3.1. Obtaining of Polyols
CSs with high molecular weights are not readily available substrates to obtain polyols. However, we previously described the methods of the polyol synthesis from a water-soluble CS
WS [
33]. Here, we replaced a chitosan with an oligomeric CSo which dissolves in water, although worse than the previously used CS
WS. The reason for the lower water solubility of the CSo is a lower degree of deacetylation (DD = 68%) compared to that of the CS
WS (DD = 85%) [
33]. The molecular mass of the CS
O was 8710 Da determined by the viscosimetric method. It seemed that this oligomer was a convenient substrate for hydroxyalkylation with GL, similar to the CS
WS. The product of glycidylation was a semisolid resin; we have further converted it by the reaction with EC to obtain a liquefied product (
Scheme 2). We tested three variants of the polyol synthesis. Firstly, we mixed the CS
O with GLYC in order to partially dissolve the CS
O and later added GL. The reaction was exothermic already upon the initial heating to 155 °C. After the consumption of GL, the semi-product was further reacted with EC in the presence of K
2CO
3 catalyst. The obtained polyol was liquid and contained the products of GLYC glycidylation (
Scheme 3). In the second method, the straightforward reaction of the CS
O with GL in the absence of GLYC was attempted. The semi-product was again hydroxyalkylated further with EC. In the third method, the CS
O was mixed both with GL and EC. In the initial step of the process, EC acted as a solvent, and GL was a primary reagent, while further addition of K
2CO
3 enabled EC to react to complete the process.
The reaction was monitored using IR and
1H-NMR spectroscopies, and the products were analyzed using the MALDI–ToF technique. Based on the IR and H-NMR spectra of polyols and the oligomeric chitosan, the functional groups of the chitosan involved in hydroxyalkylation were identified, and a polyol structure was suggested. The IR spectra of the obtained polyols were compared with those of the CS
O. In the IR spectrum of the CS
O (
Figure 1), the broad band centered at the 3380 cm
−1 origin of the hydroxyl and the amine groups’ stretching vibrations, similar to the deformation vibrations of these groups at 1330 cm
−1 and 1500 cm
−1, respectively. The valence band of ether C-O-C was observed at ca. 1070 cm
−1. The presence of non-deacetylated groups was manifested by amide bands I and II at 1629 cm
−1 and 1524 cm
−1, respectively. The latter overlapped with deformation band of the CS
O amine groups. In the IR spectra of polyols (
Figure 1), a slight increase in the intensity of the band centered at 1030 cm
−1 was observed compared to that of the CS
O. This indicates the formation of additional C-O-C bonds. Moreover, the methine and methylene bands were found at 2900 cm
−1 and within 1400–1300 cm
−1 due to the ring opening of GL and EC and introducing them into the polyol structure. In the IR spectra of polyols, the valence and deformation bands of the hydroxyl groups were present at 3400 cm
−1 and 1330 i 600 cm
−1, respectively. In the spectrum of the (CS
O + GL + EC) polyol, a low intensity band was observed at 1780 cm
−1 and attributed to the valence vibration of the carbonyl group (
Figure 1), which suggested the presence of the ester groups in a polyol, which may form upon the EC ring opening according to
Scheme 4.
In the
1H-NMR spectrum of the CSo (
Figure 2), the amine group protons showed broad resonance within 7.0–9.0 ppm, while the primary and secondary hydroxyl proton resonances were present within 5.2–5.9 ppm. The C
1-H resonance was found at 4.55 ppm, while the resonances from the methine protons attached to C
3–C
6 were located within the 3.3–3.7 ppm region. The C
2H resonance was partially overlapped with residual H
2O at 3.4 ppm. The methyl group resonance from the non-deacetylated CS
O was present at 1.7 ppm [
11]. The amine proton resonances were absent in the spectra of the obtained polyols, which was expected due to the reaction of the amine groups with the hydroxyalkylating agents. The proton resonances within 3.3–3.7 ppm showed a changed pattern in the spectra of polyols (
Figure 3). Some additional resonances attributed to the methine and methylene groups appeared in that region, indicating the GL and EC ring opening upon the hydroxyalkylation of the CS
O. Moreover, the resonances from the hydroxyl protons previously observed in the spectrum of the CSo (5.2–5.9 ppm) disappeared in the spectra of polyols, while other hydroxyl proton resonances appeared within 4.4–4.6 ppm as a consequence of the GL and EC ring opening. No considerable differences in the NMR spectra of various polyols were found, indicating that the structure of polyols is very similar.
