Sustainable Lightweight Aggregates from Diatomite Residue

Abstract

This study assessed the feasibility of producing lightweight aggregates (LWAs) using diatomite waste (DW) as a clay substitute. The research aimed to reduce the consumption of natural resources and minimise the environmental impacts caused by the disorderly disposal of DW. Chemical, physical, and mechanical tests were carried out on six formulations of mixtures containing 50% to 100% DW, sintered between 1100 and 1250 °C, resulting in 24 samples. The aggregates had a particle density between 1.14 and 2.13 g/cm³, a maximum bloating index of 5.7%, a crushing strength of up to 11.14 MPa, and a mass loss of up to 8.7%. Minimum porosity of 2.8 percent and water absorption of 2.0 percent were observed. Sixteen samples met the criteria required for commercial applications, demonstrating that replacing clay with DW is technically feasible. The high porosity of DW was found to influence the density of the LWAs. The findings of this study highlight the environmental sustainability of using DW as an alternative raw material, contributing to circular economy strategies in the construction sector.

1. Introduction

Lightweight aggregates (LWAs) are granular materials characterised by low-density that can be divided into (a) natural, such as volcanic slag, pumice, or tuff, and (b) artificial, such as expanded clay and blast furnace slag [1,2]. One of the earliest reports on using LWAs dates to approximately 2800 BC. According to Chandra and Berntsson [3], burnt bricks were crushed during the time of the Indus Valley civilisation and used as aggregates to make concrete. According to the authors, this LWA was porous and had a similar structure to LWA made from sintered clay. In Norway, commercial production of lightweight aggregates began in 1954. By 2002, Norwegian Leca produced around 1 million m³ of expanded clay annually [3].

The growing demand for low-density structural materials with high acoustic, thermal, and hygrometric performance is driving interest in using lightweight aggregates [4]. The diversity of LWAs that can be produced is another favourable factor. According to González-Corrochano et al. [5], a single plant located in Spain produces different types of LWAs, which, in turn, can be used in various applications, such as thermal and acoustic insulation, geotechnical applications, gardening, horticulture, prefabricated lightweight structures, lightweight concrete, and refractory mortars. The authors emphasise that the choice of LWA depends on the desired properties for each service.

The properties most used to differentiate and choose commercial lightweight aggregates are density, mechanical strength, water absorption, granulometric composition, porosity, and surface texture [3,5,6,7]. A systematisation of the properties of the leading commercial LWAs manufactured in Germany, Brazil, Denmark, USA, Spain, Hungary, England, and Sweden revealed that the specific mass and unit mass values of these aggregates ranged from 0.5 to 2.1 g/cm³ and from 0.26 to 1.1 g/cm³, respectively [3,5,7,8,9,10], most of which met the density requirements proposed by EN-13055-1 [11]. In addition, this literature review showed that the grain size of these commercial lightweight aggregates ranged from 0.5 to 19.0 mm and that the water absorption and crushing strength indices ranged from 0.4 to 40% and between 0.75 and 6.9 MPa, respectively [3,5,7,8,9,10].

The process of designing an artificial LWA is complex, as it depends on several factors, such as the granulometric, chemical, mineralogical, and microstructural composition of the raw materials, and the preparation process, including mixture formulation, homogenisation, shaping, and the adopted sintering method [10,12,13]. According to the published literature, the most used methodology for obtaining LWAs seeks to achieve the desired properties during the sintering process, based on the expansion of the granule and the formation of pores after the decomposition of compounds [10,12,13,14]. Despite its complexity, producing LWAs shows great potential for incorporating new materials, including waste [13,15,16,17,18,19,20,21,22,23,24,25].

The valorisation of industrial by-products, such as diatomite waste, aligns with global sustainability goals and promotes resource efficiency by reducing the extraction of raw materials and minimising landfill disposal. This approach supports the principles of circular economy, which have been identified as effective strategies for increasing sustainability in the built environment [26]. In particular, the reuse of these wastes supports the principles of circular economy and contributes directly to several United Nations Sustainable Development Goals (SDGs), including SDG 9 (Industry, Innovation, and Infrastructure), SDG 11 (Sustainable Cities and Communities), and SDG 12 (Responsible Consumption and Production) [27,28,29,30,31].

Diatomite is a natural material originating from the fossilised skeletons of diatomaceous aquatic plants. Its majority composition is amorphous silica. The diatom skeletons give this sedimentary rock a highly porous nature, with low density and considerable chemical inertia [32,33,34]. The main industrial uses of diatomite are as a filtration aid, functional filler for paints and plastics, and as an absorbent, abrasive, or insulating material [35,36,37,38].

In 2013, the world production of diatomite was approximately 2.3 million tonnes. However, environmentally hazardous waste is generated during the manufacture and subsequent use of this material [39]. In China, tens of thousands of tonnes of diatomite waste (DW) are produced annually [40,41]. According to Gong et al. [41], the main risks of DW are the waste of natural resources, the emission of CO₂ into the atmosphere, and the leaching of nitrogenous substances into the soil. Studies show that a significant amount of DW is generated during the brewing of beer, the filtration of vegetable oils, and the production of high-purity diatomite [41,42].

In the production of high-purity diatomite, a significant amount of material that does not meet commercial specifications is discarded. This low-quality diatomite waste, often mixed with impurities, accumulates near mining facilities, contributing to environmental liabilities and land occupation problems [43,44]. In addition to being classified as hazardous, its high volume and limited economic value make its reuse a significant challenge for sustainability.

