Alkalization of plant or wood fibre (WF) is the most widely used method of chemical modification to improve reinforcement in thermoplastic composites. This process involves the complete or partial removal of extractives and or modification of lignocellulosic material. While research has shown that removal of the less thermally stable extractives results in an improvement in fibre thermal stability, in the current work it has been shown through single-factor analyses, Fourier transform infrared microscopy, scanning electron microscopy, thermogravimetric analyses and wide angle X-ray diffraction that meranti WF thermal stability is largely influenced by the holistic changes in the WF structure, which itself is affected by alkalization factors. After implementing stepwise regression on a central composite design, no empirical model could be established to explain or predict thermal stability due to interaction of treatment factors. As a result, single-factor analyses of temperature, time and alkali concentration were conducted. Single-factor analyses showed that different combinations of time, temperature and alkali concentration through a central composite design result in WF with different thermal stabilities, lignocellulosic content, crystallinities, crystallite sizes, extractives content and morphology. Alkali-treated meranti WF showed lower thermal stability compared to the untreated WF. Mild treatment conditions (e.g. 50 °C/30 min/5%) were seen to result in the most thermally stable WF. Increasing temperature, treatment duration and alkali concentration increased thermal stabilities except at harsh conditions (e.g. 50 °C/90 min/15%). A combination of high alkali concentration and long treatment times showed a combined detrimental effect on WF thermal stability. Changes in the lignocellulosic structure, crystallinity, crystallite sizes and surface features explain the observed changes in thermal stabilities.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Introduction
Plant-based fibres offer both ecological and economic advantages over glass or synthetic fibres due to their biodegradability, low cost, non-abrasiveness, safe fibre handling, low energy consumption, high specific properties, relatively low density and diversity (Bagheri et al. 2018; Bledzki and Gassan 1999; Ghasemi et al. 2017; Mohanty et al. 2006; Wambua et al. 2003). As a result, the wood plastic composites (WPC) market has grown rapidly, especially in synthetic fibre replacement applications in the automotive sector (Markets & Markets 2020; Marsh 2003; Suddell 2008; Wambua et al. 2003). In the making of WPC, plant fibres are usually modified to facilitate compatibility with thermoplastics or thermosets. One of the most common methods of fibre modification is alkali treatment, commonly known as alkalization or mercerization (Farsi 2010; Liu et al. 2004; Mwaikambo and Ansell 2002). Knowledge of wood fibre structural modifications is necessary to inform and explain the observable properties such as mechanical and rheological properties of related wood-fibre products. The current study therefore seeks to contribute to the field of WPC by investigating alkalization factors and their effects on wood flour properties.
Wood flour (WF) alkalization involves the use of sodium hydroxide (NaOH) to modify the fibre structure. The process involves the removal of WF-waxy materials, oils, proteins, amorphous lignin and hemicelluloses (Bledzki and Gassan 1999; Mwaikambo and Ansell 2002). This results in fibre with altered physico-chemical properties such as crystallinity, thermal stability, surface properties and lignocellulosic content. Research has shown that WF alkalization results in a high degree of crystalline cellulose as well as rough fibre surfaces. The increase in the crystallinity index has been shown to result from the reduction of the amorphous phase and or the increased order of cellulose crystallites (Aboul-Fadl et al. 1985; Mwaikambo and Ansell 2002). Alkalization results in cellulose I transition to cellulose II type, where the cellulose polymer chains are shortened and more ordered, thereby increasing the crystallinity index (Bledzki and Gassan 1999; Mwaikambo and Ansell 2002). However, the crystallinity index then declines at high alkali concentrations due to cell wall damage. This results in a less thermally stable amorphous cellulose, with reduced crystallite length (Aboul-Fadl et al. 1985; Bisanda and Ansell 1991; Nguyen et al. 1981). On the other hand, changes in the lignin structure and content were shown to result in a decrease in the peak degradation temperature (Tpeak) of WF (Bisanda and Ansell 1991; Chen et al. 2014; Nguyen et al. 1981; Ohtani et al. 2001). Moreover, due to its wide range of thermal degradation temperatures, the degradation of lignin also influences the degradation of hemicelluloses (Tserki et al. 2005). Lignin degradation leaves the hemicelluloses exposed to subsequent degradation (Chen et al. 2014; Ohtani et al. 2001). Hemicellulose has been shown to degrade in the range of 180–350 °C, lignin in the range 250–500 °C, while cellulose in the range 275–350 °C (Kim et al. 2006). WF thermal stability has also been reported to be influenced by extractives removal (Chen et al. 2014; Poletto et al. 2010). Extractives comprise a vast array of materials that have relatively low thermal stability such as fatty acid esters (of fatty alcohols, terpene alcohols, and sterols), fats (i.e. fatty acid glycerol esters) and phenolics (i.e. stilbenes, tannins, lignans and flavonoids) (Yang and Lu 2021). As such, removal of the thermally less stable extractives generally results in an improvement in the WF thermal stability (Chen et al. 2014; Poletto et al. 2010).
