Effect of Thermal Aging on the Physico-Chemical and Optical Properties of Poly(ester urethane) Elastomers Designed for Passive Damping (Pads) of the Railway

The aim of this study consists of monitoring the effect of thermal aging on the physico-chemical and optical properties of poly(ester urethane) elastomers designed as damping materials for railways. The materials were obtained by polyaddition in two stages in melt, resulting in regular structures. The structural modifications during the thermal aging of the samples were monitored using FTIR, color changes, TGA in non-isothermal and isothermal conditions, DSC and physico-mechanical measurements. The structural regularity of the rigid and flexible segments maintained the good mechanical properties of the structures up to 200 h of thermal aging at the elevated temperatures of 40 °C, 70 °C, 100 °C and 130 °C. It was observed that at 40 °C and low exposure times, changes occur mainly to the carbonyl groups of the soft segments. At higher temperatures and longer exposure times urethane groups were affected. Extended thermal aging led to significant changes in thermo-mechanical and optical properties.


Introduction
Polyurethanes (PURs) are a class of polymers containing the urethane segments (-NHCOO-) in the main chains. This structure assures very good chemical resistance and mechanical properties [1][2][3][4][5][6]. Polyurethane elastomers (PUREs) have known increasing interest because of their efficient cost-effective production process and recyclability, as compared to vulcanized elastomers. PUREs possess elastomer characteristic properties in normal conditions and plasticize during heating, thus they may be processed by conventional methods used for thermoplastics, such as extrusion, injection, molding or calendering [7].
In a first instance, the unique properties of PURs are dictated by microphase separation, generated by the mutual immiscibility between hard and soft segments, owed to thermodynamic incompatibility. Secondly, the PURs properties are also influenced by the strong intermolecular hydrogen bonding in the hard segments, with bonding energy The hard segments constituted of 1,6-hexamethylenediisocyanate (HDI) and 4,4 -diphenyl methane diisocyanate (MDI). BD and HD were purchased from Sigma-Aldrich and used as received. PEA, MDI and HDI were purchased from Fluka (Fluka, St. Gallen, Switzerland), the latter two being used as received. PEA (M w = 2000 g mol −1 ) was dried prior to use, at 120-130 • C for 2 h under a vacuum of 2 mm Hg. Dibutyltindilaurate (DBTDL) 95%, as catalyst, was purchased from Sigma-Aldrich (Sigma-Aldrich, St. Gallen, Switzerland) and used as received.

Preparation of the PEURs
A series of six PEURs were obtained (Scheme 1). In a first stage, the precursors for the 6 PEURs were synthesized following the pre-polymer method [21]. For the synthesis of the NCO-terminated pre-polymer appropriate quantities of PEA and diisocyanate were reacted at a specified NCO/OH equivalent ratio [35]. For this purpose, these quantities were added into a 500 mL glass vessel, equipped with an oil bath, thermometer, mechanical stirrer and gas inlet and outlet through which nitrogen was purged continuously. The synthesis was conducted under stirring for 3 h at 80 • C.
The PEURs were obtained using the polyaddition technique in melt and in two steps. For this purpose, 0.01 mol of pre-polymer (polyester) was introduced into a reactor equipped with a stirrer, vacuum plant and nitrogen tube, heated up to 125 • C under stirring and a residual pressure of 2-3 mm Hg for 2 h. The temperature was lowered to 80 • C and 0.02 mol of diisocyanates (MDI or HDI) and 0.001 g of dibutyltindilaurate (DBTDL), as catalyst, were introduced under a stream of nitrogen. The reaction was continued under stirring for 1 h. Then 0.02 mol of diol or triol (BD or HD) as chain extender and Gly were introduced, with respect to MDI/PEA ratios, and stirring was continued for a short time. Films or castings were obtained from the synthesized polymer by heat treatment at 90 • C for 24 h. The reaction was monitored by FTIR (Bruker, Karlsruhe, Germany) until the band at 2200-2300 cm −1 , corresponding to free isocyanate stretching, completely disappeared. Afterwards, the PEURs were cast into a mold and cured 24 h at 100 • C, followed by postcuring at room temperature for 7 days when they were detached in the form of sheets. The composition of the PEURs is shown in Table 1. Hard segment composition was controlled through the polyester diol/MDI/diol and triol ratios used in the synthesis, these being 1:2:1 (22% hard segment) and 1:3:2 (32% hard segment), respectively. A 1:1 ratio of OH diol/OH Gly was used during the obtaining of each PEURs. The length of the soft segment was maintained constant, while the length of hard segment was varied systematically. FTIR spectra were obtained with a Bruker Vertex 70 spectrometer equipped with a Miracle TM (Karlsruhe, Germany) device with a diamond crystal for attenuated total reflectance (ATR) measurements. All spectra were acquired with a spectral resolution of 4 cm −1 .

