A novel method to develop nanocomposite bimodal foams containing expandable polymeric microballoons: microstructural characteristics

This study aims to develop a novel technique in manufacturing nanocomposite bimodal foams containing expandable polymeric microballoons. Low density polyethylene syntactic foams were prepared via injection molding process, afterwards, a batch refoaming method was utilized to create bimodal structure. The effects of microballoon and nanoclay content and foaming time and temperature on microstructure and physical properties of foams were investigated. The results revealed that refoaming leads to a considerable decrease in density due to nucleation of microcells along with re-expansion of microballoons, as well as CO2 diffusion in voids between the matrix and microballoon surfaces. Microballoon content has no significant effect on cell size of bimodal foams, while a great growth in cell density was observed as its content increased. Results also indicated that at low and high foaming temperature and time, melt strength and gas loss are the overcoming phenomena, respectively leading to an optimal processing temperature and time.


Introduction
Polymeric composites are capable of reducing the structural weight and result in improved efficiency and performance in many applications. Introduction of foaming technology in the late 1920's led to more weight reduction and new applications [1,2]. Syntactic foams are hollow particle-filled composites fabricated by filling a metal, polymer or ceramic matrix with hollow microspheres [3]. Polymeric syntactic foams are polymer matrixes reinforced with hollow glass microspheres (HGMs), polymer microballoons and carbon spheres [4,5]. Due to their low density and thermal conductivity, excellent specific compressive strength and stiffness, polymer syntactic foams are used in various fields such as underwater vehicle structures and submarine buoyancy modules [6], automotive and aerospace parts [7,8], thermal insulation or high thermal conductive materials [9,10] and materials with low dielectric constants [11].
In recent years, significant research and studies have been carried out on the improvements of physical and mechanical properties of polymer syntactic foams. Initial studies mostly focused on determining the effect of particle volume fraction [12] and radios ratio [13] on mechanical properties of syntactic foams with thermoset matrix. According to the reports, increasing the volume fraction and radius ratio leads to the reduction of compressive strength.
The method of syntactic foam preparation also affects the properties of composite. The first attempts included manufacturing resin/HGM syntactic foams using stirring casting method (SCM). For example, Gupta et al. [14] investigated tensile and compressive properties of vinyl ester/ glass microballoon syntactic foams. Wang et al. [15] and Ciardiello et al. [16] prepared the carbon nanotubes (CNTs) and Graphene nano-platelets (GNPs) reinforced resin/HGM matrix composites and studied the effect of their reinforcement on mechanical properties of syntactic foams. Due to the air trapping in viscous matrix during stirring and formation of cavity defects which affect the properties of the composite, Ding et al. [17] introduced the design of a new preparation method called vacuum resin transfer molding (VARTM) method. According to their report, epoxy/HGM resin syntactic foams, prepared by the use of VARTM, showed properties superior to syntactic foams prepared by SCM including higher compressive strength and module followed by lower water absorption.
Notwithstanding thermoset syntactic foams, less attention has been devoted to thermoplastic syntactic foams due to their different processing methods and testing protocols. Two widely used processing methods of thermoplastic composites are injection molding and compression molding. Kumar et al. [18][19][20][21][22] developed industrial injection molding technique to manufacture HDPE/cenosphere syntactic foams. In this research, they overcame the main challenges in adopting injection molding machine for fabrication of syntactic foams including minimizing hollow particle fracture and uniform mixing of light weight particle in polymer matrix resin. Jayavardhan et al. [23,24] optimized the process of compression molding method to fabricate HDPE/ HGM syntactic foams and characterized their flexural and tensile response.
In the 1950s, microcapsules with polymer shells were fabricated by the American National Cash Register Company to be used in thermal paper to encapsulate ink during polymerization [25]. Afterward, heat expandable microcapsules known as expandable polymeric microballoons (EPMs) have been industrialized by Dow Chemical Co. as foaming agents [26]. EPMs are microscopic spheres containing a low boiling point liquid hydrocarbon [27]. These microscopic spheres comprise of a thermoplastic shell, and when heated to a subsequently high temperature, they soften the thermoplastic shell. The increasing pressure of the hydrocarbon inside the thermoplastic shell cause the microsphere to expand by 50-100 times. The market size for EPM in the areas of automotive, construction, sport and leisure is rising and will reach USD 833.9 million by 2026, according to the reports [28].