The side-products accompanying polyols were identified using the MALDI–ToF spectrometry (
Table 1,
Tables S1 and S2). In all cases, the peaks from GL and GLYC were observed (
Table 1, entries 2, 3;
Tables S1 and S2, entries 1, 3, 4). In the spectrum of the (CS + GLYC + GL) + EC polyol, the peaks indicating the presence of the GLYC hydroxyalkylation with glycidol were found (
Scheme 5) (
Table 1, entries 6, 12). These peaks were also present in the spectra of the (CS + GL) + EC and (CS + GL + EC) polyols (
Table S1, entries 8, 9, 11;
Table S2, entry 6), which were obtained without the participation of GLYC. The presence of GLYC is a consequence of the reaction of GL with water (
Scheme 6), which is formed upon the dehydration of the semi-products at elevated temperatures (
Scheme 7).
A semi-product of hydroxyalkylation with GL can further react with EC (
Scheme 8) (
Table S1, entries 12, 13, 15;
Table S2, entries 7, 8 10).
The products of GL and its oligomers’ hydroxyalkylation with EC (accompanied with the CO
2 release) were present in all polyols (
Scheme 9;
Table 1, entry 5;
Table S1, entries 5, 6, 7, 16, 17;
Table S2, entries 5, 11). The side-products can also undergo dehydration in reaction conditions applied here (
Scheme 7;
Table 1, entries 1, 7, 10, 14;
Table S1, entries 2, 10, 14, 17;
Table S2, entries 2, 9, 12, 13). We concluded that using the MALDI–TOF spectrometry, we identified the products of the oligomerization of GL and the presence of the products of the hydroxyalkylation of glycerol with glycidol and then with EC.
The obtained polyols were dark brown, viscous resins. The hydroxyl number (HN) and some physical parameters such as density, viscosity, and surface tension of the obtained polyols were determined (
Table 2). The HN of polyols fell within 482–548 mg KOH/g, which suggests their applicability for obtaining rigid polyurethane foams. The obtained HN values evidenced the high functionality of the polyols, which is presumably due to the use of GL as a hydroxyalkylating agent, because every added amount of GL to the hydroxyl group of the CS
O increased the functionality of the polyols by 1 (
Scheme 10).
Therefore, the products of the reaction between the CSo and GL were semisolid resins not miscible with isocyanates. The character of these resin products is consistent with the plenty of hydrogen bonds present in the structure. When oxyalkylene chains undergo elongation upon reaction with EC, the number of hydrogen bonds decreases, and eventually, the resins show lower viscosity. The densities and surface tensions of polyols are similar and fall within the region of 1.262–1.275 g/cm
3 and 38.1–38.6 mN/m, respectively. The lowest viscosity was obtained from the polyol produced according to method 3, i.e., the all-reagent method. Presumably, the low viscosity of the product is due to the shortest time of reaction (compare
Section 2.2). In other cases, the prolonged heating led to the partial condensation of the hydroxyl groups and eventually increased the viscosity of the product. Typical dependences of physical parameters on the temperature were observed for the obtained polyols (
Figure S1). They fell within the region suitable for polyols used for the synthesis of PUFs. The polyols suitable for obtaining PUFs have the viscosity within 200–30,000 mPas [
45]. According to our experience, the surface tension of polyols suitable for PUFs should be within 30–50 mN/s. Low surface tension is a preferable property of the polyol because it enables effective foaming.
