Drying of the Natural Fibers as A Solvent-Free Way to Improve the Cellulose-Filled Polymer Composite Performance

When considering cellulose (UFC100) modification, most of the processes employ various solvents in the role of the reaction environment. The following article addresses a solvent-free method, thermal drying, which causes a moisture content decrease in cellulose fibers. Herein, the moisture content in UFC100 was analyzed with spectroscopic methods, thermogravimetric analysis, and differential scanning calorimetry. During water desorption, a moisture content drop from approximately 6% to 1% was evidenced. Moreover, drying may bring about a specific variation in cellulose’s chemical structure. These changes affected the cellulose-filled polymer composite’s properties, e.g., an increase in tensile strength from 17 MPa for the not-dried UFC100 to approximately 30 MPa (dried cellulose; 24 h, 100 °C) was observed. Furthermore, the obtained tensile test results were in good correspondence with Payne effect values, which changed from 0.82 MPa (not-dried UFC100) to 1.21 MPa (dried fibers). This raise proves the reinforcing nature of dried UFC100, as the Payne effect is dependent on the filler structure’s development within a polymer matrix. This finding paves new opportunities for natural fiber applications in polymer composites by enabling a solvent-free and efficient cellulose modification approach that fulfils the sustainable development rules.


Method parameters
In this method, the near-infrared region of the electromagnetic spectrum was used. The measurements were carried out in the range of 10000-4000 cm -1 in adsorption mode-64 scans (Thermo Scientific, Nicolet 6700). Cellulose fibres were dried for 24 h at 100 °C (Binder® oven; crystallizer 70x40 mm) before being analysed.

NIR Characterization
The processes of water adsorption and desorption has been also tracked with the employment of NIR technique which, according to literure, is more fragile to polar groups [1][2][3]. Therefore, it seems as a perfect tool in the cellulose moisture content investigation [4]. Wavenumber [cm -1 ] Chemical group Ref.
Considering the data gathered in Figure 1, which describes the changes in the water content during the moisture adsorption process, some variations are visible. Firstly, the intensity increases at 6709 cm -1 and 5174 cm -1 are revealed. These signals are assigned to the adsorbed moisture [6,7]. Nevertheless, different absorption bands are also influenced by the changes in cellulose water uptake, e.g., 4751 cm -1 (-OH moieties) [5].
Furthermore, regarding Figure 2 describing the moisture desorption process, also some changes in the intensities of 6709 cm -1 and 5174 cm -1 might be noticed. Nevertheless, differences observed at 6709 cm -1 signal are slighter than the ones visible in Figure 15. This may indicate some information about the diversified rates of water adsorption and desorption processes, as NIR technique is more fragile to changes in moisture content [9]. On the other hand, observed results might be the due to the water evolution of different states [10][11][12].
Moreover, considering data gathered in Figure 1 and Figure 2, the water content is possible to be tracked with 5174 cm -1 absorption band as the intensity changes between the analysed fibres are possible to be distinguished. Furthermore, contrary to data obtained with the FT-IR, according to presented NIR spectra, it may be said that the water desorption process rate is faster at the beginning. The confirmation of this phenomenon could be the fact that intensity at 5174 cm -1 drops significantly after 45 min of cellulose drying (Figure 2). However, regarding the same signal in Figure 1, it is clearly seen that the intensity increase caused by the water adsorption seems to be steadier and slower.

Experiment conditions
Cellulose fibres were dried for 24 h at 100 °C (Binder® oven; crystallizer 70 × 40 mm) before being analysed. Differential scanning calorimetry (DSC) investigation has been performed in a temperature range from -20-200 °C (heating rate: 10 °C /min; Ar 60 cm 3 /min) prior to analyse the water evaporation process establishing its enthalpy (ΔH) and temperature of the peak (Tpeak). Here, as well, Mettler Toledo TGA/DSC 1 STARe System equipped with Gas Controller GC10 has been employed.