Color Modification Measurements
The color changes during the thermal aging of the specimens were described using the CIEL * a * b * system, where L * is the lightness factor, a * is the redness factor and b * is the yellowness factor. The L * , a * and b * parameters were measured with a color comparison device (Pocket Spec Color QA, SUA with a sensor head of 6 mm). Measurements were made using D65 illuminant and a 10 • standard observer. The L * , a * and b * values were used to calculate the color change, ∆E as a function of the thermal aging time, according to Equation (1): In Equation (1) indexes 1 and 2 represent the values measured before and after thermal aging. Samples were analyzed for color changes according to DIN6174 (Farbaständen bei Körperfarbennach der CIELAB-Formel, 1979).

Thermogravimetric Analysis (TGA)
All TGA measurements (in non-isothermal and isothermal conditions) were performed on a STA 449F1 Jupiter device (Netzsch, Germany).

TGA in Non-Isothermal Conditions
TGA measurements in non-isothermal conditions were conducted from 30 • C up to 700 • C in nitrogen atmosphere (50 mL min −1 ) and at a heating rate of 10 • C min −1 . Approximately 20 mg of each sample were used.

TGA in Isothermal Conditions
For the isothermal measurements in air, the variation of sample mass was recorded at the elevated temperatures in the aging oven. Before thermal aging, samples were dried for 24 h at 23 • C in order to remove the moisture absorption during storage.
2.2.6. Differential Scanning Calorimetry (DSC) DSC measurements were undertaken on a DSC 200 F3 Maia device (Netzsch, Germany) at heating and cooling rates of 10 • C min −1 and -10 • C min −1 , in nitrogen atmosphere (50 mL min −1 ). Approximately 10 mg of each sample were used.

Mechanical Measurements
Mechanical measurements were undertaken on dumbbell-shaped samples (ISO 37 type 2) cut from the PEURs, obtained in the form of 2 mm thick sheets. The tests were performed at 25 • C on a Shimadzu EZTest device (Tokyo, Japan), equipped with a 5 kN load cell. The cross-head speed was 50 mm min −1 . The mechanical results were the mean of five measurements on identically shaped samples.

