Effect of LbL Deposited Chitosan-Nanosilica Bilayers on Flammability and Thermal Properties of Polylactide Materials

A processing line consisting of a Brabender DR20 feeder (RHL-Service, Poznań, Poland), a Haake Rheomex OS PTW 16/25 twin-screw extruder (RHL-Service, Poznań, Poland), a Zamak W1500 cooling bath (Zamak Mercator, Skawina, Poland) and a Zamak G-16/325 peletizer (Zamak Mercator, Skawina, Poland) was used in the high-temperature processing (Figure 1).

With this laboratory processing line, PLA and PLA/A200 samples containing 5 wt.% of filler were obtained. Standard samples for LOI and UL94 testing were obtained using a P-200 compression press (Zamak Mercator, Skawina, Poland). All processing conditions are shown in Table 1.

6.1 Thermogravimetry analysis (TGA)

The results of the thermogravimetric analysis illustrate the two-stage degradation of the biomaterials (Figure 6). The course of all thermograms is very similar to each other except for the reference sample. It can be assumed that the effect of CS and NS is small on the thermal properties of the biomatrix. However, there are two areas where differences in the mass-temperature relationship are noticeable. The first is the area measured up to the T_5% point, where the thermal stability of PLA is about 10℃ lower than samples containing bilayers (Table 2).

61.png

62.png

Figure 6. DTG (A-B) and TGA (C-D) curves for PLA and PLA/A200 samples deposited with 5, 10 and 15 CS and NS bilayers

Table 2. TGA indices determined for PLA and PLA/A200 samples based on Figure 6

In the area of highest mass loss, the thermograms almost overlapped up to the T_ENDSET point. On the other hand, the second is the area of charring afterburning noted from the T_ENDSET point to ~500℃. It is in this area that the real impact of bilayers on the PLA combustion mechanism can be seen. The differences are mainly due to the type of substrate used and the number of deposited layers. Figures 6A and 6C show graphs for PLA samples deposited with bilayers. The peak of the second stage of decomposition for PLA was at 470℃, and it shifted toward lower temperatures as the number of applied layers increased. Homogenization of NS in the PLA melt resulted in only a few degrees of shift in the peak point of the mass loss rate (Figure 6B and 6D). In contrast, the increase in the deposited layers tended to shift this peak toward higher temperatures, reaching as high as 490℃. In other words, adsorption of CS and NS bilayers onto the PLA surface leads to acceleration of the material's charring process, while adsorption of the same layers onto the PLA/A200 composite surface promotes delayed char formation.

The form of the substrate as well as the amount of deposited bilayers also affects the amount of carbon residue after thermo-oxidative decomposition. Pure PLA burns completely with the release of volatiles, yielding no residue. In contrast, covering the surface of the biopolymer with 5 and 10 bilayers leads to a solid residue of about 1%. A threefold increase in the concentration of CS and NS on the PLA surface also results in a nearly threefold increase in the char generation. Since pure PLA decomposes completely into gaseous products, it can be surmised that close to 5% of the decomposition residue of the PLA/A200 composite comes from the 5% share of NS in PLA. Meanwhile, as the layers applied onto the PLA/A200 biocomposite surface increase, the amount of carbon residue at 600℃ gradually decreases. The observed trend can be explained by the effect of entrainment of light NS nanoparticles, simultaneously by the stream of oxygen flowing in the measurement chamber and escaping volatiles. Note that NS has a specific surface area of 200 m²/g, an average particle size of 12 nm, and a low bulk density of 50 g/l [50], making it a highly dusty material, especially under fluctuating mass and temperature conditions. Thus, the higher the concentration of CS-NS on the PLA/A200 surface, the higher the residue loss.

6.2 DSC vs TGA

The objective of differential scanning calorimetry (DSC) was to study the effect of the number of adsorbed bilayers on the phase transformations of the polylactide observed with a constant rise in temperature. Importantly, DSC curves were contrasted with TGA thermograms over the same measurement range (Figure 7). This gives an idea of the scale of mass loss during the occurrence of all the phase transformations shown.

