Lightweight geopolymer-based hybrid materials

Accepted Manuscript Lightweight geopolymer-based hybrid materials Giuseppina Roviello, Costantino Menna, Oreste Tarallo, Laura Ricciotti, Francesco Messina, Claudio Ferone, Domenico Asprone, Raffaele Cioffi PII: S1359-8368(17)31039-9 DOI: 10.1016/j.compositesb.2017.07.020 Reference: JCOMB 5155 To appear in: Composites Part B Received Date: 22 March 2017 Revised Date: 14 June 2017 Accepted Date: 13 July 2017 Please cite this article as: Roviello G, Menna C, Tarallo O, Ricciotti L, Messina F, Ferone C, Asprone D, Cioffi R, Lightweight geopolymer-based hybrid materials, Composites Part B (2017), doi: 10.1016/ j.compositesb.2017.07.020. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. ACCEPTED MANUSCRIPT Lightweight geopolymer-based hybrid materials Giuseppina Rovielloa,b, Costantino Mennac, Oreste Tarallod, Laura Ricciottic, Francesco Messinaa,b, Claudio Ferone,a,b, Domenico Asproncc, Raffaele Cioffi,a,b Dipartimento di Ingegneria, Università di Napoli ‘Parthenope’, Centro Direzionale, Isola C4, 80143 Napoli, Italy b INSTM Research Group Napoli Parthenope, National Consortium for Science and Technology of Materials, Via G. RI PT a Giusti, 9 50121 Firenze (ITALY) c Dipartimento di Strutture per l’Ingegneria e l’Architettura, Università di Napoli Federico II, Napoli 80125, Italy; d Dipartimento di Scienze Chimiche, Università degli Studi di Napoli “Federico II”, Complesso Universitario di Monte * M AN US C S. Angelo, via Cintia, 80126 Napoli, Italy Corresponding Author: tel. +39-081-5476781; fax: +39-081-5476777; e-mail address: [email protected] (G. Roviello) _______________________________________________________________________________ TE D Graphical abstract G-Ep AC C EP G-Sil 1 ACCEPTED MANUSCRIPT Lightweight geopolymer-based hybrid materials Giuseppina Rovielloa,b, Costantino Mennac, Oreste Tarallod, Laura Ricciottic, Francesco Messinaa,b, Claudio Ferone,a,b, Domenico Aspronec, Raffaele Cioffi,a,b Dipartimento di Ingegneria, Università di Napoli ‘Parthenope’, Centro Direzionale, Isola C4, 80143 Napoli, Italy b INSTM Research Group Napoli Parthenope, National Consortium for Science and Technology of Materials, Via G. RI PT a Giusti, 9 50121 Firenze (ITALY) c Dipartimento di Strutture per l’Ingegneria e l’Architettura, Università di Napoli Federico II, Napoli 80125, Italy; d Dipartimento di Scienze Chimiche, Università degli Studi di Napoli “Federico II”, Complesso Universitario di Monte M AN US C S. Angelo, via Cintia, 80126 Napoli, Italy Abstract The present study reports on the preparation and characterization of new organic-inorganic geopolymer based hybrid foams obtained by reacting an aluminosilicate source and an alkalisilicate solution with mixtures of dialkylsiloxane oligomers or organic resins precursors. By using different amounts of Si0 powder as in situ foaming agent, hybrid geopolymer-based foams with densities D ranging from 0.25 to 0.85 g/cm3 were successfully prepared. These new materials are characterized by remarkable mechanical properties, good fire resistance and low thermal conductivity, TE significantly better than those shown by neat geopolymer foams reported in the literature and comparable or even better than those of typical (not geopolymer) inorganic foamed materials with 1. Introduction EP similar densities. AC C Porous materials are widely used in several applications including membranes, high-efficiency adsorption materials, catalysis, as well as in the construction industry where they find their application as insulating or sound-proof panels or for the production of lightweight structural components. To these aims, organic polymers, metals or inorganic materials can be successfully used as source raw material for the foamed products. In particular, thanks to their low mass, low heat conductivity and good sound-proofing properties, organic polymer foams are particularly suitable as insulating components and are extensively used in civil engineering. However, most polymeric systems suffer from thermal degradation under high temperatures, affecting their structural integrity and releasing toxic or flammable gases into the atmosphere. ACCEPTED MANUSCRIPT In the case of high temperature applications, as an alternative to polymeric foams, inorganic porous materials can be used. Such materials are usually divided in two main classes, relying on the specific use. One class comprises materials obtained by employing lightweight aggregates while the other one is represented by highly porous materials. Several studies have been carried out about lightweight and eco-compatible materials/aggregates [1-5] employing cold bonding techniques but RI PT the obtained densities (even less than 2000 kg/m3 defining lightweight aggregates) are, however, still considerable. Slightly lower values have been obtained by employing thermal processes [6] but such processes typically imply a high environmental impact (production of CO2). Among inorganic porous materials, great attention has been devoted to porous geopolymers due to i) the possibility of M AN US C using low cost raw materials [7,8], ii) their chemical resistance [9], iii) their good thermal properties [10] and iv) their environmental friendly nature [11,12]. The term ‘‘geopolymer’’ was originally introduced by Davidovits [13] to describe the amorphous inorganic aluminosilicate framework produced by reacting natural Si- and Al-rich materials, such as metakaolin or industrial by-products (fly ash, blast furnace slag), with highly alkaline aqueous solutions. Geopolymers exhibit a wide variety of technologically relevant properties, such as high compressive strength [14], fire resistance and low shrinkage [15]. For these reasons, geopolymers seem to be a desirable alternative to ordinary Portland cement also thanks to their environmentally D sustainable characteristics [16] mainly associated with the reduced CO2 emissions arising from the production of the raw materials from which they are obtained [17]. TE Geopolymers are produced by means of a polycondensation reaction (the so called “geopolymerization”). During this reaction, a gel network is formed, consisting of SiO4 and AlO4 EP tetrahedra sharing oxygen corners and forming rings of various sizes analogous to those found in zeolites. If foaming agents are added into the geopolymer paste during its consolidation, porous materials can be obtained. Well know blowing agents are hydrogen peroxide, metallic Al or Si AC C powder [18-20]. By following this synthesis approach, porous materials with pore size ranging from nanometers to few millimetres and a total porosity up to 90% were obtained [21,22] without using high temperature treatments (such as burn out of organics and sintering) that are, by contrast, necessary for the production of porous ceramics through conventional techniques [23]. In general, the geopolymer-based porous materials developed to date exhibit very interesting properties in terms of thermal and acoustic conductivity but the presence of pores and the extremely heterogeneous structure affect unavoidably their mechanical properties, sometimes resulting in low compressive strengths (found to be about 1 MPa or less depending on the corresponding apparent density) [20]. In detail, consolidated foams characterized by thermal conductivity values of approximately 0.15 Wm-1K-1 or less were successfully obtained with the addition of silica fume or ACCEPTED MANUSCRIPT silicon powder [21] as a foaming agent to sodium silicate and kaolin [24] or potassium silicate based [25] geopolymers. In the cited studies, the authors exploited the capability of free silicon contained