Mechanical and Thermal Properties of Geopolymer Foams (GFs) Doped with By-Products of the Secondary Aluminum Industry

The article deals with the investigation of geopolymer foams (GFs) synthesized using by-products coming from the (i) screening-, (iv) pyrolysis-, (iii) dust abatement- and (iv) fusion-processes of the secondary aluminum industry. Based on principles of the circular economy to produce new technological materials, the experimental study involves industrial by-products management through the recovery, chemical neutralization, and incorporation of these relatively hazardous waste into the GFs. The geopolymeric matrix, consisting of metakaolin (MK) and silica sand (SA) with a 1:1 wt.% ratio, and chopped carbon fibers (CFs, 1 wt.% MK), was doped with the addition of different aluminum-rich industrial by-products with a percentage from 1 to 10 wt.% MK. The gas (mainly hydrogen) produced during the chemical neutralization of the by-products represents the foaming agents trapped in the geopolymeric structure. Several experimental tests were carried out to characterize the mechanical (flexural, compressive, and Charpy impact strengths) and thermal properties (thermal conductivity, and diffusivity, and specific heat) of the GFs. Results identify GFs with good mechanical and thermal insulation properties, encouraging future researchers to find the best combination (for types and proportions) of the different by-products of the secondary aluminum industry to produce lightweight geopolymer foams. The reuse of these industrial by-products, which according to European Regulations cannot be disposed of in the landfill, also brings together environmental sustainability and safe management of hazardous material in workplaces addressed to the development of new materials.


Starting Materials
The inorganic two-component aluminosilicate binder (commercial name: Bausik LK), (České lupkové závody, a.s., Nové Strašecí, Czech Republic) [39] is a two-component aluminosilicate binder based on metakaolin (hereafter MK, part A), (commercial name: Mephisto L05), (grain size D 50 = 3 µm, D 90 = 10 µm) activated by an aqueous alkaline activator (part B). The mixing ratio of these two components was taken out according to the manufacturer requirements. In preparing the binder mixture based on the inorganic polymer, five parts by weight of part A and four parts of B (activator) are usually used. The silica sand (hereafter SA, ST 01/06), (Sklopísek Střeleč, a.s., Újezd pod Troskami, Czech Republic), (D 50 = 0.44 mm, D 90 = 0.63) [40] was used as aggregate. Chopped carbon fibers with an elastic module up to 230 GPa and tensile strength of 3500 MPa [41][42][43][44][45] were used as reinforcing materials. Table 1 shows the chemical composition of the raw materials used in this experiment to produce the geopolymer-based matrix. Various aluminum-rich by-products (Table 2) were used as additives to foam the geopolymers. The studies of the starting materials were conducted with specific analytical techniques to determine the chemical content subsequently indicated and for the planning of laboratory experiments. The chemical analyses of the by-products of the secondary aluminum industry were performed by ICP-MS with near-total multi-acids (hydrofluoric, nitric, and perchloric acids) digestion at Actlabs (Ancaster, ON, Canada). After the digestion and dehydration, only specific species of the sample were brought into solution using aqua regia and analyzed with ten duplicates and eight reference materials through Perkin Elmer Sciex ELAN ICP-MS.
The data processing enabled a quantitative assessment of the dangerous compounds in the aluminum processing slags, which are critical when reused [46,47]. The samples were classified under the normative requirements ( Figure 2) of the Decree of environmental assessments and authorizations n.31/VAA (30 April 2015) [48], which were used by the European industries to issue the integrated environmental authorization (AIA) (EU directive 2010/75 and Legislative Decree 152/2006) [49,50], on the environmental safety and pollution control. The normative requirements provide the classification of hazardous substances on the CE Reg. 1272/2008 [51] and limits and characteristics of danger (HP) on the CE Reg. 1375/2014 [52].   A macroscopic overview of the aluminum-rich by-products is given in Figure 3. The materials used as fillers into the geopolymers derive from the main processes of the secondary aluminum industry: (i) screening process, (ii) pyrolysis process, (iii) fusion process. FG and UBC acronyms are from coarse-grained domestic appliance scrapes and urban beverage cans, the primary materials used for recycling.
The powder X-ray analyses of the aluminum-rich by-products were determined with a Bruker D8 Advance diffractometer at CRI.ST (Centro di Servizi di CRIstallografia STrutturale, Florence, Italy), and a Philips X'Change PW1830 powder diffractometer at University of Urbino (Urbino, Italy). The grain size analyses ( Figure 4) were performed through a Laser beam particle analysis (Hydro 2000MU analyzer, University of Milano-Bicocca, Milan, Italy).