Structural Characterization by FTIR Spectroscopy
The FTIR spectra of the obtained PEURs ( Figure 1) show characteristic absorption bands: 3500-3200 cm −1 with a peak around 3340 cm −1 (free and bound -NH urethane bonds), 3000-2700 cm −1 (CH 2 ), 1750-1650 cm −1 with a peak at 1730 cm −1 (both C=O disordered and ordered urethane bonds (1706 cm −1 )), 1250-1050 cm −1 (C-O). The NH entity in PURs is characteristic to hard segments and may be conventionally used in the study of hard blocks orientation. The CH 2 group is used for orientation of soft segments, however CH 2 groups may also be present in low concentrations in the hard segments. Other entities include the aromatic group at 1600 cm −1 and -C-O-C-in the ester group at 1033 cm −1 . The absorbance ratio 1706 cm −1 /1730cm −1 increases with increasing concentration of rigid segments suggesting an increase in the proportion of bound urethane groups.
PURs is characteristic to hard segments and may be conventionally used in the study of hard blocks orientation. The CH2 group is used for orientation of soft segments, however CH2 groups may also be present in low concentrations in the hard segments. Other entities include the aromatic group at 1600 cm −1 and -C-O-C-in the ester group at 1033 cm −1 . The absorbance ratio 1706 cm −1 /1730cm −1 increases with increasing concentration of rigid segments suggesting an increase in the proportion of bound urethane groups.   Table 2 show the color parameters variation of the PEURs as a function of temperature and heating time.   Table 2 show the color parameters variation of the PEURs as a function of temperature and heating time.      - Red and yellow degradation intermediates accumulate on the surface of samples during the thermal treatment, as it can be seen from Figure 3, Figure 4 and Table 2. The displacements of ∆a * and ∆b * to either positive or negative values do not show any pattern. This is an indication of the formation of highly instable colored intermediates during thermal aging. Red and yellow degradation intermediates accumulate on the surface of samples during the thermal treatment, as it can be seen from Figure 3, Figure 4 and Table 2. The displacements of Δa * and Δb * to either positive or negative values do not show any pattern. This is an indication of the formation of highly instable colored intermediates during thermal aging.       In order to estimate the lifetime of the PEURs a simplified form of an Arrhenius equation based on time-temperature relation was used (Equation (2)) [36]. The time (t) required for 20% deterioration of a physical property, such as ΔE, at various temperatures is used in Equation (2), where t is the lifetime, t0 is the pre-exponential factor, E is the activation energy, T is the aging temperature and R is the gas constant.
t =t0 e E/RT (2) Aging times that correspond to an increase in ΔE value from 0 to 5, when the color differences became visible, are shown in Table 3, together with kinetic parameters values obtained from the Arrhenius plots.  In order to estimate the lifetime of the PEURs a simplified form of an Arrhenius equation based on time-temperature relation was used (Equation (2)) [36]. The time (t) required for 20% deterioration of a physical property, such as ∆E, at various temperatures is used in Equation (2), where t is the lifetime, t 0 is the pre-exponential factor, E is the activation energy, T is the aging temperature and R is the gas constant. In general, during thermal aging, polymers usually undergo chain scission, crosslinking or both. In order to obtain more in-depth information concerning the PEURs structures after thermal aging, thermal analysis was also implied, through correlating the TGA and DSC data. TGA experiments in air for the six PEURs were reported in a previous paper [21]. In general, as expected, depending on the chain extender, in air atmosphere the samples thermally degraded at lower temperature ranges (250-500 • C) and in three stages, compared to the TGA study in nitrogen, described below. The global mass loss values were also considerably higher in air atmosphere, leaving significantly lower residue values. Figure 6 shows the TGA curves ( Figure 6a) and their corresponding first derivatives (DTG) (Figure 6b). Experiments were conducted in inert atmosphere (nitrogen) in order to avoid oxidation and thus secondary undesired reactions during thermal degradation of the aged samples. Data extracted from these curves are given in Table 4. The temperature corresponding to 5% mass loss (T 5% ) was used as criterion for evaluating the thermal stability of the samples. In general, during thermal aging, polymers usually undergo chain scission, cross-linking or both. In order to obtain more in-depth information concerning the PEURs structures after thermal aging, thermal analysis was also implied, through correlating the TGA and DSC data. TGA experiments in air for the six PEURs were reported in a previous paper [21]. In general, as expected, depending on the chain extender, in air atmosphere the samples thermally degraded at lower temperature ranges (250-500 °C) and in three stages, compared to the TGA study in nitrogen, described below. The global mass loss values were also considerably higher in air atmosphere, leaving significantly lower residue values. Figure 6 shows the TGA curves ( Figure 6a) and their corresponding first derivatives (DTG) (Figure 6b). Experiments were conducted in inert atmosphere (nitrogen) in order to avoid oxidation and thus secondary undesired reactions during thermal degradation of the aged samples. Data extracted from these curves are given in Table 4. The temperature corresponding to 5% mass loss (T5%) was used as criterion for evaluating the thermal stability of the samples. Regardless of their composition, PURs are generally known as being thermally unstable, since the urethane bond's initial degradation temperature is dependent on the structure. In light of this aspect, the urethane bond may undergo three general thermal decomposition pathways, either individually or simultaneously. In a first instance, the urethane moiety witnesses scission to generate isocyanate and alcohol entities. A second thermal degradation pathway consists of the yielding of olefin and primary amines. The third instance resides in the formation of secondary amine and carbon dioxide evolvement [37,38]. Figure 6 and Table 4 show that the samples respect a four-stage decomposition pattern, except for PEUR4, which degrades in three stages and exhibits the highest thermal stability (T5% = 335 °C). Depending on their structures and the processes occurring during thermal aging, the PEURs thermal stability increases in the following order: PEUR5 < PEUR6 < PEUR3 < PEUR2 < PEUR1 < PEUR4. According to the thermal degradation data, Regardless of their composition, PURs are generally known as being thermally unstable, since the urethane bond's initial degradation temperature is dependent on the structure. In light of this aspect, the urethane bond may undergo three general thermal decomposition pathways, either individually or simultaneously. In a first instance, the urethane moiety witnesses scission to generate isocyanate and alcohol entities. A second thermal degradation pathway consists of the yielding of olefin and primary amines. The third instance resides in the formation of secondary amine and carbon dioxide evolvement [37,38].  