3.1. Filler Characterization
The cumulative size distribution Q3(x) and adequate histograms dQ3(x) made for the inorganic fillers used in this study are presented in
Figure 1. An analysis of the graphs shows that the ground and sieved expanded W is characterized by larger particle sizes compared to WO. Additional treatment using hydrogen peroxide allows increasing the content of particles with smaller sizes, which is probably due to the destruction of the concertina-shaped vermiculite structure and an additional exfoliation effect, confirmed by XRD evaluation. Both vermiculite grades (W and WO) exhibit two particle size distribution modes due to the fraction of finely divided filler sheets formed during the grinding process.
In
Figure 2, X-ray diffraction patterns of expanded vermiculite and vermiculite are also treated with hydrogen peroxide. The peaks at 2θ = 9°, 21.0°, 26.8°, and 34.3° in both vermiculites correspond to d-spacing of 9.8 Å, 4.2 Å, 3.3 Å, and 2.6 Å, respectively. The peak at 9° was shifted slightly to 8.96° after the treatment, which caused a d-spacing shift from 9.82° to 9.86°. This may be understood as an additional exfoliating effect of a chemical treatment [
6]. The phenomenon may improve the dispersion of the filler in a polyol composition and the efficiency of modification of the final PUR foam.
BET surface area (S
BET), t-Plot external surface area (S
EXT), t-Plot micropore area (S
MIC), desorption average pore width (4 V/A), single point desorption total pore volume of pores less than 12.1 nm diameter at p/p
0 = 0.98325121 (V
P), t-Plot micropore volume (V
MIC) values are presented in
Table 2.
Figure 3a shows N
2 adsorption-desorption isotherms of the filler before and after the H
2O
2 treatment. The course of the a(p/p
0) curve is typical of expanded hydrous phyllosilicates [
20,
21]. The physicochemical properties found in our experiment are in good agreement with the literature [
20]. S
BET increased after the two-step treatment (
Table 2). Both fillers correspond to similar courses of the curves without other inflections and hysteresis loops, which may be related to significant changes in the modified material structure.
Figure 3b presents a pore volume vs. pore diameter plot. It can be concluded that for both fillers, the pore diameter is below 40 nm. Therefore, the primary mechanism of adsorption results from mesopores adsorption. When p/p
0 increases above 0.8, the adsorption increases significantly, suggesting the presence of micropores [
22]. Both vermiculites can be described as type II in the Brauner classification. The measured adsorption in the whole considered range is higher for WO. The additional chemical treatment improved the physicochemical properties of the filler, including the specific surface area, which should cause an improved reactivity toward the chemically hardened polymer.
3.4. Properties of Rigid PUR Foams Modified with Vermiculite
The cellular structure of porous materials has a significant effect on their properties and depends on many factors such as: premix viscosity, their modification by fillers, method of foaming etc. [
26].
All the tested foams exhibited well-developed hexagonal cell structures (
Figure 6). The modification of the reference system (PUR) with the vermiculite and modified vermiculite improved the morphology of the PUR composites, generally reducing cells’ average diameters (
Table 3). However, the changes are insignificant and within the measurement error. The contents of closed cells were higher for the foams modified with vermiculite (
Figure 7). The content of closed cells is important from the thermal insulation properties point of view. Closed-cell foams are characterized by lower values of the thermal conductivity coefficient compared to open-cell foams.
Based on the results presented in
Figure 7, it can be observed that the content of closed cells increased from 76% for the reference material (PUR) to 91% for the foams with the highest vermiculite content (PUR15W). However, the value of the thermal conductivity coefficient is characterized by the highest value for a given material (PUR15W) despite the highest closed cell content. Such an effect may be related to the highest apparent density and relatively high AI of PUR15W material. It was observed that the material into which the WO was introduced has a higher apparent density, while the thermal conductivity of this material is lower than that of the material containing the same amount of W. This can be explained by the lowest AI, which means this PUR15WO foam has a less anisotropic structure. This structure limits heat transport through the foamed material. Depending on the type of filler, the effect on the variation of the foam’s apparent density may be different. Natural fillers, such as flax and hemp fibers, can decrease the apparent density of foams as a result of the moisture present in them (carbon dioxide is generated in the reaction of water and isocyanate). In the case of fillers characterized by high density, e.g., carbon fibers, montmorillonite, or other inorganic fillers, the apparent density of PUR foams is increased [
27].
The PUR foams modified with vermiculite were characterized by an apparent density (
Figure 8) in the range of 35–39 kg/m
3. The foams with modified vermiculite (PUR15WO) had the highest apparent density. However, differences among tested foams are not significant, taking into account the standard deviation. The compressive strength (
Figure 8) measured in the direction parallel to the direction of the foam growth is characterized by greater values than when measured in the perpendicular direction. These differences are due to the anisotropic nature of the cellular structure of the PUR foams obtained. The compressive strength results are comparable with those obtained for the reference material. There is a slight increase in the mechanical strength of the foams with the highest vermiculite content. This effect can be related to a slight increase in the apparent density of the materials with the highest filler content.
Figure 9 shows the collectively presented SEM images made for the reference sample (PUR), fillers (W, WO), and composites with the highest concentration of fillers (PUR15W, PUR15WO). The analysis was performed with SE and BSE modes to increase the visibility of the filler particles in the matrix. It can be concluded that the filler is well distributed in the polymer matrix. There are no torn-out inorganic fractions that could result from improper adhesion. Moreover, it should be emphasized that there are no agglomerated structures. Filler particles smaller than 1 µm are evenly distributed, while larger particles are embedded in the walls of the foam cells, especially in the nodes. Therefore, their localization does not weaken the cell structure, and no voids were noted in the interphase area, which could suggest a lack of adhesion between a polymer and a filler.
