Silicate Sawing Sludge Recovery in Thermo Eco-Mortar for Macroporous Plaster

The aim of this research was to develop a new eco-product to find a correct recovery of silicate sawing sludge by means of waste management according to European criteria. To reach this goal, a thermal eco-mortar for a macroporous plaster was developed. The main characteristics of a plaster that influence the correct choice of the mortar are good adherence with underlying support, impermeability, thermal and acoustic insulation, mechanical resistance and ability to allow transpiration processes through the wall’s perimeters. Plaster is a mortar composed of a binding part that incorporates sand with a selected particle size distribution, not greater than 2 mm. The sludge, to be used as plaster, must satisfy requirements related to thermal insulation, resistance to moisture, mechanical resistance and good injection. For this purpose, low-content metals sludge, derived from the Luserna stone flaming and cutting slabs, are to be reused as a substitute for the sands and fine particles, respectively, that are normally used to produce plasters. The laboratory tests carried out on the finished product, in accordance with European standards, are as follows: water absorption, specific density, flexural and compressive strength, before and after freeze and thaw cycles, pull out, salt crystallisation cycle resistance and thermal conductivity. Chemical and leaching tests were carried out to verify the possible release of heavy metals into the environment after installation. The product quality was demonstrated as the cement mortars, incorporating the metals, did not allow their release in nature. A sludge recovery, in an unaltered state, was provided to reduce any costs connected to a pre-treatment and to make recovery economically advantageous for the stone sector.


Introduction
Typically, stone quarries and stone processing plants are designed to extract and produce a marketable product with the maximum output and profit, but stone waste production and environmental management are a consequence of this process. One of the main options for managing rock processing waste, according to European Strategies on Best Available Technique (BAT) [1], is for land use, e.g., as aggregates or for restoration. The three factors: cost, environmental performance and risk of failure must be considered for the choice of the waste management method. The directive 2008/98/EC [2] defines the concept of "end-of-waste" and the priority order actions hierarchy for waste management. Prevention on products is a measure aimed at reducing the amount of waste and reducing the content of hazardous materials; re-use, recycling and recovery of waste are all actions aimed at resource efficiency, for reducing the environmental impact.