Table 4 show that the samples respect a four-stage decomposition pattern, except for PEUR4, which degrades in three stages and exhibits the highest thermal stability (T 5% = 335 • C). Depending on their structures and the processes occurring during thermal aging, the PEURs thermal stability increases in the following order: PEUR5 < PEUR6 < PEUR3 < PEUR2 < PEUR1 < PEUR4. According to the thermal degradation data, in the first stage, samples start decomposing in the temperature range 304-335 • C, not uncommon for PURs, ending in the range 322-390 • C and with mass loss values in the interval 7.97-33.5%, the highest being registered for PEUR4, attributed to urethane bonds cleavage. Cleavage of urethane bonds further initiates depolymerisation with hard segment decomposition in the second stage [39]. This is followed by complex degradation through random scissions of polymer chains together with soft segments in the third and fourth stages [40,41]. For sample PEUR4, the second and third stages coalesce into a single one and with a very high cumulated mass loss (54.21%). This aspect, together with the highest recorded T 5% value is an indication of more intense cross-linking, compared to the other PEURs. The residual mass (M rez ) is very dependent on the hard segment structure, rich in chains of 6 CH 2 from HDI and HD which exhibit different behavior during thermal degradation. M rez values vary in the range 3.52 (PEUR4)-13% (PEUR5) ( Table 4). Figure 7 shows the DSC first heating (Figure 7a) and second heating curves (Figure 7b) of the studied compounds.
Polymers 2021, 13, x FOR PEER REVIEW 12 of 21 ment decomposition in the second stage [39]. This is followed by complex degradation through random scissions of polymer chains together with soft segments in the third and fourth stages [40,41]. For sample PEUR4, the second and third stages coalesce into a single one and with a very high cumulated mass loss (54.21%). This aspect, together with the highest recorded T5% value is an indication of more intense cross-linking, compared to the other PEURs. The residual mass (Mrez) is very dependent on the hard segment structure, rich in chains of 6 CH2 from HDI and HD which exhibit different behavior during thermal degradation. Mrez values vary in the range 3.52 (PEUR4)-13% (PEUR5) ( Table 4). Figure 7 shows the DSC first heating (Figure 7a) and second heating curves ( Figure  7b) of the studied compounds. Data extracted from the DSC measurements are given in Table 4. The DSC heating curves present only the glass transition temperature (Tg), associated with the soft segments, from −33 to −5 °C on the first heating, and from −35 to −7 °C on the second heating, due to little influence from hard segment domains and/or cross-linking. The lowest Tg values were recorded for the sample PEUR4. This signifies that PEUR4′s longer chain components endow the structure not only with enhanced cross-linking capacity, as shown from TGA data, but also with flexibility, due to more free volume between chain segments. As expected, the highest Tg values correspond to the more rigid aromatic structures of the PEUR5 and PEUR6 structures.  Data extracted from the DSC measurements are given in Table 4. The DSC heating curves present only the glass transition temperature (T g ), associated with the soft segments, from −33 to −5 • C on the first heating, and from −35 to −7 • C on the second heating, due to little influence from hard segment domains and/or cross-linking. The lowest T g values were recorded for the sample PEUR4. This signifies that PEUR4 s longer chain components endow the structure not only with enhanced cross-linking capacity, as shown from TGA data, but also with flexibility, due to more free volume between chain segments. As expected, the highest T g values correspond to the more rigid aromatic structures of the PEUR5 and PEUR6 structures.  T 5% -temperature corresponding to 5% mass loss; T onset -onset thermal degradation temperature; T max -temperature that corresponds to the maximum rate of decomposition for each stage evaluated from the peaks of the DTG curves; T f -final thermal degradation temperature; M m -mass loss rate corresponding to each thermal degradation stage; M rez -percentage of residue remained at the end of the thermal degradation process (700 • C); T g -glass transition temperature; T m -melting temperature; ∆H m -melting enthalpy.
Furthermore, the two PEURs do not exhibit any other transition than the T g specific to the soft segment domains. All the other structures show an endothermic peak, associated with a melting process (T m ) [42], in the range 46 • C-52 • C. However, these neither exhibit corresponding crystallization peaks on the cooling curves (not shown), nor melting reproducibility on the second heating curves, this being another aspect related to cross-linking phenomena. The highest T m (52 • C) and melting enthalpy (∆H m = 36.84 J g −1 ) values registered for sample PEUR4 demonstrate that its aliphatic chains from hard domains possess the highest capacity to form structural crystalline domains. The other three structures show significantly reduced ∆H m values, compared to PEUR4, with almost negligible differences between them (10.67 J g −1 for PEUR2, 9.039 J g −1 for PEUR3 and 8.141 J g −1 for PEUR1). Another aspect which may be due to cross-linking phenomena resides in the fact that none of the PEURs structures exhibited the so called exothermic "cold crystallization" transition on either of the heating curves. This transition is related to the minor structural reordering in soft segments [43].