inside the silica fume to generate porosity by releasing molecular hydrogen from silicon oxidization in alkaline solution, achieving significant percentages of pore volume (> 60% ) [26]. Porous geopolymer materials were also produced starting from fly ash precursors and in RI PT combination with hydrogen peroxide [27] or sodium perborate [28] as foaming agent, resulting in consolidated foams with a porosity of ≈ 80 %, thermal conductivity of ≈ 0.08 Wm-1K-1 and compressive strengths ranging between 0.80 and 0.40 MPa. Higher compressive strengths were recorded in the case of fly ash based geopolymer foams with bulk densities in the range of 400–800 M AN US C kg/m3 and obtained with the use of aluminium powder as blowing agent [29] . Furthermore, in the case of geopolymer and fly ash foam concretes, the higher densities of the obtained materials (typically from 720 to 1600 kg/m3) led to very high compressive strengths ranging from 3 to 48 MPa, with thermal conductivity values in the range of 0.15–0.48 Wm-1K-1 [30] which are reasonably low for concrete applications. Recently we succeeded in obtaining organic-inorganic hybrid materials by reacting an aluminosilicate source and an aqueous alkali hydroxide and/or alkalisilicate solution with mixtures of dialkylsiloxane oligomers or organic resins precursors [31-38]. Compared to neat geopolymers D with analogous Si/Al ratio, these materials are characterized by enhanced mechanical properties, along with good temperature and fire resistance [39-42]. TE In this study, for the first time, we report on the synthesis and characterization of high performance organic-inorganic hybrid foams obtained by in situ foaming of the hybrid materials described before EP through the addition of silicon powder. These foams show remarkable mechanical properties, good fire resistance and low thermal conductivity, significantly enhanced in respect to those AC C characterizing neat geopolymer foams reported in the literature and comparable, or even better, than those of typical (not geopolymeric) inorganic foamed materials with similar densities. Considering that commonly used organic insulating materials are flammable while inorganic ones need complex processing conditions and/or high sintering temperatures (increasing the manufacturing cost), the hybrid porous geopolymer-based materials described in this study have good application potential as an effective alternative for thermal insulation or fire-resistant sealant materials. 2. Experimental Section 2.1. Materials ACCEPTED MANUSCRIPT Metakaolin was kindly provided by Neuchem S.r.l. (Milan, Italy) and its composition is reported in Table 1. Sodium hydroxide with reagent grade, was supplied by Sigma-Aldrich. The sodium silicate solution was supplied by Prochin Italia S.r.l (Caserta, Italy). with the composition reported in Table 1. A commercial oligomeric dimethylsiloxane mixture was purchased from Globalchimica S.r.l (Turin, Italy) with the name of Globasil AL20. The epoxy resin used in this paper, called Epojet®, RI PT was purchased by Mapei S.p.A (Milan, Italy) [43]: it is a commercial two-component epoxy adhesive for injection, which, after the mixing, takes the aspect of a low viscosity liquid and it is usable for 40 minutes at room temperature. Silicon powder ~325 mesh was purchased from Sigma- M AN US C Aldrich. Additional experimental details are reported in references 32 and 33. Table 1. Chemical composition (weight %) of the metakaolin and sodium silicate solution used in this paper. Metakaolin Al2O3 SiO2 K2O Fe2O3 TiO2 MgO CaO others 41.90 0.77 1.60 0.19 0.17 0.67 52.90 Sodium silicate solution Na2O H2O 27.40 8.15 64.45 EP 2.2 Specimen Preparation TE D SiO2 1.80 2.2.1. Geopolymer (G-MK) AC C The alkaline activating solution was prepared by dissolving solid sodium hydroxide into the sodium silicate solution. The solution was then allowed to equilibrate and cool for 24 h. The composition of the solution can be expressed as Na2O 1.4SiO2 10.5H2O. Metakaolin was then incorporated into the activating solution with a liquid to solid ratio of 1.4:1 by weight, and mixed by a mechanical mixer for 10 min at 800 rpm. As revealed by EDS analysis on the cured samples, the composition of the whole geopolymeric system can be expressed as Al2O3 3.5SiO2 1.0Na2O 10.5H2O, corresponding to a complete geopolymerization process. Neat geopolymer sample was indicated as G-MK. 2.2.2. Foamed Geopolymer Composites (G-Ep) Geopolymer-based foamed composites were obtained by adding 10% by weight of Epojet® resin to the freshly-prepared geopolymeric suspension, and quickly incorporated by controlled mixing (5 ACCEPTED MANUSCRIPT min at 1350 rpm) [32]. Before being added to the geopolymeric mixture, Epojet® was cured at room temperature for 10 minutes, when it was still easily workable and long before its complete crosslinking and hardening (that takes place in about 5–7 hours at 23 °C). In order to obtain a smooth but effective foaming process, silicon powder was then added as foaming agent with different wt% ratios, ranging between 0.03 and 0.24%; afterwards, the system RI PT was mixed for further 5 min at 1000 rpm. The consolidated geopolymer/organic resin samples obtained through the above mentioned procedure are hereafter indicated as G-EpXX, where XX refers to the decimal units by weight (percentage) of silicon foaming agent added to the geopolymer composite paste (e.g. G-Ep03 refers to a Geopolymer-Epojet composite with 0.03% by weight of Si M AN US C content; G-Ep12 corresponds to 0.12% wt% of Si content). The composite sample that did not undergo the foaming process was indicated as G-Ep. The mix design details of G-Ep specimens are reported in Table 2. 2.2.3. Foamed Geopolymer Hybrids (G-Sil) Hybrid polysiloxane-geopolymer foamed samples were prepared by incorporating 10% by weight of a commercial oligomeric dimethylsiloxane mixture into the freshly prepared geopolymeric suspension under mechanical stirring, when the polycondensation reaction of both geopolymer and D dimethylsiloxane were already started but far to be completed. In order to obtain a set of samples with a different degree of porosity, silicon powder was added to the geopolymer-hybrid suspension TE as foaming agent in different wt% ratios, ranging between 0.03 and 0.24%; then, the system was mixed for further 5 min at 1000 rpm. These samples are hereafter indicated as G-SilXX, where XX EP refers to the decimal units by weight (percentage) of silicon foaming agent added to the geopolymer paste (e.g. G-Sil03 refers to a polysiloxane-geopolymer hybrid sample with 0.03% by weight of Si content while G-Sil12 corresponds to 0.12% wt% of Si content). The hybrid sample that did not AC C undergo the foaming process was indicated as G-Sil. The mix design details of G-Sil specimens are reported in Table 2. Table 2. Mix composition (wt%), apparent density and open porosity of the studied samples. Sample MK SS NaOH Epoxy Resin G-MK 41.55 50 8.45 G-Ep 37.4 45.0 7.6 10 G-Ep03 37.4 45.0 7.6 10 Open DMS Si porosity (%) - 0.03 Apparent density (g cm–3) 15% 1.524 22% 1.425 38% 0.860 ACCEPTED MANUSCRIPT 37.4 45.0 7.6 10 - 0.06 41% 0.629 G-Ep12 37.4 45.0 7.6 10 - 0.12 48% 0.500 G-Ep14 37.4 45.0 7.6 10 - 0.14 49% 0.431 G-Ep18 37.4 45.0 7.6 10 - 0.18 50% 0.363 G-Ep24 37.4 45.0 7.6 10 - 0.24 52% 0.312 G-Sil 37.4 45.0 7.6 - 10 7% 1.339 G-Sil03 37.4 45.0 7.6 - 10 0.03 27% G-Sil06 37.4 45.0 7.6 - 10 0.06 41% G-Sil12 37.4 45.0 7.6 - 10 0.12 46% G-Sil14 37.4 45.0 7.6 - 10 0.14 54%% G-Sil18 37.4 45.0 7.6 - GSil24 37.4 45.0 7.6 - M AN US C RI PT G-Ep06 0.701 0.508 0.396 0.365 10 0.18 60% 0.320 10 0.24 75% 0.252 MK = metakaolin; SS = sodium silicate solution; epoxy resin = Epojet® resin; DMS = oligomeric dimethylsiloxane mixture; Si = silicon powder 2.2.4. Curing treatments D As soon as prepared, all the specimens were casted in cubic molds and cured in >95% relative humidity conditions at room temperature (≈ 22°C) for 24 h and then at 60 C for further 24 h. TE Subsequently, the specimens were kept at room temperature for further 5 days in >95% relative AC C 2.3 Methods EP humidity conditions and then for further 21 days in air. 2.3.1 Physical testing and microstructure SEM analysis was carried out by means of a Nova NanoSem 450 FEI Microscope. Hydrostatic weighing for apparent density and open porosity measurements was carried out by means of a balance OHAUS-PA213 provided by Pioneer. Average pores diameter was determined by optical image analysis of polished surfaces. The analysis was performed by means of an in-house routine developed using a commercial software package (Matlab R2015a) [44]. 