The characterisation data of diatomite residue found in previous studies show that this residue (a) has a high SiO₂ concentration rate [45,46,47], (b) has quartz as the main crystalline phase [39,48], and (c) has significantly low mass-loss rates [46,47]. These properties make DW a promising option for various applications. Given this, Galán-Arboledas et al. [39] used DW to manufacture ceramic bricks. Letelier et al. [49] evaluated the mechanical properties of concrete made with DW and recycled aggregates. Man et al. [44] produced porous bricks using DW and sugar filter sludge. In another study, porous ceramics were made with DW and oyster shells [48]. However, to date, no research publication has been found on the use of DW in manufacturing lightweight expandable aggregates. The primary impediment appears to be the chemical composition and thermal stability of the material, which hinder its capacity for bloating. Accordingly, this study aims to develop LWAs by reusing diatomite residue.

2. Materials and Methods

2.1. Materials

Diatomite waste (DW) was collected from a waste disposal site of the company Mineralite Mineração Indústria e Comércio Ltd.a, located in the municipality of Rio do Fogo, Brazil. During the processing of high-purity diatomite, diatomaceous earth is collected and sent for drying in the open air. The material is then transported to the burning process, which operates at temperatures up to 1000 °C. The calcined mass is then cooled and sent to the final stages of milling and classification. At this stage, the densest material is separated and discarded, resulting in the generation of the waste under consideration. DW consists predominantly of low-grade diatomite mixed with silica sand, which is denser and does not meet the morphological and chemical purity requirements for commercial use. Similar residues have been identified in the literature, including low-grade diatomite, which typically has higher levels of mineral impurities and lower filtration efficiency. Its reuse contributes to waste valorisation and aligns with sustainable material management practices. Moreover, the extraction of clay for ceramic production purposes is associated with environmental degradation, including land destruction and resource depletion. The substitution of clay with industrial waste, either partially or in its totality, has been demonstrated to contribute to the conservation of natural resources. The clay (RC) was provided by Cerâmica do Gato Ltd.a. situated in Acauã district, in the city of Açu, Brazil. The material, red clay, is the same as that used to make the ceramic pieces sold by the company in question. The clay was supplied in its natural state, comprised of air-dried clods of varying granulometry. It should be noted that this material has already been used in previous studies [50].

Grinding and sieving techniques were employed in the processing of the raw materials used in this study. This processing aims to develop samples with an ideal particle size distribution for manufacturing lightweight aggregates, following a methodology previously used in successful studies [10,51,52]. Before use, the DW and RC were dried in an oven at 110 ± 5 °C until they reached constant mass and then fragmented in a ball mill to obtain fractions that passed through a sieve with an aperture of 150 µm.

Characterisation of Raw Materials

A CILAS 1090 high-resolution laser analyser (Orléans, France), was used to characterise the particle size distribution of the raw materials. The milled powders were analysed dry in a particle size range of 0.10 to 500.00 µm. Figure 1 shows the particle size distribution of the DW and RC. The red clay had an average particle diameter of 17.60 µm. The material also had the following particle percentages: D₁₀ = 1.01 µm, D₅₀ = 8.35 µm, D₉₀ = 49.78 µm.

In contrast, the results obtained for the diatomite waste were D10 = 0.38 µm, D50 = 5.83 µm, D90 = 76.36 µm, and an average diameter of 21.80 µm. In all the materials analysed, most of the particle size curves were located outside the ideal region. However, according to Cougny [51], it is still possible for granules to expand, as long as the sample has an appropriate chemical composition [10].

The chemical analysis of the raw materials was carried out using the energy-dispersive X-ray fluorescence (XRF) method with a Thermo Scientific Niton XL3T portable spectrometer (Waltham, MA, USA).

Table 1. Chemical composition of raw materials.

Sample	SiO2	Al2O3	Fe2O3	CaO	MgO	Na2O	K2O	Others	LOI (%)
DW	90.54	1.61	2.81	1.12	0.00	0.00	0.43	3.49	0.91%
RC	41.45	21.02	27.99	1.17	2.09	0.00	2.96	3.33	8.6%

LOI = loss on ignition.

presents the chemical composition of the initial raw materials. These data were used as input parameters during the formulation of the samples to develop materials that can form molten granules with sufficient viscosity to trap gases [51,53].

No potentially hazardous elements (such as Pb, Cd, Cr, or As) were detected above the detection limits of the equipment used in this study (Thermo Scientific Niton XL3T). However, given the limited sensitivity of the equipment for trace elements, additional studies using more sensitive techniques may be necessary to fully confirm the absence of environmentally hazardous impurities.

According to Table 1, the main compound identified in DW is SiO₂ (90.54%). This result is very similar to the data presented in previous research [39,47,49]. This significant concentration of silica oxides can raise the sample’s melting point, especially in samples with higher DW contents. Therefore, obtaining a liquid phase with adequate viscosity will require increasing the sintering temperature, as previously observed in the sintering behaviour of silica-rich materials [10,54]. The main RC constituents are SiO₂ (41.45%) and Fe₂O₃ (27.99%). In addition, significant quantities of alumina and potassium oxides were also identified. According to the relevant literature, it can be surmised that this high content of iron oxides can promote a reduction reaction of iron oxide with carbonaceous or organic materials, thus enhancing the expansion of the samples [4]. On the other hand, this high iron content can accelerate the viscous melting of the sample [55].

The mineralogical analyses of the starting materials were carried out using a Shimadzu X-ray diffractometer, XRD-7000 (Kyoto, Japan). The crystalline phases were identified using Cu Kα radiation at 40 mA and 40 kV in the range of 10 to 80° (2θ), using an angular step of 0.02° and a scanning speed of 1.20°/min. These test conditions are the same as those in a previous study [10].