TGA in Isothermal Conditions
The percent of mass variation as a function of isothermal aging time is given in Figure 8. Figure 8 shows that the PEURs are generally thermally stable up to 200 h aging time. Slight mass losses (0.1-0.6%) occur in the range 50-100 h, followed by avery slight increase. This is most probably due to some oxidation phenomena. In this case, no specific pattern was found in mass variation during thermal aging. Additionally, mass loss percentages are generally very low. Furthermore, the DSC data show seemingly increased T g values generally compared to non aged PEURs [21]. This means that cross-linking processes compete with chain scissions, owed to complex degradation phenomena.

Structural Modifications During Thermal Aging
The FTIR analysis showed that the changes in structural properties are generated by some structural modifications in the PEURs. As TGA indicated high thermal stability (approx. 300 °C), no major structural changes were observed during the thermal aging of PEURs (Figure 9). In this regard, Figure 10 shows an example of the specific changes in FTIR spectrum of PEUR1 after aging 200 h at the specified temperatures. It can be generally observed from Figure 10 that loss of urethane linkages occurs during heating, accompanied by the enlargement of the band from 3340 cm −1 , characteristic to the urethane NH moiety, and the slight decrease in the signal from 1730 cm −1 , specific to the urethane carbonyl group. The urethane group was cleaved during the thermal aging process. The difference FTIR spectra generated mainly medium intensity negative signals (the downward signals) due to the loss of carbonyl groups during thermal aging (1200-1100 cm −1 ). These highly unstable moieties are responsible for both color and mass variations.