3.2. Analysis of Reaction between CS with GL and EC
The reactions of the CS
O with GL and EC has not been described in the literature until now. The polyols obtained in such reaction can be used to produce PUFs (vide infra). Therefore, the reaction of hydroxyalkylation was studied here in details. The system of reactants was chosen so that the mass ratio of the substrates was analogous to that in the described syntheses of polyols (see
Section 2 and
Section 2.2). The reaction in GLYC was executed with 1.2 g CS
O, 9 g GLYC, and 12 g GL (experiment 1). The mixture was heated to 160 °C within 10 min in order to dissolve the CS
O, and then the temperature was decreased to 130 °C. At this temperature, the consecutive consumption of GL could be monitored by the determination of EN. Then, the CSo was replaced with an inert substance, not reacting with GL, for which cyclohexanone was chosen, and again, the EN was determined to follow the consumption of GL (experiment 2). A fast decrease in the EN was observed for the system with the CS
O (
Figure 4). The amount of GL (the epoxide groups) after dissolving the CS
O in GLYC and increasing the temperature to 130 °C (ca. 10 min) decreased from 0.7305 mol/100 g (initial EN) to 0.3817 mol/100 g, which corresponded to the 47.7% reaction of GL, while in the presence of cyclohexanone, the degree of the GL reaction was 0.5385 mol/100 g, which corresponded to the 26.3% consumption. The reaction of the CS
O with GL in GLYC ended after ca. 60 min (EN = 0.00); at the same time period, in the system without the CS
O, only 58.9% of GL reacted with GLYC (EN = 0.3020). This indicated that GL reacted both with the CS
O and GLYC, which corroborated the analytical data obtained using the MALDI–ToF technique (
Table 1).
For comparison, the same reactions were monitored using the methods previously employed in a study of the CS
WS and GL in GLYC [
33] (
Figure 4, experiments 3 and 4). In this case, the amount of GL after dissolving the CS
WS and increasing the temperature to 130 °C (10 min) diminished to 0.1858 mol/100 g, which corresponded to the 74.6% GL consumption, while in absence of the CS
WS, the consumption was 26.3%. The reaction of the CS
WS with GL in GLYC took ca. 30 min at a temperature of 130 °C; at the same time, in the presence of inert cyclohexanone, instead of the CS only 40% of GL reacted with GLYC (EN = 0.4388 mol/100 g). Based on the numbers obtained in these experiments, we can conclude that the CS
WS was more reactive towards GL compared to the CSo. It was an expected result due to the higher percentage of the amine groups in CS
WS (DD = 85%) compared to the CSo (DD= 68%).
The reaction between the CS
WS and GL in water, described previously in [
33], was different (
Figure 5).
The experiment was performed as follows: 12 g GL was added to 1.2 g CS
WS in 9 g water, and the mixture was heated to 90 °C for 10 min. At this temperature, the first measurement was made, and then the changes in the GL content over time were studied (experiment 5). For comparison, the same protocol was used with dioxane instead of the CS
WS (experiment 6). Faster GL consumption was observed in water compared to GLYC. As it has been mentioned previously [
33], the amount of distilled water after the completion of reaction corresponded to the amount of water left in the reaction mixture, assuming that GL reacts completely only with water but not with the CS
WS. Thus, the fast consumption of GL is not due to the reaction with the CS
WS but to the presence of the amine groups in the structure of the CS
WS merely, which catalyzed the reaction between water and GL (compare with
Scheme 6). Such catalysis by the amines in the hydroxyalkylation reactions with epoxide is known [
46].
Based on the results, we can conclude that GLYC promotes the hydroxyalkylation of chitosan better compared to water. In the latter case, the CS
WS remains dissolved in the oligomer obtained by the reaction of GL with water. The hydroxyalkylation of the CS
WS occurs in the next step of the reaction with EC [
33].
The reason for the prevailing hydroxyalkylation of the CS in GLYC vs. water is the lower amount of GLYC (0.13 mol) compared to that of water (0.5 mol) in the reaction mixture, even if we take into account the contribution of the three hydroxyl groups in the GLYC (0.39 mol). Moreover, the viscosity of the mixture plays its role. It is much lower in water, which in turn increases the mobility of the CSWS and GL molecules and the probability of their collisions.