Despite the ease of access to wide range of literature on syntactic foams with HGM, studies on syntactic foams containing EPM are rather scarce. Easily breaking of glass microballoons in shear, makes it very difficult to be implemented in manufacturing processes that cause shearing of the material. EPMs are substitutes by which it is possible to produce a foamed product using commercial extruders or injection molding machines. Moreover, Lawrence and Pyrz [30] investigated the viscoelastic properties of extruded PE/EPM syntactic foams. Also, Mea et al. [31] studied the effects of strain rate and density on tensile behavior of injection molded PP/EPM syntactic foams. Kawaguchi et al. [29] optimized rheological properties of EPMs in PP and studied both injection molding and extrusion molding of syntactic foams. Peng et al. [32] compared injected PLA/EMP syntactic foams and PLA/SC-N2 (super critical nitrogen) foams and witnessed higher elongation at break of syntactic foam. Compression molding of Epoxy/EPM syntactic foam was studied by Wang et al. [33] demonstrating the drop of mechanical properties in elevated molding temperatures due to thinner microballoon walls. In another study, the comparison between epoxy/HGM and epoxy/EPM syntactic foams was investigated by Dando et al. [34]. They acclaimed a better height recovery after compression test in the presence of thermoplastic microballoons. Uchio et al. [35] examined the effect of resin viscoelastic behavior on EPM expansion rate by conducting an extrusion foaming of different type of PP and graphing the relation between the shear viscosity and average microballoons diameter. According to their findings, a lower viscosity of the surrounding base polymer, results in higher diameter of the EPMs.
Bimodal foams (also known as bi-cellular foams) consist of two kinds of cells with significant difference in cell size, which are fabricated to increase thermal insulation and sound absorption of polymeric materials [36]. The most commonly used methods to obtain a bimodal cellular structure applying two different blowing agents [37], two-step depressurization foaming approach [38] and using a polymer blend [39] or nanoparticles [40] as nucleation sites. There is a very great interest on bimodal syntactic foams with metal matrix [41][42][43][44] while polymer matrix bimodal syntactic foams have received insignificant scientific consideration. Eom and Kim [45] fabricated syntactic bimodal foams of polysiloxane/LDPE blends with two different types of polymer microbeads with an average diameter of 20 and 50 μm. Blends were manufactured in a twin-screw extruder and then batch foamed via scCO 2 . They acclaimed that adding LDPE increased porosity of foams due to the lower amount of microbeads which acted as a nucleating site for the cells during foaming [46]. Their study revealed a bimodal structure containing small spherical cells of polymer microbeads and relatively large cells nucleated and grown via foaming with CO 2 . Ozkutlu et al. [47] combined syntactic foam extrusion and scCO 2 foaming to produce bimodal PMMA foams. PMMA/HGM/nanoparticle syntactic foams were manufactured via extrusion and then bimodal structure was obtained through batch foaming process. In the foam matrix, large cells of CO 2 diffused in voids between the matrix and HGM surfaces, and small cells nucleated and grew in nanoparticle nucleation sites. They achieved a 69% decrease in foam density compared to bulk polymer, while shore D hardness decreased by only 12%. Galvagnini et al. [48] studied the thermophysical properties of multifunctional epoxy/HGM bimodal syntactic foams containing phase change microcapsules (PCM). The average diameter of HGM and PCM were 60 and 20 μm, respectively. They witnessed that due to the higher packing efficiency of bimodal filler distribution, the systems containing both HGM and PCM had lower viscosity than those containing only one filler type.
Nowadays the use of carbon dioxide (CO 2 ) as a physical blowing agent has received a great attention both academically and scientifically, due to its non-toxic and environmentally friendly nature [49,50]. In this study, a novel method is introduced to fabricate low density polyethylene bimodal foams. In first step, LDPE/EPM nanocomposite syntactic foams are produced via injection molding and then a CO 2 gas is utilized as physical blowing agent to nucleate small cells besides the large microballoons. The effect of EPM percentage, nanoclay addition, foaming temperature and foaming time on microstructure and physical properties of syntactic foams and bimodal samples is investigated. It is noteworthy that despite the potentialities of such LDPE/EPM bimodal foams, no similar studies can be found in the open scientific literature to the best of the authors' knowledge.

Sample preparation
LDPE, P501E1, Cloisite 30B and montmorillonite were hand mixed and loaded to a NBM HXF-128 injection machine (Nekoo Behine Machine Co. Iran) according to the conditions listed in Table 1. Eight different samples were produced with the codes presented in Table 2.