Experimental Procedure for the Geopolymer Synthesis
Several geopolymers were synthesized to investigate the influence of the aluminumrich by-products on several physical properties: flexural strength, compressive strength, Charpy impact strength, thermal conductivity, specific heat, and thermal diffusivity.
For this purpose, metakaolin (MK), (Al 2 O 3 40.1 wt.%; SiO 2 : 54.1 wt.%) have been used during the alkaline activation process as precursor materials, using a potassium hydroxide aqueous solution (A) (pH 11) [53,54]. In addition, chopped carbon fibers, which show evidence to increase the mechanical properties of the materials [55], are employed in the REF-2 geopolymer and in the geopolymer foams where aluminum waste materials represent additives for foaming.
The previously described aluminum-rich by-products would play the role of foaming agent, generating H 2 -enriched gas pockets inside the geopolymer structure and making the material more porous and therefore lighter. The foaming process regards the aluminum and alkaline aqueous solution interaction, where the potassium hydroxide reacts, forming tetra hydroxy aluminate (III) and hydrogen gas, and aluminum undergoes oxidation. The primary reaction involved is described by the Reaction (1): The experimental procedure reported in Figure 5 shows how the raw materials were mixed to prepare all the references and geopolymer foams. The metakaolin (MK) and alkaline activator (A) were mixed for about 5 min to obtain a homogenous mortar. Next, chopped carbon fibers (CFs) were added, mixing for 2 min. After that, silica sand (SA) was added and mixed for 3 min. Finally, each industrial byproducts (marked as V.FG, V.UBC, D.FG, D.UBC, C.FG, C.UBC, FF.FG, or FF.UBC) were mixed for 2 min in order to prepare different GFs (Table 3). Table 3. The ratio of the main components used to synthesize the geopolymer foams with respect to MK content. After the mixing, the geopolymer mortar was decanted into molds with the dimension of 30 × 30 × 150 mm (for three-point bending test and compression test), 19 × 20 × 60 mm (for Charpy impact test), and 100 × 100 × 100 mm (for thermal analysis). These samples were covered using a polypropylene film and cured at room temperature for about 24 h. After that time, the samples were pulled out of the molds, wrapped again using a polypropylene film, and kept at room temperature for 28 days before being analyzed (standard EN 12390-3:2019) [56].

By
Two types of reference samples were used. The first, labeled as REF- A name coding system was introduced to distinguish the geopolymers ( Table 4). The first part indicates the type of the added industrial by-product (e.g., V.FG), the second its percentage (1, 2, 3, 5, and 10 wt.%) referred to the metakaolin (MK) (e.g., V.FG-1).

Methods for the Mechanical Tests
The samples were cured for 28 days before being tested to characterize the mechanical properties of the GFs and the influence of the different by-products used as foaming agents. Figure 6 shows the three main laboratory instruments (at the Department of Material Science, University of Liberec, Liberec, Czech Republic) and techniques to carry out analyses for mechanical properties: (a) three-point bending test, (b) compressive strength test, (c) Charpy impact test.  [57]. Tests were carried out on six 30 × 30 × 150 mm specimens (Figure 6a) at room temperature with a crosshead speed of 6.0 mm/min and a span length of 100 mm. The flexural strength (σ f ) was calculated by the Equation (2): where: F max -the maximum applied load indicated by the machine (N); L-the span length (mm); b-the width of the sample (mm); h-the depth of the sample (mm). As for flexural strength determination, the compressive tests were performed employing the INSTRON (Model 4202) Testing Machine (standard EN 196-1:2016) [58]. The broken parts from the samples used in the bending test were used (Figure 6b). In this way, twelve samples with dimensions 30 × 30 × 30 mm were obtained for each composition.
The tests were conducted at room temperature with a 6.0 mm/min crosshead speed. The compressive strength (σ c ) was obtained by the Equation (3): where: A c -the cross-sectional area of the sample (mm 2 ). The impact tests were carried out using a PIT-C Series Pendulum Impact Testing Machine (standard EN ISO 148-1:2010) [59] with a pendulum capacity of 150 J, energy losses compensation of 0.23 J, and estimated absorbed energy of 150 J. Six samples with the dimensions 19 × 20 × 60 (mm) were tested (Figure 6c). The tests were performed at room temperature. The impact strength (σ i ) was calculated by the Equation (4): where: E-the absorbed energy indicated by the machine (J); V-the sample volume (mm 3 ).