Mix Design Components Characterisation
The processing waste of Luserna stone was used to prepare the design mix of the mortar. Luserna stone is a Piedmont gneiss characterised by a flat-schistose texture and by a heteroblastic structure with a micro-occhiadine tendency, due to the presence of millimetric micro-porphyroblasts (0.03-3 mm) ( Figure 1). It is composed of Quartz (50%); Plagioclase (15%); Alkaline Feldspar (20%); White mica (5%); Chlorite (5%); Epidote (5%). The two types of aggregate, derived from Luserna stone processing, were considered in this study: • waste after flaming used as a sand aggregate. • waste from sawing blocks into slabs through diamond tools, used as a filler aggregate.
A preliminary investigation was carried out for the particle size distribution analysis, performed according to UNI EN 933 Part 1 [18]. The results obtained are shown in Figure 2. Luserna sand, from the flaming process, contains only mineral fractions that compose Luserna stone. Sawing sludge, from diamond blade cutting, contains a fraction from the mineral part and a metal fraction derived from the diamond tool's wear. For this reason, magnetic separation, chemical analysis, leaching tests and XRPD (X-Ray Powder Diffraction) were performed on the Luserna sawing sludge sample. The quantification of the metal content in the sludge is necessary to define what is now a waste, a secondary raw material that can be used in other production processes (in accordance with Italian Legislative Decree 152/2006 [19], implementation of European Directive 2008/98/EC). Chemical analysis was performed, according to Italian Legislative Decree 152/2006, by means of Fkv Ethos Easy microwave mineraliser and ICP-Mass analyser. The results obtained, with the law threshold concentration limits, are reported in Table 1. The two types of aggregate, derived from Luserna stone processing, were considered in this study: • waste after flaming used as a sand aggregate. • waste from sawing blocks into slabs through diamond tools, used as a filler aggregate.
A preliminary investigation was carried out for the particle size distribution analysis, performed according to UNI EN 933 Part 1 [18]. The results obtained are shown in Figure 2. The two types of aggregate, derived from Luserna stone processing, were considered in this study: • waste after flaming used as a sand aggregate. • waste from sawing blocks into slabs through diamond tools, used as a filler aggregate.
A preliminary investigation was carried out for the particle size distribution analysis, performed according to UNI EN 933 Part 1 [18]. The results obtained are shown in Figure 2. Luserna sand, from the flaming process, contains only mineral fractions that compose Luserna stone. Sawing sludge, from diamond blade cutting, contains a fraction from the mineral part and a metal fraction derived from the diamond tool's wear. For this reason, magnetic separation, chemical analysis, leaching tests and XRPD (X-Ray Powder Diffraction) were performed on the Luserna sawing sludge sample. The quantification of the metal content in the sludge is necessary to define what is now a waste, a secondary raw material that can be used in other production processes (  Luserna sand, from the flaming process, contains only mineral fractions that compose Luserna stone. Sawing sludge, from diamond blade cutting, contains a fraction from the mineral part and a metal fraction derived from the diamond tool's wear. For this reason, magnetic separation, chemical analysis, leaching tests and XRPD (X-Ray Powder Diffraction) were performed on the Luserna sawing sludge sample. The quantification of the metal content in the sludge is necessary to define what is now a waste, a secondary raw material that can be used in other production processes (in accordance with Italian Legislative Decree 152/2006 [19], implementation of European Directive 2008/98/EC). Chemical analysis was performed, according to Italian Legislative Decree 152/2006, by means of Fkv Ethos Easy microwave mineraliser and ICP-Mass analyser. The results obtained, with the law threshold concentration limits, are reported in Table 1. The results showed an excess of the concentration limits for the elements cobalt and Cr VI, but only for its reuse in the green area (Column A).
Leaching testing was performed according to UNI EN 12457-2 [20], taking into account the threshold limits concentration defined in the appendix A of UNI 10802 [21]. The results obtained are shown in Table 2. On the basis of the chemical analysis, it could be asserted that Luserna sawing sludge could be re-cycled for plaster applications in buildings.
Magnetic separation, by means of Magnetic Separator (Eriez magnetics manufacturer, L Series, Model 4, Brasil) (Kolm-type high gradient), was performed to identify the quantity of metal content in the sludge sample. One of the main requirements for plaster application is thermal insulation. A low metal content can produce a lower thermal conductivity since metals are good thermal conductors. Table 3 shows the low percentage of magnetic (metals) fraction in the considered sludge sample. The X-ray powder diffraction was performed with Rigaku model Geigerflex (model smartlab SE, Rigaku, Japan). The sample preparation was carried out by means of grinding, through a pestle, the sludge to obtain a fine powder (with dimensions between 25 and 10 m). This procedure is necessary to improve the homogeneity of the material.
Only the sample deriving from diamond blades sawing was analysed since the flaming sample has only components of the Luserna stone of which the composition is known. The analysis was carried out both on the sample as it is and on the sample after the magnetic separation on the magnetic fraction. This choice was due to the complexity in the interpretation of the results. The sample was composed of several minerals and several metals: for this reason, the peaks were overlapped. With magnetic separation, there was an elimination of some minerals phases, so the peaks of the metal were more evident. A comparison of spectrum for samples as it is and with the only magnetic fraction is shown in Figure 3.
Materials 2020, 13, x FOR PEER REVIEW 5 of 16 magnetic separation, there was an elimination of some minerals phases, so the peaks of the metal were more evident. A comparison of spectrum for samples as it is and with the only magnetic fraction is shown in Figure 3. The Luserna diamond blade sawing sample showed minerals like Quartz, Feldspar and Mica and some peaks of metals that compose both the steel core of the disc and the matrix of the diamond segments. The metals identified were: Cobalt Molybdenum Oxide (CoMoO3), Aluminum Nickel (Al3Ni), Nickel Phosphide (NiP) and Nickel Chromium Oxide (NiCrO3).
In the analysis performed on magnetic fraction, we could observe a peak related to the Chlorite and other metals that were not identified in sample analysis before magnetic separation. CuS, Cobalt Iron (Co3Fe7), Nickel Molybdenum Phosphide (NiMoP2) and Silicon Carbide (SiC) were identified. All elements were contained in high-speed tool steels.