7.png

72.png

Figure 7. DSC curves for the samples: (A) PLA; (B) PLA/A200; deposited with 5, 10, and 15 bilayers along with the corresponding mass loss curves from 30-200℃

Heat flow curves for all analyzed biomaterials over the entire measurement range were characterized by the dominant phenomenon of energy absorption by the system. The biomaterials showed a negative energy effect throughout the entire cycle of the conducted test. As the number of bilayers applied onto the surface of PLA and PLA/A200 increased, the negative energy effect decreased relative to pure PLA. At 61-64℃, the midpoints of inflection of the curves were recorded, which were noted as the glass transition temperature of the material (Table 3). The addition of NS to the biopolyester matrix as well as the coating of the surface with CS and NS layers had virtually no effect on the shift of the transition point from the glassy state. However, the surface modification carried out had a weak but noticeable effect on the energy effect accompanying this transition. This means that covering the surface of PLA, ten or more bilayers makes it more difficult to melt the sample. In turn, to melt composite samples covered with bilayers, less energy must be supplied than is the case with the reference sample.

Table 3. DSC indicators where: melting temperature (T_m), melting enthalpy (ΔH_m), change in the heat capacity (ΔC_P), cold crystallization temperature (T_cc), and glass transition temperature (T_g)

With a further increase in temperature, cold crystallization appeared. This phenomenon observed in polyester materials, among others, is associated with a rather delayed response of the material to changes in its environment. The fact is that with the number of applied bilayers, there is a shift in the peak of cold crystallization by up to several tens of degrees towards lower temperatures. This may be a rationale for supposing that the presence of CS microparticles and NS nanoparticles not only within the condensed phase but also on its surface, may contribute to the nucleation of crystalline phases within the near-surface layer rather than the core of the sample (accelerating the occurrence of the cold crystallization effect at lower temperatures).

A further phase transformation observed was the melting of the samples. PLA and PLA/A200 samples containing 10 and 15 bilayers on the surface were characterized by two characteristic melting peaks, the others by only one. PLA can crystallize in three crystalline forms [51-56]. The most common and stable polymorph is the α form, which is most often formed when PLA crystallizes from a melt. The β form can be obtained using high tensile and high-temperature conditions. The γ form can be obtained by epitaxial crystallization on a hexamethylbenzene substrate. Focusing on the crystallization of PLA from the melt as we have in this case, PLA can also crystallize into a disordered crystal called α'. This α' metaphase has an identical conformation to the α form, but looser packing and lower density [57, 58]. When crystallization occurs at below 115℃ both α and α' phases are formed. At the same time, if cold crystallization occurs below 110℃ the dominant phase is the α phase, while when crystallization occurs at temperatures of 110-115℃ the dominant phase is the metastable α' phase [58, 59]. According to the thesis, deposition above 10 bilayers onto PLA promotes nucleation of the dominant α' metaphase, while adsorption above 10 bilayers on the PLA/A200 surface mainly results in nucleation of the stable α phase. The number of applied bilayers or the type of substrate used does not affect the melting temperatures of these phases.

It is also worth referring to the mass loss in the measurement range in question. Pure PLA, with a constant increase in temperature, loses mass linearly until it reaches 1.3% at 200℃, which may be related more to moisture loss and small initial degradation of the sample. Adsorption of bilayers on PLA initially promotes a small loss of mass, followed by a small increase in mass in the melting and degradation range. This swelling effect was observed only for PLA samples (the greater adsorption, the greater swelling). PLA/A200 composites practically in the range up to 200℃ were stable in terms of mass (∆M=0.3%), and the number of deposited layers did not affect its loss.

6.3 Micro Combustion Calorymetry (MCC)

Measurement of the Heat Release Rate (HRR) as a function of time and temperature (Figure 8) is characterized by several indicators (Table 4): PHRR - the peak of the HRR curve, TTI - Time To ignition, TOF - Time Out Flame.