The aesthetics of the final products often plays an important role in selecting materials by design teams. The analysis of color, which is one of the primary criteria for the qualitative assessment of products, is essential from the point of view of the potential of selected product groups [
28,
29].
Table 4 summarizes the L*, a*, and b* chromatic parameters, describing the color in the CIELab space of the produced foams with different vermiculite contents. Additionally, the results of the total color change were calculated according to Equation (1). Even the smallest addition of the filler caused significant color changes, taking into account the criteria described in the ISO 2813 standard and the literature [
29]. Based on the low values of standard deviations, all the foams were characterized by lower luminescence and had a brown shade with a uniform color. This also confirms the good compatibility and miscibility of the PUR-W/WO. It should be emphasized that in the case of the foams with the highest filler concentration, no significant changes in ΔE between the batches made with expanded and H
2O
2 treated with vermiculites were noted.
The results of the thermogravimetric analysis are presented in the form of TG, and DTG graphs in
Figure 10.
Table 5 collectively shows thermal parameters, such as temperature at 5%, 10%, 25%, and 50% mass loss, residual mass at 900 °C, and data describing peaks observed at the first derivative of the group. The courses of the TG and DTG curves (
Figure 10) indicate the three-step course of the thermal degradation process of the rigid PUR foams. The first stage of degradation in the temperature range from 110 to 220 °C, with the maximum between 170 and 190 °C, is related to the evaporation of residual water and low molecular weight products in the PUR foam [
30]. The observed dominant degradation stage with the maximum process intensity observed around 320 °C is associated with hard-segment decompositions. As demonstrated by Jiao et al. [
30], in a narrow range between 320 °C and 350 °C, isocyanate monomers almost disappear. First, N-H bonds are degraded, resulting in the degradation of hard segments, then C-H bonds from methyl and methylene groups. The last step of decomposition observed in the temperature range of 370–420 °C corresponds to the degradation of ester bonds in polyols [
31,
32]. It should be emphasized that the introduction of the filler caused shifts in decomposition stages; however, it did not affect its mechanism, which proves that there were no significant changes in the chemical structure of the PUR composition. Based on the conducted research, it can be clearly stated that adding unmodified and hydrogen peroxide-modified vermiculite improved the thermal properties of the composites as compared to the unmodified PUR foam (enhanced T
5% and yield of residue). In the case of introducing expanded vermiculite into the PUR matrix, it is difficult to find a clear trend related to the amount of the filler and only an apparent effect of increasing thermal stability, especially distinct as evaluated at a residue (
Table 5). On the other hand, using a two-step treatment based on thermal development and subsequent immersion in concentrated peroxide significantly improved thermal stability of PUR-based composite. This may be related to the improvement in the filler dispersion resulting from its structure modification described in the earlier paragraph, with the formation of sheets of reduced size.
The cone calorimeter test is a small-scale test employed to observe a comprehensive set of fire features in a well-defined fire scenario [
33]. The measurement provides the value of parameters, such as time to ignition (TTI), heat release rate (HRR), the maximum average rate of heat emission (MARHE), total heat release (THR), effective heat of combustion (EHC), specific extinction area (SEA), and total smoke release (TSR). The HRR curves for the materials investigated in this study are illustrated in
Figure 11, while detailed data are summarized in
Table 6.
The HRR curves suggest that all PUR foams ignited at a comparable time, which was confirmed by the time to ignition values presented in
Table 6. Their cellular structure and low thermal conductivity strongly influence the burning behavior, and TTI reached only 5 ± 1 s. The heat release rate is an essential parameter to estimate fire development, intensity, and spreading. The trend of the HRR curve demonstrates the burning behavior of the materials as a function of time. The curve of the PUR exhibits quite a broad peak with a maximum average value of 288 kW/m
2. It can be observed that vermiculite led to a change in the curves trend from characteristic for thick non-charring to thick charring ones [
33]. The lowest pHRR of 237 kW/m
2 (reduction of 18%) was obtained for the composites PUR3W, so the values were independent of the filler content and its modification. MARHE, as an indicator determined from HRR, is used to estimate the hazard of developing fires. Consequently, lower MARHE was obtained for samples with lower pHRR. Vermiculite is a filler known for its flame-retardant effects [
34,
35,
36]; however, no change in LOI values as a result of W or WO addition was observed.
The integral of HRR over time expresses the total heat output, i.e., the THR [
33]. The vermiculite addition caused a non-linear decrease in THR, suggesting incomplete combustion affected by char formation or reduced combustion efficiency [
37]. Since there was no change in EHC, as well as according to the increased yield of residue (
Table 6,
Figure 10a), action probably occurred in the condensed phase. Moreover, the content of triethyl phosphate, which is a phosphorus flame retardant active mainly in the gas phase, was the same for all materials. The analysis of the photographs confirms that the presence of vermiculite facilitated the formation of a more compact structure, and the number of holes decreased with an increase in the amount of vermiculite (
Figure 12). Similar to the carbonaceous char, inert residue from inorganic fillers works as a barrier and additionally replaces polymer, reducing the fuel release [
30]. Probably, the residues of the investigated materials were the origin of both effects.
During a fire, smoke is of great importance as it reduces visibility and makes an escape more challenging [
38,
39,
40]. Considering the SEA values together with the standard deviation, it can be concluded that the use of vermiculite did not change this parameter. The lowest SEA, which corresponds to the surface light-absorbing particles of smoke [
39], was recorded for composites PUR3W and amounted to 787 m
2/kg. In turn, the TSR of all composites was reduced compared to the unmodified foam, and the highest decrease reached approximately 13% (PUR15W). Presumably, this is due to the increased amount of the material remaining in the condensed phase.