Bimodal foams were prepared using batch foaming technology. Syntactic foam samples derived from injection molding, were saturated with CO 2 gas at ambient temperature for a specific time under a constant saturation pressure.
To obtain the sufficient saturation time, 6 samples containing EPM were saturated in scCO 2 at a saturation pressure of 5 MPa for 12 h. Then, the pressure was released and the samples were weighted. This process was repeated in certain time intervals to ensure a constant weight variation [51]. Figure 1 illustrates the absorption behavior of samples. As it can be observed, the amount of absorbed gas is constant by increasing the saturation time from 36 to 48 h. Accordingly, 48 h was chosen as absorption time. The reason of the increase in absorption time compared to previous studies, is the barrier effect of microballoons in polymer matrix.
After the full saturation of syntactic foam samples, the pressure was discharged and samples were immediately removed from the pressure vessel. Foam expansion was inducted by immersion of the saturated sample in the hot oil bath (hot glycerol) at a certain temperature (foaming temperature) and time (foaming time). Finally, foam structure was stabilized by cooling in ice/water solution. Three different temperatures of 120, 130 and 140 °C and three times of 80, 100 and 120 s were considered as foaming temperature and foaming time, respectively.
Bimodal samples are coded as A-B-C, where the first part presents the polymer blend section according to Table 3,

Sample characterization
An Attenuated Total Reflection-Fourier Transform Infrared Spectrometer (ATR-FTIR UNICAM Maston 1000, UK) was used to identify chemical bonds of P501E1 pellets. To study the thermal behavior of microballoons during manufacturing process, a Differential Scanning Calorimeter (DSC-200, F3 Maia, NETZSCH, Germany) was undertaken on P501E1 films. DSC test was performed in temperature range of 20-250 °C with a heating rate of 10 °C/min. Foam density was measured using a water displacement method according to the following equation: where W f is the weight and V is the volume of foamed sample. The relative density of the foam was measured using Eq. 2 in accordance with ASTM-D3575, as follows: where * and s are the relative density and polymer matrix density of the foam, respectively. Also, the foaming degree was calculated by: Cell microstructure of syntactic and bimodal specimens was observed using a Tescan MIRA3 FEG-SEM (Czech Republic) scanning electron microscopy (SEM) operating at 20 kV. The foamed samples were cryogenically fractured in liquid nitrogen to ensure an intact cellular structure and then the fracture surfaces were coated with a thin layer of gold according to ASTM-F1372 before SEM observations to enhance electrical conductivity. The analysis of images was conducted by Image Pro. Express software, which enables determination of average cell size. Cell density of foams is defined as the number of cells per unit volume: where N 0 and N are cell density and number of cells in the micrograph of area A with magnification of M.

ATR-FTIR analysis
ATR-FTIR test was performed in the range of 500-4000 cm −1 to get necessary information about the chemical composition of P501E1 pellets and the graph is illustrated in Fig. 2. Table 4 presents the data related to the main absorption bands and vibrations of the functional groups. The vibration peaks of C = O, C-H, C-O and C-O-C bands are related to PMMA signals and other peaks denote LDPE [52,53]. The observed peak of 3500 is probably associated with the O-H group of crosslinking agent  Figure 3 shows the DSC diagram of P501E1 pellets. Three main endothermic peaks and one exothermic peak can be identified in this diagram. The first peak in the temperature range of 50-70 °C is vaporizing temperature of free hydrocarbon gas, dissolved in the base polymer during the masterbatch production process. The second peak observed in the range of 105-115 °C represents the T g of the shell polymer. Since the manufacturer did not provide information about the composition of shell polymer alloy, it is speculated that this peak is related to Acrylonitrile copolymer [56]. The third peak in the temperature range of 170-180 °C illustrates the vaporization temperature of liquid hydrocarbon inside the microballoon. Although the evaporation temperature of hydrocarbon at atmospheric pressure is below 100 °C, due to the high pressure inside unexpanded microballoon shell compared to the atmospheric pressure, hydrocarbon will evaporate at higher temperatures [57]. Finally, the last exothermic peak observed in the temperature range of 230-240 °C can be related to shell copolymer crystallization [58]. Other smaller peaks are probably associated with other components of the shell material. The results of the DSC test completely conform to the expansion diagram of the polymer microballoons provided by the manufacturer (Fig. 4), where it claims 167 °C as the expansion start temperature of microballoons. Data of Fig. 4

Foam density
Foam density is a parameter that determines the relative amount of the solid phase and gas, and serves as one of the major parameters to govern physical and mechanical properties of polymeric foams [59]. Density results of syntactic and refoamed samples are given in Fig. 5. It is noteworthy that data were derived from three iterations of measurement, with the polymer matrix density being considered as 0.92 g/ cm 3 to calculate the foaming degree.