Anzeige
Inarguably, studies have made clear the role of alkali treatment of plant fibres; however, most studies have looked more into the effects of varying alkali concentration at specific temperatures, or vice versa. This study aims to contribute towards existing literature by investigating the individual and (possible) combined effects of alkalization conditions of time, temperature and alkali concentration following a mixture design. In this study, alkalization is conducted through a central composite design (CCD) where temperature, time and alkali concentration are varied while morphology, thermal stability and crystallinity are used to monitor the alkalization process. Implementation of a CCD will allow for probing into alkalization effect at mid- and extreme points of experimental factors, thereby giving a clear understanding of factorial effects on response variables. Morphological imaging will be done through scanning electron microscopy (SEM), thermal stability tests through thermogravimetric analyses (TGA) and the degree of crystallinity using wide-angle X-ray diffraction. Alkalization was conducted on waste meranti wood flour (Shorea spp.), obtained from a local manufacturing furniture shop. Meranti WF was chosen for our project as it was readily available as waste, at no cost, and has been identified to exhibit good mechanical properties (Database 2021) thereby possessing the potential of beneficial reinforcement properties as a filler in WPC. In addition, compared to many hardwoods, meranti wood has a relatively lower extractives content compared to other hardwoods (Database 2021) and thus severe treatment conditions to remove extractives may not be necessary.
Method and materials
Materials
Waste wood flour from light red meranti (Shorea spp.) was obtained from a local furniture shop called Creative Timber Mouldings, reagent grade sodium hydroxide (NaOH) pellets and acetic acid (CH3COOH) from Sigma-Aldrich.
Method
Wood fibre treatment
Waste meranti WF was sieved using ASTM sieves in the range 1 mm and 400 mesh. After drying at 105 °C for an hour in a forced air ventilation oven, WF samples were treated according to the following steps.
WF treatment procedure:
soaking of 10 g wood flour in 100 mL alkali at concentration, temperature and time determined by the central composite design shown in Tables 1 and 2.
adding dilute acetic acid (1%) to quench the reaction once set reaction time is reached, followed by washing of the wood flour with distilled water until pH of the mixture reaches neutral, and
drying the treated wood flour in a forced air ventilation oven at 50 °C until constant mass is reached.
Table 1
Central composite design experimental domain
Design points
Temperature/°C
Time/min
Concentration/wt% NaOH
Maximum
50
90
15
Minimum
25
30
5
Mid-point
37.5
60
10
Step size
12.5
30
5
Table 2
Central composite design points showing all experimental runs
Sample
Temperature
(°C)
Time
(min)
Concentration
(wt% NaOH)
1
25
30
5
2
25
30
15
3
25
90
5
4
25
90
15
5
50
30
5
6
50
30
15
7
50
90
5
8
50
90
15
9
38
60
10
10
22
60
10
11
55
60
10
12
38
24
10
13
38
96
10
14
38
60
4
15
38
60
16
Alkali treatment followed the CCD shown in Tables 1 and 2. Temperature conditions chosen were close to ambient conditions but varied enough to show differences in the effects on alkalization. The use of moderate temperatures allows for energy conservation during the alkalization process.
Analyses
Extractives content
The extractives content in the WF was determined using TAPPI T 204cm-07 method (TAPPI 2007), where about 5 g samples of dried WF were extracted in thimbles using 150 mL of absolute ethanol/toluene solvent mixture (1:2 volume ratio) via Soxhlet extraction. Due to the relative toxicity of benzene compared to toluene, toluene was used as an alternative for this research (Chen et al. 2014). The extraction was conducted for 5 h, ensuring a minimum of 24 cycles of extraction by setting the heating mantle to an appropriate heating rate. The extracts were rotor-vapored to a small volume (ca. 5 mL), which was transferred to a watch glass for drying and weighing. Drying of the extracts was done in an air forced ventilation oven for 1 h at 105 ± 3 °C followed by cooling in a desiccator prior to weighing. Determination of % extractives was done in triplicates following the same method and conditions and calculated according to equation (Eq. 1).
where We = oven-dry weight of extract in grams (g), dWF = oven dry weight of wood flour (g), and Wb = oven-dry weight of blank residue (g).
The Wb was obtained by drying up the fresh extraction solvent of volume equal to the extraction volume (i.e. 150 mL) and then weighing for any possible residues (i.e. blank determination).
Crystallinity analyses using X-ray powder diffraction (XRD)
Structural analyses of the meranti WF were done on a D2 PHASER bench-top powder X-ray diffractometer (BRUKER Billerica). The X-ray diffraction (XRD) pattern of the samples was measured over the angular range of 2θ = 5°–70° with a scanning step of 0.2°/min and 4 s per step with rotations. The diffractograms acquired were analysed using the DIFFRAC.EVA software.
The crystallinity index (Ic) was determined using the empirical method by Segal et al. (1959) as shown in Eq. 2
The crystalline phase diffraction intensity I(002) was determined from the maximum height of the (002) lattice diffraction intensity at 2θ range of 21–23°. The amorphous diffraction intensity I(am) was determined at the minimum intensity close to 2θ = 18°.