DSC analysis
According to the performed differential scanning calorimetry analysis, the water evaporation process may be observed. The endothermic peak visible in Figure 3 is assigned to the moisture desorption from the surface of the analysed biopolymer [13,14].
What should be underlined, the significant difference in the sample thermal behaviour between the dried and not dried cellulose may be detected. Considering UFC100/ND, a crucial increase in the enthalpy assigned to the water evaporation might be noticed, while in case of UFC100/D/1440 the endothermic peak is only slight. This proves the good efficiency of thermal drying. Moreover, according to the data given in Table 2, the difference in the enthalpy value is also significant. ΔH in case of UFC100/ND specimen is approximately 4 times higher in comparison with UFC100/D/1440. Elevated enthalpy change value means more energy required to desorb the water from the cellulose fibre surface as dehydration heat increases considerably when the sample exhibits raised moisture level [15]. What is also interesting, the Tpeak of water evaporation shifts from 98°C (UFC100/ND) to 105 °C has been detected (UFC100/D/1440). This could be due to some structural changes caused by the thermal treatment [16,17]. Thermogravimetric analysis has been employed in order to investigate the thermal decomposition of analysed polymer composites. In the Figure 4, TGA curves which are typical for ethylene-norbornene copolymer filled with cellulose fibres might be observed. It consists of two significant decomposition steps. First one is considered to be connected with the biopolymer thermal degradation and the second one-assigned to the polymer matrix disintegration. Moreover, also a TGA curve of neat polymer matrix (TOPAS) is presented. In this case only one decomposition step around 350 °C is evidenced.
According to the data given in Table 14, it may be observed that filled systems exhibit lower T05%, which is considered to be the initial decomposition temperature, in comparison with the neat polymer matrix. Furthermore, it is clearly visible that the most dynamic mass loss of TOPAS is detected from 350-500°C. It proves higher thermal resistance of neat polymer matrix while comparing with the other analysed composite samples.
On the other hand, in case of filled systems, the most dynamic mass loss for investigated composite samples is detected between 300 °C and 500 °C. This phenomenon has been reported before in literature [19]. Furthermore, around 360 °C cellulose reveals a significant drop (mass loss of almost 80%) [20]. Nevertheless, all investigated samples of cellulose-filled TOPAS, according to data gathered in Table 3 and Figure 4, follow the same degradation path and the differences between the specimens are not relevant. Furthermore, it may be noticed that from approximately 450 °C on, all analysed samples begin to degrade in the same rate. This is the moment that cellulose fibres are fully degraded.  Moreover, Table 4 reveals activation energies assigned to cellulose degradation (EA1) and polymer matrix decomposition ( 2 ) which have been calculated according to the Broido's approach [18]. It may be observed again that drying has almost no impact on the degradation behaviour of analysed composite samples. Moreover, what should be underlined, the activation energy of neat TOPAS degradation is comparable with the activation energy values presented for different composites. Therefore, it may be stated that the degradation of polymer matrix itself is not affected by the presence of cellulose fibres and the reason for lower cellulose-filled polymer composites thermal stability is the presence of biopolymer. Decreased thermal resistance of cellulose-filled samples has been also reported in other research study [21]. Moreover, this could be explained by the lower thermal stability of cellulose in comparison with the polymer matrix [22] which is a consequence of -OH groups presence [23].

Experiment parameters
Differential scanning calorimetry (DSC) investigation has been performed in a temperature range from −40-200°C (heating rate: 10 °C /min; argon atmosphere) prior to analyse changes in glass transition temperature of ethylene elastic segments (Tg1), glass transition temperature of rigid norbornene segments (Tg2) and softening enthalpy (ΔH). Here, as well, Mettler Toledo TGA/DSC 1 STARe System equipped with Gas Controller GC10 has been employed.

DSC analysis
Differential scanning calorimetry (DSC) was used in order to assess the filler influence on the glass transition temperature process of ethylene (Tg1) and norbornene (Tg2) segments, as well as, the phenomenon of material softening and its enthalpy change value (ΔH). In Figure 5 DSC curves of composites filled with neat cellulose fibres dried for different times are presented. Firstly, one may observe that the curve shapes are similar to each other. However, some differences in softening enthalpy value and shifts in glass transition temperatures between neat polymer matrix and filled TOPAS are detected. Giving a closer look on data gathered in Table 5, it may be observed that there are some major differences between the neat polymer matrix and cellulose-filled TOPAS, namely, Tg1 fall from 9°C (TOPAS) to approximately 3.5°C (filled composite), enthalpy change from 54 J/g (TOPAS) to roughly 41 J/g (filled composite). Yet, the impact of UFC100 incorporation on norbornene segments glass transition temperature (Tg2) is not relevant and there are no variations considering temperature of this transition (Tpeak). This phenomenon has been observed before in different studies [24][25][26]. Moreover, described above decrease in softening enthalpy (ΔH) also has been reported before [27,28].

. SFE investigation
Due to the cellulose drying time, some variations of composite samples surface free energy have been detected. Similar changes have been also reported in literature [30]. Regarding Figure 6, neat polymer matrix exhibit the surface free energy which is approximately E = (40 ± 1 ) mJ/m 2 . Consequently, it may be claimed that E drops significantly while TOPAS is loaded with cellulose fibres. All polymer composites filled with natural fibres dried for different times exhibit a surface free energy of almost 30 mJ/m 2 . Figure 6. Surface free energy of investigated composites filled with modified cellulose fibres: a) not dried before the modification process, b) dried before the modification process.
Moreover, the polar part of surface free energy, in case of unfilled TOPAS, is (0.10 ± 0.06) mJ/m 2 . Ep values obtained in case of the composites filled with not dried cellulose fibres, either dried for 45 min and 180 min, are lower than the surface free energy polar part assigned to neat polymer matrix. However, TOPAS + UFC100/D/1440 specimen exhibits the elevated polar part of surface free energy value of approximately 0.2 mJ/m 2 .
Described above changes may be important considering the wetting of prepared composite materials, e.g., the lower the surface energy is, the higher contact angle value in case of water [31]. What is more, the polar and dispersive components of surface free energy are also crucial in case of wetting [32]. Due to their ratio, prepared composite would be more or less fragile to polar either nonpolar solvents [33].