Methods for the Thermal Measurements
The thermal analyses were conducted at the Faculty of Civil Engineering, Mechanics and Petrochemistry, Warsaw University of Technology, Płock, Poland. After 28 days of curing, six measurements for each specimen were performed using the Isomet 2114 device (standard ASTM D5334-08) [60], a microprocessor-controlled commercial instrument with interchangeable probes.
A known heat source produced a wave propagating radially into the specimen. The dissipation of electrical energy generates the heat flow through the probes in direct contact with the material, and a serial port (RS-232C protocol) [61] records the signal. Semiconductor sensors at specific points on the materials sampled the temperature change in function of time: the temperature rises linearly with the logarithm of time [62][63][64][65].
According to the 2nd law of thermodynamics, the thermal conductivity (λ) was determined by the Equation (5): where: Q-the amount of heat transferred, d-the distance between the two isotherms, A-the surface, and ∆T-the temperature gradient. The specific heat capacity (C p ) is the heat needed to increase the temperature of 1 g of a substance by 1 • C and is given by: {\displaystyle ∇T} (6) where: m-the mass. The thermal diffusivity (α) quantifies the heat transfer rate of the material from the hot side to the cold side, and it was computed by the Equation (7): where: ρ-the density of the geopolymer (obtained dividing the sample mass by volumestandard EN 1936:2006) [66].

Mechanical Properties
Mechanical properties are the most relevant parameters for evaluating geopolymer performances and understanding the applications [67,68]. The results of the three-point bending, compressive and Charpy impact strengths are shown in Table 4 The reactivity of the industrial by-products used as fillers and foaming agents during the geopolimerization can be mainly attributed to the chemical composition (aluminum content), mineralogy, and grain size [69,70]. These features influence the physical and mechanical characteristics of the geopolymers thanks to the porosity formed during the aluminum oxidation [71][72][73][74][75].
It is highlighted that by adding the aluminum-rich by-products and increasing their percentage, the flexural and tensile strengths of the geopolymers decrease (Table 4) due to the gas bubbles formed in their structure during the consolidation process. On the other hand, most of the impact strengths data mainly increase. Figure 7a,b illustrates the gas bubbles distribution of the geopolymer foam FF.UBC-3 that appear not homogeneous and characterized by different size holes. The areas of these bubbles were quantitatively estimated on the breaking section after the three-point bending tests by an open-source software analysis (ImageJ), applying a color threshold for the analysis. 13.2% of the total surface (900 mm 2 ) consists of bubbles that, of course, define the overall geopolymer structure and shape the surface along which the break occurs.

GFs with the Addition of the Aluminum-Rich By-Products of the Pyrolysis Processes
The aluminum contents of the pyrolysis by-products D.FG and D.UBC are 32,204 and 40,198 ppm, respectively ( Table 2). As shown in Figure 9, the mechanical strengths are better performed than the scraps of the screening processes. In this case, the impact strength of D.FG-1 is around four times higher than the reference sample REF-1 and two times more than REF-2. Moreover, also D.UBC-2 shows the same behavior with a σ i of 0.71 MPa. This increase in performance is directly related to the aluminum content and finer-grained and more homogeneous particles of this kind of by-products.

GFs with the Addition of the Aluminum-Rich By-Products of the Fusion Processes
The best mechanical performances for the geopolymers obtained with the addition of the by-products of the fusion processes ( Figure 11) are found in FF.UBC where compressive, flexural, and Charpy impact strengths are almost similar to the reference samples. In particular, FF.UBC-1 is the best GF in term of mechanical performance with σ f = 7.48 ± 0.22 MPa; σ c = 44.67 ± 0.31 MPa; σ i = 0.54 ± 0.02 MPa. We can conclude that FF.UBC slag, having the lowest aluminum content (6636 ppm) is the most suitable by-product to be trapped into the geopolymeric structure keeping unchanged the fundamental mechanical properties of the reference geopolymers.

Densities versus Thermal Conductivity, Diffusivity, and Specific Heat
The density (ρ) and the thermal conductivity (λ), diffusivity (α), and specific heat (C p ) of the obtained geopolymer foams are reported in Table 5. A clear relationship between the density and the represented thermal properties can be observed. Table 5. Summary of density (ρ, g/cc) and thermal properties (thermal conductivity, λ; specific heat, C p ; diffusivity, α) of the synthesized geopolymer foams, by adding (1, 2, 3, 5, 10 wt.% of MK) the by-products from the screening (V.FG and V.UBC), pyrolysis (D.FG and D.UBC), abatement dust (C.FG and C.UBC) and fusion (FF.FG and FF.UBC) processes.