2.3.2 Rheological measurement ACCEPTED MANUSCRIPT Simple flow measurements related to pastes can be performed by means of the minislump cone test [36]. This testing technique represents a simple procedure that is employed quite frequently in literature since it allows making both qualitative and quantitative observations on fresh slurries. However, the geometry of the experimental setup is not standardized and varies in a wide range [45-49]. In this work, the geometry of the minislump cone was had the following dimensions: top RI PT diameter equal to 5.0 cm; bottom diameter equal to 6.8 cm; height equal to 6.5 cm (corresponding to a paste volume equal to 179.1 cm3). Freshly prepared slurries were slowly poured in the cone placed on a horizontal plane and carefully compacted by means of a thin steel rod. Slump orthogonal diameters were measured and the average diameter was used to calculate minislump M AN US C area. Mini-slump measurements were repeated at 0, 15, 30, 60 and 120 minutes in order to enlighten workability loss depending on time, by keeping the fresh slurries in controlled environment (T=20±2°C; sealed container). Apparent viscosity of geopolymeric mixtures was assessed by means of a Brookfield viscometer DV2T. Test were carried out with different spindles considering the variation of rheological features of investigated mixtures. Spindles used are standardized for the indicated device and are indicated by numbers, namely n° 4, 5, 6 and 7. Additional tests were carried out in order to understand the capability of geopolymer slurries to retain liquid phase. Hence, washout tests were performed by means of a modified experimental D setup with respect to the one used in literature [50]. In particular, instead of 750 mL beakers, 500 compared to the initial mass. EP 2.3.3 Mechanical testing TE mL ones were used. Washout loss after dilution in distilled water was expressed in percentage Uniaxial compression tests were carried out according to of ASTM D 1621 on 50×50×50 mm cubic specimens by using a MTS 810 servo-hydraulic universal testing machine. For each sample type, AC C three specimens were tested under displacement control in order to obtain the corresponding stressstrain curve, compressive strength and Young’s modulus. Compression tests were performed until the sample densified and/or ruptured at a constant displacement velocity of 0.60 mm/min. The measurement of the displacement was performed through the crosshead displacement while the Young’s modulus of each sample was computed from the initial linear stress-strain response recorded during the test. All values presented in the current work are an average of three samples. 2.3.4 Thermal characterization The thermal conductivity measurement was performed in accordance with UNI EN 1745 and UNI EN 12664, by means of the Guarded Hot Plate Method on air-dry samples with conditioning at ACCEPTED MANUSCRIPT 50°C. The measurement equipment used follows ASTM E1530. For each sample examined, three specimens with diameter 50.8±0.3 mm and thick 4.3 mm at 10°C were tested. The test specimens were installed between heating and cooling plates. A constant heat flow flowed through the test specimens in the stationary temperature state. Thermal conductivity was determined by the heat flow and the temperature difference between the sample surfaces. The calibration was conductivity λ (W/mK) can be determined according to the following: = (1) M AN US C where s is the specimen thickness and Rs is the thermal resistance. RI PT performed on a series of reference samples with certified thermal characteristics. The thermal 2.3.5 Fire testing Flame tests were performed by a cone calorimeter in accordance with the procedure described in ISO5660 standard method. The heat flux produced was 50kW/m2 on the specimen, which had an exposed surface of 100×100 mm. The testing equipment consisted of a radiant electric heater in trunk-conic shape, an exhaust gas system with oxygen monitoring and instrumentation to measure the gas flux, an electric spark for ignition, and a load cell to measure the weight loss. The test was Results and discussion 3.1 Foaming process EP 3. TE D terminated after 600 s of exposure. As described in the previous section, different amounts of silicon powder were added as foaming AC C agent to the neat geopolymer slurry (G-MK) and to the two hybrid organic-geopolymer systems GEp and G-Sil. As far as G-MK slurry is concerned, due to its very low viscosity (see section 3.2), the foaming process yielded the rapid collapse of the foamed structure initially formed. For this reason, foamed G-MK samples were not considered in this discussion. On the contrary, the addition of the foaming agent to G-Ep and to G-Sil slurries produced homogeneously foamed structures. Volume expansion and density of the different G-Ep and G-Sil foamed samples obtained are reported in Figure 1. M AN US C RI PT ACCEPTED MANUSCRIPT D Figure 1. Density values (red bars) and volume expansion (black bars) of G-Ep and G-Sil foamed TE samples. Volume expansion was evaluated by the percentage ratio between the volume of the specimen after the foaming process and the starting volume of the slurry. EP As expected, the effectiveness of the foaming process was strictly dependent on the Si amount added to the G-Ep and G-Sil slurry as foaming agent. The volume expansion of the not cured slurry AC C increases with increasing the amount of the foaming agent while the density of the cured porous materials decreases as the amount of foaming agent used increases. For silicon amount ranging from 0.03% to 0.24 wt%, high resolution optical photographs of the specimens after expansion and curing, are reported in Figure 2. EP TE D M AN US C RI PT ACCEPTED MANUSCRIPT G-Sil G-Ep AC C Fig. 2: Optical images of polished section surfaces of some representative G-Sil (left column) and G-Ep (right column) specimens. By examining the morphologies shown in Figure 2, it is clearly apparent that macropores are observable in all samples and, for each composition, they are rather uniformly distributed in the specimens, resulting in good homogeneity also in terms of dimension and shape. Only a minor amount of egg-shaped pores was detected, parallel to the expansion direction, probably attributable to the shape and dimensions of the used mold. As expected, a low quantity of silicon powder added as foaming agent led to poorly expanded structures with small and regularly distributed rounded pores while high Si content produced the ACCEPTED MANUSCRIPT coalescence of the pores. Average diameters of pores were equal to 0.70, 1.33, 2.28 and 4.84 mm for G-Sil03, G-Sil06, G-Sil12 and G-Sil24, respectively while were equal to 