Plaster Mix Design
The preparation of plaster mix design was carried out at TEK.SP.ED s.r.l. company (Casandrino, Napoli). The plaster types differ according to the kind of binder used. In this research, the binders were hydrated lime and portland cement, with a predominance of portland cement. The mix design involves the presence of lime, cement, aggregates, water, reinforcing fibres and foam. It is known from the literature [22] that the foam improves the thermal conductivity characteristics of the plaster but worsens the mechanical characteristics, such as compression. The foam, in this case, was used to give the final material a certain lightness, while the fibres gave greater mechanical resistance (tensile resistance). Table 4 shows the components of the mix design and the quantities used compared to 1 m 3 . Table 5 shows the mix design with the addition of Tufo's powder (commercial ''zeolitite", rocks with zeolite content higher than 50 wt.%, coming from a quarry site located near Comiziano, Nola-Naples (Campanian Ignimbrite)), used mainly for chemical resistance test [23,24] compared to specimens produced with waste derived from the Luserna stone only. Adding of Tufo powder has been carried out since the Roman Empire period [25,26]. In the analysis performed on magnetic fraction, we could observe a peak related to the Chlorite and other metals that were not identified in sample analysis before magnetic separation. CuS, Cobalt Iron (Co 3 Fe 7 ), Nickel Molybdenum Phosphide (NiMoP 2 ) and Silicon Carbide (SiC) were identified. All elements were contained in high-speed tool steels.

Plaster Mix Design
The preparation of plaster mix design was carried out at TEK.SP.ED s.r.l. company (Casandrino, Napoli). The plaster types differ according to the kind of binder used. In this research, the binders were hydrated lime and portland cement, with a predominance of portland cement. The mix design involves the presence of lime, cement, aggregates, water, reinforcing fibres and foam. It is known from the literature [22] that the foam improves the thermal conductivity characteristics of the plaster but worsens the mechanical characteristics, such as compression. The foam, in this case, was used to give the final material a certain lightness, while the fibres gave greater mechanical resistance (tensile resistance). Table 4 shows the components of the mix design and the quantities used compared to 1 m 3 . Table 5 shows the mix design with the addition of Tufo's powder (commercial "zeolitite", rocks with zeolite content higher than 50 wt.%, coming from a quarry site located near Comiziano, Nola-Naples (Campanian Ignimbrite)), used mainly for chemical resistance test [23,24] compared to specimens produced with waste derived from the Luserna stone only. Adding of Tufo powder has been carried out since the Roman Empire period [25,26]. The components were mixed with the S8 EVM (single-phase electric with mixer) machinery ( Figure 4) manufactured by TEK.SP.ED. s.r.l. company (Casandrino, Napoli, Italia). This machine is low power consumption, has a mixing system and a system for pumping plaster directly on site.  The components were mixed with the S8 EVM (single-phase electric with mixer) machinery ( Figure 4) manufactured by TEK.SP.ED. s.r.l. company (Casandrino, Napoli, Italia). This machine is low power consumption, has a mixing system and a system for pumping plaster directly on site. Once the mix design was prepared, the material was poured onto a vertical panel to simulate conventional industrial practice ( Figure 5). In this way, both the adhesion of the material to the support and its behaviour with regards to hot summer temperatures could be tested. This critical weathering condition could cause cracks and fractures, on normal plasters, that could worsen over time. Once the mix design was prepared, the material was poured onto a vertical panel to simulate conventional industrial practice ( Figure 5). In this way, both the adhesion of the material to the support and its behaviour with regards to hot summer temperatures could be tested. This critical weathering condition could cause cracks and fractures, on normal plasters, that could worsen over time. After two days, curing time (critical weathering condition), the results of plaster pouring were optimal. Figure 6 shows the homogeneity of the product obtained and the absence of fractures and cracks. Using the same mix design, specimens were prepared for the physical-mechanical tests (Figure  7.). The number of specimens and their size were in accordance with the standards for each test reported in Table 6. After two days, curing time (critical weathering condition), the results of plaster pouring were optimal. Figure 6 shows the homogeneity of the product obtained and the absence of fractures and cracks. After two days, curing time (critical weathering condition), the results of plaster pouring were optimal. Figure 6 shows the homogeneity of the product obtained and the absence of fractures and cracks. Using the same mix design, specimens were prepared for the physical-mechanical tests (Figure  7.). The number of specimens and their size were in accordance with the standards for each test reported in Table 6. Using the same mix design, specimens were prepared for the physical-mechanical tests (Figure 7.). The number of specimens and their size were in accordance with the standards for each test reported in Table 6.

