Influence of polypropylene residues on the physico-mechanical and water-resistance properties of gypsum plasters

Journal of Cleaner Production 371 (2022) 133674 Available online 20 August 2022 0959-6526/© 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Influence of polypropylene residues on the physico-mechanical and water-resistance properties of gypsum plasters M.I. Romero-G´ omez a , * , R.V. Silva b , ** , I. Flores-Colen b , J. de Brito b a Departamento de Construcciones Arquitect´ onicas 1, Escuela T´ ecnica Superior de Arquitectura, Universidad de Sevilla, Avenida Reina Mercedes, No. 2, 41012, Sevilla, Spain b CERIS, DECivil, Instituto Superior T´ ecnico, Universidade de Lisboa, Av. Rovisco Pais, 1049-001, Lisboa, Portugal ARTICLE INFO Handling Editor: M.T. Moreira Keywords: Polypropylene waste Gypsum plaster Mechanical performance Water-resistance properties Eco-friendly material ABSTRACT The main objective of the present research is to determine the feasibility of using polypropylene (PP) waste to produce eco-friendly gypsum-based composites. Four different replacement levels of plastic waste were selected for analysis - 2.5%, 5%, 7.5% and 10% by weight of gypsum. The first stage of the experimental campaign consisted of analysing the resulting composites based on density, mechanical performance (flexural and compressive strength) and surface hardness (Shore C). In the second stage, samples were evaluated based on their water absorption by capillary action or exposing them to continuous moisture in a wet chamber, and water resistance by subjecting them to several wetting-drying cycles. After that, scanning electron microscopy (SEM) analyses were carried out to evaluate the quality of the cohesion at the interfacial transition zone (ITZ) between the PP waste particles and the gypsum matrix. The results showed equivalent compressive strength and even a slight improvement (up to ~9% when using 2.5% of PP waste when compared to the control material) most likely due to the greater compactness from the improved particle size distribution. In addition, a considerable reduction in water absorption and retention capacity was observed, due to the increasing incorporation of recycled PP into the gypsum matrix. The new PP-containing composites subjected to water absorption by capillary action and continuous moisture tests presented lower water absorption and adsorption capacity, wherein specimens with 10% of PP waste showed a decline of up to ~37% when compared to the reference material. It was generally observed that the waste material can be considered as a viable partial substitute to the current commercial gypsum for the production of gypsum plasterboards. 1. Introduction In recent years, the consumption of disposable plastic has been increased exponentially. According to recent findings put forward by the Minderoo Foundation (2019), 130 million tonnes of single-use plastics were discarded throughout the world in 2019. The uncontrolled production of this type of plastic is leading to a considerable widespread problem, due to its low recycling ratio, as almost all of it is burnt, accumulating in landfills or illegally discarded directly into the environment. Among the different sources of plastic waste that are causing this problem, disposable coffee capsules (CCs) are one of those presenting an alarming increasing consumption trend and, at the same time, with very limited solutions for their sustainable management (The Business Research Company, 2021). About 35 thousand million of CCs are thrown away each year worldwide, corresponding to approximately 576.000 t of global waste (Life-Plan4 Project Coffee, 2018; Marinello et al., 2021). Diversity of flavours, affordability and speed joined in a very simple system to use, are the main characteristics that promote the use of one-dose CCs instead of making coffee traditionally. According to the Spanish National Statistics Institute, 2.829.744 CC units were consumed in Spain in 2020. This corresponds to an average of 60.3 units per person (INE, 2021). The Directive 2019/904 has been established by the European Parliament to limit the use of some single-use plastic products starting from 2021 (European Union. Directive (EU) 2019/904), but disposable CCs are not included in the legislation, which increases the need for looking for new recycling alternatives for this type of waste. Indeed, this waste presents a discouraging recycling ratio; barely 10% of aluminium-containing CCs are recycled in Spain, whilst plastic-based do not have specific treatment methods within the * Corresponding author. ** Corresponding author. E-mail address: [email protected] (M.I. Romero-G´ omez). Contents lists available at ScienceDirect Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro https://doi.org/10.1016/j.jclepro.2022.133674 Received 1 December 2021; Received in revised form 8 August 2022; Accepted 14 August 2022 Journal of Cleaner Production 371 (2022) 133674 2 standard recycling system currently followed by municipal entities (´ Alvarez, 2019). The large majority of plastic-based CCs are composed of polypropylene (PP) and aluminium, apart from the coffee itself, which makes it difficult to carry out adequate mechanical recycling methods. For this reason, their destination is usually common landfills (Domingues