In syntactic foam samples, it is noted that the increase in the percentage of microballoons decreases the sample density, which is due to the increased number of microballoons in the polymer matrix and enhanced amount of gas in the system that reduces the density.
In the batch foaming process, the expansion is composed of three stages: nucleation, cell growth and cell stability [60]. Cell growth is controlled by factors including system's temperature, compressive or tensile hydrostatic stress applied to the polymer matrix, viscoelastic properties of the polymer/gas solution, etc. [61][62][63]. When the melt strength is high, the cellular walls will not have enough strength to withstand the force applied from the surrounding polymer and internal gas expansion, thus easily collapse during the foaming process. Cell destruction can, on one hand, affect the cell density, and on the other hand, the foaming  Fig. 4 Expandability of P501E1 microspheres expansion coefficient will be severely affected as a result of increased rate of gas diffusivity from the open cells. At higher temperatures, the foaming agent has high diffusivity, and the gas inside the foam can easily leave the foam due to the high rate of diffusion at the elevated temperature. Correspondingly, increased expansion coefficient mitigates the cell wall thickness and rises the diffusion of the gas inside the cells. As a result, the rate of gas that escapes towards the surrounding environment increases. Gas escape phenomenon can mitigate the amount of gas required to grow the cells, resulting in lower foam expansion. Furthermore, the mobility of molecular chains reduces at low temperatures, with the polymer's viscosity, followed by melt strength rising. Excessive increment of melt strength, inhibits the appropriate growth of the bubbles and increases the density. In fact, there is a conflict between gas diffusivity and melt viscosity. The followings are descriptions of foam density trend based on Fig. 5.
In batched LDPE sample, in foaming time of 80 and 100 s, the increase in foaming temperature helps growing the microcells of batch foaming process, thus decreasing the density. Unlike 80 and 100 s, in 120 s the P sample density increased with an ascending trend of foaming temperature, which is due to the reduced melt strength and gas loss at upper foaming time.
In forming time of 80 s, the addition of nanoparticles (PN samples) in all temperatures, reduces the density, compared to the pure polyethylene. Nano clay agglomerates are suitable cell nucleation and growth sites, which result in higher number of microcells and lower density. While at higher foaming times, the presence of nanoparticles in the PN sample prevents the cellular growth due to the higher melt strength, and thus, increases the density.
In PN samples, at foaming time of 80 s, the variation of density trend along with foaming temperatures is first upward and then downward. As temperature increases, the number of nucleated cells tend to be higher around the nano clay particles, and in addition, cells coalescence and density reduction take place, due to the close distance of nucleation spots. At higher temperatures, however, the presence of nano particles cannot overcome the low melt strength, and thus, the gas loss from the system, lessen the amount of gas accessible for cell growth and result in lower density. The same phenomenon happens in higher foaming times (100 s and 120 s).
A comparison of bimodal samples with batched polymer samples, suggests that at foaming times of 80, 100 and 120 s, the presence of relatively large microballoons alongside the highly-fine cells, caused by the batch foaming process, reveals lower density values. Also, the increased amount of the microballoons from 2.5 to 5 wt%, due to the higher number of the microballoons in the polymer volume, mitigates the density, while the addition of more microballoons to the LDPE matrix, increases the density due to the fact that neighbor balloons inhibit each other from expanding. This inhibition even overcomes the diffusion of CO 2 gas in voids between polymer matrix and microballoon shell, and leads to development of cavities in matrix and causes density reduction [47].
Generally, in bimodal foam specimens, increase in temperature leads to a lower density even in high foaming time, because despite the CO 2 gas escape from the system, the elevated foaming time helps microballoons expand. Some exceptions are observed as following: In F2.5 and F5 samples at foaming time of 80 s, increased temperature first increments the density, while higher temperature mitigates the density. As temperature rises from 120 to 130 °C, the residual shells made from PMMA increase the density, due to the softening of the microballoons' shell and escaping of the hydrocarbon gas. The rise of temperature from 130 to 140 °C helps the secondary cells, caused by batch foaming process, to grow more, and the surviving microballoons expand more, leading to a reduction of density.