Physical Properties
The following laboratory tests were performed to investigate the physical properties of the plaster: bulk density, spreading test, thermal conductivity and water absorption.

Bulk Density
Bulk density testing was carried out, weighing one cubic meter of material in the fresh state. For dry condition bulk density evaluation, three samples of plaster were placed in an oven at 105 • C. This was followed by two weighings, two hours apart, to verify that the mass was constant. Once the mass of the dry specimen was obtained (Ms, dry), the formula used to calculate the apparent bulk density for each specimen is as follows: bulk density [kg/m 3 ] = (Ms, dry)/Vs (1) where Vs is the volume of the mold (m 3 ). The average value of the three measurements was then calculated, approximating the result to 10 kg/m 3 .

Spreading Test
A spreading test was carried out using a Hagerman cone, according to ASTM D 6103 of 2017. The cone has a chamfer of 45 • , an upper diameter of 70 mm and a bottom diameter of 100 mm.

Thermal Conductivity
Thermal conductivity testing was performed using a KD2 Pro (Meter group manufacturer, Munich, Germany) with probe RK-1 model, 6 cm long and 3.9 mm diameter. Thermal conductivity was carried out on three samples, at 15 and 68 days of curing time and in dry condition (placed in the oven for 48 h at 60 • C).

Water Absorption
Water absorption was carried out on 12 samples, with 6 samples not subjected to a freeze and thaw cycle, and the other 6 samples after a freeze and thaw cycle. The procedure involved placing specimens in an oven at 60 • C to obtain a constant mass. Once the dry weight (Md) of the specimens was obtained, the specimens were placed in water until saturation. Subsequently, the specimens were weighed to obtain the saturated mass (Ms). The results were expressed as the percentage of absorption obtained according to the equation: where Ms = saturated mass (g); Md = dry mass (g).

Mechanical Properties
The following laboratory tests were performed to investigate the mechanical properties of the plaster: flexural strength, compressive strength, pull out.

Flexural Strength
Flexural strength was performed on three samples, according to UNI EN 1015-11 of 2007. A constant speed load of 50 N/s was applied to obtain a fracture between 30 s and 90 s. Flexural strength, after the freeze and thaw cycle, was performed according to UNI EN 12371 of 2010, which is related to natural stone, as it is not a test required by plaster standards. Twenty-five freeze and thaw cycles were performed with a four-hour duration of each cycle, consisting of two hours at −15 • C (in a saturated solution of sodium chloride) and two hours at 21 • C. The flexural test was performed seven days after the last cycle, according to UNI EN 1015-11:2007.

Compressive Strength
Compressive strength was performed on six samples, according to UNI EN 1015-11 of 2007. Freeze and thaw cycle was performed, according to UNI EN 12371 of 2010, related to natural stone, as it is not a test required by plaster standard.

Pull out
Pull out test was carried out, according to UNI EN 1015-12 of 2016, by TEK.SP.ED company on plaster pouring directly on the vertical panel after 28 days of curing.

