1. Introduction
The enormous amount of waste generated worldwide challenges global sustainability. Turning waste into a new raw material is one key way to achieve integrated ecological productivity in all economic sectors. This should avoid the negative impacts of waste disposal at the local level, such as landscape deterioration, and water and air pollution. These are two crucial and current demanding tasks for mankind. The construction sector could have a key role in incorporating waste in cement-based materials, such as concrete or mortar [
1].
This industry is constantly evolving and searching for more efficient technologies that allow optimizing natural resources. Self-compacting concrete (SCC) has been one of the most important advances in construction materials, and its use has become widespread in recent years in civil works and buildings. These fluid mixes can be placed and compacted under their own weight, without vibration, segregation, blockage of coarse aggregate, bleeding, or paste exudation [
2]. The coarse aggregate content in SCC is lower than that of conventional concrete, so part of the performance of SCC is defined by the behavior of the SCM phase [
3]. The use of cement to improve fluidity increases the cost and CO
2 emissions of SCC, so the use of mineral additions, such as fly ash, blast furnace slag, or filler, modifies the properties of the paste and improves the fluidity of the mix without increasing its cost and environmental impact [
4]. Hence, the SCM phase is characterized by a high amount of powder content. Previous evaluations of the SCM to analyze the characteristics of SCC have been carried out by different authors [
5,
6]. SCM can also be used for rehabilitating and repairing reinforced concrete structures [
7]. The incorporation to SCM of different raw materials such as dolomite powder [
8], metakaolin and zeolite [
9], pumice [
10], or recycled waste materials, such as treated marine sediments [
11], calcined foundry sand from the metal casting industry [
12], fly ash [
13,
14], silica fume [
15,
16], and ground clay bricks [
17], has been proven to be viable.
Globally, among the industrial by-products generated in the world with a powdery size that can be incorporated as additions in SCM production, fly ash (FA) is commonly used. After pulverizing coal is burned to produce electricity in coal-power plants, different coal combustion by-products are generated [
18]. These by-products amount to approximately 780 million metric tonnes in the world per year [
19]. FA is the one that accounts for the greatest quantity of the total, at 68% [
20]. In Spain, the annual reports provided by Spanish electrical companies declare that around 60% of the FA is meant for the cement industry, as a supplementary cementitious material in concrete and in road construction [
18]. FA must conform to the standards EN 450-1: 2013, EN 450-2: 2006, and EN 14227-4 to be used as cementitious materials in concrete. The part of this by-product that does not comply with the fineness criteria (UNE-EN 933-10: 2010), which establishes a maximum of 40% retained (in mass) in the 0.045 mm sieve, is also known as non-confirming fly ash (NcFA). In Spain, this is landfilled and ranges between 30 and 40% of the FA generated in the last decade [
18]. This waste disposed at landfills can cause environmental issues [
21,
22]. Cheerarot & Jaturapitakkul [
23] studied FA disposed of for 6, 12, and 24 months in landfills as pozzolanic material in cement replacement for mortars, in comparison with FA grinded with a finer size particle distribution. Fineness was found to be an important factor for compressive strength. Torres-Gómez et al. [
18] studied the replacement of siliceous filler with NcFA in mortars, finding that there was an improvement in compressive strength. The studies on the incorporation of NcFA in SCC carried out by Esquinas et al. [
24,
25] have proved its feasibility in terms of durability and mechanical properties.
Another type of industrial by-product that can be incorporated in cement-based materials results from heating and drying the aggregates in hot-mix asphalt plants. This waste, hereinafter called recovery filler (RF), drops from the rotating drum along with combustion gases and is collected by means of baghouse filters to avoid being vented into the atmosphere. Lin et al. [
26] studied the application of this waste; despite the fact that it comes from natural aggregates and as such, its properties are expected to be similar, they commented that some authors found diverging results in which the addition of RF in asphalt mixes led to the decline of some properties. Its reincorporation in the manufacturing hot-mix asphalt process is limited to 3–4% [
27]. In 2017, in Europe, the production of hot and warm-mix asphalt was near 300 million tonnes, whereas in Spain it was 15.2. The generation of RF is estimated at 4% by weight of the total hot-mix asphalt production [
27,
28]. Martin et al. [
28] studied the use of RF in SCC finding no signs of segregation or exudation, as well as satisfactory results regarding resistance to reinforcement corrosion by carbonation and compressive strength regarding the technical provisions and standards. Esquinas et al. [
27,
29] evaluated the behaviour of SCC with RF through mechanical properties and durability. Despite the fact that the mechanical properties of SCC with RF were slightly lower than the ones of a reference mix with silicious filler, drying shrinkage at early ages also presented lower values. Martinez-Echeverría et al. [
30] also found a loss of mechanical properties in SCC with RF relative to the reference one. However, Shahidan et al. [
31] found that compressive strength in SCC with RF presented greater values than SCC with natural aggregates.