We also studied the reaction between the CS
O and GL in the absence of a solvent (
Figure 6, experiments 7 and 8) according to synthesis 2. The mass proportions of reagents were the same as in the synthetic protocol, namely 4 g CS
O and 15 g GL, and in the control experiment with cyclohexanone instead of the CS
O. We followed the consumption of epoxides. The control experiment was aimed at investigating the possibility of the oligomerization of the GL itself. The starting point was taken after dissolving the CSo at 130 °C (10 min). Immediately after mixing the reagents, the EN was 1.0067 mol/100 g, while after reaching the 130 °C temperature, it dropped to 0.6711 mol/100 g, corresponding to 33.4% GL consumption. The full consumption of GL was achieved after 50 min. In the same time scale, the reference sample showed the EN equal to 0.8879 and 0.8195 mol/100 g, which corresponded to the 11.8% and 18.6% GL consumption. The result indicated that the oligomerization of GL occurs simultaneously with the CS
O hydroxyalkylation, although to a small extent (see
Scheme 5), which corroborated the MALDI–ToF results (
Table S1).
The reaction of the CS
O with GL at the initial molar stoichiometry 1:1 was monitored in DMSO because the CS
O did not dissolve completely in GL. The reaction was performed at 130 °C, and the product was isolated and analyzed using the IR and HNMR spectroscopies. The IR spectrum of the semi-product CS
O:GL = 1:1 (
Figure 7) compared to that of the CS
O (
Figure 1) showed the change in the hydroxyl-bond valence pattern at 3400 cm
−1. The intensity of the amine deformation band at 1520 cm
−1 decreased significantly, which indicated the reaction of GL with the amine groups. The high reactivity of the amine groups against the GL ring was attributed to the higher nucleophilicity of the amine groups compared to that of the hydroxyl. In the spectrum of the consecutive semi-product obtained from CS
O:GL = 1:2, the reaction did not show any additional changes. Based on that, we can conclude that a further reaction took place between the hydroxyl groups of the substrate and the semi-product with GL. It was confirmed by the H-NMR spectrum of the semi-product of CS
O:GL = 1:1 (
Figure 8). The disappearance of the broad band at 7.0–9.0 ppm for the amine groups, which was present in the spectrum of the CS
O, and other O-H resonances in the range of 5.2–6.0 ppm, and the simultaneous growing of resonances at 4.1–5.1 ppm was observed, as identified by the selective deuteration with D
2O. Moreover, in the spectrum of the CS
O:GL = 1:1 semi-product, there were no resonances from GL at 2.7 and 3.0 ppm [
47], which are characteristic for the GL spectrum. Upon the reaction, the shape of the resonances at 3.2–3.5 ppm changed, and new resonances from methine and methylene protons grew, indicating the GL ring opening.
According to synthesis 3, all the three reagents, CS, GL, and EC, were mixed at the following ratio: 1.2 g CS
WS + 15 g GL + 18 g EC (initial EN = 0.5927 mol/100 g) or 1.2 g CSo + 4.5 g GL + 12 g EC (initial EN = 0.3436 mol/100 g). The EC was the solvent for that process. The components were heated analogously to synthesis 3 to 190 °C for 15 min to completely dissolve the chitosan and determine the EN (
Figure 9). In this condition, 37.5% of GL was consumed upon the reaction with the CS
WS, while the 79.6% GL reacted with the CSo. The differences between the two reaction systems depend on the amount of GL vs. CS and the molecular mass of the latter. The chitosan of a lower molecular mass (CSo) dissolved faster, had lower viscosity, and greater mobility of the molecules, which makes them more likely to collide with the GL molecules. The higher molecular mass and viscosity of the CS
WS imposed a slower reaction. After the reaction with GL was complete, the amount of EC in the post-reaction mixture was monitored to answer the question whether it was already involved in the hydroxyalkylation in the absence of K
2CO
3. It was found that in the case of the system with the CS
WS, the percentage of the EC drops from 52.6% after mixing the components to 48.5% after the total consumption of GL (40 min). In the case of the CS
O, this parameters were 67.8% and 63.2% (after 20 min). Thus, in both cases EC reacted to a small extent. The reaction with EC was further triggered by the addition of the catalytic amount of K
2CO
3.
The analysis of the reaction of chitosan in various environments (water, GLYC, or GL) allowed us to determine the reactivity of the substrates and understand the reaction mechanism. Understanding the reaction mechanism based on the reactivity is important to optimize the synthetic reaction conditions, namely the temperature, concentration of the reagents and catalyst. This can further help to design and optimize the reactors in a scaled-up process.