Chemical Properties
The following laboratory tests were performed to investigate the chemical properties of the plaster: resistance to salt crystallisation, chemical analysis and leaching test.

Resistance to Salt Crystallisation
Resistance to salt crystallisation was carried out on two samples with Luserna stone mix design (L1 and L2) and two samples of the mix with added Tufo Powder (T1 and T2), according to UNI EN 12,370 of 2001, which is related to ornamental stone standard, as it is not a test required by plaster standards. The test procedure included the use of a 14% solution of sodium sulphate decahydrate with the solution only being used for a single test cycle. The specimens were dried until a constant mass was attained, then weighed. The procedure was carried out for a total of 15 cycles. Each cycle involved the immersion of the specimens in the solution for two hours, at a temperature of 20 • C, and then drying in the oven at 105 • C for a time ranging from 10 to 15 h. After the 15th cycle, the specimens were placed in water for 24 h at 23 • C and then rinsed in water. They were then weighed, after drying to constant mass, to evaluate the mass lost during the test. The results were expressed as a percentage and obtained using the following formula: where Mf is the final mass (g); Md is the initial mass (g).

Chemical Analysis and Leaching Test
Chemical analysis and leaching tests were performed according to D.Lgs. 152/2006. These tests are not required for plaster, but they were nevertheless performed to verify the content of the metals in the sludge after the recovery.

Physical Properties
The average values obtained are reported in Table 7. Plaster regulations state that plaster can be considered a light plaster if it has a dry density less than 1300 kg/m 3 . The eco-plaster can, therefore, be defined as a light plaster. Once the cone was removed, the plaster had not lost its shape (Figure 8).
Materials 2020, 13, x FOR PEER REVIEW 11 of 16 Plaster regulations state that plaster can be considered a light plaster if it has a dry density less than 1300 kg/m 3 . The eco-plaster can, therefore, be defined as a light plaster. Once the cone was removed, the plaster had not lost its shape (Figure 8).   The results of water absorption tests showed that there was no change in the porosity of the specimens after the freeze and thaw cycles.
Comparing our plaster with some commercial lightened plasters [38,39], in terms of thermal conductivity, our plaster obtained better values: 0.302 (W/m·K) after 15 curing days and 0.201 (W/m·K) in dry condition, with respect to 0.54 (W/m·K) or 0.47 (W/m·K) of another product.

Mechanical Properties
The average values obtained by means of mechanical tests are shown in Table 8.   Plaster regulations state that plaster can be considered a light plaster if it has a dry density less than 1300 kg/m 3 . The eco-plaster can, therefore, be defined as a light plaster. Once the cone was removed, the plaster had not lost its shape (Figure 8).   The results of water absorption tests showed that there was no change in the porosity of the specimens after the freeze and thaw cycles.
Comparing our plaster with some commercial lightened plasters [38,39], in terms of thermal conductivity, our plaster obtained better values: 0.302 (W/m·K) after 15 curing days and 0.201 (W/m·K) in dry condition, with respect to 0.54 (W/m·K) or 0.47 (W/m·K) of another product.

Mechanical Properties
The average values obtained by means of mechanical tests are shown in Table 8. The results of water absorption tests showed that there was no change in the porosity of the specimens after the freeze and thaw cycles.
Comparing our plaster with some commercial lightened plasters [38,39], in terms of thermal conductivity, our plaster obtained better values: 0.302 (W/m·K) after 15 curing days and 0.201 (W/m·K) in dry condition, with respect to 0.54 (W/m·K) or 0.47 (W/m·K) of another product.