As shown in Figure 2, the only crystalline phases identified in DW were cristobalite and quartz, the latter being denoted by some high-intensity signals. These findings align with those reported in earlier studies [39,55,56]. Accordingly, it is reasonable to infer that DW may facilitate gas entrapment during the granule sintering process, helping to produce mixtures with a viscosity suitable for this purpose.

In turn, the red clay mainly consisted of biotite, illite, and quartz. These data also reinforce the chemical composition results obtained in the XRF analysis. The addition of RC in the preparation of mixtures can promote a chemical–mineralogical correction of the samples, resulting in granules with significant proportions of clay minerals. Studies show that illitic clays favour the expansion of granules, form a substantial amount of liquid phase, are effective in trapping gases, and contribute to developing the mechanical strength of LWAs [4,10,13,53].

Initially, mass loss tests were conducted using an SDT650 thermal analyser (Hewlett-Packard Company, Palo Alto, CA, USA).At this stage, the temperature range was 30 to 1200 °C, following a heating ramp of 8 °C/min in a nitrogen atmosphere. Due to the limitations of the thermogravimetric analyser made available for this research, the mass loss analyses of the starting materials had to be carried out in two different ways. Additionally, other samples were subjected to mass loss analysis through the sintering process using a three-phase chamber furnace manufactured by JUNG Fornos Industriais Ltd.a. (São Leopoldo, Rio Grande do Sul, Brazil). At this stage, the starting materials were subjected to temperatures of 1250 °C, a value similar to the threshold that will be applied during the manufacture of the LWAs. The difference in raw material mass calculated the mass loss index at each point before and after sintering [10,57].

Regarding the red clay (Figure 3), a mass loss of approximately 4.69% was observed at 100 °C. This can be attributed to the evaporation of adsorbed water. An 8.88% mass loss was observed in the 100 to 700 °C range. According to Souza [10], this behaviour is related to the combustion of organic matter, the decomposition of carbonates, and the dehydroxylation of illite. For a heating level of 1250 °C, analysed in a complementary manner, the red clay exhibited a total mass loss of more than 16.0%. According to Souza [10], this favours the release of gases from the decomposition of illite and reactions with Fe₂O₃.

The DW showed strong thermogravimetric stability (Figure 4). This is due to the temperature used during diatomite processing. In the case of the DW, the total mass loss at 1000 °C was approximately 2.30%. This behaviour results from minor dehydration, decomposition of residual carbonates, and traces of organic or industrial residues not entirely removed during diatomite processing [10,39]. Complementary mass loss measurements, carried out by sintering the raw materials at 1250 °C due to the temperature limitation of the TG-DTA equipment, revealed an additional 0.51% mass loss for the DW. This result suggests a lower potential for gas release during firing when compared to the other raw material evaluated under the same thermal conditions.

Scanning electron microscopy (SEM) studies were carried out on all the raw materials used in this research using a Tescan Vega-3 LMU microscope (TESCAN ORSAY HOLDING, Brno, Czech Republic) at an accelerating voltage of 20 kV. Figure 5 shows that the RC particles are composed of irregular grains with a rough texture and fine granulometry. In general, a rougher surface tends to impair the workability of mixtures when making LWAs; however, according to Souza et al. [50], red clay has adequate plasticity for forming granules when mixed with water in the appropriate proportions.

The diatomite residue exhibited a multiphase structure (Figure 6). According to Figure 6a, the DW sample consists of large grains with a smooth texture and varied shapes, as well as tiny particles with an irregular and rough structure. The larger particles appear solid and dense, formed from the quartz contained in the diatomaceous earth. In turn, the morphology of the smaller particles consists of a highly porous structure in the form of cylinders, boats, and straight-chain diatoms (Figure 6b). The coarser phase of the material tends to improve the workability of the mixture and increase the density of the LWAs manufactured. In contrast, the finer-grained phase will influence these properties in the complete opposite way.

2.2. Methods

2.2.1. Mixtures Composition

All the mixtures produced in this study were produced using a protocol based on characterisation tests of the raw materials. This methodology aimed to develop suitable formulations for producing lightweight aggregates. The characterisation protocol was developed, and the results of the variables considered fundamental by Riley [53] and Cougny [51], such as (a) chemical composition, (b) mineralogy, and (c) granulometry, were determined. The six mixtures produced are shown in Figure 7. The DW content in the set in question ranged from 50% to 100%. In addition, the ideal plasticity for homogenising the samples was obtained by adding 36 to 50% water by mass. The replacement rates employed in this study were selected not only to assess technical performance but also to explore the potential for maximising the use of waste by sustainable production practices.

Chemical composition data from the formulations developed in this study were represented on the Riley diagram [53]. Figure 8 shows that a significant proportion of the samples were outside the expandable region. In general, the low Al₂O₃ content identified in DW prevented the samples from approaching the expandable region. Therefore, a large proportion of the samples may not reach a viscosity suitable for capturing gases during the sintering stage [53,58]. However, previous studies have shown that a suitable mineralogical composition can remedy this deficiency by making up for a certain lack of aluminosilicates [10,13].