3.3. Polyurethane Foams
Synthesized polyols were used to obtain PUFs on a laboratory scale. The kind of the polyol, the amount of isocyanate (pMDI), the catalyst, and the foaming agent were varied to optimize the PUFs. Gaseous CO
2 was used as a foaming agent in the reactions of water with pMDI. The optimized compositions are shown in
Table 3. The cream and rise times for all compositions were similar and fall within 41–44 s and 33–43 s, respectively. They depended slightly on the kind of the polyol and the amount of the catalyst. Exceptional was the PUF obtained from the (CS
O + GL + EC) polyol, which had shorter cream and rise times: 38–40 s and 30–39 s, respectively. This can be attributed to the presence of a large number of reactive hydroxyl groups in the (CS
O + GL + EC) polyol, as determined by HN = 510 mg KOH/g. Generally, high hydroxyl numbers also result in a short drying time for the composition. The PUFs were optimized in terms of the amount of isocyanate. It was found that the best PUFs were obtained at mixtures with a molar ration of the isocyanate groups to hydroxyl groups (isocyanate coefficient) of 1.0–1.3 (
Table 3).
Crucial properties of PUFs were determined, such as apparent density, volume water uptake, dimension stability, thermal conductivity coefficient, pore size, thermal resistance, compressive strength, and glass transition temperature. All these parameters are relevant for the quality and functionality of PUFs, depending on their type and application. One of these parameters of PUFs is, for instance, their apparent density. It also influences other physical and mechanical properties such as compressive strength. The apparent density of PUFs is illustrated in
Figure 10.
The apparent density of PUFs was dependent on the amount of water in the foaming composition. At a higher limit (3% water), more CO
2 was evolved, resulting in better foaming, and eventually, low-density PUFs were formed, with a 51–64 kg/m
3 value. At a lower water limit (2%), the apparent density of PUFs was 63–69 kg/m
3. Further increase in water in the foaming mixtures led to the deterioration of the mechanical resistance of PUFs and irregular, large pores. The exception was the PUF obtained from the (CS
O + GL) + EC polyol and 4% water, which had an apparent density of 52 kg/m
3 and high compressive strength (vide infra). The values of the apparent density of the obtained PUF correspond to those of rigid PUFs. The water uptake after the 24 h exposition is low and fall within 1.9–4.6%, except for the PUF obtained from the (CS
O + GL + EC) polyol, for which the uptake was 5.8% (
Figure 11). Low water uptake of PUFs is preferable, because it allows for the application of such PUFs as isolating materials in humid surroundings without the loss of their isolation capability [
48]. The water uptake of the PUFs used in the building construction industry should be within 3–5%. Such conditions are fulfilled by the PUFs described here. Low water uptake values suggest the presence of closed pores in the PUFs, which is an advantageous property of the PUFs used as thermal isolation materials.
The pores of the PUFs obtained from the (CS
o + GLY + GL) + EC and (CS
o + GL) + EC polyols and 2% water were close to spherical (
Table 4,
Figure 12a,c) with the longest diameter of 129–148 µm and the shortest diameter of 95–123 µm. The increase in the water content to 3% or 4% in the foaming composition resulted in considerable elongation of the pores (
Table 4,
Figure 12b,d). In such cases, the longest diameter was within 109–436 µm, while the shortest diameter was within 71–160 µm. This explained why the compressive strength of the PUFs obtained with the 3 and 4% water in the foaming composition was lower. The thickness of the pore walls was between 12 and 18 µm. The PUFs obtained from the (CS
o + GL + EC) polyol had oval pores, independently of the amount of the foaming agent; the pores obtained with the 3% water were smaller. This can be attributed to the fast crosslinking related to the high values of HN of the used polyol. Moreover, a known relationship was noticed: the PUFs with smaller and oval pores showed a higher compressive strength compared to that of the PUFs with the elongated pores.
In the case of rigid PUFs, the thermal conductivity coefficient is usually within the 0.020–0.035 W/m·K region. The thermal conductivity coefficient of the PUF obtained from (CS
O + GLYC + GL) + EC and 2% foaming agent was typical of rigid PUFs (0.0269 W/m·K) [
45]. For other PUFs, it fell within 0.0314–0.0351 W/m·K (
Figure 13). The thermal conductivity coefficients (λ) depends on amount of open and closed pores in foams. Generally, more open pores result in low λ. Such situation was observed in the studied PUFs. The increase in the open pores measured by the water uptake corroborated the increase in the thermal conductivity coefficients.