On the other hand, the management of hazardous waste is one of the most complex current challenges, since incorrect management can have irreversible environmental impacts. World steel production in 2017 amounted to nearly 1.7 billion tonnes of crude steel, including carbon, stainless, and other alloys. The amount produced in electric furnaces (arc and induction) was above a quarter of the total. These figures represented the greatest numbers in the last decade [
32]. One of the by-products generated in this industry is electric arc furnace dust (EAFD), which is generated as a result of the vaporization of molten iron with nonferrous metals, CO bursting bubbles, and the ejection and dragging of particles from the metal bath, slag, and other materials in the oven [
33]. Its production ranges between 10 and 30 kg per tonne of steel for scrap smelting and direct reduced iron [
34]. It is composed of heavy metals with leaching potential, which leads to negative impacts, such as Pb, Zn, Cd, Cr, and Ni [
35]; because of this, it has been classified as hazardous waste in the European Waste Catalogue [
36]. Therefore, its management implies a very high cost for steel mills since it cannot be deposited in landfills without prior treatment. The Spanish steelmaking industry generates an amount of 115,000 tonnes of EAFD per year [
37]. Its particle size distribution matches the requirements for its use as filler. Different researchers have studied the influence of EAFD on mortars in terms of mechanical properties and leaching behaviour in the solid state [
38,
39,
40,
41]. In these studies, the heavy metals Zn, Se, Cd, Mo, and Pb, and the chloride anion were identified as the elements of EAFD that exceeded the limits to not be classified as hazardous waste. Additionally, it is said that the formation of the phase CaZn
2(OH)
6·2H
2O in mortars with EAFD was found, and it inhibited hardening during the early stages of curing [
40]. Based on this issue, Massarwe et al. [
42] confirmed their hypothesis of the addition of EAFD as a set retarder in concrete, and Magalhães et al. [
43] successfully studied the possibility of applying a pre-treatment with a NaOH solution to enhance strength gain. Santamaría et al. [
44] studied the use of another type of by-product in steelmaking, called electrical arc furnace slag, which presents a particle size distribution similar to a regular fine aggregate, in SCM. In this study, the gain of compressive strength over time was postponed when this by-product was used. Other authors have studied the solidification/stabilization of EAFD with different matrix compositions. Salihoglu & Pinarli [
45] studied the use of paste samples based on cement, lime, or both for EAFD S/S. Samples with 35% of cement and lime and 30% of EAFD presented optimum composition for minimizing leaching test results. Fernández-Pereira et al. [
46] studied the EAFD S/S with geopolymers based on low calcium fly ash. Fares et al. [
47] compared the usage of EAFD and different conventional binders, such as silica fume and FA, with Portland cement in mortars. One of the features of SCMs that make it adequate for the encapsulation of potential harmful components is the higher density of the cement matrix than in regular mortars [
40]. This, along with the need for filler material in SCMs, makes EAFD and the other industrial by-products excellent candidates for this purpose.
This study analyses the influence of incorporating different industrial by-products with a powdery size, such as EAFD, RF, FA, and NCFA, as fillers on the performance of self-compacting mortars (SCMs). To the extent of the authors’ knowledge, none of the existing studies focused on the simultaneous use of FA, NcFA, or RF and EAFD as fillers in SCMs. This study addresses the topic regarding the use of ternary blends combining the use of EAFD and other industrial by-products, along with cement, to produce sustainable SCMs; this appears to be highly suitable due to its appropriateness as a filler and capability of limiting the release of potential pollutant elements. For this purpose, this investigation presents the characterization of the materials used, the performance of sustainable SCMs, and the leaching behaviour in the monolithic state. Therefore, this study contributes to the circular economy, since it avoids landfilling through its incorporation in SCMs and, consequently, reduces the consumption of raw materials.