F5 sample in foaming time of 100 s presents a rising density with rising temperatures while at t = 120 s, the results suggest a reduction in density as the temperature continues to elevate. This is due the fact that, high soaking time in the hot oil bath, softens the microballoons shell and results in the hydrocarbon gas to escape the cell; in addition, lower melt strength of the polymer helps in carbon dioxide gas escaping and density increasing. However, in t = 120 s the procedure is contrariwise, which is totally similar to the 80-s foaming time, due to the higher amount of microcells nucleation and growth.
A comparison of the density of P, PE, and refoamed samples at different foaming times shows that the increasing time of hot bath helps reduce the density since it helps microballoons and secondary cells to expand more.
An evaluation of results indicates that there is an optimal processing temperature and time to achieve an appropriate level of expansion. If the time and temperature are excessively high, the volumetric expansion coefficient will be controlled by the gas escape process. On the other hand, if the melt temperature is too low, or the sample stays in the hot oil bath for a short time, the volumetric expansion coefficient is controlled by higher melt strength, and increases with the increase of temperature and time. Foaming results suggested that the F5-120-100 sample showed the highest level of expansion with 67% of foaming degree. Figure 6a illustrates the effects of microballoon content on foaming degree under constant conditions of foaming temperature of 120 °C and foaming time of 100 s. In the sample containing 5 wt% of microballoon the effects of foaming temperature and time on foaming degree are also given in Figs. 6b and c, respectively. As expected, an increase in the percentage of the microballoon rises the foaming degree, while with more increase in microballoon content, foaming degree is decreasing.

Microcellular structure
To investigate foam microcellular structure, SEM method was used and cell size and cell density were studied. 100-200 cells in a rectangular area of 150,000-500,000 μm 2 , depending on the specimen, were investigated using Image Pro. Express software and the average cell size and cell density were measured. Each of these two parameters are governed by specific factors. Cell size is directly controlled by viscoelastic properties of the polymer. This property affects the cell coalescence and cell growth, which are two major reciprocal parameters. The higher melt viscosity is, the more it will prevent the growth and coalescence of the cells. In the contrast, the less the melt viscosity is, the better it will be an agent for the coalescence and growth of the cells. Thus, to achieve a specific structure, there may be an optimal value needed for the viscoelastic properties [61]. Figure 7, compares the cell size of syntactic foam samples with those of the microballoons. Microballoons of P501E1 has an average cell diameter of 24 ± 5 μm which is in acceptable agreement with the producer's datasheet. A comparison of unexpanded P501E1 cell size with the S2.5 sample reveals an average expansion of 2.3 times for microballoons. Obviously, this value does not confirm the one provided in Fig. 4. It is noteworthy that values of Fig. 4 were derived for free expansion of microballoons, while Fig. 7 is presenting microballoons under the injection pressure. Figure 7 indicates an increase in the percentage of microballoons from 2.5 to 5 then to 10 wt% would reduce the cell size, either in the presence of nano clays or their absence. The increased volumetric number of the microballoons can cause the microcapsules to have a barrier effect on each other's growth, thus resulting in lower cell size. Also, in nanocomposite samples, the cell size declines due to the elevated melt strength and inhibited growth of microcapsules.
SEM images of syntactic foam samples are given in Fig. 8. The key point is the presence of tiny cells in the matrix, which are clearly observable in Fig. 8d. The average cell size of the syntactic foam samples is also calculated by taking these tiny cells into account. Authors believe that at injection temperature of 205 °C, as shown in Fig. 4, the softening of the cells helps the hydrocarbon gas escape the cells and be dissolved in the polymer melt inside the chamber. Furthermore, during the injection process, this gas results in nucleation and helps to grow new cells, thereby reducing the density of the samples and average cell size. In nano samples, these microcells tend to be observed more which is due to their nucleation and growth around the nanoparticles. Figure 8g illustrates the image of these micro cells at higher magnification.