The PUFs had good dimensional stability. The thermal exposure of PUFs for 40 h at 150 °C caused distortions within (−5.56)–(+2.04)%. Typically, the acceptable dimension stability of polyurethane foams is 3–5%. In some cases, the shrinkage was observed in one direction and elongation in another (
Table 5).
The thermal resistance of the PUFs was analyzed by the mass loss during one-month thermal exposure of the PUF samples at 150 and 175 °C with the concomitant measuring of compressive strength before and after the thermal exposure. Thermal resistance means that a material maintains the physical and mechanical properties at the testing temperature. Generally, the mass loss can be related to physical conversions, such as evaporation of TEA or water, and chemical degradation leading to volatile side-products. The PUFs obtained here were rigid at room temperature, and they remained as such after the thermal exposure. The PUFs’ mass loss is the largest in the first day of the exposure. When heated to 150 °C, the mass loss was within 12–19% (
Figure 14) in the case of the PUFs obtained from the (CS
O + GL) + EC or (CS
O + GL + EC) polyols, i.e., slightly larger compared to that of the PUFs obtained from the CS
WS in the same conditions (10.4–17.4%) [
33]. The lowest mass loss was observed for the PUFs synthesized in GLYC (5.9–6.1%). It was comparable to that of the thermoresistant PUFs obtained from the polyols containing azacyclic components such as isocyanuric acid and purine (6–9%) [
49]. The PUFs based on the (CS
O + GLYC + GL) + EC polyol showed the highest thermal resistance also at 175 °C (
Figure 14). Their mass loss after the one-month thermal exposure was 22.6–28.7%, while in other PUFs, it was 6–10% higher. Such a high thermal stability of these PUFs can be related to the presence of the multifunctional oligomeric products of glycerol with GL during the synthesis of (CS
O + GLYC + GL) + EC polyol. These oligomeric side-products can be responsible for additional crosslinking reactions.
The compressive strength of the obtained PUFs is typical of rigid PUFs (0.200–0.241 MPa,
Figure 15).
In the case of the PUF obtained from the GLYC solution and 2% water, the compressive strength was twice as high (0.403 MPa). This can be explained by the additional crosslinking and spherical pores (
Figure 12a). In other cases, the kind of the polyol used did not influence the compressive strength. It was a little lower for the PUFs obtained with a larger amount of foaming water (3% and 4% water/100 g polyol) due to a lower apparent density (
Figure 10) and larger and elongated pores (
Table 4). The PUFs showed a greater compressive strength after the thermal exposure at 150 °C or 175 °C compared to that before the exposure. This can be explained by the additional crosslinking of the PUFs upon heating. It was especially pronounced in the case of the PUFs obtained from the (CS + GL) + EC and (CS + GL + EC) polyols and 2% water; in those cases, the compressive strength increased after the thermal exposure at 150 °C and 175 °C by 41.1 and 28.8% and 60.6 and 14.6%, respectively. In some cases, the compressive strength decreased upon annealing at 175 °C compared to that upon annealing at 150 °C. Still, however, the compressive strength of the PUFs after the thermal exposure was larger compared to that of the untreated materials. The decrease in the compressive strength at 175 °C was probably related to the partial degradation of the samples. It is characteristic that the decrease in the compressive strength at 175 °C was incomparably low in relation to the loss of mass at this temperature. It should be remembered that the classic rigid PUFs are thermally resistant up to 90–110 °C, and above this limit, they lose their functional properties [
1,
2]. The PUFs described here were not designed as thermoresistant materials. The large mass loss at 175 °C was not surprising, because the degradation of PUFs starts in such condition, which is accompanied by graphitization. The latter is the reason for small changes in the compressive strength, which, according to our earlier experience, can even grow after annealing.
The obtained PUFs were subjected to a thermogravimetric analysis. The TG measurements confirmed the high thermal resistance of the PUFs obtained from the (CSo + GLYC + GL) + EC polyol.