0.57, 0.85, 1.28 and 3.13 for G-Ep03, G-Ep06, G-Ep12 and G-Ep24, respectively. Although not reported in this study, it’s worth pointing out that the addition of Si in amounts of more than 0.24 wt% caused the unstable collapse of the foamed microstructure. The morphological characterization is very useful to receive RI PT information about the micromechanical model [51-53]. By examining the morphologies of the foams by SEM micrographs (Figure 3), it is worth pointing out that, as expected, the surfaces of both samples appeared to be very different from the typical microstructure of an unfoamed geopolymer matrix [32, 33]. In particular, in the case of G-Ep series M AN US C of samples, a 3D pore structure was detected and characterized by interconnected channels passing throughout the specimen. In the case of G-Sil foams, the microstructure was characterized by isolated pores uniformly distributed within the specimens, showing good homogeneity, also in terms of dimensions and shape. This different morphology was rationalized in terms of rheological properties of the two slurries (next paragraph). As discussed in paragraph 3.4, these different microstructures of the two sets of samples turned out AC C EP TE D in different mechanical behaviours. TE D M AN US C RI PT ACCEPTED MANUSCRIPT Figure 3 SEM images at 100 (A, C) and 2000 (B, D) magnification of G-Ep03 (A, B) and G-Sil03 AC C EP (C, D). In B and D, images of the wall between large pores are shown. 3.2 Rheology Slump test results are reported in Figure 4 and refer to the G-MK, G-Sil and G-Ep slurries tested as soon as prepared and for different resting times. The system exhibiting the largest flow is G-MK, while an intermediate performance was obtained for G-Sil. At variance, G-Ep was able to keep almost unchanged the shape of the slump cone after cone lifting due to very high values of viscosity and yield stress, thus indicating a strongly different rheological performance respect other investigated materials. In terms of workability loss, the two “flowable” systems, i.e. G-MK and GSIL, exhibited a similar behaviour. Particularly, the system G-MK was able to keep its characteristic flow area for at least 30 minutes; this represents a significant parameter for most of the technological processing procedures on laboratory and industrial scale. G-Sil exhibited a faster ACCEPTED MANUSCRIPT workability loss with respect to G-MK, since after flow measured after 30 minutes was about 10% of initial slump value. G-Ep did not experience any significant change over time since the starting value of mini slump area, which was substantially similar to the bottom area of the cone, was kept almost constant for all subsequent measurements. 500 RI PT G-MK G-Sil G-Ep 450 350 300 250 200 150 100 50 0 20 40 M AN US C 2 Slump Area (cm ) 400 60 80 100 120 D t (min) TE Figure 4. Variation of slump area with time for G-MK (black squares), G-Sil (red circles) and G-Ep EP (blue triangles) slurries. Lines are a guide for the eyes. Apparent viscosity measurements related to G-MK, G-Sil and G-Ep slurries are reported in Figure 5. Measurements were performed with different rotating spindles in order to provide further AC C information and ensure a better reproducibility of the results. The possibility of using several spindles depends on the viscosity of the investigated suspension: for relatively high viscosity values, the number of suitable spindles is limited. In Figure 5, it is evident that G-Sil is characterized by higher viscosity than G-MK since the number of suitable spindles (spindle n°5 and spindle n°6) is limited with respect to those that are suitable for G-MK (spindles n°4, n°5 and n°6). In the torque range investigated, the two mixtures exhibited a very different rheological behaviour. First, for G-MK, a plateau of apparent viscosity was clearly detected, while for G-Sil, a slight decrease was detected as a function of torque growth. ACCEPTED MANUSCRIPT G-MK slurry, spindle n°4 G-MK slurry, spindle n°5 G-MK slurry, spindle n°6 G-Sil slurry, spindle n°5 G-Sil slurry, spindle n°6 G-Ep slurry, spindle n°7 5 8x10 5 5 6x10 4 RI PT 1.8x10 4 1.6x10 4 1.4x10 4 1.2x10 4 1.0x10 3 8.0x10 3 6.0x10 3 4.0x10 3 2.0x10 0.0 M AN US C Apparent Viscosity (cP) 7x10 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 Torque (%) Figure 5. Apparent viscosity of G-MK (squares), G-Sil (circles) and G-Ep (triangles) slurries as D measured with the spindle indicated. Lines are a guide for the eyes. TE With regard to the apparent viscosity related to G-Ep slurry, it must be pointed out that measurement was strongly influenced by the applied shear regime. Hence, a significant variation of apparent viscosity was experienced depending on different torque applied. In the same range of EP investigated torque for other systems, G-Ep slurry exhibited an apparent viscosity equal to 6.8·105±0.4·105 cP. The presence of a plateau viscosity value was shifted at higher torque (torque > AC C 80%). The apparent viscosity values are evidently much higher than in the other cases corresponding to G-MK and G-Sil slurries. This fact corresponds also to the restrained choice in terms of suitable spindles, which was limited only to spindle n°7. Finally, Figure 6 shows washout loss values for the three investigated slurries: for all the systems, a highly promising water retention capability was detected thus indicating a good processability of the material before its hardening. ACCEPTED MANUSCRIPT 12 8 6 4 2 0 G-Sil G-Ep M AN US C G-MK RI PT Washout loss (%) 10 Figure 6. Washout loss values for G-MK, G-Sil, G-Ep as determined on their freshly prepared slurries. The different rheological behaviours herein discussed allowed for a possible interpretation of the different foaming performances observe for the three investigated systems. Actually, we can infer that the neat geopolymer prepared in this study was not able to develop a properly foamed D microstructure, due to the very low viscosity of the slurry. The addition of silicone or epoxy resin to TE the geopolymer slurry turned out in a significant increase of the viscosity of the system that allows the development of stable pores, thus resulting in a properly foamed microstructure. Moreover, the different rheological behaviour of the G-Sil and G-Ep systems is probably connected with the EP different microstructure of the two samples [32, 33]. In fact, G-Ep is a geopolymer based composite system containing discrete microspheres of epoxy resin (mean diameter 10-20 µm), homogenously AC C dispersed in the geopolymer matrix (see Figure 3B). At variance, G-Sil is a hybrid geopolymer material obtained through the chemical bonding between silicon atoms of the geopolymer structure with the silicon atoms of the dimethylsiloxane oligomer moieties, characterized by a single phase, with compact and continuous microstructure, up to nanometric level (see Figure 3D) [32, 33]. 