2.2.2. Manufacture of LWA

Initially, the raw materials defined for each formulation were carefully homogenised by hand. Subsequently, water was added in predetermined quantities. At this juncture, an empirical approach was adopted to determine the water content required to achieve the requisite plasticity for granule formation. Each granule was moulded using approximately 1.0 ± 0.1 g of the wet mixture, resulting in spheroidal particles with varying diameters. Despite the absence of mechanical compaction, manual pressure was standardised using a single operator for all the samples. The technique employed for the processes of mixing and moulding was informed by existing research [10,59,60].

The granules were moulded manually, resulting in approximately spherical pellets. The raw granules were dried for 24 h at room temperature and then placed in an oven at 110 ± 5 °C until they reached a constant mass. The samples were then sintered from room temperature to temperatures of 1100, 1150, 1200, and 1250 °C, always at a heating rate of 8 °C/min and with an isothermal plateau of 15 min. Finally, the granules were cooled naturally to room temperature.

2.2.3. Characterisation of Lightweight Aggregates

Comprehensive characterisation analyses were carried out on the specimens produced in this study. Bloating control was analysed based on the variation in the average volume of the particles before the sample was fired (V1) and after the sintering process (V2), following the equation BI = 100 (V2 − V1)/V1 [61]. The loss of mass (LOI) was determined using the equation LOI = 100 (Mi − Mf)/Mi [57]. In this experiment, the diameters of the granules were measured with a digital calliper at three points equally spaced along the perimeter. The mean value was then used to calculate the volume of the granules using the formula for a sphere. This index was calculated from the difference between the mass of the sample dried in an oven at 110 ± 5 °C (Mi) and the mass of the respective sample after the sintering process (Mf). During these experiments, the quantity of granules analysed was subject to variation depending on the volume of the moulded mixture, with a typical range of 10 to 15 units of each formulation produced at each of the sintering temperatures employed.

The particle density of the samples (ρd), the density of the dry samples, excluding permeable pores (ρs), and the water absorption of the samples (WA24H) were evaluated according to the methods described in the Brazilian standard NBR 16917 [62], using the equations ρd = mA/(mB − mC), respectively; ρs = mA/(mA − mC) and WA24H = 100 (mB − mA)/mA, where mA = mass of the dry sample, mB = mass of the sample saturated surface dry, and mC = mass of the sample submerged in water.

The procedure involved submerging the oven-dried samples in water at room temperature for 24 h. Following this period, the samples were extracted, dried superficially with a cloth, and weighed to obtain the saturated superficial dry mass. Subsequently, the samples were immersed in a container attached to a hydrostatic balance to determine the submerged mass. The samples were then dried in an oven at a temperature of 105 ± 5 °C until a constant mass was achieved, yielding the dry mass.

The actual density of the samples (Dt) was determined by a water pycnometer based on the recommendations described in ME 093 [63]. In this test, the samples were ground to pass through a 150 µm sieve, dried to a constant mass, and placed in a pycnometer. After boiling to remove air, the pycnometer was cooled and weighed at each stage to obtain the pycnometer masses: empty (P1), with sample (P2), with sample and water (P3), and with water (P4). The actual density was then calculated using Dt = (P2 − P1)/[(P4 − P1) − (P3 − P2)]. To ascertain the value of Dt, 10 g of ground material were utilised in each test [63]. The porosity analyses of the samples followed the methodology commonly adopted in previous studies [64,65]. Thus, for each of the samples prepared, the total porosity (PT), open porosity (PO), and closed porosity (PC) were calculated according to the following equations: PT = 100 [1 − (ρd/Dt)]; PO = 100 [1 − (ρd/ρs)]; and PC = PT − PO, respectively.

The crushing strength of the granules (S) was determined individually using a hydraulic press. In this test, the granule was subjected to an increasing linear load until it reached the point of failure. According to the equation S = 2.8 Pc/πx², from the maximum load applied at the point of failure (Pc) and the diameter of the approximately spherical particle (x), it was possible to determine the individual crushing strength of the granule. Subsequently, the S result was obtained by averaging the results found in three samples extracted from the exact formulation and burned at the same temperature. It should be noted that this methodology is frequently used in studies related to the manufacture of LWAs [10,66,67]. For the sake of clarity and to facilitate understanding of the procedures adopted, the equations utilised for the calculation of the bloating index, densities, porosities, and resistance to crushing are presented in Appendix A.

2.2.4. Commercial Potential of the Formulations

The potential for commercial application was assessed by comparing the S and WA24H data with the categorisation parameters established by Souza [10], complemented by international benchmarks, such as EN 13055 [11], which sets density and particle size requirements for lightweight aggregates used in concretes. These categorisation parameters are grounded in standard requirements [11], technical datasheets from commercial LWA suppliers (CINEXPAN, ARLITA), and comparative analysis of mechanical and physical performance reported in the literature [7], providing a comprehensive reference framework. In the work in question, the treatment of data extracted from commercial LWA catalogues, technical standards, and the relevant literature resulted in the creation of five possible application groups for lightweight aggregates. A similar methodology has already been adopted in previous studies [5,50]. It should be noted that specimens with ρd greater than or equal to 2.00 g/cm³ were not analysed, thus meeting the density requirement for LWAs [11].

3. Results and Discussion

3.1. Bloating Index (BI)

Table 2. Bloating index of mixtures.

Samples	Bloating Index (BI) (%)
	1100 °C	1150 °C	1200 °C	1250 °C
D100	−5.2	−1.4	−5.1	0.3
D90	2.1	−0.4	−0.6	−7.5
D80	4.5	−4.9	−11.5	−7.9
D70	1.5	5.7	−7.5	−15.3
D60	−6.9	−14.0	−22.7	−24.9
D50	−2.5	−8.9	−20.4	−13.1

shows the BI data for the six binary samples sintered at four different temperatures. Analysing the results, it is possible to see significant variability between the data found. In this set of samples, the bloating index ranged from −24.9% to 5.7%, identified in samples D60 when sintered at 1250 °C and D70 when sintered at 1150 °C, respectively. A particular potential for bloating is evident, as four of the six samples expanded (BI > 0) at least one of the adopted sintering temperatures.