Mechanical Properties
The average values obtained by means of mechanical tests are shown in Table 8. It could be seen that there was no appreciable difference between flexural strength values before and after the freeze and thaw cycle. The excellent results obtained placed the eco-plaster in the M10 compressive strength class (in accordance with UNI EN 998-2) and in the category of compressive strength CS IV (in accordance with UNI EN 998-1). Figure 10 shows the test methodology for the pull-out test. Figure 11 shows the batched back test area and the good adhesion between plaster and panel support.  It could be seen that there was no appreciable difference between flexural strength values before and after the freeze and thaw cycle. The excellent results obtained placed the eco-plaster in the M10 compressive strength class (in accordance with UNI EN 998-2) and in the category of compressive strength CS IV (in accordance with UNI EN 998-1). Figure 10 shows the test methodology for the pull-out test. Figure 11 shows the batched back test area and the good adhesion between plaster and panel support.    It could be seen that there was no appreciable difference between flexural strength values before and after the freeze and thaw cycle. The excellent results obtained placed the eco-plaster in the M10 compressive strength class (in accordance with UNI EN 998-2) and in the category of compressive strength CS IV (in accordance with UNI EN 998-1). Figure 10 shows the test methodology for the pull-out test. Figure 11 shows the batched back test area and the good adhesion between plaster and panel support.    Table 9 shows the results of the tests for the four specimens L1, L2, T1 and T2. Samples did not show big weight and aesthetical variations after the salt cycle. Upon verification of the physical and mechanical characteristics, water absorption and flexural strength were carried out on specimens L1, L2, T1 and T2.
Flexural strength test reported good values, higher than eco-plaster, before the salt cycle results (Table 10). While imbibition testing carried out after flexural strength caused the breaking of the specimens during the saturation step ( Figure 12). This aspect is to be explored in the next study.  Table 9 shows the results of the tests for the four specimens L1, L2, T1 and T2. Samples did not show big weight and aesthetical variations after the salt cycle. Upon verification of the physical and mechanical characteristics, water absorption and flexural strength were carried out on specimens L1, L2, T1 and T2.
Flexural strength test reported good values, higher than eco-plaster, before the salt cycle results (Table 10). While imbibition testing carried out after flexural strength caused the breaking of the specimens during the saturation step ( Figure 12). This aspect is to be explored in the next study.

Chemical Analysis and Leaching Test
The results of the two tests are shown in Table 11 for chemical analysis and Table 12 for leaching test.

Chemical Analysis and Leaching Test
The results of the two tests are shown in Table 11 for chemical analysis and Table 12 for leaching test.  All metal values were below the standard limits threshold. These chemical analysis results demonstrated the possibility of recovery of this type of sludge for plaster applications, instead of landfill disposal.

Conclusions
At present, there is no market, or regulation, in favour of the recovery of silicate sawing sludge for use as a plaster. This research aimed to contribute to the re-use of this "waste" as a "by-product". The studied recoveries were foreseen on sludge without any treatment for economic advantages and to avoid landfill disposal. The recovery of sludge could turn a unidirectional system into a circular system. Thermal-eco-mortar for plaster application, sawing sludge with low metal content from Luserna stone cutting with a diamond blade, improved the rheological, thermal and physical performance, conferring a light macroporous cellular structure by means of adding an organic foam. This characteristic facilitates the plaster installation, even for high thickness. The mechanical strength and thermal conductivity results exceeded the values of standard plaster, making this eco-friendly plaster an excellent product, ideal for use in an energy-saving role in buildings and in environments with high presence of humidity. Resistance to salt crystallisation showed a breakdown of samples after the water absorption test. This feature should be investigated by comparison with the results from a normal production plaster.
The first aim for future studies is to improve the mix design used for the plaster by increasing the amount of sludge used compared to the cement content. Furthermore, the use of sludge obtained immediately after the filter press process could be possible, taking into account the water content in this condition, to avoid a possible drying step in the oven, which would lead to high energy consumption. Finally, pozzolanic cement could be added to the mix design to avoid the disintegration of the plaster due to the action of the salts.
The circular economy approach, with the decrease in costs related to the disposal, the improvement of environmental conditions and the retrieval of, still exploitable, secondary raw materials must be necessary for an era in which the quantity of waste continues to increase. A collaboration between entities that produce waste and administrations, which provide guidelines for its recovery, is necessary to make what is now a unidirectional system, a circular system.