SEM images of the P501E1 sample (Fig. 9a) indicate the appropriate dispersion of the micro capsules on the LDPE matrix. Also, a free injection sample containing 10 wt% of the microballoons was also prepared and its SEM image is shown in Fig. 9b. It is worth noting that the presence of an external force hinders the uniform growth of the microballoons, as the pressure inside the barrel suppresses the expansion of EPMs in the injection machine and in absence of such pressure, a uniform expansion carried out [29].Average cell size of the polyethylene and nanocomposite batch foaming samples are given in Table 5. It should be noted that a closed cell microstructure containing ultramicrocells in range of 0.2-0.8 μm with a cell density variety of 0.1-2.64 × 10 6 cells/cm 3 is achieved during batch foaming. Microstructure of two selected samples P-130-100 and PN-130-100 is illustrated in Fig. 10. As it can be seen, the increased melt strength in presence of nanoparticles, has a barrier effect on cell nucleation and growth leading to smaller cells as well as lower cell density. Figures 11 and 12 also give the cell size and cell density of the refoamed specimens. The addition of the microballoons to the pure polyethylene increases the average cell size considerably, which is expected considering the significant difference of microballoons size with the one of microcells appearing through batch foaming process. An increase in percentage of microballoon from 2.5 to 5, and then to 10 wt%, declines the cell size, and significantly increases the cell density. The higher the number of the microballoons, the higher they inhibit each other's growth, which reduces the cell size.
In the pure polyethylene sample at 80 and 100 s of foaming, with the increase of foaming temperature, the cells grow resulting in cell size increase and cell density decrease. At very high foaming temperature, the cell size remains relatively the same as the 120 °C due to the gas loss through system; but cell density increases considerably. This is because of the nucleation and growth of the microcells at higher foaming temperature which increases the number of the cells, and mitigates the density, while the average cell size does not change. When the samples are stored at a hot At foaming temperatures of 120 °C, as the foaming time rises, the cells become larger and cell density drops down which is due to the lower viscosity and easier growth of cells. At higher temperatures, as the foaming time increases, the cell size increases initially and then decreases, with cell density experiencing a contradictory trend. Increased cell density along with increased cell size is due to the formation and growth of large number of new microcells at foaming temperature of 140 °C. With the addition of nanoparticles in the batch foamed and refoamed samples, the cell size witnesses an upward trend compared to samples without nano clay, resulting in lowered cell density. Agglomerated nanoparticles serve as sites for the nucleation and growth of the cells and thus form large number of cells in adjacence to each other and around the nanoparticles. Because of high number of nuclei in a small area, these tiny cells coalesce and cause the average cell size to increase and the density to mitigate. It is noted that in these samples the high hot  oil bath temperature helps to grow both microballoons and micro cells, thus, increasing the cell size and decreasing cell density. Generally, in bimodal samples the increase in foaming time results in more growth of microballoons and microcells, thus the cell size and cell density increase and decrease, respectively. This trend is not observed in all samples. Followings are the explanation of the effect of foaming time and temperature on foam cell size and cell density.
In bimodal samples containing 2.5 wt% of microballoons, at foaming time of 80 s, the increased temperature forms a large number of tiny microcells in the batch process, while at higher temperatures, the cells tend to grow more, thus the cell size initially decreases then increases. At foaming time of 100 s, as the temperature increases, the cell size increments, and cell density mitigates, which is due to the growth of the cells. More increment in temperature lessens cell size and rises cell density. At foaming time of 120 s, the increased temperature increases the cell size and decreases the density. However, a growth in cell density is noted. Raised cell density is due to the formation of a great number of new cells in batch foaming process, but these microcells failed to overcome the large microballoons, resulting in high cell density along with high cell size.
At foaming temperature of 120 °C, as the foaming time increases, the cell size and density both mitigate with the cell density first experiencing an upward and then a downward trend. At high foaming times, lower viscosity and escape of the gas from matrix lead to microballoon's gas loss, and form a large number of new cells. Thus, the average cell size and the density decrease. Also, due to the nucleation of the secondary microcells, cell density increases. At higher foaming temperatures, the melt strength reduction helps to create unwanted cavities in the system which then reduces both density and cell density. At lower temperatures in nanocomposite specimens, cell size grows and cell density declines, while at elevated temperatures, a contradictory trend is observed. At low temperatures, nucleus is formed around the nanoparticles coalesce, which, increase the cell size. Raised temperatures seem to form large number of new cells, which overcome the phenomenon of coalescence, thereby decreasing the cell size, and increasing the cell density.
In the bimodal sample containing 5 wt% of microballoons, and at foaming time of 80 s, with the rise of foaming temperature, the cell size descends and then ascends. Density and cell density first go up and then decline. Although increased temperature helps to grow the cells, contrary trend was observed, which is due to the nucleation of the secondary microcells that diminish the average cell size.