et al., 2020). One of the possible solutions for this unrecyclable waste could be the availability of consumers to separate the different components of the CCs before discarding them. However, it should be the initiative of the original producers to manufacture a product that can be easily used and recycled at the end of its life, instead of relying on consumers to take on this task. Alternatively, some companies have launched a collecting campaign to recycle PP-CCs, by providing collection points, to produce new plastic products for the company or urban furniture (NESCAF´ E Dolce Gusto, 2021). Taking into account that PP is one major plastic that is used to produce single-use products and the high impact of disposable CCs, several research studies have been developed in order to analyse their recycling possibilities and evaluate their degradation during different processes of recycling (Campos de Bomfim et al., 2019). In the study of Domingues et al. (2020), a proposal was presented for a manual separation approach of all the components (body, lid, aluminium filter and plastic filter) of a disposable CC. Similarly, Campos de Bomfim et al. (2021) used a manual separation process, followed by cleaning of the plastic components of expresso CCs, to make a homemade compost product using the waste coffee grounds. However, a more industrialized recycling process should be developed to achieve a viable solution to the growing environmental problem produced by single-use plastics. Given the volume of materials within the construction sector’s supply chain, it could offer an adequate solution for the aforementioned waste, because of the fairly high versatility of many of the already commercialized products (Plastic Europe, 2020). Several researchers have focused their efforts on the development of new building materials incorporating different types of waste plastic (e.g. Polyethylene Terephthalate-PET; Polycarbonate-PC; Extruded Polystyrene-XPS; Expanded Polystyrene-EPS; Polyurethane Foam-PUR) as a partial replacement for raw material. Apart from offering a second life to this waste, they occasionally improve some of the properties of cementitious composites. Concerning this issue, Kim et al. (2021) quantitatively evaluated the economic and environmental effects in the replacement of natural sand with ocean-borne plastic flakes up to 20% by volume, to develop cement composites. Also, the use of PET waste as partial substitute for sand in concrete was examined by Almeshal et al. (2020), achieving a reduction in unit weight in spite of the mechanical strength and fire resistance of concrete decreasing. On the other hand, the incorpoation of PUR waste to gypsum mortar mixes was studied by Vicario et al. (2020), leading to porous and lighter mortars, with enhanced resistance to water-vapour permeability, but lower mechanical strength and higher levels of absorbency. Within existing research, PET, PE and PP are the main types of plastic waste used in such composites as: recycled reinforcement fibres to improve the mechanical properties of the composites (Araya-Letelier et al., 2019); an aggregate replacement to achieve lightweight materials or to improve their water-resistance, thermal or acoustic behaviour (Coppola et al., 2018). Some significant works in this line were published by Ohemeng and Ekolu (2019), who developed a strength prediction model for cement mortar with different type of plastic waste as fine aggregate. Likewise, other researchers proposed using chemically pre-treated plastic waste as aggregates in cement mortar mixes, getting a more porous matrix to the detriment of the mechanical properties, as the plastic dosage increased (Kaur and Pavia, 2021). Some studies on gypsum composites incorporating recycled plastic have been developed. Pedre˜ no-Rojas et al. (2019) studied the influence of crushed PC waste from shredded CDs and DVDs on the performance of gypsum composites. The authors assessed the use of PC with different particle sizes at different replacement levels up to 60% by weight of gypsum, wherein a ~15% density decrease was observed when compared to the control material. Despite the overall negative impact on the mechanical performance, improved behaviour in terms of flexural strength had been observed for mixes with up to 10% by weight of smaller PC particles (<4 mm). In addition, enhanced compressive strength was observed for mixes with up to 40% by weight of PC in comparison with the control specimens. Other interesting works have been published by the same authors, Pedre˜ no-Rojas et al. (2020), based on the mixture of two different types of waste, recycled gypsum and PC waste, to developed new eco-friendly gypsum-based materials, entirely recycled, wich improved flexural and compressive strength (up to ~15% and ~27% increase, respectively), as well as lower thermal conductivity (~43% decrease), related to the reference material. Romero-G´ omez et al. (2021) analysed the influence of adding PP-based reinforcement fibres, sourced from single-use wet-wipes, on the performance of gypsum composites. Level replacements up to 3.5% by weight of gypsum binder were used. The results showed that there was an adequate adhesion between the plaster and the recycled PP fibres resulting in a 19% increase in flexural strength for mixes with 