81.png

82.png

Figure 8. Heat release rate versus temperature and time for: A) PLA deposited by 5, 10 and 15 bilayers; B) PLA/A200 composites PLA deposited by 5, 10 and 15 bilayers

The PHRR for PLA was 584 W/g. The addition of 5 wt.% NS to PLA resulted in a 16% loss of PHRR. Virtually the same effect can be obtained by applying 10 bilayers of CS and NS onto the PLA surface. The LbL technique used a 10 times lower weight of NS than during homogenization in the melt. Interestingly, adsorption of 5 and 10 bilayers on PLA/A200 did not affect the flammability of the composite sample at all. The best PHRR reduction effect was obtained for all samples coated with 15 bilayers (22% flammability reduction).

Table 4. Summary of basic flammability indices for PLA/A200 samples

The presence of CS and NS also affected the flash point and quench point of the burned samples. All of the samples with bilayers ignited faster than PLA and generally extinguished later. Thus, the consequence of lowering PHRR relative to PLA was to increase the combustion time of biomaterials. No less the effect of the number of layers applied on the combustion time was small, as very similar results were obtained for both (5BL) and (15BL) samples. However, it is worth noting that much better results were obtained by the LbL technique deposition than by homogenization of the nanofiller in the biopolymer melt.

6.4 MCC vs TGA

Using ASTM D7309, additional flammability indices such as fire growth capacity (FGC) and heat release capacity (ηc), specific heat released (hc also known as Total Heat Released - THR), pyrolysis residue, and specific heat of combustion gases were determined. The values of these indicators are summarized in Table 5.

The addition of NS to the polyester melt did not significantly change the start fire ability, but the application of 5, 10, and 15 bilayers onto the surface of PLA and PLA/A200 increased this ability twofold. This confirms the previously stated characteristic of the bilayers to ignite earlier than the core of the sample. At the same time, all samples showed a reduced heat release capacity compared to PLA. The higher the number of applied layers, the lower the η_c values. As a result, a nearly 25% reduction in η_c was recorded for the (15BL)PLA and (15BL)PLA/A200 samples. An identical relationship was observed during h_c analysis. With the application of 15 bilayers (regardless of the substrate) resulted in a 20% reduction in h_c relative to PLA.

Table 5. Flammability factors for polylactide biomaterials calculated according to ASTM D7309 standard

As indicated by thermogravimetric studies, PLA decomposes completely into gaseous products, leaving no solid residue. The incorporation of NS nanoparticles into the biopolymer resulted in the highest recorded values of the solid residue after thermo-oxidative decomposition. Also, the results of the MCC analysis confirm the previous relationship. With that said, a 6-fold increase in the heating rate of the PLA/A200 composite generates a 16-fold higher solid residue after combustion. Covering the surface of PLA and PLA/A200 biocomposite with 5 bilayers of CS and NS is ineffective, as these samples also burn completely to gaseous products along with which light NS particles leave the system. Similar conclusions were made for the (10BL)PLA/A200 sample. The most effective solution turned out to be covering both surfaces with 15 bilayers of CS and NS. In both cases, there was a change in the combustion mechanism of the continuous phase and the generation of solid char. This may mean that the real impact on the formation of solid char has not only the synergistic action of Si-O-Si and NH-CO groups alone but de facto their respective amounts.

One of the most important characteristics of a material that controls its fire properties is the amount of gas phase and condensed phase released. Reducing the amount of gas phase released is one of the most effective approaches to blocking combustion. Increasing carbonization reduces the amount of gas and condensed phase released and increases the amount of residue. Chitosan is mainly responsible for the charring that occurs, which was also confirmed in a publication [60].