Structural Modifications during Thermal Aging
The FTIR analysis showed that the changes in structural properties are generated by some structural modifications in the PEURs. As TGA indicated high thermal stability (approx. 300 • C), no major structural changes were observed during the thermal aging of PEURs (Figure 9). In this regard, Figure 10 shows an example of the specific changes in FTIR spectrum of PEUR1 after aging 200 h at the specified temperatures. It can be generally observed from Figure 10 that loss of urethane linkages occurs during heating, accompanied by the enlargement of the band from 3340 cm −1 , characteristic to the urethane NH moiety, and the slight decrease in the signal from 1730 cm −1 , specific to the urethane carbonyl group. The urethane group was cleaved during the thermal aging process. The difference FTIR spectra generated mainly medium intensity negative signals (the downward signals) due to the loss of carbonyl groups during thermal aging (1200-1100 cm −1 ). These highly unstable moieties are responsible for both color and mass variations.

Mechanical Properties
The mechanical behavior of PUREs depends on the intermolecular interactions between the hard segments. Physico-mechanical analyses show the extent of supramolecular changes in PURs occurred depending on their structural changes. Hard segments act as multifunctional connecting points that function both as physical and chemical crosslinks, while soft segments form an elastomeric matrix determining the elastic properties of PEURs. Three-component PEURs are organized into hard and soft segments consisting of isocyanate and chain extender on the one hand, and PEA chains on the other. The sequence distribution of these segments may also change with the composition and separation of the influence phases, as well as with the formation of the hard and soft phases. In addition, physical cross-linking through the rigid parts of diphenyl methylene also contributes to the maximum stress. In general, the superior mechanical properties of PEURs are the result of microphase separation. The rigid segment acts as a multifunctional cross-linker and is very important for the mechanical properties of the material. In general, the behavior of these materials depends on the size and concentration of rigid segment domains, the strength of the aggregation of rigid segments and the ability of elastic segments to crystallize under deformation. The length of the elastic segments also has a strong influence on the mechanical properties. Short elastic segments generally yield rigid materials with high modulus while long elastic segments give highly elastic materials that can induce crystallization of elastic segments under applied stress. The synthesized compounds, subjected to mechanical deformation in static regime, have both elastomeric and plastic behavior. The mechanical parameters depend on the structure and defects of the analyzed sample. The rigid domains of elastomers are responsible for their "strength". Their influence is strong because they are deformable and evenly distributed in the sample volume. The plastic deformation of the rigid domains reduces the concentration of internal energy, the result being the increase in the resistance and hardness of the tested material.
Strength, modulus and elongation are important for characterizing polymers depending on their structure. At room temperature, the PEA polymer is in a glassy or partially crystalline state (T g = −42 • C). Thus, flexible PEA-based segments induce greater elongations than other PURs. The introduction of chemical cross-linking in the PEUR structure prevents the mutual movement of the macromolecular chains subjected to deformation. Increasing the concentration of polymer bonds leads to an increase in the interpenetration of macromolecules and an overlap of chain fragments with the intensification of intermolecular bonds.
The mechanical properties of cross-linked PEURs were analyzed following the influence of the structure, quantity and length of the rigid segments (diisocyanate and chain extender). PEA at ambient temperature is in the solid state due to the fact that it possesses ester bonds in the structural unit. Its use in the synthesis of urethane copolymers should contribute to increasing the flexibility of chains undergoing mechanical testing. The tensile properties of PEURs block copolymers with various cross-linked hard segment content are shown in Table 5.  The physico-mechanical measurements whose results are presented in Table 5 show that an increase in the ratio of rigid segments in the polyesters leads to a decrease in tensile strength. Thus, an increase in the ratio of diisocyanate groups from 2 to 3 and chain extender groups from 1 to 2 leads to a decrease in breaking strength of about two times. The same can be seen in the case of the elongation at break where there is a decrease from 500% to 200%. This can be explained by the formation of hydrogen bonds due to the structure of the hard domains.
Therefore, continuing to increase the deformation for PEA, the forces that prevent the deformation by rearranging the macromolecules also increase. In tetramethylene monomer units, only hydrogen atoms are attached to the molecular chain.
The high molecular symmetry and mobility of the chain segments make them partially crystalline at room temperature. Because crystalline regions play a similar role to crosslinking in improving mechanical properties, the properties of crystallizable materials are superior to non-crystallizable ones. Varying the molecular masses of the small diol segment affects the physical and mechanical properties of PURs. In PUR segments, the mechanical properties are generally credited as a result of the pseudo reticular effects of hard segment aggregation. The domains of the hard segments develop different degrees of ordering or semicrystalline structures, which have been increased in the case of PURs by adding the chemical cross-linking effect.
The presence of amorphous materials can also be established within the domain of hard segments. Pseudo-cross-linking failure can begin in this region. From an engineering point of view, the final mechanical parameters of the synthesized PEURs (strength, elongation and breaking modulus) are of particular interest.
These are indicators that recommend the use of PEURs at static or dynamic loads. Long elastic segments and the appropriate hard segment content are required for PEURs with high final shape recovery. This confirms the existence of interactions between soft and hard segments in the molecular chain. The interaction becomes strong with the growth of hard segments. These indicate that there is a strong interaction or existence of physical crosslinks that lead to shape memory. Stabilization of PEURs by hydrogen bonds and induced dipole-dipole interactions of hard segments appear to be responsible for shape memory. It can be concluded that 96-98% of the maximum shape recovery can be obtained at a concentration of 15-35 wt% hard segments when the PEUR copolymer has enough tightly interacted hard segments capable of restoring it to its original shape. Shape memory PURs can be used to manufacture products with both deformation control and damping capacity ( Figure 11).
These are indicators that recommend the use of PEURs at static or dynamic loads. Long elastic segments and the appropriate hard segment content are required for PEURs with high final shape recovery. This confirms the existence of interactions between soft and hard segments in the molecular chain. The interaction becomes strong with the growth of hard segments. These indicate that there is a strong interaction or existence of physical crosslinks that lead to shape memory. Stabilization of PEURs by hydrogen bonds and induced dipole-dipole interactions of hard segments appear to be responsible for shape memory. It can be concluded that 96-98% of the maximum shape recovery can be obtained at a concentration of 15-35 wt% hard segments when the PEUR copolymer has enough tightly interacted hard segments capable of restoring it to its original shape. Shape memory PURs can be used to manufacture products with both deformation control and damping capacity ( Figure 11).