3.4. Physical and Mechanical Properties Physical and mechanical properties of foamed hybrid and composite geopolymers are herein discussed in terms of material apparent density and uniaxial compressive behaviour, including compressive strength and Young’s modulus results. ACCEPTED MANUSCRIPT The average apparent densities of G-Sil (blue symbols) and G-Ep (red symbols) foams are reported in Figure 7 as a function of the wt% of Si powder added to the geopolymer slurry during foaming procedure. The two foamed materials exhibited a decrease of bulk density with increasing the Si content in the range 0.03-0.24 wt%, following a 2nd order polynomial law. The obtained trends suggested that lower densities were achieved as a consequence of the coalescence of individual cells RI PT of roughly constant smaller pore size (see Figures 2 and 3) into large voids. However, G-Sil foamed material was able to reach lower density values than G-Ep foamed material when using an equal content of foaming agent. In particular, G-Sil foamed material exhibited an average apparent density varying between 0.25 and 0.70 g/cm3 whereas, in the case of G-Ep foamed material, the M AN US C density values varied between 0.31 and 0.86 g/cm3. In other words, within the density range obtained with the same amount of foaming agent, G-Sil foam was averagely less dense by 20%. This feature is mainly related to the rheological proprieties of the two different initial materials which affected cell nucleation, growth and coarsening mechanisms taking place during the foaming process. Indeed, as reported in the previous section, the G-Ep system was characterized by very high values of viscosity and yield stress. During the expansion process, the gas volume in the growing cell increases driven by the gas pressure generating inside the cell itself, until the pressure inside the cell equals to its surrounding pressure, which is a function of surface tension and D viscosity of the fresh matrix material. Because of greater viscosity of G-Ep material, a lower surface area is required to make the cell in equilibrium with its surrounding matrix which turns in smaller TE cell sizes of the foamed material (Figure 2). Furthermore, the dense spheres of epoxy resin forming the heterogeneous microstructure of G-Ep system may have acted as physical obstacle for cell EP growth phase. On the contrary, in the case of G-Sil material, the lower viscosity of the system allowed for a larger volume growth, which led to a decrease of the stability of the foam, causing, in AC C turn, individual cells to coalesce producing larger voids (Figure 2). ACCEPTED MANUSCRIPT 1.00 G-Sil G-Ep 0.60 0.40 0.20 0.00 0.10 0.20 0.30 M AN US C % wt of Si RI PT Density [g/cm3] 0.80 Figure 7: Average apparent densities (over three samples) of G-Sil (blue symbols) and G-Ep (red symbols) foamed materials as a function of the foaming agent content (wt% of Si powder). Because of the large number of specimens tested, only the stress-strain curves of the samples that revealed a representative mechanical performance were reported in Figure 8. The stress–strain curves for each tested foam type were found to be similar kind. In details, all the investigated foamed materials exhibited a well defined elastic regime, which was noticeable at the early stages of D stress (stage 1). The linear elastic regime remained with an almost constant slope until reaching the TE yield or unstable collapse point (stage 2) characterized by a sharp load loss on the stress strain-curve up to a plateau or softening region in the case of G-Sil and G-Ep foam, respectively (stage 3). In EP stage 3, the compression stress was almost constant with strain for higher densities of G-Sil foams (i.e. for density values of 0.50 and 0.70 g/cm3) as the cells deformed plastically; in those cases, after the yield stress was reached, the specimen continued to carry 3 MPa of stress up to relatively large AC C strains, i.e. 25% of axial deformation. On the contrary, G-Ep foams showed a softening behaviour in stage 3 up to lower ultimate strains compared to G-Sil foams. The recorded discrepancy was probably due to the heterogeneous microstructure [32] of G-Ep matrix which was not able to accommodate the increased local deformation associated with the progressive collapse mechanism of the cells. A further region of rapidly increasing load (densification region, stage 4) was recorded only in the case of higher density G-Sil foams. ACCEPTED MANUSCRIPT G-Sil06 12 G-Sil12 G-Sil24 9 6 3 0 0.00 0.05 0.10 0.15 0.20 0.25 G-Ep03 G-Ep06 12 Stress [MPa] Stress [MPa] 15 G-Sil03 0.30 Strain [mm/mm] G-Ep12 G-Ep24 9 6 3 0 0.00 0.05 RI PT 15 0.10 0.15 0.20 0.25 0.30 Strain [mm/mm] (b) M AN US C (a) Figure 8: Stress-strain curves in compression of G-Sil (a) and G-Ep (b) samples. The compressive strength of tested foamed materials was calculated from the maximum load applied to the specimens (corresponding to the yield or unstable collapse point - stage 2) while the Young's modulus under compression was derived from the slope of the initial linear region of the stress-strain diagram. The average values of compressive strength and Young’s modulus (computed from measurements on three samples) are shown in Figure 9a and Figure 9b for G-Sil (blue D symbols) and G-Ep (red symbols) foamed materials as a function of the relative density ρ*/ρs, i.e. the ratio between the actual apparent density of the foam, ρ*, and the density of the fully dense neat TE G-Sil and G-Ep materials, ρs, equal to 1.34 and 1.42 g/cm3, respectively. For both foam material types, the overall results showed that the foaming process was able to produce satisfactory EP mechanical performances in reference to similar materials [20, 22, 30]. In particular, considering that average compressive strengths of fully dense neat G-Sil and G-Ep materials was 65 and 41 AC C MPa, respectively [33], the corresponding compressive strengths of foamed samples varied between 11 and 0.67 MPa in the density range of 0.25-0.86 g/cm3 (corresponding to 0.19 and 0.60 of relative density, respectively). As general tendency, G-Sil foamed material developed higher compressive strengths than G-Ep foam type in the density range of 0.45-0.70 g/cm3 (0.34-0.52 ρ*/ρs). The greatest difference was estimated in correspondence of a density value 0.70 g/cm3 and was equal to 47%. Below the density threshold of 0.45 g/cm3 (corresponding to a relative density between 0.350.40), this tendency was reversed in favour of G-Ep foams which exhibited, for instance, a compressive strength 40% greater than G-Sil foam in the case of 0.30 g/cm3 of density (1.62 vs 1.15 MPa, respectively). This behaviour can be correlated to the deformation response of the two different cellular materials which is strongly influenced by corresponding value of relative density. At higher densities, the mechanical performance of the foamed material under compression ACCEPTED MANUSCRIPT typically approaches the compressive behaviour of the full dense material. Indeed, as reported in a previous study by the authors [27], in the full dense state, G-Sil material has averagely improved mechanical performances (either stiffness and strength) compared to G-Ep systems. On the other side, at low densities, compressive failure is typically caused by tensile failure of the bent cell walls of a given diameter. In this respect, full dense G-Sil was found to behave as a brittle material under RI PT failure [33] while G-Ep exhibited a moderate toughening mechanism (with regard to fracture) due to the presence of the discrete resin particles. Consequently, the tougher G-Ep was able to accommodate larger tensile strains under bending loads and, in turn, provide a more effective resistant mechanism to failure in compression at lower densities. M AN US C Stress-strain curves obtained by compressive tests allowed determining also the Young’s modulus in compression of all samples (Figure 8b). The elastic stiffness of the two foamed materials showed a similar dependency on the density values, resulting in Young’s modulus values approximately ranging between 100 and 1400 MPa in the density interval of 0.25-0.86 g/cm3 (corresponding to 0.19 and 0.60 of relative density, respectively). Moreover, the overall trend of the results as a function of material density, resulted very similar to the one observed for the compressive strength, i.e. with an inversion point in correspondence