At foaming time of 100 s, increased temperature helps grow all the cells and leads to increased cell density and reduced cell size. At 120 s, however, cell size and cell density experience a trend which is completely opposite of the 80 s foaming time. At low temperatures of 120 °C, the cell size increases as the cells enlarge, but since the foaming time at 140 °C is high, gas diffuses from microballoons and they shrink, thereby reduce the average cell size. On the other hand, as the melt strength decreases, a large number of new microcells are formed, which leads to the high cell density.
As the foaming time increases to 120 °C, cell size first increments and then mitigates. Naturally, cell density also goes through a similar trend. At lower foaming times, cell size increases with the growth of cells; however, at high foaming times, the gas escapes the microballoons which make the microcapsules shrink. Therefore, the average cell size gets smaller. On the other hand, the reduced melt strength, and the formation of a large quantity of new microcells help increase the cell density. At 130 °C and 140 °C, higher soak time at a hot oil bath can increase the cell size and decreases the density. At high temperatures, the melt strength seems not to be sufficient to control proper cell growth, and the gas loss from the system being a factor which controls the cell morphology. In the nano composite sample containing 5 wt% microballoon, the increased viscosity of nano clays, leads to the hard growth of microballoons and secondary microcells which declines the cell size. On the other hand, new cells are not capable of nucleation or growth, which may fall the cell density.
In bimodal samples containing 10 wt% of microballoons, in all three foaming times, the increased temperature helps cell size to remain almost constant, but decreases the cell density. These samples seem to have a higher melt strength due to the presence of large number of the microballoons; even at upper temperatures, secondary microcells are not capable of nucleating, which explains the reduced cell density at elevated temperatures. As the foaming time increases from 80 to 120 s, the cell size increases slightly with the growth of the cells. However, on the other hand, the cell density gets also higher. Here, high foaming time overcomes the large number of microballoons. Meanwhile, with the decrease of the melt strength, new cells are formed and cell density increases. Figures 13a and b illustrate the cross-section of a syntactic polyethylene foam and bimodal sample containing 5 wt% of EPM, respectively. The height of syntactic foam samples is 2 mm and the height of bimodal foam samples varies from 2 to 4 mm depending on the process parameters. To avoid confusion, images of all samples are not presented. It is worth noting that in the syntactic foam sample, the microballoons grow in the hot center of the workpiece, and almost no microballoon growth takes place in area close to the mold wall due to the high cooling rate of the mold. During the refoaming process, the microballoons located in center of the workpiece tended to grow more and the microcells from nucleation of scCO 2 gas appeared in zones near the mold wall. The effect of increasing the weight percentage of microballoons is given in Figs. 14a-c. The findings show that in the F2.5 sample, the low number of microcapsules in polymer matrix has helped their high growth, as with the higher number of microballoons, their size has declined in F10 specimen, the neighbor microballoons are clearly seen to have prevented the growth of each other. Correspondingly, The CO 2 diffusion in voids between the matrix and EPM surfaces can be clearly observed which is totally compatible with the previous findings [47].

Conclusion
A novel technique as a combination of injection molding process and batch foaming method in presence of expandable polymeric microballoons were utilized to produce nanocomposite LDPE bimodal foams. Density and cellular structure were investigated in different foaming conditions. The effect of adding nanoclay was also studied. A bimodal structure containing ultramicrocells in range of 0.2-0.8 μm and microballoons in cell size variety of 30-130 μm is achieved. The results indicate F5-120-100 sample as the highest level of expanded foam with 67% of foaming degree. It is noteworthy that there are optimal processing temperature and time to achieve appropriate level of expansion. In high foaming time and temperature, density and expansion coefficient will be controlled by the gas diffusion rate, while in low foaming conditions the overcoming phenomenon is melt strength which has a barrier effect on the well growth of microcells and microballoons. The results also determine foaming time as the most effective parameter governing bimodal foam cell size and cell density and F10 specimen as the optimum sample in terms of microstructural view. Introducing the nanoparticles to the polymer matrix has a dual effect whereas in low foaming time and temperature, nanoclay cell nucleation leads to higher cell size and cell density, while smaller cells and lower cell density are achieved in elevated foaming conditions due to the increased melt viscosity. Polymeric bimodal syntactic foams have a huge number of advantages in light weight, low cost, proper