PEUR3 PEUR4
Polymers 2021, 13, x FOR PEER REVIEW 19 of 21 PEUR5 PEUR6 Figure 11. Variation of physico-mechanical properties with temperature and heating time.

Conclusions
Six PEURs were subjected to thermal aging at temperatures of 40 °C, 70 °C and 130 °C for 200 h. The FTIR spectra have indicated urethane bond scission during the thermal treatment. The destruction of the urethane bond advanced with temperature and expo- Figure 11. Variation of physico-mechanical properties with temperature and heating time.

Conclusions
Six PEURs were subjected to thermal aging at temperatures of 40 • C, 70 • C and 130 • C for 200 h. The FTIR spectra have indicated urethane bond scission during the thermal treatment. The destruction of the urethane bond advanced with temperature and exposure time increase. The lifetime prediction of the studied samples was conducted following the color changes during the thermal aging. The lifetime calculation was made with an Arrhenius type relation. The structural changes were evidenced by FTIR spectroscopy. The best results were obtained for PEURs synthesized from 4,4 -diphenyl methane diisocyanate and 1,4-butanediol. The evolution of the mechanical properties coincides perfectly with the structural variations, chemical changes deduced from the FTIR and thermal characterization. The degradation of the PEURs was essentially manifested by modification in the intensity of some urethane specific signals. From the mechanical analysis, it was concluded that maximum shape recovery, up to 96-98% may be obtained at a concentration of 15-35 wt% hard segments, when the PEURs develop sufficient physical interactions capable of restoring their original shapes after mechanical deformation. Sample mass variation of PEURs during isothermal aging was analysed. All characterization results were in good agreement and showed a variation in thermal, mechanical, optical properties and mass. This led to a competition between scission and cross-linking reactions during thermal aging.
Author Contributions: All authors have contributed to the ideas, methods, data interpretation, results dissemination and writing of the manuscript. All authors have read and agreed to the published version of the manuscript.