Produced From Plant Biomass . Part II : Evaluation of Crystallinity and Degradation Kinetics of Cellulose

In this study Eucalyptus grandis (CEG) and Pinus taeda (CPT) cellulose fibers obtained from kraft and sulfite pulping process, respectively, were characterized using Fourier transform infrared (FTIR) spectroscopy and thermogravimetry (TGA). The degradation kinetic parameters were determined by TGA using Coats and Redfern method. FTIR results showed that CPT presented a more ordered structure with higher crystallinity than CEG. Thermogravimetric results showed that CPT had a higher thermal stability than CEG. The kinetic results revel that for CEG the degradation mechanism occurs mainly by random nucleation, although phase boundary controlled reactions also occurs while for CPT the degradation process is more related with phase boundary controlled reactions. Results demonstrated that differences between thermal stability and degradation mechanisms might be associated with differences in the cellulose crystalline structure probably caused by different pulping processes used for obtaining the cellulose fibers.


Introduction
Cellulose fibers are extensively used to reinforce polymeric composite materials [1][2][3][4] .The considerable interest in these fibers is due to their biodegradability, renewability, low density and mechanical properties comparable to those inorganic fibers 5 .Cellulose is a natural polymer consisting of D-anhydroglucose (C 6 H 11 O 5 ) repeating units joined by 1.4-β-D-glycosidic linkages at C1 and C4 position 6 .Each repeating unit contains three hydroxyl groups.These hydroxyl groups and their ability to hydrogen bond play a major role in directing the crystalline packing and also govern the physical properties of cellulose 6 .Solid cellulose forms a microcrystalline structure with regions of highly order i.e. crystalline regions and low order regions i.e. amorphous regions.
In native cellulose, the crystalline regions are cellulose I, which consists of two allomorphous phases.The cellulose Iβ, characterized by a monoclinic unit cell with two polymer chains in a parallel arrangement, and cellulose Iα, with a shift of the polymer chains along the chain axis, resulting in a triclinic unit cell 7,8 .The relative amounts of phases Iα and Iβ depend on the origin and the chemical treatment of the cellulose.
Changes in cellulose fibers and their aggregates occurring during pulping process and have impact on the fiber properties and crystallinity that are important factors when cellulose was used in composite formulations.Hult et al. 9 investigated the hierarchic organization of cellulose microfibrils in kraft and sulfite pulp fibers and showed that the kraft fibers exhibit higher ordered cellulose regions and more aggregated fibrils in contrast to the sulfite fibers.This may contribute to higher crystallinity in kraft fibers.However, pulping conditions as cooking temperature may influence on cellulose properties.The increase in cooking temperature during pulping process can promote more chain scission reactions increasing the amorphous character of the cellulose, which reduce the total amount of cellulose crystalline regions 9,23 .
Cellulose crystallinity is one of the most important crystalline structure parameters in cellulose fibers 10 .The thermal stability of cellulose was found to depend mainly on its crystallinity 11,12 .Thus, cellulose crystallinity and hydrogen bond between cellulose chains play an important role in the mechanical and thermal properties of composite materials reinforced with cellulose fibers 1,13 .Several techniques as X-ray diffraction and FTIR spectroscopy were used to evaluate the wood and cellulose crystallinity 5,7,8,9,10,12 .However, FTIR spectroscopy has been used as a simple technique for obtaining rapid information about the structure of cellulose and chemical changes taking place in cellulose due to various treatments 10,21,22,[25][26][27] .
As mentioned earlier, the crystalline structure of cellulose affects the physical, mechanical and thermal properties of cellulose fibers.Thus, it is paramount to ascertain the structure of cellulose and its crystallinity whenever cellulose fibers are intended for use as a reinforcing agent in composite materials.The aim of this study was to investigate the differences between two cellulose samples resulting from two different pulping processes and evaluate how these differences influence the thermal properties and decomposition kinetics of the cellulose fibers analyzed.