2.5% of the waste by weight of gypsum. Interesting findings were also observed in the study of Vidales-Barriguete et al. (2020) regarding the water-resistance of plastic-containing gypsum composites. The addition of plastic cable waste (particle size >3 mm) up to 70% in weight of gypsum led to a significant decrease in water absorption and retention capacity, whilst keeping their mechanical properties within the required standardized limits. From a thermal behaviour point of view, some studies have been published, wherein those developed by and Guti´ errez-Gonz´ alez et al. (2012) and Villela et al. (2020) presenting encouraging findings. Both of them show a decrease of the thermal conductivity of gypsum composites containing recycled polyamide powder and crushed multilayer packaging, respectively. Concerning the use of PP waste as replacement in conventional mortars, Rubio-De Hita et al. (2018) analysed the influence of adding mixed PP plastic as partial substitute for sand. The composites with a content of 12.5% PP by weight of natural aggregate were selected as the optimal, as they exhibited the lowest density and maintained a mechanical performance within the standardized minimum requirements, with marginal changes to workability. Martínez-L´ opez et al. (2021) studied the impact of using PP and polyethylene (PE) from waste bottles on the mechanical properties of mortars. Different levels of sand replacement (1%, 2%, 3%, by weight) and particle sizes (0.7 mm; 1.4 mm; 2.38 mm) were tested. The results showed that the mixes with 1% of 0.71 mm-sized particles from both Abbreviations list CoV Coefficient of variation CC Coffee capsule E Modulus of elasticity EPS Expanded polystyrene EU European Union G Gypsum ITZ Interfacial Transition Zone PC Polycarbonate PE Polyethylene PET Polyethylene terephthalate PP Polypropylene PUR Polyurethane foam RH Relative humidity SEM Scanning electron microscopy W/G Water/gypsum XPS Extruded polystyrene M.I. Romero-G´ omez et al. Journal of Cleaner Production 371 (2022) 133674 3 types of waste led to an increase up to 27% in compressive strength when compared with the control specimens. Chen et al. (2019) studied an approach to reprocess PP waste (2 mm-sized pellets) from recycling plants as partial substitution of sand to make cement mortar composites. The authors observed decreasing density and improved toughness after replacing the sand up to 60% by volume. Nevertheless, the composite also showed worse mechanical performance as well as increased open porosity, water absorption and shrinkage, when compared with the control mortars. Even though some studies have been carried out on plasticcontaining gypsum/cement-based composites, none were found on the use of recycled waste PP granules from discarded organic mattercontaminated PP-based packaging waste as partial replacement of the binder to produce new eco-friendly construction materials. For this reason, this research presents innovative results of an experimental campaign analysing the behaviour of gypsum composites containing PP waste particles sourced from disposable CCs. Despite the relatively low amount of discarded plastic CCs in comparison with the large amount of undifferentiated plastic waste, this waste was chosen to be representative of a real waste material rather than using pure and uncontaminated PP. Additionally, since this product is showing a growing consumption trend over the last few years, its proportional environmental impact should not be neglected. Thus, the main purpose of this work is analysing the influence of PP waste granules on the physical (i.e. dry state density and surface hardness), mechanical (i.e. dynamic Young’s modulus, flexural and compressive strengths) and water-resistance properties of new gypsum composites according to standard EN 13279-2 (2014). The quality of the interfacial transition zone (ITZ) between the paste and PP particles was also evaluated. Concurrently, this study intended to carry out an approach to separately recycle the different components of CCs, focusing on the plastic fraction. 2. Materials and methods 2.1. Materials The gypsum used in this study was controlled setting plaster for construction (B1), according to standard EN 13279-1 (2009). Its properties are shown in Table 1. The water used for production was regular tap water in compliance with Council Directive 98/83/EC (European Union, 1998). The PP waste was sourced from used CCs, collected from different collection points around the Municipality of Almargen, Malaga, Spain. The composition of the CCs is shown in Fig. 1, as previously analysed by other researchers who used the same type of capsules in their research (Domingues et al., 2020). The mechanical performance and water absorption of these capsules are presented in Table 2. The real density of the PP components was found to be between 0.895 and 0.920 g/cm 3 , and 2.7 g/cm 3 for aluminium particles, by applying the pycnometer method. After the collection campaign, the residual organic matter (e.g. coffee, chocolate or tea powder) was manually removed from the capsules to simulate an eventual beneficiation step for this waste. In this cleaning step, the capsules were washed with water and air-dried. After that, the capsules were crushed in LAPIST lab of CERENA research unit by using a RETSCH