Among the formulations prepared by mixing RC with DW, three exhibited expansions: D90 and D80 at 1100 °C and D70 at 1100 and 1150 °C. In many cases, raising the temperature resulted in shrinkage for this group of samples. In general, there was a tendency for the shrinkage of the samples to decrease as the DW content increased. These results confirm that DW is highly resistant to thermal degradation. The strong contraction identified in the samples richer in RC, especially at 1200 and 1250 °C, suggests that the viscous mass of these samples did not retain the gases released by the decomposition of the clay minerals. Therefore, this loss of mass resulted in shrinkage [68].

As shown in Figure 9, the external appearance of the D100 samples confirms that obtaining viable LWAs using only DW is not feasible. The surfaces of these samples were not vitrified, regardless of the sintering temperature used. This compromises essential properties for LWAs, such as mechanical strength and water absorption. It is, therefore, clear that excessive use of DW in LWAs can be detrimental to the samples, making them unsuitable for commercial use.

Unlike the samples mentioned above, the D70 formulation exhibited bloating at 1100 and 1150 °C, while shrinkage occurred at 1200 and 1250 °C. Among the samples reported in Figure 9, this mixture was the one with a chemical composition closest to the expandable region proposed by Riley [53]. Thus, it is possible to state that the chemical composition of D70 allowed the formation of a molten material with the appropriate viscosity to capture the gases released at 1100 and 1150 °C. According to Quina et al. [14], this temperature range is similar to what the industry uses to manufacture commercial LWAs.

The surface texture of most of the samples is visibly rough. The sintering temperatures adopted were not high enough to generate a glassy surface layer. However, some studies show that the flexural strength of concrete at early ages can be increased by using aggregates with a rougher texture, as these improve the physical bond with the cement paste [68].

Alternatively, an analysis of the data in Table 2 indicates the existence of a relative bloating index. Taking the D50 mixture as an example, there is a constant contraction while the material is sintered up to 1200 °C. However, after sintering at 1250 °C, there is a significant increase in grain size. This suggests that the bloating potential of this sample is more significant than that observed through BI. From this point of view, it can be said that samples D100 and D80 also experienced bloating during the sintering process. Figure 10 shows a simulation of the behaviour of sample D50 to increase understanding of the existence of relative bloating. The data for the other samples are presented in Appendix B.

The D50 mixture exhibited a negative BI at all the sintering temperatures used in this study. However, D50 showed a relative bloating index of approximately 9.2% between 1200 and 1250 °C temperatures. These results suggest that (a) the sample above, at a certain point, was able to form an adequate amount of liquid phase and consequently a glassy surface, which allowed the generation and retention of gases, causing an increase in the volume of the grains [69]; and (b) the ideal temperature range for this specimen is higher than that applied in this study.

In the D100 formulation, the values obtained for BI were −5.2%, −1.4%, −5.1%, and 0.3% after sintering at 1000, 1150, 1200, and 1250 °C, respectively. In practical terms, raw granules with a diameter of 10 mm resulted in spheres with final dimensions of around 9.83 mm, 9.95 mm, 9.84 mm, and 10.29 mm, as shown in Appendix A. Interpreting the BI values in isolation could wrongly suggest a low expansion potential for the mixture. However, the dimensional growth observed from 1100 °C onwards indicates that there was significant swelling, estimated at approximately 15.0%.

3.2. Loss of Ignition (LOI)

The loss of ignition of the samples after sintering at 1250 °C was measured, and the results are shown in Figure 11. Invariably, all the mixtures produced showed a significant loss of mass. The LOI of this group showed a minimum value of 6.1% and a maximum of 8.7%, identified in samples D100 and D50, respectively. These results are encouraging from the perspective of gas generation involved in bloating, as they align with the findings of Moreno-Maroto et al. [64]. In some cases, the amount of gas needed to swell the sample is derived from a mass loss of approximately 0.1% of its original weight.

On the other hand, the bloating results of samples D90 and D80 at 1100 °C show an expansion of 2.1 and 4.5%, respectively. This suggests that a significant portion of the mass loss occurred at a temperature range lower than or at least equal to 1100 °C. According to Moreno-Maroto et al. [64], such samples have difficulty expanding as the temperature rises. Many of the gases were lost before their mass reached an adequate viscosity. This thesis is consistent with the bloating index data for D90 and D80 at 1150, 1200, and 1250 °C.

An LOI of almost 9.0% is noted when the DW content used is 50%. For DW proportions of around 70%, the mass loss rates are reduced to around 7.0%. Finally, when the clay is entirely replaced by diatomite residue, the LOI is reduced to almost 6.0%. Therefore, in this group of samples, the increase in DW tends to reduce the amount of gases generated due to its significant thermal stability. Thus, RC would have a greater influence on releasing gases, while DW could affect the viscosity of the liquid phase obtained during sintering.

3.3. Density (ρd)

Table 3. Particle density (ρd) of the mixtures.