Materials
Bleached sulfite cellulose fibers from Pinus taeda (CPT) were supplied by Cambará S.A. (Cambará do Sul, Brazil) obtained at cooking temperature of 140 °C and bleaching with hydrogen peroxide.Bleached kraft cellulose fibers from Eucalyptus grandis (CEG) were supplied by CMPC S.A. (Guaíba, Brazil), obtained at cooking temperature of 155 °C, bleaching with hydrogen peroxide.The samples were dried at 70 °C for 24 hours in a vacuum oven before the tests.The average fiber particle length for CTP and CEG is around 150 µm.

Fourier transform infrared (ftir) spectroscopy
Fourier transform infrared spectroscopy spectra were obtained using a spectrometer Nicolet IS10-Thermo Scientific.Samples of the finely divided celluloses (5 mg) were dispersed in a matrix of KBr (100 mg) followed by compression to form pellets.The sample collection was obtained using 32 scans, from 4000 to 400 cm -1 , at a resolution of 4 cm -1 .Care was taken to ensure all samples remained dry during sample preparation and FTIR analysis.

Thermogravimetric analysis (TGA)
Thermogravimetric analysis (TGA50-Shimadzu) was carried out under N 2 atmosphere, from 25 up to 610 °C.Approximately 10 mg of each sample was used.The analysis was carried out at four different heating rates (5, 10, 20 and 40 °C/min).The results obtained were used to calculate the kinetics parameters.

Coats and redfern method
The conversion rate (α) is defined according to Equation (1) 12,14,15 as: where m 0 , m f and m i are the initial and final weights of the sample and its weight at temperature (T), respectively.For determining the degradation kinetic mechanism, the reactions are considered irreversible and the reaction rate is both dependent of the activation energy and reaction order 16 .These parameters can be estimated by combining the Arrhenius equation with non-isothermal experiments using the relationship described in Equation ( 2): where f(α), A, E a , R and T are the reaction model, pre-exponential factor, activation energy, universal gas constant (8.314J.mol -1 K) and temperature (in Kelvin), respectively.The Coats and Redfern method 17 is not based on derivates of dα/dt.However, in the non-isothermal method the temperature increases linearly, the time (t) variable is eliminated and the Equation ( 2) can be rewritten as 18 : ( ) where, β is the heating rate.As the term exp(-E a /RT) does not have exact solution the Taylor expansion is used 17,18 .After integration of Equation ( 3) the Equation ( 4) can be obtained: Considering the possible degradation mechanisms in solid state reactions, the solution for Coats and Redfern Equation can be obtained replacing the function g(α) by the mechanisms described in Table 1 [16,18] .When the values of n were different from one, the method can be described by Equation (5) and when the values of n were equal to one the Equation ( 4) is summarized by Equation ( 6) 17 : ( ) Therefore, to evaluate which degradation mechanism is closer to the solid state reaction under study, it is possible to replace the values of the functions showed in Table 1 by isolating the term on the left side of Equation (5) or Equation ( 6) as a function of inverse temperature (1/T).
The plot of ( ) ( ) against 1/T should result in a straight line with a slope to (-E a /2.3R).In fact, the calculations have to be performed using several values of n in order to retain the value leading to the best linear relationship.