SM 2000 cutting mill (Fig. 2), with an output of particles measuring up to ~4 mm (particle size distribution of the ground PP waste in Fig. 3). This material contained a mix of PP and aluminium. To adequately separate and recycle each of the materials, the mixed particles were submerged in water for 1 h to allow a density-based separation. After scrapping the PP granules from the surface, they were dried at 100 ◦C ±2 ◦C to constant mass. 2.2. Sample preparation The guidelines established in EN 13279-2 (2014) were followed for the preparation of samples. According to the standard, a water/gypsum (W/G) ratio of 0.55 was fixed for all samples. The mixes were produced with the following procedure: the gypsum powder and PP waste granules were dry-mixed by hand to obtain a homogenous mix and avoid any potential segregation during the mixing process; after that, water was added to the mix and the components were mixed manually for 2 min to a homogenous state. To keep the workability constant for all mixes without the need to adjust the W/G ratio or add plasticizers, the content of PP waste was limited to 10% by weight of gypsum. The resulting composites were placed in moulds producing 40 ×40 ×160 mm 3 prismatic specimens. The material proportions used for each of the mixes are summarized in Table 3. The mix code is the following: gypsum (G); type of waste used - polypropylene (PP); percentage of plastic waste addition (e.g. G/PP/2.5). After demoulding the specimens, they were placed in a dry chamber at a temperature at 24 ◦C and relative humidity (RH) of 50 ±1% to cure for seven days. Afterwards, they were placed in an oven at 40 ±2 ◦C, until constant mass and cooled to laboratory temperature in a desiccator. 2.3. Test methods 2.3.1. Fresh state properties - Setting time: according to procedure set by standard EN 13279-2 (2014) for pre-mixes gypsums, the influence of PP waste on gypsum mixes’ setting time was measured with Vicat needle method. Table 1 Properties and limits of B1 gypsum used as binder according to standard EN 13279-1 (2009) (Provided by manufacturer). Purity [%] Particle size [mm] Surface hardness [Shore C] Performance [kg/m 2 /cm thickness] Flexural strength [N/ mm 2 ] Compressive strength [N/ mm 2 ] Adherence [N/ mm 2 ] pH >75 0–1 ≥45 10–12 ≥2 ≥2 >0.1 >6 Fig. 1. Composition of coffee capsules (adapted from Domingues et al., 2020). Table 2 Mechanical properties of the plastic components of a disposable coffee capsule - Tensile strength ( σ ), Young’s Modulus (E), Yield point ( σ f ), ductility ( ε ), toughness (u t ) and resilience (u r ) (Domingues et al., 2020). Sample σ max [MPa] E [MPa] σ 1.max [MPa] ϵ [%] u t [J/ m 3 ] u r [J/ m 3 ] Body +plastic filter +lid 29.5 ± 1.4 655 ± 9 25.8 ± 1.5 120.3 ± 20.6 17.1 ±2.9 2.1 ±0.7 M.I. Romero-G´ omez et al. Journal of Cleaner Production 371 (2022) 133674 4 2.3.2. Physical and mechanical properties - Dry bulk density: after measuring the dry weight and volume of the test samples, as per standard EN 13279-2 (2014), the dry density of each specimen was then calculated as the relation between its dry mass and its volume. The final data of new composites density were obtained by arithmetic mean of the six measurements performed per replacement level. - Superficial hardness (Shore C): this property was determined for three samples per replacement level, prior to any destructive mechanical tests. Following the procedure in EN 13279-2 (2014), a durometer was used to measure the Shore C hardness in two longitudinal and parallel faces of each prismatic sample (Fig. 4a). A perpendicular force to each specimen was applied, choosing a minimum of five points placed in the central plane and separated at least 20 mm from the edges of the sample. Finally, the superficial hardness value of the compounds was expressed as the mean value of 30 measurements obtained per replacement level. - Flexural strength: based on the procedure explained in EN 13279-2 (2014), the pieces were supported in the Form and Test Seidner Strength testing machine-D7940, centred on rollers distanced by 100 mm (Fig. 4b). After that, a load was applied in the central plane of the sample with a speed of 10 N/s, until the specimen’s failure. For each of the specimens, the value of flexural strength was calculated by the arithmetic mean of the six measurements per replacement level. - Compressive strength: the 12 broken sections of samples obtained from the previous test were subjected to the compressive strength test, according to standard EN 13279-2 (2014). The section of the same testing equipment previously described, was used to apply a progressive centred load with a speed of 20 N/s to each of the 12 pieces per replacement level, after which the average value was calculated. - Dynamic Young’s modulus: this test was carried out according to the standard ASTM E1876-21 (2021), by using the GrindoSonic MK5 equipment to measure the fundamental resonant frequency of the test samples. In order to achieve this, the specimens were mechanically excited by a singular elastic strike using an impulse tool. The main value of the dynamic Young’s modulus of the new composites was calculated by using the appropriate fundamental resonant frequencies, dimensions and mass of the specimens. - Ultrasonic pulse velocity: carried out in accordance with the