91.png

92.png

Figure 9. Reaction Rate calculated from TGA and MCC results for A) PLA deposited by 5, 10, and 15 bilayers; B) PLA/A200 composites PLA deposited by 5, 10, and 15 bilayers

Figure 9 shows the decomposition reaction rates (RR) determined for PLA and PLA/A200 composites based on MCC and TGA measurements. The differences between the observed peaks of the RR curves for the different measurement techniques are no more than 20℃. This means that regardless of the measurement method used, and the heating rate of the sample, all materials undergo complete decomposition in virtually the same temperature range. In contrast, the presents of filler and applied bilayers manifests a slight variation in these values.

Comparison of the results obtained from the two measurement methods (TGA and MCC), is also possible by determining the values of Overall Thermal Stabilization Effect (OSE) and Overall Flame Retardancy Effect (OFRE) in the same temperature range using the following Eqs. (1) and (2) [61-63]:

$\begin{aligned}

& \text { OSE } =\sum_{T=150}^{600}\left(\left(mass \ percent \ of \ polymer \ material { }_T\right)\right.  \left.-\left(mass \ percent \ of \ polylactide{ }_T\right)\right)

\end{aligned}$     (1)

$\begin{aligned}& \text { OFRE } =\sum_{T=150}^{600}\left(\left(H R R \ percent \ of \ polylactide { }_T\right)\right.  \left.-\left(H R R \ percent \ of\ polymer \ material { }_T\right)\right)

\end{aligned}$     (2)

10.png

Figure 10. Overall Thermal Stabilization Effect (OSE) and Overall Flame Retardancy Effect (OFRE) for all tested PLA materials

The resulting OSE and OFRE ratios relating to the reference sample are summarized in Figure 10. Studying the effect of LbL-deposited bilayers on two types of surfaces, pure PLA and PLA/A200 nanocomposite, an identical relationship was found: the more layers adsorbed, the lower the OSE, and vice versa the more layers, the higher the OFRE. In contrast, the lower the thermal stability, the faster the ability to form char, and the more solid residue, the lower the flammability and longer the burning time.

6.5 UL-94 test

Table 6 shows the results of the UL-94 test in the vertical mode. When this measurement was carried out, there was a decrease in the viscosity of the biomaterial due to the rapid increase in temperature. As a consequence, hot melt sprang from all the tested samples without exception, which ignited the cotton placed under the samples. The base sample burned for less than a minute, and the addition of 5 wt.% nanofiller extended the average burning time of the composite by only 30 s relative to PLA. In general, as the number of applied layers increased, the combustion time of biomaterials increased. Applying 15 bilayers onto the PLA surface resulted in a doubling of the combustion time of the samples. Although among PLA/A200 composites the adsorption of 15 bilayers also proved to be the best, it did not achieve the same effect as on the PLA surface. The classification carried out quite transparently shows that in the case of composite samples containing applied bilayers, it was not possible to classify into any of the given standard categories. Based on this summary, it can be concluded that the adsorption of CS and NS in the form of layers did not reduce the flammability class of both polylactide and biocomposite materials.

The vertical orientation of the sample promotes rapid movement of the flame from the bottom to the top. Virtually all of the heated pure PLA flows with the flame onto the fabric beneath it. The sudden temperature change reduces the viscosity of the melt, which lacks the additional forces that maintain the geometry of the sample. The use of NS prolongs the cohesion effect but is still not sufficient to effectively counteract gravitational forces. The presence of bivariate layers on the PLA/A200 sample or 10BL and 15BL on the PLA, leads to visible effects in a relatively short time. As the number of falling droplets decreases, the burning time of the samples increases. This is because the NS particles in the CS combination form a physical insulator of the sample core against access to the flame. When the fabric is heated, both the biomatrix and low-molecular-weight organic compounds are rapidly decomposed into gaseous products. The gaseous products cause the mineral layer in between to clump together, creating temporary scaffolds that hold the sample geometry together for a much longer time.

Table 6. Measurement results using the UL-94 VB mode

Table 7. Measurement results using the UL-94 HB mode