a high degree of compactness within the matrix. Although XRD (Figure 20) and FTIR (Figure 22) confirm the inert chemical nature of the scale (with no new hydration products formed), the interface between the scale and the cement paste is tight and free of significant detachment voids. This effective physical interlocking at the Interfacial Transition Zone (ITZ) explains why the compressive and tensile strengths were maintained (Figure 14 and Figure 16) despite the replacement of natural sand. The scale acts as a high-stiffness inclusion that mechanically reinforces the matrix, provided it is well-encapsulated.
• The Mechanism of Permeability Reduction: The capillary absorption results (Figure 12), which showed a decrease in permeability for the SUB-A group, are directly explained by the EDS findings. The iron-rich scale particles are denser (5.5 g/cm³) and less porous than the natural sand they replaced. In the microstructural images, these particles appear as dense blocks that interrupt the continuity of capillary pores. Consequently, they act as impermeable physical barriers, increasing the tortuosity of the pore network and hindering water transport, both of which are critical to the enhanced durability potential of these composites.

4. Conclusions

The results demonstrated that the use of industrial lamination waste (lamination scale) in mortars is a technical and environmentally viable alternative, aligning with the principles of sustainable development and circular economy in civil construction. The technical feasibility, however, depends directly on the method used to incorporate waste. The study demonstrated that the scale primarily functions as an inert filler, exhibiting no significant pozzolanic activity. Microstructural (SEM, XRD, FTIR) and chemical (XRF, EDS) analyses confirmed this nature.

The main scientific findings were:

• Cement Replacement (SUB C): This approach compromised the performance of the mortar.

  ○

  Mechanical Properties: There was a progressive decrease in axial compressive strength and tensile strength at flexure as the replacement percentage increased. XRD and FTIR analyses revealed reduced formation of hydration products, as evidenced by decreases in the portlandite (P) and alite (C₃S) peaks and in Si–O bonding bands. Chemically, a reduction in CaO content and an increase in Fe₂O₃ content were observed.

  ○

  Physical Properties: Although workability (consistency index) was maintained or slightly increased, water absorption by capillary increased in intermediate proportions (20%, 30% and 40%) compared to references.

  ○

  Conclusion (SUB C): The cement replacement route should be avoided for applications that require mechanical performance.

• Sand Substitution (SUB A): This approach offered the best balance for hardened state properties but presented a clear trade-off regarding fresh state performance.

  ○

  Mechanical and Physical Balance: Unlike cement replacement, substituting sand did not compromise the hydration process, maintaining compressive and flexural strength levels comparable to the reference up to 40% replacement. Furthermore, due to the high density of mill scale (5.5 g/cm³), the matrix became denser and less permeable, with 20% and 30% proportions exhibiting the best durability indicators (i.e., the lowest capillary absorption).

  ○

  Practical Limitations: However, a significant reduction in workability (consistency index) was observed as the scale content increased. This drastic drop represents a practical challenge for on-site applications.

Final Considerations: Consequently, the use of mill scale as a fine aggregate is technically feasible and environmentally beneficial, particularly within the 10–40% range. However, this recommendation is contingent on effective management of rheological properties. For high replacement levels, the use of water-reducing admixtures (plasticizers) is strongly recommended to recover workability without increasing the water-to-binder ratio, thereby preserving the observed mechanical gains.

Study Limitations and Future Research: It is important to note that this study focused on the physical–mechanical feasibility over 28 days. Although the reduced capillary absorption observed suggests potential for high durability (by limiting the ingress of water and oxygen), specific long-term behaviors—such as the risk of iron oxide leaching (rust staining), volume stability under thermal cycles, and resistance to carbonation—were not evaluated. Therefore, future research must conduct accelerated aging and durability tests to fully validate the service life of these sustainable composites before their widespread structural application.