2. Materials and Methods
2.1. Materials
For SCM production, two fractions of aggregates of siliceous nature were used: fine sand 0/1 (NS 0–1) and coarse sand 0/4 (NS 0–4).
Figure 1 shows the particle sieve distribution of both sands. NS 0–1 and NS 0–4 presented water absorption after 24 h of 0.7% and 1.1% and specific gravity of 2.58 and 2.55, respectively. The cement used was CEM I-42.5 R (CEM) and it complied with NP EN 197-1. It had a specific gravity of 3.14. The high-performance/strong water reducer superplasticiser used, SikaPlast 898 (SP), is based on a combination of modified polycarboxylates in an aqueous solution that works by electrostatic and steric repulsion.
The different by-products used as fillers were conforming fly ash (FA), non-conforming fly ash (NcFA), recovery filler (RF), and electric arc furnace dust (EAFD). The FA was supplied by SECIL Company (Lisbon, Portugal) and complied with the NP EN 450-1 and NP EN 450-2 standards; on the contrary, the other FA used did not comply with the fineness standard UNE-EN 933-10: 2010, therefore named NcFA. It was obtained from the Puente Nuevo coal-fired power station (Córdoba, Spain) of Viesgo electric company. The origin of the RF was from waste powder from asphalt mixture sands supplied by PAMASA (Málaga, Spain). The hazardous waste, EAFD, came from a steelwork located in Zumárraga (Guipúzcoa, Spain) that uses an electric arc furnace (EAF) for steel production.
2.2. Test Methods for Materials Characterization
The chemical composition of CEM and EAFD was found by X-ray fluorescence (XRF) using 4 kW of power and S4PIONEER, BRUKER equipment, whereas for FA, NcFA, and RF it was done by analysis of the dispersive energy of X-rays (EDAX).
The mineralogical composition of the raw materials was determined by X-ray diffraction (XRD) using a Bruker D8 Discover A 25 with Cu Kα radiation (λ = 1.54050 A; tube voltage: 40 kV; Tube current 30 mA). For EAFD, goniometric scanning was used from 10° to 70° (2θ°) at a speed of 0.00625°/min, whereas for the rest of the materials, the speed was 0.05°/min. The identification of the main minerals was done by comparison with the JCPDS Powder Diffraction File database [
48].
The specific surface area of the CEM, FA, NcFA, RF, and EAFD samples was analysed by the Brunauer-Emmett-Teller method (BET), using Micromeritics ASAP 2010 equipment, Norcross, GA, USA. The single-point pore volume was determined from the amount adsorbed at a relative pressure of ∼0.99. The real particle density was estimated according to UNE 80103:2013. The particle size distribution of CEM, FA, NcFA, RF, and EAFD was measured using a Mastersizer S 2000 device (Malvern Instruments, Malvern, UK) with a previous ultrasonic homogenization for sample particle dispersions. Furthermore, the adsorption-desorption isotherms of nitrogen were determined to study the pore size distribution in CEM, FA, NcFA, RF, and EAFD by means of the DFT (density function theory).
Thermogravimetric and differential thermal analysis (TGA-DTA) was performed in a Setaram Setsys Evolution 16/18 apparatus under nitrogen at a heating rate of 5 °C/min in FA, NcFA, RF, and EAFD samples.
The standard UNE-EN 12457-4:2003 [
49] is a basic characterization procedure which was conducted on 0.090 kg of each filler (FA, RF, NcFA, and EAFD), by means of a single-step batch leaching test, in which the solution was shaken for 24 ± 0.5 h at an L/S ratio of 10 L/kg. After the contact phase, the samples were left to decant, then filtered and a subsample of 40 mL of eluate was collected for testing and analysed within 24 h, for several elements and/or anions using ICP-MS and ionic chromatography, respectively.
2.3. Mortar Mix Proportions
The composition of the mortar mixes (
Table 1) was designed following the Nepomuceno method [
5], separating powders (CEM, FA, RF, NcFA, and EAFD) and fine aggregates (NS-0/1 and NS-0/4). The self-compacting parameters used based on the absolute volume proportions were: powder materials and fine aggregates (V
p/V
s), which were kept constant at 0.70 for all the mixes; and water and powder materials (V
w/V
p), with a value of 0.83 and 0.78 for mixes with 70% and 40% of cement with respect to powder materials, respectively. The mass of the superplasticizer and mass of the powder materials ratio (Sp/%p) were adjusted for each mix. This increased as the EAFD ratio incorporation rose since the EAFD had finer particle size distribution and greater specific surface area than the powders replaced (RF, FA, and NcFA).