of a material density of approximately 0.45 g/cm3 (corresponding to a relative density between 0.35-0.40). In particular, for densities below 0.45 D g/cm3, the Young's modulus of the foams is largely controlled by the volume fraction of large coalesced voids. In the case of G-Sil foam, brittle micro-cracking of bent coalesced voids may act TE as an additional contribution to deformation within the foam matrix, decreasing the initial stiffness. For densities above 0.45 g/cm3, the volume fraction of large coalesced voids is small and the 9 6 3 Analytical 3 2 1 0 0.14 0 0.10 Young's Modulus [MPa] 12 G-Ep EP G-Sil AC C Compressive Strength [MPa] Young's modulus of the G-Sil foam is controlled by the stiffer matrix compared with G-Ep one. 0.30 0.26 0.50 Relative Density, ρ*/ρs [-] a) G-Sil 1000 450 500 300 150 0 0.14 0.38 0.70 Analytical G-Ep 1500 0 0.10 0.30 0.50 0.26 0.38 0.70 Relative Density, ρ*/ρs [-] b) ACCEPTED MANUSCRIPT Figure 9: Average (over three samples) compressive strengths (a) and Young’s modulus (b) for GSil (blue symbols) and G-Ep (red symbols) foamed materials as a function of the relative density. It’s worth noting that within the density range 0.20-0.45 g/cm3 both foamed materials exhibited interesting compression properties, especially with reference to that density interval which RI PT represents an important range for technological applications of lightweight materials [23] In terms of analytical interpretation of the results, we refer to the classical relationships between the strength (eq. 1) or Young’s modulus (eq. 2) of a cellular material and its relative density proposed  ρ * σ c* = C1   s σc  ρs  M AN US C by Gibson-Ashby [54]: m  ρ * Ec* = C2   s Ec  ρs  (1) n (2) where σ c* , Ec* and ρ * are the compressive or yield strength, the Young’s modulus (in D compression) and the relative density of the foam material, respectively; σ c , Ec and ρs are the TE corresponding properties referred to the fully dense solid of which the foam is made or, in an equivalent manner, the properties of the solid cell wall material; C1, C2 are dimensionless constants EP and the exponents m, n depend on the cell morphology and can be established both by experiment and by numerical computation. These relationships can be successfully used to model the mechanical properties of most foams, whether open or closed cell, with a bending-dominated AC C deformation behaviour. Dotted lines in Fig. 9 a,b represent the best fit of equations (1) and (2), the parameters of which are reported in Table 3. Table 3: Parameters of data fitting reported in Fig. 9 a,b (dotted lines) according to equations (1) and (2). Compressive Strength G-Ep-foam Young's modulus C1 [-] 0.79 C2 [-] 1.52 m [-] 2.05 n [-] 1.95 σ cs [MPa] [27] 40.8 Ecs [MPa] [27] 2600 ACCEPTED MANUSCRIPT G-Sil-Foam C1 [-] 1.15 C2 [-] 2.35 m [-] 2.85 n [-] 2.90 σ cs [MPa] [27] 62.0 Ecs [MPa] [27] 4160 As discussed in this section, the G-Sil and G-Ep foams behave as a linear elastic material up to the RI PT elastic limit under compression, at which point the cell edges failure takes place. The variations of Young’s modulus with the relative density follows equation (2) with the exponent n equal to 2 (GEp) or 3 (G-Sil), in agreement with previous results dealing with inorganic (cementitious) foams [55]; the different values of the exponent, n, between the two foamed materials can be related to the M AN US C elastic properties of the corresponding fully dense solid ones. A similar analysis can be conducted for the variation of the compressive strength with the relative density (equation (1)). The exponents m obtained from the best fit of the experimental data resulted higher that the ones suggested by Gibson-Ashby [54] (i.e. m = 3/2) probably due to several reasons. Primarily, the possibility of having mixed closed and open cells along with the presence of pores in most of the cell walls may have affected the failure initiation under compression deformation. Secondly, the hybrid nature of the foams’ microstructure may act as positive stabilization for the collapse limit especially at higher densities, for which the contribution of axial and shear stresses in D the cell wall is dominant. TE 3.3 Thermal conductivity Thermal conductivity tests were carried out on G-Sil12 and G-Ep12 samples. Table 4 shows the Sample Density (kg/m3) λ10, dry (W/mK) G-Ep12_1 480.1 0.101 ± 0.002 G-Ep12_2 500.5 0.103± 0.002 G-Ep12_3 523.6 0.105± 0.002 G-Sil12_1 396.2 0.101 ± 0.002 G-Sil12_2 395.5 0.105± 0.002 G-Sil12_3 398.0 0.103± 0.002 AC C values. EP values of the volumetric density (kg/m3) for each examined specimens and the corresponding λ10,dry Table 4: Density and λ10, dry of three different specimens of G-Ep12 and G-Sil12. ACCEPTED MANUSCRIPT In both cases, the values of thermal conductivity are in the range 0.101-0.105 W/mK. Figure 10 shows the relation between the experimental thermal conductivity and the measured density for G-Ep12 and G-Sil12 specimens. These results are compared with the values reported in Tables A of EN 12664 for some representative lightweight materials used for masonry products. Each line refers to 50% fractiles for each category of materials. As far as G-Sil12, the results RI PT obtained are in the same range with those found for reference materials with the same density, while in the case of G-Ep12 the conductivity values obtained are far lower than 50% of the values TE D M AN US C reported in the reference materials with the same density thus indicating better insulating properties. Figure 10: Relation between λ10, dry and density for G-Ep12 (black square); G-Sil12 (black circle); EP masonry mortars, Tab.A12 (red square); autoclaved aerated concrete, Table A10 (blue circle); concrete with polystyrene aggregates, Table A5 (magenta triangle); concrete with expanded clay AC C aggregates, Table A6 (green rhombus). All the straight lines represent P = 50% of the data. 3.2.5. Flame Tests In order to obtain information on the combustion behaviour of the investigated materials under ventilated conditions, fire resistance tests were performed on G-Ep18 and G-Sil12 (i.e. samples showing similar apparent density values, see Table 2). The optical images of these samples before and after the cone calorimeter test are reported in Figure 11. M AN US C RI PT ACCEPTED MANUSCRIPT Figure 11 Images of G-Ep18 (A, B) and G-Sil12 (C, D) before (A, C) and after (B, D) the cone TE D calorimeter test. The cone calorimeter showed that these specimens did not ignite, burn or release appreciable smoke EP even after extended heat flux exposure. Moreover, the heat release rate (HRR), that represents the contribution in terms of heat released by the material in case of fire, was minimal and also the CO, CO2 and the produced fumes were of a negligible amount. In particular, G-Sil12 sample showed specimen. AC C HRR and THR values (see Table 5), CO and CO2 production (see Figure 12) lower than G-Ep18 Therefore, the foamed organic-inorganic materials described in the present paper can be considered as not-flammable, with a not significant production of toxic fumes and smokes. Thus they can efficiently replace the neat geopolymers in all the applications where exposure to flames and lightness are desired. Table 5 summarizes the main outcomes of cone calorimeter tests conducted on the investigated foamed specimens. ACCEPTED MANUSCRIPT HRR (peak) HRR (60 s) HRR (180 s) THR (kW/m2) (kW/m2) (kW/m2) (MJ/m2) G-Ep18 16.6 3.9 9.4 6.0 G-Sil12 7.4 2.1 1.4 1.1 RI PT Samples Table 5 Flame tests results of G-Ep18 and G-Sil12, HRR= heat release rate, THR= total heat AC C EP TE D M AN US C release. AC C EP TE D M AN US C RI PT ACCEPTED MANUSCRIPT Figure 12 Flame test results: (A) HRR-Heat Release Rate (kW/m2) vs time (s); (B) THR-Total Heat Release (MJ/m2) vs time; (C) SPR-Smoke production rate (m2/s) vs time (s); (D) CO production (%) vs time (s); (E) CO2 production (%) vs time (s) for G-Ep18 and G-Sil12. 