Samples	Particle Density (pd) (g/cm3)
	1100 °C	1150 °C	1200 °C	1250 °C
D100	1.14	1.18	1.22	1.25
D90	1.33	1.37	1.39	1.40
D80	1.45	1.51	1.58	1.60
D70	1.52	1.62	1.74	1.80
D60	1.69	1.85	2.00	2.10
D50	1.82	2.00	2.13	1.92

presents the particle density results. There was a significant variation in the ρd data due to changes in the proportions of materials and the heating rate. The densities of the samples varied from 1.14 to 2.13 g/cm³. The formulations met the density requirement proposed by EN-13055-1 [11] for lightweight aggregates (ρd < 2.0 g/cm³) in at least two of the sintering ranges applied.

With the exception of D50, all the mixtures exhibited a gradual increase in density with rising temperature. In sample D60, for example, there is a strong negative correlation between the progression of the bloating index and particle density. On the other hand, the D50 mixture showed a significant reduction in ρd as the heat rose from 1200 to 1250 °C. This shows the importance of relative bloating in the behaviour of LWAs since, according to Figure 10, the increase in grain size seen in this temperature range contributed to the decrease in ρd and, consequently, to the suitability of D50 at 1250 °C for the proposed density requirements [11].

According to Figure 12, the results generally show a strong negative correlation between the DW content and ρd data. Except sample D50 at 1250 °C, reducing the amount of DW resulted in gains in density at all the sintering intervals adopted. The sample in question comprises 50% RC and 50% DW, and this result can be attributed to the relative bloating suffered at this temperature level.

On the other hand, considering that in this group of samples, there is a negative correlation between the mass loss data and the DW content (Figure 11) and that few samples in this set showed a BI > 0 (Table 2), it is possible to assume that the high porosity of the diatomite residue exerted a strong influence on the results presented, helping to reduce the density of the samples and consequently obtaining LWAs. This proposition is consistent with the results of the DW microstructure (Figure 6) and the reports presented by Lynn et al. [70].

3.4. Closed Porosity

The results of the closed porosity of the samples are shown in

Table 4. Closed porosity (Pc) of the mixtures.

Samples	Closed Porosity (Pc) (%)
	1100 °C	1150 °C	1200 °C	1250 °C
D100	3.3	3.1	2.8	2.8
D90	5.8	5.8	5.5	6.0
D80	5.4	6.3	6.1	7.7
D70	5.1	6.1	7.1	9.3
D60	5.5	7.5	10.2	12.4
D50	7.3	10.0	11.1	20.4

. The values of PT, PO, Dt, and ρs used to calculate Pc are available in Appendix C and Appendix D.

According to Table 4, the closed porosity values ranged from 2.8 to 20.4% and were identified in samples D100 and D50, respectively. Considering that the sintering temperature applied was 1250 °C in both cases, it can be assumed that the influence of the raw material content on Pc is stronger than the effect caused by the firing method. However, in most cases, there is a strong positive correlation between the sintering temperature and the Pc data. This indicates that the heating rate also significantly influences these results.

In the D50 mixture, for example, the increase in heating rate from 1100 to 1250 °C resulted in an increase in Pc from 7.3 to 20.4%, in that order. This strongly correlates with the relative bloating of the sample in question (Figure 10). In general, the increase in closed porosity in lightweight aggregates affects fundamental properties of LWAs, such as strength and water absorption [64,71]. This will be discussed in the following sections.

The results demonstrate the high potential of these formulations for manufacturing LWAs. According to Moreno-Maroto et al. [64], increasing the temperature typically leads to lower porosity as a result of enhanced sintering. However, an inverse effect is observed in most samples. The increase in closed porosity is related to the synergy of the mixture of RC with DW and the retention of gases. In addition, the open porosity data (Appendix D) indicate an increase in liquid phase formation as the temperature rises, since the PO values almost always decrease after the heating rate increases.

The data show that an increase in the proportion of DW usually leads to a reduction in the Pc of the granules (Figure 13). A more complete assessment of Pc development in this set of samples can be made from a joint assessment with Appendix C. At 1250 °C, for example, the open porosity increases significantly with the increase in DW. In other words, high DW contents favour the formation of open porosity due to the loss of gas generated at higher temperatures [72]. Additionally, the increase in Po can significantly enhance the water absorption of the samples.

3.5. Water Absorption (WA_24H)

The water absorption results of the formulations are shown in

Table 5. Water absorption for the mixtures.

Samples	WA24H (%)
	1100 °C	1150 °C	1200 °C	1250 °C
D100	44.7	41.6	39.4	37.2
D90	30.7	28.6	27.7	26.9
D80	25.1	22.3	19.3	17.6
D70	22.7	18.3	13.5	10.6
D60	16.1	10.9	5.3	2.0
D50	11.4	5.8	2.5	2.0

. Thirteen of the twenty-four specimens produced had WA_24H < 20.0%. These results are encouraging since, according to the literature, the WA_24H required for lightweight aggregates used in high-strength concrete and lightweight structural concrete is usually less than 20.0% [50]. Furthermore, in 20.8% of the specimens, WA_24H was less than 10.0%, indicating that the surface of these samples was well vitrified [73,74].

According to Table 5, the formulations showed a downward trend in WA_24H as the temperature rose. Sample D90, for example, showed a slight variation in the results, decreasing from 30.7% at 1100 °C to 26.9% at 1250 °C. However, these results jeopardise its application potential. An increase in DW content appears to increase the porosity associated with the surface, thereby reducing the impermeability of the granule shell. This is supported by the open porosity data (Appendix D).