standard EN 12504-4 (2006), this non-destructive test was applied to check the relationship between the density and porosity of the compounds. By using the Proceq Pundit Lab +equipment connected to a pair of probes, an emitter and another receiver, with a frequency of 54 kHz, one longitudinal and three transverse measurements were carried out to six specimens per replacement level, whereby the average value then calculated (Fig. 4c). - Water absorption by capillary action: as determined by standard RILEM TC 25-PEM (1980), after the curing period, three prismatic specimens per replacement level were placed in a vertical position inside a container, with a distance of 10 ±1 mm of water from the bottom. Throughout a 10-min testing period, the water level in the specimens’ surface was measured every minute. The final value was expressed in mm/min (Fig. 5a). - Hygroscopicity range: this non-standardized test, designed by Del Rio-Merino (1999) in her thesis, consisted of subjecting prismatic specimens to constant moisture, by placing them in a wet chamber for 5 day at a temperature of 21 ±2 ◦C and a RH of 72 ±5% (high RH value over the healthy humidity range 40–60%, without achieving the dew point) (Fig. 5b). After that, the weight gain of the different samples was measured and then placed on a table in a Fig. 2. Selective recycling procedure: a) RETSCH SM 2000 mill; b) density-based separation of PP from aluminium; c) Recycled PP, crushed and ready to be used in mixes. Fig. 3. Particle size distribution of shredded PP waste from coffee capsules. Table 3 Mix proportions for six specimens of 40 ×40 ×160 mm 3 . Mix code Gypsum [g] Water [g] W/G ratio PP waste [g] G/CM 1800 990 0.55 – G/PP/2.5 1755 965 0.55 45 G/PP/5 1710 941 0.55 90 G/PP/7.5 1665 916 0.55 135 G/PP/10 1620 891 0.55 180 M.I. Romero-G´ omez et al. Journal of Cleaner Production 371 (2022) 133674 5 laboratory environment for 7 days. Finally, three specimens per replacement level underwent the surface hardness (Shore C), flexural and compressive strength tests with the purpose of comparing the results with those of the specimens with constant moisture. - Deterioration from water-oven cycles: water absorption and drying capacity, as well as the loss of mechanical strength due to the deterioration of compounds after being saturated and dried for a short period at standard environmental conditions, was evaluated based on the non-standardized procedure carried out by Del Rio-Merino (1999). Specimens were placed in a container, distanced from the bottom, and covered with water for 2 days (Fig. 5c). Afterwards, the water was removed, and the weight of the specimens measured. The specimens were then placed in an oven for 2 days at 40 ±5 ◦C. Three specimens per replacement level were evaluated based on their surface hardness (Shore C), flexural and compressive strengths after the first cycle, while the rest of the samples was subjected to two additional cycles of the aforementioned procedure. These tests were developed in the Construction Laboratory of CERIS (Civil Engineering Research and Innovation for Sustainability) at the Instituto Superior T´ ecnico of Lisbon, Portugal. 2.3.3. Other test methods - Scanning electron microscopy (SEM): this method was used to check the quality of the ITZ between the gypsum matrix and PP particles in order to clarify the behaviour of the materials developed from a microscopic level. Two different samples were analysed and compared: one of them without plastic addition (G/CM) and the other with a 7.5% PP waste content (G/PP/7.5). Due to the low conductivity of the samples studied, the specimens were coated with gold to improve the output’s quality. To carry out this analysis, the FEI TENEO field emission SEM belonging to the General Research Services area of the University of Seville, Spain, was employed. 3. Results and Discussion After evaluating the physico-mechanical properties of the proposed PP-containing composites in the fresh and hardened state, the results are presented and discussed in this section. Next, the water-resistance behaviour of the new materials after mixing a waterproofing waste (polypropylene) as partial replacement in the gypsum matrix, as well as how this affects its hygrothermal properties, are evaluated. 3.1. Setting time According to the data summarized in Table 4, the setting time of gypsum composites decreased as the PP waste content added to the mixes increased. Higher setting time reduction (−33.75%) compared to the reference material was presented by mixes with a 10% PP waste addition. This performance as previously analysed by other researchers (Vidales-Barriguete et al., 2018). Fig. 4. Physical-mechanical tests: a) superficial hardness (Shore C); b) flexural strength; c) ultrasonic pulse velocity (longitudinal measurement). Fig. 5. Water exposure tests: a) water absorption by capillary action; b) hygroscopicity range; c) deterioration from water-oven cycles. Table 4 Setting time of PP-containing composites (min). Specimens Time [min] Δ time compared to reference sample G/CM 20.00 – G/PP/2.5 16.00 −20.00% G/PP/5 15.25 −23.75% G/PP/7.5 14.00 −30.00% G/PP/10 13.25 −33.75% M.I. Romero-G´ omez et al. Journal of Cleaner Production 371 (2022) 133674 6 3.2. Dry bulk density The results of the specimens’ density and modulus of elasticity are