Geopolymer
By-Products (wt.% of MK)  The linear regression of λ with ρ shows a R 2 of 0.7766 (Figure 12a), so the thermal conductivity depends on the density of the geopolymers. Moreover, also C p (Figure 12b) and α (Figure 12c) are strongly related to the density with R 2 of 0.5951 and 0.8193, respectively. For low densities, the porosity of the GFs increases, and consequently λ, Cp, and α significantly decrease. Definitively, the lower densities of these materials are a great advantage compared to the traditional building materials such as Portland cement. They are lightweight materials, and the thermal insulation properties are better performed. λ, C p and α decrease by adding the industrial by-products which act as foaming agents.
REF-1 and REF-2, with a density of 1.81 ± 0.06, and 2 ± 0.08 g/cc show a λ of 1.2981 ± 0.0606, and 1.4607 ± 0.0167 W/mK, a C p of 1.8518 ± 0.0855, and 1.9078 ± 0.0194 J/KgK, an α of 0.7056 ± 0.0295, and 0.7667 ± 0.0124 mm 2 /sec, respectively. The higher values in the REF-2 are due to the chopped carbon fibers (CFs), which improve the mechanical properties, but on the other hand, increase the thermal properties by around 5-10%.
The densities decrease because of the foaming agents and range from 0.95 ± 0.04 g/cc (C.UBC-5) up to 1.99 ± 0.08 g/cc (D.FG-1). The lowest thermal conductivity (Table 5) was measured with the industrial by-products C.FG and C.UBC from the dust abatement collectors (cyclons). The geopolymer foam C.FG-3 ( Figure 13a) recorded a thermal conductivity of 0.3306 ± 0.0069 W/mK and a density of 1.05 ± 0.08 g/cc. C.UBC-10 ( Figure 13b) has an even lower λ of 0.3265 ± 0.0150 W/mK, and a density of 1.08 ± 0.05 g/cc.

Classification of the GFs
The GFs were classified into six groups following the physical parameter of density versus compressive strength and thermal conductivity ( Figure 14) to highlight which material has the best thermal insulation and mechanical properties. Group A shows the lowest thermal conductivity values and the lowest densities from 0.95 to 1.16 g/cc. This population of data shows relatively low σ c ranging between 2.96 and 4.05 MPa. Group B has relatively higher densities than group A and, consequently, higher thermal conductivities. The compressive strengths are slightly higher, with an average value at around 5 MPa. Group C is characterized by σ c at around 10 MPa and λ that corresponds to 0.7 W/mK. The compressive strength of Group D range between 10 and 20 MPa, with thermal conductivity with an average value of 0.9 W/mK and a mean density of around 1.8 g/cc. Group E (density between 1.6 and 2 g/cc) is between 20 and 30 MPa for the compressive strength, with thermal conductivity of 1.1 W/mK. Finally, group F exhibits similar performance as the reference standard geopolymers (REF-1 and REF-2) concerning mechanical and thermal properties thanks to its higher density. The group F population shows a density between 1.8 and 2.0 g/cc, a mean λ of 1.3 W/mK, and mean σ c of around 42 MPa.

Conclusions
The present study deals with the mechanical (flexural, compressive, Charpy impact strengths) and thermal (thermal conductivity, specific heat, thermal diffusivity) properties of GFs obtained by adding aluminum-rich by-products of the secondary aluminum industry. According to the European Regulations, these industrial by-products cannot be disposed to landfills because they are classified as special hazardous wastes which can develop flammable gases and form explosive mixtures with air. The hazard mainly comes from hydrogen production due to metallic aluminum oxidation. Nevertheless, if the reaction producing hydrogen occurs when geopolymers are synthesized, the by-product themselves undergo a chemical neutralization, and the hydrogen-rich gas is used as foaming agents modifying the structure of standard geopolymers (REF-1, REF-2).
In particular, the work highlight that FF.UBC by-product coming from the fusion processes of the secondary aluminum industry is the most suitable material to improve the mechanical properties of geopolymers compared to REF-1 and REF-2, and it, therefore, is the appropriate raw material to foam lightweight geopolymers. In addition, significant decreases in thermal conductivity, specific heat, and thermal diffusivity, thus emphasizing good thermal insulation properties, are observed in the GFs doped with by-products C.FG and C.UBC from the dust abatement (cyclons) processes of the secondary aluminum industry.
The study unravels that using geopolymer foams as an alternative building material finds a compromise to balance the mechanical and thermal properties and guarantee the usability of the composite materials. For this reason, future studies will focus on mixing the three by-products (FF.UBC, C.FG, C.UBC), maintaining good mechanical performance for building material, and giving to GFs excellent thermal insulation properties those characterizing groups A-D of geopolymer foams with thermal conductivity ≤ 0.9 W/mK. Accordingly, the final remarks are addressed to (i) recovery and process several byproducts of the secondary aluminum industry, most of them not suitable to be disposed of in landfills; (ii) development of building materials with good mechanical and thermal insulation properties trapping the hazardous industrial by-products through the synthesis of GFs; (iii) reuse of the industrial by-products as a resource for new technological materials combining environmental sustainability and safety in the secondary aluminum industry workplaces, in the framework of a circular economy.