Fourier transform infrared (ftir) spectroscopy
FTIR spectroscopy has been used as a simple technique for obtaining rapid information about the chemical structure and crystallinity of celluloses samples 5,19 .Contrary to conventional chemical analysis, this method requires small sample sizes, short analysis time besides being non-destructive 20 .
Because of their complexity, the spectra were separated into two regions, namely: the OH and CH stretching vibrations in the 4000-2700 cm -1 region, showed in Figure 1a, and the "fingerprint" region which is assigned to different stretching vibrations of different groups in 1800-800 cm -1 , Figure 1b.It can be observed in Figure 1a a strong broad band in the region of 3700-3000 cm -1 which is assigned to different OH stretching modes and another band in the region of 3000-2800 cm -1 is ascribed to the stretching of asymmetric and symmetric methyl and methylene CH cellulose groups 19 .The band at around 3360 cm -1 related with OH stretching modes is more prominent for CPT than for CEG.This is probably due to a larger number of hydroxyl groups in CPT which may be associated with an increase in the number of hydrogen bonds formed.Thus, a mixture of intermolecular and intramolecular hydrogen bonds is considered to cause the broadening of the OH band in the IR spectra 20 .
Figure 1b shows that in the "fingerprint" region the spectra revealed several bands.The band at 1642 cm -1 is associated with adsorbed water in cellulose 19,22,23 .The bands at 1430, 1370, 1335 and 1320 cm -1 are attributed to CH 2 symmetric bending, CH bending, OH in plane bending, CH 2 rocking vibration, respectively [21][22][23] , and the bands at 1162, 1111, 1057, 1033, 898 cm -1 are assigned to asymmetric C-O-C bridge stretching, the anhydroglucose ring asymmetric stretching, C-O stretching, C-H in plane deformation, C-H deformation of cellulose, respectively [21][22][23][24]26 . On he other hand, the peaks for xylan from hemicelluloses are derived from molecular vibrations in the uronic acids at 1730 and 1600 cm -1 [28] .CEG sample showed a more prominent peak than CPT at around 1730 cm -1 while the band at 1600 cm -1 overlap with the band at 1642 cm -1 associated with adsorbed water in cellulose 19,22,23 .This may indicate that xylan is more alkali resistant and probably hemicelluloses can be precipitation on the surface of the eucalyptus cellulose fibers during kraft cooking 28,29 .
The ratio between the heights of the bands at 1372 and 2900 cm -1 proposed by Nelson and O'Connor (1964) as total crystalline index (TCI) 22 was used to evaluate the infrared (IR) crystallinity ratio.The band at 1430 cm -1 is associated with the amount of crystalline structure of Table 1.Algebric expressions commonly used for solid thermal decomposition processes.

Mechanism -Solid state process g(α) f(α)
A (n) -Nucleation and growth 1) ] ( 1) cellulose, while the band at 898 cm -1 is assigned with the amorphous region in cellulose 25 .The ratio between the areas of the bands at 1429 and 897 cm -1 are used as a lateral order index (LOI) 22 .Considering the chain mobility and bond distance, the hydrogen bond intensity (HBI) of cellulose is closely related to the crystal system and the degree of intermolecular regularity, that is, crystallinity 27 .The ratio of the absorbance bands at 3400 and 1320 cm -1 was used to study the cellulose samples HBI.The obtained results are displayed in Table 2.
The TCI is proportional to the crystallinity degree of cellulose 20 and LOI is correlated to the overall degree of order in cellulose 22,26 .Based on this fact, the CPT showed the highest TCI and LOI value indicating highest degree of crystallinity and more ordered cellulose structure than the CEG.On the other hand, CEG presented lowest TCI and LOI value which may indicate that this cellulose is composed for more amorphous domains in the cellulose structure when compared with the CPT.The HBI value is higher for CPT than for CEG sample.This result might be indicated that CPT contain much more cellulose chains in a highly organized form which can lead to higher hydrogen bond intensity between neighbor cellulose chains and result in more packing cellulose structure and higher crystallinity than CEG.The crystallinity of cellulose is closed related with the thermal stability 12,30 .Therefore, it is possible that cellulose samples with highest TCI, LOI and HBI may exhibit higher thermal stability.

Thermogravimetric analysis
Figure 2 shows the TGA and derivative thermogravimetric (DTG) curves of the cellulose samples studied.A small weight loss for both samples occurs between 40-70 °C which is attributed to the removal of absorbed water in cellulose 14,31 .
As depicted in Figure 2a, the CEG sample initiates a more pronounced degradation process at around 280 °C while for CPT a more pronounced degradation process occurs at 292 °C.At this reaction time, the cleavage of the glycosidic linkages of cellulose reduces the polymerization degree leading to the formation of organic compounds like alkenes and other hydrocarbon derivatives 32,33 and after 400 °C the residual decomposition process can lead to formation of CO, CO 2 , H 2 O and char 5,31 .The DTG curve of CPT was shifted to higher temperatures than CEG.The DTG peak occurs at 353 °C for CEG and 360 °C for CPT, as shown in Figure 2b.This result suggests that celluloses with higher TCI, LOI and HBI have higher thermal stability probably due to the much more hydrogen bonds between cellulose chains that can result in more ordered and packing cellulose regions, this in turn possibly increasing the thermal decomposition temperature of cellulose.Moreover, the CEG sample presented a small shoulder between 250 and 300 °C.This behaviour can be associated with the hemicellulose degradation in this sample.According Yang et al. 31 hemicellulose starts a more prominent decomposition between 220-315 °C.However, for both CEG and CPT samples the start of hemicellulose decomposition is difficult to distinguish and overlaps with the degradation of cellulose.