The self-compacting parameter regarding the fresh state of SCM was determined through the
Gm. It is obtained by means of the slump-flow test; a truncated cone with a height of 60 mm and diameters of 70 and 100 mm at the top and bottom, respectively. It was filled up and then it was immediately lifted to let the fresh mortar spread out on a plate with a smooth and dampened surface. The average of the two perpendicular diameters of the spread mortar was used in the formula for
Gm, provided by Nepomuceno et al. [
5], to obtain the
Gm parameter (Equation (1)). This value decreased when the EAFD content increased, even with the addition of superplasticiser. Above the amount used in mixes with EAFD incorporation (
Table 1), the phenomena of segregation and water exudation were found. Although the
Gm values in some mixes were lower than what is recommended [
5], none of the aforementioned phenomena occurred.
where
Dm is the mean of the diameters of the spread mortar, in mm, and
D0 stands for the initial diameter at the base of the cone, in mm.
The designation of the mortars depends on the composition by volume of the powders as shown in
Table 1.
To produce the SCM specimens, a standard mortar mixer was used. The constituents in solid state were homogenised 30 s after 80% of the water was poured in the mixer. Two minutes later, the remaining 20% of the water was added with the superplasticiser previously diluted. Then, the mixer was stopped for 1 min to ensure that all the constituents were mixed and to clean the mixer-blades. Finally, the mix went on for another 1 min and prismatic specimens, 40 × 40 × 160 mm3, were cast without agitation or mechanical compaction. Before 24 h, 48 h, and 72 h for mix specimens with no EAFD, with 10% EAFD, and with 20% EAFD, respectively, were demoulded, and afterwards stored in a climatic chamber at 50 ± 5% relative humidity and 20 ± 2 °C. This increasing time between manufacturing and demoulding as EAFD percentage rose was due to the retardation of hardening of cement matrixes in the presence of EAFD, as aforementioned in the introduction section.
2.4. Test Methods for SCM
The mineralogical composition of the SCM was determined using the equipment used for raw materials with 0.05°/min of speed. For this purpose, samples were collected after testing compressive strength, and then immersed in ethylic alcohol and stored at least 72 h to inhibit any setting reaction and to dry by solvent replacement [
6]. The XRD patterns of the crystalline mineral phases in SCM were compared with the JCPDS powder diffraction file database [
50].
The mechanical properties of the SCM in hardened state were studied at 28 and 91 days of curing. The flexural strength was obtained using a standard three-point-bending test with a 100 mm span, and compressive strength was measured on perpendicular edges of each of the two residual pieces obtained from the flexural test, following the standard UNE-EN 1015-11:2000 [
50]. The rate of loading for flexural and compressive strength were 0.04 kN/s and 0.4 kN/s, and their value corresponds to the average of the three tests.
The following physical properties related to the durability were studied: water absorption by immersion and capillarity, and dry bulk density; these tests were carried out after 91 days of curing in three specimens each. Water absorption by immersion was obtained by measuring the water absorbed after immersion until mass became constant, and its apparent volume was determined by subtracting the saturated mass and weighing the sample under water. To study the water absorption by capillary, the capillarity coefficient was obtained in accordance with UNE-EN 1015-18:2003 [
51]. Dry bulk density was yielded in the relationship between the dry mass of the specimen and its apparent volume, following the standard UNE-EN 1015-10:2000 [
52].
With the aim of analysing the capability of encapsulating heavy metals, with potential hazardous impacts to the environment, in SCM due to the presence of EAFD, the test conducted was the tank-leaching test according to XP X31-211:2012 [
53]. It determines the leachability of a solid waste material, in this case generated in a solidification process. Prismatic sample halves were introduced in a PET container with a liquid to solid ratio of 10 with de-ionized water being agitated for 24 h. After that, the pH and electrical conductivity were measured and the amount of elements and anions was determined using ICP-MS and ionic chromatography, respectively.