A 5% mass loss was observed at 221 °C and 247 °C for the PUFs obtained with 2 and 3% foaming water, respectively (
Table S3). Some PUFs showed a 5% mass loss already at 75–87 °C. This is consistent with high moisture uptake from the air. This was also confirmed by the results of the water absorption determined in these foams (see
Figure 11) and by the DSC analysis which evidenced the presence of volatiles in the PUFs (
Figure S2).
There were three peaks observed in the DTG curves of the PUFs, which were clearly pronounced at ca. 200, 280, and 400 °C (
Figure 16). The first peak was due to the thermal dissociation of the urethane and urea bonds [
1], the second one was due to the decomposition of the chitosan rings [
11], while the third one was related to the decomposition of polyurethane to amine and carbon dioxide [
1]. The total decomposition was completed at a temperature of ca. 600 °C.
The DSC analysis of the PUFs evidenced the presence of volatiles in the unheated samples. The endothermic peak was observed within 15–110 °C in the first cycle, which was absent in the second cycle of the DSC profile (
Figure S2). This was caused by the remnant TEA used as a catalyst and the moisture absorption from the atmosphere. The glass transition temperature could be determined from the second cycle and was found to be in the range of 72–113 °C (
Table S3). The glass transition temperature allowed us to consider the materials as rigid PUFs.
The PUFs obtained here were compared with those obtained from the CS
WS [
33]. The PUFs obtained from the CS
WS were synthesized from the polyols obtained in GLYC by the hydroxyalkylation with GL and EC, following a two-step protocol, in the reactions with GL and then EC and in the reactions, the environment of which was initially EC. The results are collected in
Table 6. The PUFs obtained from the polyols obtained from the CS
O had a lower apparent density compared to that of the PUFs form the polyols obtained from the CS
WS. This led to a slightly lower compressive strength of the former.
More favorable functional properties of the foams based on both chitosans were obtained for the products synthesized in the GLYC, because it leads to the high functionality of polyols. The best mechanical and thermal properties had the PUFS obtained from the (CSo + GLYC + GL) + EC polyols with 2% water. They had low apparent density, water uptake, and heat conductance coefficient. They also showed the lowest mass loss upon heating, while their compressive strengths increased upon the thermal exposure. In the case of other foams, we conclude that the chitosan-containing PUFs obtained from the water-soluble chitosan had better functional properties compared to those of the PUFs obtained from the oligomeric chitosan. Mainly, they had a higher thermal resistance measured by the mass loss, compressive strength, and water uptake. This is probably related to the higher molecular weight of the water-soluble chitosan, which translates into longer chains of the chitosan built into the PUF, rendering a more rigid and thermally resistant PUF. Thus, the best chitosan substrate for the PUF is the water-soluble chitosan due to its higher average molecular weight.
In summary, it can be stated that so far, the limitation of the use of chitosan for the production of polyurethane foams was the lack of appropriate solvents and difficulties in carrying out the chitosan hydroxyalkylation reaction with the use of oxiranes towards obtaining polyols. We found that glycerol and glycidol as hydroxyalkylating agents and then ethylene carbonate solved this problem in the case of the water-soluble chitosan.
Another problem that needs to be solved is the assessment of the possibility of using higher-molecular-weight chitosans for the production of polyols and polyurethane foams. We found here and in the previous studies that chitosan with a higher molecular weight could lead to PUFs of enhanced mechanical and higher thermal resistance. The synthesis of a polyol from a chitosan with a high molecular weight and GL as a reagent and a solvent was not possible due to insolubility of chitosan in such conditions, as was determined in the previous study [
33]. However, it is possible to convert a high-molecular-weight chitosan into a polyol using a mixture of glycerol and glycidol as hydroxyalkylating agents and further hydroxyalkylation with EC to obtain a liquid polyol (similar to synthesis 2). The second method of the chitosan conversion into a polyol is based on preliminary dissolving the chitosan in a mixture of glycidol and ethylene carbonate and executing its hydroxyalkylation in the EC solvent. The second step is the further hydroxyalkylation of the product with the excess of EC at a higher temperature (similar to synthesis 3). These two synthesis routes were tested by us and made it possible to obtain polyurethane foams from the derived polyols. Detailed studies of the properties of the polyurethane foams based on chitosans with higher molecular weights will be conducted and published separately. These materials can find industrial applications due to their enhanced thermal resistance as thermal isolating materials for hot water or even overheated water vapor tubing.