Kinetics results
Figure 3 illustrates the application of the Coats and Redfern method for the kinetic models listed in Table 1.The values of n used were 3.42 and 2.45 for CEG and CPT, respectively according to a previous work 12 .As can be seen in Figure 3a, the best fits using the Coats and Redfern method for CEG indicated that for this cellulose sample the degradation process occurs by first order reaction with random nucleation resulting in an F1 mechanism, which is  in agreement with the results describes by Conesa et al. 38 and Dahiya et al. 39 .On the other hand, phase boundary controlled reactions with contracting area and contracting volume also occurs, in agreement with other results from the literature [33][34][35][36][37] .This degradation behavior may be associated with the lower crystallinity of the CEG sample observed by FTIR results.The degradation process probably starts sporadically on the cellulose amorphous domains from the entire sample.When the crystalline domains degradation takes place the degradation initially occurs on the crystallites surface and phase boundary controlled reactions also described the degradation mechanism.For the CPT sample the best fit occurs by a degradation mechanism corresponded to R n , i.e. the phase boundary-controlled reaction, as shown in Figure 3b.Similar results were described by Wu and Dollimore 34 .Once again, the FTIR results may be explained better the degradation mechanism.The CPT sample had higher crystallinity than CEG, as can be seen in Table 2.
The more packing cellulose chains in CPT sample might be difficult the heat transfer by diffusion through the cellulose chains and then the degradation process may be initially occurs by degradation of the surface of the cellulose crystallites with contraction of the area and volume, resulting in R2 and R3 degradation mechanisms.
The activation energy values for CEG were found in the range of 173-190 kJ.mol -1 and for CPT between 124-126 kJ.mol -1 , as presented in Table 3.The values found for both cellulose samples are consistent with the values reported in the literature 38,39 .However, the CEG sample had a higher activation energy range than the CPT sample.
The lower activation energy values observed for CPT might be attributed to the thermal decomposition of this sample being controlled by dehydration which results in anhydrocellulose.The CPT sample showed a higher band at 1642 cm -1 assigned to the adsorbed water into the cellulose structure, see Figure 1b.The more quantity of adsorbed water into the cellulose structure of CPT sample may be responsible for initiate the degradation process by dehydration which probably results in lower activation energy values.The higher activation energy for CEG may be indicating that the thermal decomposition occurs by depolymerization of cellulose with production of levoglucosan, according to the classic model proposed by Broido-Shafizadeh 18,40 as shown in Figure 4. Therefore, the two parallel or competitive reactions occurs for the two cellulose samples study, however dehydration reactions are more prominent for CPT sample while for CEG  of the cellulose fibers studied.For the CEG sample the degradation processes occurs by random nucleation and probably starts on the cellulose amorphous domains while for CPT, that had more crystallinity regions than CEG, the more packing cellulose chains might be difficult the heat transfer by diffusion through the cellulose chains and then the degradation process may be occurs by degradation of the surface of the cellulose crystallites with a phase boundary controlled reaction.The lower activation energy values observed for CPT than CEG might be attributed to the thermal decomposition of this sample being controlled by dehydration while for CEG the higher activation energy values probably indicated that the thermal decomposition occurs by depolymerization of cellulose.In general, the crystallinity and thermal stability were more affected by the kraft pulping conditions than by those of sulfite pulping.This behavior may be associated with the higher cooking temperature used during the cooking process employed for the CEG sample.

Figure 2 .
Figure 2. TGA (a) and DTG (b) curves of the cellulose samples studied.

Figure 3 .
Figure 3. Linear fits of the Coats and Redfern method using different degradation mechanism (g(α)) determined through the thermogravimetric curves for β = 10 °C/min where: (a) CEG e (b) CPT samples.

Table 2 .
Cellulose infrared crystallinity ratios and hydrogen bond intensity

Table 3 .
Activation energy values from Coats and Redfern method