4. Conclusions ACCEPTED MANUSCRIPT Two different types of novel organic-inorganic geopolymer based foams were successfully manufactured by mixing metakaolin and an aqueous alkalisilicate solution with mixtures of dialkylsiloxane oligomers or organic resins precursors, respectively. The density of the obtained materials (ranging from 0.25 to 0.85 g/cm3) was tailored by simply controlling the amounts of silicon powder added to the slurry as in situ foaming agent. RI PT In both cases, the addition of the organic components or of the dialkylsiloxane oligomers, turned out in a significant change of the intrinsic viscosity of the geopolymeric slurry, thus allowing the obtainment of a homogeneous and very regular foaming process. In particular, in the case of the composite foamed materials obtained by adding the organic resin precursor to the geopolymeric M AN US C slurry (G-Ep), the very high viscosity of the fresh mixture (in respect to neat geopolymeric one) has allowed obtaining a microstructure characterized by pore cells with a mean diameter up to 3 mm, in which, the dense spheres of epoxy resin forming the heterogeneous microstructure of the cell walls could have likely acted as physical hindrance for cell growth phase. At variance, in the case of GSil foams, the lower viscosity of the slurry allowed for a larger volume growth of the cells, which led to a decrease of the stability of the foam, causing, in turn, individual cells to coalesce thus producing larger voids. These new porous materials, despite their cellular morphology, are still characterized by interesting mechanical properties. In particular, it is worth pointing out that, for both foamed materials, average D compressive strengths up 5 MPa were recorded within the density range 0.2-0.5 g/cm3, which TE represents an important range for technological applications of lightweight materials. Moreover, within the same density range, both G-Sil and G-Ep foams exhibit good fire resistance EP and low thermal conductivity with λ values in the range 0.101-0.105 W/mK. These properties are significantly better than those shown by neat geopolymer foams reported in the literature and comparable or even better than that of typical (not geopolymeric) inorganic foamed AC C materials with similar densities and could be related to the hybrid nature of the materials used. In fact, as already shown in the case of their dense form, the synergistic effect of the organic and inorganic components of the developed materials has a beneficial effect on their properties. For these reasons, it can be reasonably concluded that these new hybrid materials could represent a valid alternative to commonly used inorganic foams (e.g. Portland Cement foams) for insulation and lightweight applications, since they combine performance benefits and operational energy savings. Acknowledgement ACCEPTED MANUSCRIPT The authors thank Neuvendis S.p.A. for the metakaolin supply and Prochin Italia S.r.l. for the silicate solution supply. Dr. Giovanni Morieri and Dr. Luciana Cimino are warmly acknowledged for assistance in laboratory activities. Università di Napoli “Parthenope” is acknowledged for financial support to research activities with a grant within the call “Support for Individual Research RI PT for the 2015-17 Period” issued by Rectoral Decree no. 727/2015. References AC C EP TE D M AN US C [1] Colangelo, F., Messina, F., Di Palma, L., & Cioffi, R. Recycling of non-metallic automotive shredder residues and coal fly-ash in cold-bonded aggregates for sustainable concrete.Composites Part B: Engineering,116 (2017) 46-52. [2] Colangelo, F., & Cioffi, R. . Mechanical properties and durability of mortar containing fine fraction of demolition wastes produced by selective demolition in South Italy. Composites Part B: Engineering,115 (2017) 43-50. [3] Colangelo, F., Cioffi, R., Liguori, B., & Iucolano, F. . Recycled polyolefins waste as aggregates for lightweight concrete. Composites Part B: Engineering106, (2016) 234-241. [4] Colangelo, F., Messina, F., Cioffi, R. . Recycling of MSWI fly ash by means of cementitious double step cold bonding pelletization: Technological assessment for the production of lightweight artificial aggregates. Journal of Hazardous Materials 299 (2015) 181-191. [5] N. Singh, D. Hui, R. Singh, I.P.S. Ahuja, L. Feo, F. Fraternali, Recycling of plastic solid waste: A state of art review and future applications. Composites Part B: Engineering 115 (2017) 409-422. [6] Qiao, X. C., Tyrer, M., Poon, C. S., & Cheeseman, C. R. . Characterization of alkaliactivated thermally treated incinerator bottom ash.Waste management, 28(10) (2008) 19551962. [7] L.Y. Zhang, Production of bricks from waste materials - A review, Constr. Build. Mater. 47 (2013) 643–655 [8] M. Safiuddin, M.Z. Jumaat, M.A. Salam, M.S. Islam, R. Hashim, Utilization of solid wastes in construction materials, Int. J. Phys. Sci. 13 (2010) 1952–1963. [9] A.M. Rashad, Alkali-activated metakaolin: a short guide for civil engineer - an overview, Constr. Build. Mater. 41 (2013) 751–765. [10] H.Y. Zhang, V. Kodur, S.L. Qi, L. Cao, B. Wu, Development of metakaolin–fly ash based geopolymers for fire resistance applications, Constr. Build. Mater. 55 (2014) 38-45. [11] J.L. Provis, S.A. Bernal, Geopolymers and related Alkali-activated materials, Annu. Rev. Mater. Res.44 (2014) 299–232. [12] B.C. McLellan, R.P. Williams, J. Lay, A. van Riessen, G.D. Corder, Costs and carbon emissions for geopolymer pastes in comparison to ordinary portland cement, J. Clean. Prod. 19 (2011) 1080–1090. [13] J. Davidovits, Geopolymers: inorganic polymeric new materials, J. Therm. Anal., 37 (1991) 1633–1656. [14] M. Rowles, B. O'Connor, Chemical optimisation of the compressive strength of aluminosilicate geopolymers synthesised by sodium silicate activation of metakaolinite, J. Mater. Chem. 13 (2003) 1161–1165. [15] J. Davidovits Geopolymer. Chemistry and Applications, 3rd ed. Institut Geopolymere, Saint Quentin, France, 2011. [16] G. Habert, C. Ouellet-Plamondon, Recent update on the environmental impact of geopolymers, Rilem Technical letters, 1 (2016) 17-23. ACCEPTED MANUSCRIPT AC C EP TE D M AN US C RI PT [17] P. Duxson, J.L. Provis, G.C. Lukey, J.S. Van Deventer, The role of inorganic polymer technology in the development of ‘green concrete’, Cem Concr Res 37(12) (2007) 1590– 1597. [18] J.L. Bell, W.M. Kriven (2009). Preparation of ceramic foams from metakaolin-based geopolymer gels. In: Lin HT, Koumoto K, Kriven WM, Garcia E, Reimanis IE, Norton DP (eds). Developments in strategic materials: ceramic engineering and science proceedings, 29(10) 97–111. [19] E. Prud'homme, P. Michaud, E. Joussein, C. Peyratout, A. Smith, S. Rossignol, In situ inorganic foams prepared from various clays at low temperature, Applied Clay Science 51 (2011) 15–22. [20] V. Medri, A. Ruffini, Alkali-bonded SiC based foams, J. Eur.Ceram. Soc 32 (2011) 1907– 1913. [21] V. Medri, E. Papa, J. Dedecek, H. Jirglova, P. Benito, A. Vaccari, E. Landi, Effect of metallic Si addition on polymerization degree of in situ foamed alkali-aluminosilicates, Ceramics International, 39(7) (2013) 7657-7668. [22] L. Verdolotti, B. Liguori, I. Capasso, A. Errico, D. Caputo,M. Lavorgna, and S. Iannace: Synergistic effect of vegetable protein and silicon addition on geopolymeric foams properties. J. Mater. Sci.50 (2015) 2459. [23] R. Studart, U.T. Gonzenbach, E. Tervoort, L.J. Gauckler, Processing