To assess the influence of the increase in DW on WA_24H, the data for replacing 50 to 100% of RC with DW were analysed (Figure 14). Unchanged, the addition of DW increased the water absorption rates. In other words, samples containing 50% or 60% DW had a significantly lower WA_24H than specimens with 80 or 90 percent of the waste in question. The use of higher levels of DW likely formed a less dense and more permeable surface, favouring water penetration in these samples.

3.6. Crushing Strength (S)

Table 6. Crushing strength (S) of the mixtures.

Samples	Crushing Strength (S) (MPa)
	1100 °C	1150 °C	1200 °C	1250 °C
D100	0.44	0.28	0.35	0.38
D90	0.98	1.16	1.00	1.10
D80	2.13	2.13	2.19	1.70
D70	3.17	3.20	3.57	3.03
D60	6.09	6.27	7.41	7.15
D50	8.27	9.76	11.14	6.98

presents the results of the samples’ crush resistance. In general, the mixtures showed encouraging results in terms of crushing resistance. Except for sample D100, the other specimens showed S greater than 1.0 MPa in at least three of the four sintering temperatures adopted, thus demonstrating the potential for some application in engineering works or services [50]. In addition, 50% of the specimens produced had a crushing strength greater than 2.3 MPa, a value typically found in Brazilian commercial LWA [75].

The samples showed a loss of crushing resistance as the volume of diatomite residue increased. For example, the D90 and D80 formulations showed lower S values than the commercial aggregate reported by Souza [10], regardless of the adopted temperature range. It is challenging to produce commercial lightweight aggregates with DW contents exceeding 80%.

Similarly to the ρd data (Figure 12), a strong negative correlation exists between the proportion of diatomite residue and the S results (Figure 15).

For example, during sintering at 1200 °C, the S values progressively decreased from 11.14 to 0.35 MPa with an increase in DW content from 50% to 100%, respectively. The increase in DW content tends to impair the viscosity of the liquid phase due to the reduction in Fe2O3 content, resulting in a poorly vitrified surface. This consequently contributed to a reduction in crushing resistance [76].

3.7. Mineralogical Composition

The D70 formulation, sintered at 1250 °C, was chosen for analysis of its mineralogical composition. According to Figure 16, the most critical crystalline phases in this sample are cristobalite, mullite, and quartz. Based on the literature, the preservation of some peaks of quartz minerals strongly correlates with the viscosity of the liquid mass of D70 [10]. It seems that the sample in question could generate a molten material with a viscosity suitable for trapping gases. This proposition is strengthened when analysing the BI results at 1100 and 1150 °C (Table 2).

The amount of molten mass generated in D70 can also be attributed to the partial fusion of mullite since the insufficiency of fusing oxides did not compromise the formation of molten material [77]. The generation of mullite in D70 is related to the reactions that caused the quartz peaks in this sample to decrease [78]. According to Piszcz-Karaś et al. [79], mullite formation is essential for producing lightweight aggregates with a mechanically durable structure.

3.8. Microstructure

Finally, Figure 17 shows the internal morphology of the D70 sample. According to Figure 17a,b, the microstructures of the fracture surfaces of D70 at 1150 and 1200 °C still preserve part of the large grains identified in DW. In general, the formation of the liquid phase facilitated the coverage of these particles. However, the low amount of molten material favoured a higher water absorption rate, especially in the lower temperature ranges. For example, the WA24H found in D70 at 1100 °C was approximately 22.7%.

The DW grains become more fused as the temperature rises to 1250 °C. At this stage, the molten material makes the structure of D70 relatively denser. Unlike the other samples, the rise in temperature in D70 was not accompanied by a significant increase in the number of large pores. In addition, there was a gradual reduction in open porosity. As a result, the sample in question progressively showed shrinkage (Table 2), densification (Table 3), and a decrease in water absorption rates (Table 5).

To further facilitate comprehension of the microstructural evolution, additional annotated SEM images are appended in Appendix F. The images presented here underscore the preservation of substantial diatomite grains at 1150 °C and their subsequent fusion at 1250 °C, thereby reinforcing the established correlation between an increase in temperature and the degree of vitrification.

3.9. Commercial Potential of the Formulations

Based on Souza’s parameters [10], 16 of the 24 specimens produced had potential for future commercial applications. It is evident that formulation D100, which is composed entirely of DW, has repeatedly failed to satisfy the minimum physical and mechanical criteria that are requisite for any commercial application. The findings of this study unequivocally suggest that the utilisation of pure DW aggregates is not advised. On the other hand, specimens D50, sintered at 1100 °C, and D60, fired at 1150 °C, showed potential for simultaneous use in all the applications catalogued by Souza [10].

Specimens D60, sintered at 1100, 1150, and 1200 °C, and D50, sintered at 1100 and 1250 °C, showed low WA_24H (≤20.0%) and high S (>5.0 MPa), characteristics typically required of LWAs used in high-strength concretes [50]. For example, all the specimens mentioned had a higher crushing strength than the values reported by Franus et al. [80] in two different types of commercial LWAs, Arlita and LECA. On the other hand, aggregates used in structural lightweight concrete usually have low WA_24H indexes and moderate values for S [10]. In this context, eight specimens exhibited WA_24H lower than 20% and S higher than 2.30 MPa, thus resembling the primary Brazilian commercial lightweight aggregate [75].

Lightweight aggregates used in non-structural lightweight concrete and lightweight mortars do not require low water absorption levels (WA_24H ≤ 34.0%), nor do they need high crushing strength levels (S > 1.8 MPa) [50,81]. Thus, 12 specimens from this grouping proved suitable for the application. For example, sample D80, burnt at 1200 °C, proved ideal for this application. In addition, ten samples could be used in geotechnical applications (10.0% ≤ WA_24H ≤ 34.0% and S > 1.8 MPa), and 14 samples showed potential for use in gardening, landscaping, and thermal and acoustic insulation (10.0% ≤ WA_24H ≤ 34.0% and S > 1.0 MPa) [10].