presented in Table 5. Unexpected values of dry bulk density were obtained. Despite the PP’s lower density when compared to gypsum (around 1000 kg/m 3 vs. 2650 kg/m 3 , respectively), the new composites showed higher density values, peaking at 2.5% PP content (4% increase) and decreasing linearly after that. This behaviour could be attributed to the improved particle size distribution of the components leading to densification of the mix when recycled PP particles under 2 mm are incorporated. However, the declining linear trend observed after exceeding 2.5% of recycled PP, pointed out that this is an optimum content, after which the compactness started to deteriorate. Vidales-Barriguete, (2019, 2021) also observed decreasing density values with increasing plastic cable waste content to the binder matrix with replacement levels of up to 70% by weight of gypsum. In this study, it was observed that the reference sample presented the highest volume of porosity, which decreased with the increasing percentage of plastic cable waste. This suggests an inverse relationship between the porosity of the material and the bulk density was demonstrated, which implies a reduction of porosity and bulk density due to the waste’s impervious nature and lower density. This fact was also found in another previous research (Ahmed et al., 2011; Flores-Medina and Barbero-Barrera, 2017). Furthermore, a linear reduction of Young’s modulus was noticed in both, the work developed by Vidales-Barriguete (2019) and the present study, which pointed out a loss of the mixes’ stiffness of up to ~50% and ~19% when plastic cable waste and PP waste were added, respectively, compared to the reference composite. 3.3. Mechanical properties Concerning the flexural strength behaviour of the novel composites, the results are shown in Fig. 6. As expected, the incorporation of recycled PP particles had a negative influence on the gypsum matrix, leading to a decline in flexural strength of up to ~35% (G/PP/10) with respect to the control. However, it should be noted that, in all cases, the values were over the minimum 1 MPa set by EN 13279-2 (2014), thus showing the applicability of the waste despite the decline in performance. The previous findings of the present study are consistent with results in the literatures (Rubio-De Hita et al., 2018; Ohemeng and Ekolu, 2019), which showed major reductions in flexural strengths of mortars, as the incorporation of plastics waste to the mixes increased. Fig. 7 shows the cross-sections of two specimens with different PP contents (i.e. G/CM and G/PP/7.5) after the flexural strength test. It shows a random and homogenous arrangement of the PP particles throughout the gypsum matrix. The compressive strength showed an unexpected trend (Fig. 8); it increased when PP waste was incorporated. The highest compressive strength was achieved by mixes with a 2.5% PP content (i.e. 9.19 MPa), after which a slight downwards linear trend was observed for subsequent replacement levels (G/PP/2.5, G/PP/5 and G/PP/7.5 presented values of 8.42, 8.36 and 8.24 MPa, respectively). Moreover, considering that the minimum requirement is of 2 MPa for gypsum binders and plasters (EN 13279-2, 2014), all mixes were well within the limit. The Table 5 Summary table with the results obtained for density and modulus of elasticity tests developed in this work. Specimens Density [kg/m 3 ] (CoV [%]) E [MPa] (CoV [%]) G/CM 1204 (1.58) 6139 (0.45) G/PP/2.5 1251 (0.49) 5771 (3.56) G/PP/5 1240 (0.73) 5595 (2.50) G/PP/7.5 1236 (0.70) 5183 (3.89) G/PP/10 1230 (0.72) 4975 (3.20) Fig. 6. Flexural strength (MPa) of gypsum/polypropylene composites. Fig. 7. Cross-section of the (a) G/CM and the (b) G/PP/10. Fig. 8. Compressive strength (MPa) of gypsum/polypropylene composites. M.I. Romero-G´ omez et al. Journal of Cleaner Production 371 (2022) 133674 7 aforementioned behaviour presented by the new PP-containing composites is supported by published works. Martinez-L´ opez et al. (2021) noticed improvements of about 27% in compressive strength when plastic waste (i.e. PP and PE) replacement ratios up to 1% by weight of sand and a particle size lower than 0.71 mm were used. However, after exceeding this point, the values of compressive strength decreased gradually when increasing of either the size or concentration of the waste particles. Similarly, Pedre˜ no-Rojas et al. (2019) observed the effectiveness of using granulated PC waste with a particle size smaller than 4 mm to improve the compressive strength behaviour in samples up to 40% plastic waste added (~37% increase with 5% of PC waste addition). Moreover, after carrying out X-ray computed tomography test, these authors found that the compressive strength increase was due to the reduction of the studied mixes’ porosity when plastic waste particles smaller than 4 mm were used. 3.4. Superficial hardness (Shore C) The superficial hardness of gypsum materials is a property strongly associated with the compressive strength. The specimens showed improved superficial hardness up to 1.36% and all close to 95 (corresponding to mix G/PP/5) when compared with the control gypsum (Fig. 9). Thus, it can be inferred that, to improve the superficial hardness of gypsum-based composites, the optimal replacement level range of PP waste is of 2.5–7.5%, similar to that