routes to macroporous ceramics: a review, Journal of the American Ceramic Society 89 (2006) 17711789. [24] J. Bourret, E. Prud’homme, S. Rossignol, D.S. Smith, Thermal conductivity of geomaterial foams based on silica fume, J. Mater. Sci. 47 (2012) 391–396. [25] E. Prud’homme, P. Michauda, E. Joussein, C. Peyratout, A. Smith, S. Arrii-Clacens, J.M. Clacens, S. Rossignol, Silica fume as porogent agent in geo-materials at low temperature, Journal of the European Ceramic Society 30 (2010)1641–1648. [26] J. Henon, A. Alzina, J. Absi, D.S. Smith, S. Rossignol, Potassium geopolymer foams made with silica fume pore forming agent for thermal insulation, J Porous Mater 20 (2013) 37–46 [27] J. Feng , R. Zhang, L. Gong, Y. Li, W. Cao, X. Chenget, Development of porous fly ashbased geopolymer with low thermal conductivity, Materials and Design 65 (2015) 529–533. [28] Z. Abdollahnejad, F. Pacheco-Torgal, T. Félix, W. Tahri, J. Barroso Aguiar, Mix design, properties and cost analysis of fly ash-based geopolymer foam, Construction and Building Materials 80 (2015) 18–30. [29] P. Hlavácek, V. Šmilauer, F. Škvára, L. Kopecký, R. Šulc, Inorganic foams made from alkali-activated fly ash: Mechanical, chemicaland physical properties. Journal of the European Ceramic Society 35 (2015) 703–709. [30] Z. Zhang, J.L. Provis, A. Reid, H. Wang, Mechanical, thermal insulation, thermal resistance and acoustic absorption properties of geopolymer foam concrete. Cement & Concrete Composites 62 (2015) 97–105. [31] C. Ferone, G. Roviello, F. Colangelo, R. Cioffi, O. Tarallo Novel Hybrid OrganicGeopolymer materials, Appl. Clay. Sci. 73 (2013) 42-50. [32] G. Roviello, L. Ricciotti, C. Ferone, F. Colangelo, R. Cioffi, O. Tarallo Synthesis and Characterization of Novel Epoxy Geopolymer Hybrid Composites, Materials, 6 (2013) 3943-3962. [33] G. Roviello, C. Menna, O. Tarallo, L. Ricciotti, C. Ferone, F. Colangelo, D. Asprone, R. Di Maggio, E. Cappelletto, A. Prota, R. Cioffi Preparation, structure and properties of hybrid materials based on geopolymers and polysiloxanes, Materials and Design 87 (2015) 82-94. [34] F. Colangelo, G. Roviello, L. Ricciotti, C. Ferone, R. Cioffi, Preparation and characterization of new geopolymer-epoxy resin hybrid mortars, Materials 6 (2013) 29893006. ACCEPTED MANUSCRIPT AC C EP TE D M AN US C RI PT [35] C. Ferone, F. Colangelo, G. Roviello, D. Asprone, C. Menna, A. Balsamo, A. Prota, R. Cioffi, G. Manfredi, Application-Oriented Chemical Optimization of a Metakaolin Based Geopolymer, Materials, 6 (2013) 1920-1939. [36] Gottardi, S., Toccoli, T., Iannotta, S., Bettotti, P., Cassinese, A., Barra, M., Ricciotti, L., Kubozono, Y. Optimizing picene molecular assembling by supersonic molecular beam deposition. Journal of Physical Chemistry C 116 (2012) 24503-24511. [37] C. Menna, D. Asprone, D. Forni, G. Roviello, L. Ricciotti, C. Ferone, A. Bozza, A. Prota, E. Cadoni, Tensile behaviour of geopolymer-based materials under medium and high strain rates, EPJ Web of Conferences, 94, (2015) Article number 01034. [38] G. Roviello, L. Ricciotti, O. Tarallo, C. Ferone, F. Colangelo, V. Roviello, R. Cioffi, Innovative fly ash geopolymer-epoxy composites: preparation, microstructure and mechanical properties Materials, 9(6) (2016) 461-475. [39] G. Roviello, L. Ricciotti; C. Ferone; F. Colangelo; O. Tarallo, Fire resistant melamine based organic-geopolymer hybrid composites, Cem. Concr. Compos. 59 (2015) 89-99. [40] Ricciotti, L., Borbone, F., Carella, A., Centore, R., Roviello, A., Barra, M., Roviello, G., Ferone, C., Minarini, C., Morvillo, P. Synthesis of highly regioregular poly[3-(4alkoxyphenyl)-thiophene]s by oxidative catalysis using copper complexes. Journal of Polymer Science, Part A: Polymer Chemistry, 51 (2013), 4351-4360. [41] L. Ricciotti, G. Roviello, O. Tarallo, F. Borbone, C. Ferone, F. Colangelo, M. Catauro, R. Cioffi, Synthesis and characterizations of melamine-based epoxy resins, Int. J. Mol. Sci. 14 (2013) 18200–18214. [42] A. Strini, G. Roviello, L. Ricciotti, C. Ferone, F. Messina, L. Schiavi, D. Corsaro, R. Cioffi, TiO2-Based Photocatalytic Geopolymers for Nitric Oxide Degradation, Materials 9(7) (2016) 513-525. [43] www.mapei.com/public/COM/products/367_epojet_gb.pdf. [44] Matlab release R2015a - The MathWorks, Inc. [45] Kantro, D. L. (1980). Influence of water-reducing admixtures on properties of cement paste—a miniature slump test. Cement, Concrete and aggregates, 2(2), 95-102. [46] Collins, F., Sanjayan, J. G. (1998). Early age strength and workability of slag pastes activated by NaOH and Na2CO3. Cement and Concrete Research, 28(5), 655-664. [47] Bouvet, A., Ghorbel, E., Bennacer, R. (2010). The mini-conical slump flow test: Analysis and numerical study. Cement and concrete research, 40(10), 1517-1523. [48] Dubey, R., Kumar, P. (2013). An experimental study for optimization of high range water reducing superplasticizer in self compacting concrete. Frontiers of Structural and Civil Engineering, 7(1) (2013) 62-71. [49] Messina, F., Ferone, C., Colangelo, F., Cioffi, R. Low temperature alkaline activation of weathered fly ash: influence of mineral admixtures on early age performance. Construction and Building Materials 86 (2015) 169-177. [50] Lachemi, M., Hossain, K. M. A., Lambros, V., Nkinamubanzi, P. C., Bouzoubaa, N. Performance of new viscosity modifying admixtures in enhancing the rheological properties of cement paste. Cement and Concrete Research, 34(2) (2004) 185-193. [51] M. Talò, B. Krause, J. Pionteck, G. Lanzara, W. Lacarbonara, An updated micromechanical model based on morphological characterization of carbon nanotube nanocomposites. Composites Part B: Engineering, 115 (2017) 70-78. [52] H. Khezrzadeh, A statistical micromechanical multiscale method for determination of the mechanical properties of composites with periodic microstructure. Composites Part B: Engineering, 115 (2017) 138-143. [53] A.P. Fantilli, B. Frigo, B. Chiaia, Comparing multi-scale cracking mechanisms in manmade composites and natural materials. Composites Part B: Engineering, 115 (2017) 369375. ACCEPTED MANUSCRIPT AC C EP TE D M AN US C RI PT [54] Gibson, L. J., Ashby, M. F. Cellular solids: structure and properties. Cambridge university press (1999). [55] Tonyan, T. D., L. J. Gibson. "Structure and mechanics of cement foams." Journal of Materials Science 27.23 (1992): 6371-6378. ACCEPTED MANUSCRIPT Lightweight geopolymer-based hybrid materials Giuseppina Rovielloa,b, Costantino Mennac, Oreste Tarallod, Laura Ricciottic, Francesco Messinaa,b, Claudio Ferone,a,b, Domenico Asproncc, Raffaele Cioffi,a,b Dipartimento di Ingegneria, Università di Napoli ‘Parthenope’, Centro Direzionale, Isola C4, 80143 Napoli, Italy b INSTM Research Group Napoli Parthenope, National Consortium for Science and Technology of Materials, Via G. RI PT a Giusti, 9 50121 Firenze (ITALY) c Dipartimento di Strutture per l’Ingegneria e l’Architettura, Università di Napoli Federico II, Napoli 80125, Italy; d Dipartimento di Scienze Chimiche, Università degli Studi di Napoli “Federico II”, Complesso Universitario di Monte * M AN US C S. Angelo, via Cintia, 80126 Napoli, Italy Corresponding Author: tel. +39-081-5476781; fax: +39-081-5476777; e-mail address: [email protected] (G. Roviello) Highlights 1. Hybrid geopolymer-based foams with densities ranging from 0.25 to 0.85 g/cm3 were D obtained. process was obtained. TE 2. By using Si0 powder as in situ foaming agent, a homogeneous and very regular foaming EP 3. Mechanical and thermal performances significantly better than those shown by neat AC C geopolymeric foams and inorganic foamed-materials with similar densities.