The findings indicate that a partial substitution of RC for DW, for instance, in D60 formulations sintered at 1100–1200 °C, results in lightweight aggregates that demonstrate satisfactory performance. These samples were found to meet the requisite standards for strength and absorption, as well as fulfilling the criteria for multiple proposed applications [10,11]. The effective bloating and mechanical behaviour exhibited are consistent with prior observations [51]. Despite the suboptimal technical performance exhibited by the D100 formulation, the outcomes of the intermediate mixes substantiate the technical viability of integrating diatomite tailings in the fabrication of LWA. This finding provides further support for the notion that the material can serve as a sustainable alternative raw material, thereby adding value to waste that would otherwise be discarded. To facilitate a comparative visual analysis of the formulations’ performance, Appendix E presents radar graphs. These graphs summarise the normalised values for particle density, crushing strength, water absorption, and closed porosity for each sintering temperature (1100, 1150, 1200, and 1250 °C). The graphs presented offer a clearer perspective on the behaviour of each formulation concerning the primary properties under various thermal conditions.

The present study incorporates a sustainability indicator related to material efficiency, in addition to technical analysis. In the formulation containing 90% DW, for instance, industrial waste was utilised as the primary raw material, almost entirely replacing clay. This substitution signifies a direct and substantial deviation from the conventional disposal of DW in landfills, thereby preserving a significant quantity of natural clay. Despite its modest scale in a laboratory setting, this outcome substantiates the viability of integrating substantial amounts of waste into the fabrication of LWAs. When extrapolated to an industrial context, the use of 1 tonne of 90% DW-based aggregates would result in the recovery of 900 kg of waste and a corresponding reduction in the extraction of natural clay. This approach is consistent with the principles of circular economy and contributes to the more sustainable management of raw materials in the construction sector [24,26].

From a sustainability perspective, incorporating diatomite waste into the production of lightweight aggregates presents a promising alternative for mitigating the environmental impacts associated with traditional clay extraction. The diversion of diatomite waste from disposal has been demonstrated to have a dual benefit, reducing pollution and pressure on land use, while concomitantly enhancing the environmental performance of building materials. Moreover, the development of technically viable aggregates from waste has the potential to improve the prospects for sustainable construction and to support circular resource flows in the construction sector [26].

In addition to the outcomes attained at the laboratory scale, the process of transitioning to industrial implementation necessitates meticulous deliberation. While DW is generated in significant quantities and its particle size and chemical composition are compatible with LWA production, scaling up would demand investments in drying, grinding, and precise sintering control. Furthermore, the high porosity and thermal stability of DW may require adjustments in sintering curves and formulation strategies. Logistics, encompassing the collection and transportation of DW from processing industries to aggregate plants, also plays a crucial role. However, given that numerous ceramic and construction material manufacturers already operate rotary kilns and other necessary infrastructure, partial replacement of clay by DW in industrial production appears technically feasible with minor adaptations. It is recommended that future pilot-scale trials be conducted to validate these findings under real-world conditions.

In order to provide a comprehensive overview of the experimental procedures and key outcomes of this study, please refer to the workflow diagram, which is presented in Appendix G. The visual summary delineates the sequence of raw material characterisation, mixture preparation, sintering stages, testing, and commercial classification.

4. Conclusions

In general, the results obtained in this work show that it is technically possible to produce lightweight aggregates by replacing clay with diatomite waste. In many cases, the use of this waste has resulted in LWAs with absolutely encouraging properties from a technical, commercial, and sustainable perspective. The following specific conclusions can also be drawn from the experimental data:

• To manufacture LWAs with good physical and mechanical properties, it is recommended that the addition of DW to the mixtures be limited to levels of less than 90%. In samples richer in diatomite residue, the effect of DW on the physical and mechanical characteristics of LWAs is more significant than the influence of the firing method adopted. Regardless of the sintering temperature applied, all the samples developed with 100% DW had a crushing strength of less than 1.0 MPa, preventing these specimens from being recommended for any commercial application.
• The high porosity of DW was found to affect the particle density of LWAs. Typically, the partial replacement of RC with DW resulted in a gradual reduction of ρd. On the other hand, the addition of DW also increased open porosity, negatively affecting the strength and water absorption of the samples. At 1250 °C, when the DW content was increased from 50 to 100 percent, the particle density decreased by almost 35 percent. However, S was reduced by nearly 95%, while WA24H was increased by more than 1200%.
• In many cases, the LWAs manufactured had physical and mechanical properties higher than the values required by national and international standards, as well as the results found in commercial LWAs. In total, 16 of the 24 specimens manufactured met the criteria for particle density, crushing strength, and water absorption indices required for using LWAs in commercial applications.

This research has demonstrated the potential of utilizing diatomite waste in lightweight aggregates (LWAs). If applied in producing lightweight aggregates, the quantities of DW available are reduced and gain market value. In this way, the results of this study demonstrate that LWAs possess technological properties that could soon become indispensable in sustainability plans within the construction industry and for environmental preservation. Consequently, the environmental benefits associated with the reuse of diatomite waste serve to reinforce the sustainability potential of this approach. The present study contributes to reducing resource depletion and promoting circular and sustainable construction practices by transforming a low-value industrial by-product into a functional material for construction.