observed in the compressive strength test. 3.5. Scanning electron microscopy The SEM method was applied to a specimen with 7.5% PP content (G/PP/7.5) and to the control (Fig. 10). The image revealed a good involvement of PP waste particles throughout the gypsum matrix, thus explaining the positive results previously observed. The shredding process of waste CCs resulted in PP particles with reasonably rough surfaces that counteract the hydrophobic property of PP surface, thereby providing additional sites for gypsum crystal precipitation and growth. Nevertheless, despite the acceptable link between the gypsum matrix and the PP particles, the cohesion of the ITZ is not as strong as that in between the products of hydration and is likely to be a point of failure, especially given the different moduli of elasticity of both materials. According to the reference material, it can be confirmed that no alteration in the nature of the crystalline network of plaster was noticed after Fig. 9. Surface Hardness (Shore C) of gypsum/polypropylene composites. Fig. 10. Stereo microscopy image to G/PP/7.5 (a-20x); SEM analysis to G/PP/ 7.5 (b - 150 ×; c - 600 ×) and G/CM (d - 600 ×). M.I. Romero-G´ omez et al. Journal of Cleaner Production 371 (2022) 133674 8 adding PP waste to the mixes. Other studies showed similar results (Pedre˜ no-Rojas et al., 2019; Vidales-Barriguete, 2019). 3.6. Ultrasonic pulse velocity In order to study the relationship between the density and mechanical performance of the composites, ultrasonic pulse velocity was studied. The results (Fig. 11a) show the same trend as that observed in Section 3.1; the densification of the mixes with the incorporation of 2.5% PP led to an increase in velocity, followed by a steady decrease after that with increasing PP content. A comparison of the ultrasonic pulse velocity values and the corresponding densities showed a strong linear relationship (Fig. 11b). On the other hand, according to the behaviour developed by the new composites after they were subjected to ultrasonic velocity test, Fig. 11a shows that the new gypsum compounds could be classified into two groups. Mixes that incorporated a PP waste content between 2.5 and 5% and those that used a percentage equal to or greater than 7.5%. The first group presented similar values of longitudinal and cross ultrasound pulse velocities, which suggests that, to get a good compactness of the material and thus better control of the material’s properties, the maximum ratio of PP addition should be no more than 5%. From this point, the more pronounced variability between the values of the longitudinal and cross ultrasound velocities infers that the higher content of PP particles added to the mixes would cause greater porosity, loss of cohesion and a weakened gypsum matrix. 3.7. Water absorption by capillary action The results of the water absorption by capillary action are presented in Table 6. The water uptake decreased by around 25%, 26%, 34% and 36% for composites with 2.5%, 5%, 7.5% and 10% PP content, respectively, when compared to the control specimens. Regarding to the average water absorption by capillary action (Fig. 12), the lowest value was that of G/PP/10, (i.e. 5.5 mm/min). Since the incorporation of PP particles did not cause a porous ITZ, as observed in the SEM, their increasing presence imparts an impervious property thereby leading to the slower advance of water. 3.8. Hygroscopicity range The results from the process of exposing the samples to continuous moderate humidity until 72% RH are presented in Table 7. The hygroscopicity of the gypsum composites was determined based on the weight increase after exposure. The water adsorption capacity was lower as the amount of PP particles increased. After subjecting the G/PP/10 composite to the test, it presented the lowest values for water adsorption, about 38% lower than the control specimens. This trend is equivalent to that observed in the previous properties. As the incorporation of an impervious material increased, the composites’ resistance to water movement also increased. 3.9. Deterioration from wetting-drying cycles The results of the samples subjected to wetting-drying cycles are summarized in Table 8. All PP-containing composites presented lower water absorption capacity when compared to the control (~29% regardless of PP content vs. 33% of the control). The water absorption capacity fell by 12.9% and 10.1% after one and two cycles, respectively, when compared to the control mixes. In addition, it should be noted that the water absorption in mixes with 2.5%, 5%, 7.5% and 10% PP, after two cycles of water saturation, decreased by 5.63%, 5.23%, 2.70% and 3.56% compared with the same values from the first cycle. Fig. 13 shows the loss of gypsum material presented by the sample subjected to the wetting-drying cycles test. No changes on samples dimensions were perceived after being subjected to this test. 3.10. Comparative analysis of performance before and after water exposure In order to evaluate the influence of water exposure on the mechanical behaviour of the new gypsum compounds, a comparison was carried out between the values of Shore C hardness, flexural and compressive strength tests obtained before and after exposure to water (Figs. 14–16). A decrease of surface Shore C hardness was noticed after subjecting samples to water tests. Those specimens tested to wettingdrying cycles presented a decline in this property between 0.7% and 3%, while the samples exposed to continuous moisture showed values similar to the control mixes without exposure. This could be explained by the surface layer deterioration suffered by the samples after undergoing water saturation and drying cycles. Regarding the flexural strength behaviour of the composites, mixes with a PP waste content lower than 5% showed a decline between 15% and 20% (regardless of the water test) when compared to the samples without exposure to water. Specimens with higher PP content that underwent the hygroscopicity range test did not show relevant changes to flexural strength behaviour when compared to the original unaffected samples. In contrast, specimens with more than 5% PP content showed reduced flexural strength by approximately 19%, when exposed to continuous wetting and drying. Nevertheless, all specimens presented values well above the minimum 1 MPa indicated in standard EN 13279-2 (2014). The improvement in compressive strength of PP-containing samples maintained even after the water exposure tests (Fig. 16). Still, this Fig. 11. Ultrasonic pulse velocity of (a) composites with increasing PP content and (b) correlation with density. M.I. Romero-G´ omez et al. Journal of Cleaner Production 371 (2022) 133674 9 property declined by 9–17% in all composites subjected to the hygroscopicity range test and by 11–19% in those that underwent the wettingdrying cycles. It should be noted that the maximum decrease in compressive strength of up to 25% was exhibited by the control sample without PP waste. Again, it should be noted that all values obtained after the water exposure test exceeded the minimum 2 MPa requirement (EN 13279-2, 2014). 4. Conclusions After carrying out the complete characterization of gypsum composites containing recycled polypropylene (PP), obtained from disposable CCs, in accordance with the standard requirements established by EN 13279-2 (2014), the following conclusions were drawn: – The workability of the gypsum mixes was unaffected by the use of PP waste addition up to 10%, which allowed a constant W/G ratio without the use of water reducing admixtures; – The addition of PP waste with a particle size lower than 4 mm to gypsum matrix, increased densification of the PP-containing gypsum composite when compared with the control mix. In addition, it has been demonstrated that, from a 2.5% PP content, which presented maximum density, there was a downwards linear trend of this property with increasing PP content; – The compressive strength of the PP-containing composites improved by ~9% when 2.5% PP was used but showed a linear decline for higher levels. Even though the flexural strength decreased between ~15% (2.5% PP) and ~30% (10% PP), all values were over the minimum standard requirements; – The SEM analysis showed adequate involvement of recycled PP particles within the gypsum matrix. The rough surface of the PP particles due to the crushing process facilitated the homogeneous integration of the gypsum matrix, resulting in an acceptable adhesion on ITZ; – The use of impervious PP waste in gypsum mixes led to a reduction of the water absorption and adsorption capacity. This decrease was more noticeable with increasing PP content. Concerning the mechanical behaviour of the samples after subjecting them to water exposure tests, apart from the acceptable decline in performance in all mixes, the values of flexural and compressive strengths were above the minimum indicated in the standard. Nevertheless, more Table 6 Capillary water absorption values of new gypsum composites, measured for 10 min. Specimens Min. 1 (mm) Min. 2 (mm) Min. 3 (mm) Min. 4 (mm) Min. 5 (mm) Min. 6 (mm) Min. 7 (mm) Min. 8 (mm) Min. 9 (mm) Min.10 (mm) Total (mm) Δ Capillary absorption compared to ref. G/CM 39.0 6.3 7.3 5.3 6.0 5.0 3.3 6.0 2.7 4.3 85.3 – G/PP/2.5 29.7 4.7 4.7 4.3 3.3 4.0 3.7 3.3 3.0 3.0 63.7 −25.39% G/PP/5 25.7 4.7 6.3 6.0 2.0 5.3 3.0 4.3 3.0 2.7 63.0 −26.17% G/PP/7.5 23.3 5.7 4.7 3.3 4.0 4.0 2.7 2.7 3.3 2.3 56.0 −34.38% G/PP/10 25.0 4.3 3.7 4.7 4.0 2.0 3.7 2.7 2.7 2.3 55.0 −35.55% Fig. 12. Average water uptake by capillarity of the Series II composites. Table 7 Hygroscopicity of the composites exposed to moderate humidity. Sample series Weight before the test-dry (g) Weight after the test- 72% RH (g) Water adsorption (%) G/CM 311.81 314.18 +0.87% G/PP/2.5 316.17 318.03 +0.59% G/PP/5 317.63 319.53 +0.60% G/PP/7.5 312.53 314.28 +0.56% G/PP/10 307.57 309.25 +0.54% Table 8 Values of water absorption of PP-gypsum composites after 1 and 2 wettingdrying cycles. Cycle no. Mix Initial weightdry (g) Weight after saturation (g) Water absorption (%) Weight after drying (g) Δ weight compared to initial weight 1 G/ CM 297.40 395.60 +33.02% 334.57 +12.50% G/ PP/ 2.5 316.73 410.87 +29.72% 341.15 +7.71% G/ PP/ 5 314.80 407.55 +29.46% 340.48 +8.16% G/ PP/ 7.5 310.92 398.98 +28.32% 335.65 +7.95% G/ PP/ 10 309.07 398.00 +28.77% 324.57 +7.82% 2 G/ CM 297.40 389.20 +30.87% 306.78 +3.15% G/ PP/ 2.5 316.73 405.57 +28.05% 315.17 −0.49% G/ PP/ 5 314.80 402.70 +27.92% 312.37 −0.77% G/ PP/ 7.5 310.92 396.60 +27.56% 307.90 −0.97% G/ PP/ 10 309.07 394.83 +27.75% 305.50 −1.16% M.I. Romero-G´ omez et al.