The sizes of crystallites were calculated using the Scherrer equation (Langford and Wilson 1978) shown in Eq. 3;
where Dhkl = crystallite size in the direction normal to the hkl lattice planes, λ = 1.542 Å (i.e. the X-ray radiation wavelength), β = full width at half-maximum (FWHM) of the diffraction peak in radians, and θ = corresponding Bragg angle.
Morphological analyses
Morphologies of the untreated and alkali-treated fibres were examined using a JSM-IT100 InTouchScope™ scanning electron microscope, SEM (JEOL Ltd.). WF surfaces were sputter-coated with gold to avoid electrostatic charging and poor image resolution. A backscattered electron detector (BEC), secondary electron detector (SED), accelerating voltage of 10.0 kV, and working distance of 11 mm were used for analyses.
Thermal stability analyses
Alkali-treated and untreated samples were analysed for thermal stability using a simultaneous differential scanning calorimetry-thermogravimetric analyser (SDT, TA Instruments Q600). Temperature was ramped at 20 °C/min to 600 °C in N2 (g) atmosphere. At 600 °C the gas was switched to O2 (g) and heating done at a rate of 10 °C/min to 700 °C. Sample weight was maintained at 5 mg in alumina pans.
FTIR
FTIR was used for chemical structural analysis of the wood fibre before and after alkali treatment. The spectra were recorded with an OPUS 7.0 software on a Bruker Tensor II spectrometer using the attenuated total reflection (ATR) mode. Sample scan time of 32 scans was used within the range of 4000-400 cm−1 at a resolution of 1 cm−1. The background was measured at background scan time of 16 scans before testing each sample.
Results and discussion
Alkalization versus extractives content
A significant decline in the extractives content of the WF samples was observed when the WF samples were subjected to alkalization for all treatment conditions. According to Fig. 1, the extractives content dropped from 11% for the neat WF, to a lowest of 0% for severe treatments, while most samples showed remnant extractives contents of 2% on average. This clearly shows the cleaning effect or efficiency of alkalization in removing extractives from WF samples. No visible trend was observed between treatment conditions and the extent of extractives removal (Table 2 and Fig. 1); however, a combination of severe factors seems to play a role in the degree of extractives removal as evidenced by samples 11 and 13–15.
Fig. 1
Changes in the extractives content following alkalization
×
Different types of extracts of the neat WF sample were run through FTIR to (1) compare the degree of extraction, (2) see if there is any lignocellulosic material in the extracts and (3) to see if there is any chemical modifications of the WF. The extracts’ constituents will thus inform of any chemical modifications on the WF, thereby aiding the understanding of any changes in WF behaviour due to alkalization. Extractions of neat WF sample were conducted using 10% NaOH (median concentration) and ethanol/toluene mixture. The extracts’ spectra in Fig. 2 show the presence of fatty acids (COOH, 3200–3600 cm−1), esters (COOC, 1736–1744 cm−1) and hemicelluloses (1245–1259 cm−1) in the ethanol/toluene mixture. The alkali spectrum seems enhanced and slightly shifted to the right for lignocellulosic materials. This is probably due to hydrolysis reactions effected by the NaOH on lignocellulosic materials, thereby slightly altering their chemistry (Mohanty et al. 2006). The ester peak in the range 1736–1744 cm−1 is visible in the neat sample as well as the ethanol/toluene extracts, but disappears upon alkali treatment. This is due to hydrolysis of the ester bond, known as de-esterification (Aboul-Fadl et al. 1985). The hemicellulose peak in the 1012 cm−1 region is split upon alkali treatment which is indicative of chemical reactions or disruption of the native hydrogen bonds in hemicellulose by the NaOH reacting with the hydroxyl groups (Mwaikambo and Ansell 2002).
Fig. 2
FTIR spectra of neat WF extracts using 10% NaOH, methanol and ethanol/toluene mixture
×
Anzeige
The degree of extraction is clearly depicted by the SEM images obtained for the WF samples in Fig. 3. SEM images of the untreated and treated WF samples were taken at magnifications of ×250, ×500 and ×1000; however, for the sake of illustration only the ×1000 magnifications are shown. The untreated WF (neat sample) showed a smooth surface, indicating the presence of a waxy layer of extractives on its surface. Upon treatment, the surface appearance of all WF samples were altered regardless of treatment conditions (Fig. 3). The changes in surface morphology as evidenced by the formation of rough, grooved, and serrated surfaces confirms extractives removal. In some cases, the SEM images clearly indicate intense surface changes like pits, holes and fibrillation (e.g. samples 7, 10–15) which are likely due to the removal of lignocellulosic material. Inasmuch as morphological imaging clearly revealed the degree of sample treatment, thermal stability further showed the effects of alkali treatment on the WF structural integrity.
Fig. 3
SEM images of untreated (neat) and treated WF surface features (×1000 magnification) identified by designation according to Table 3
×
Thermal properties of treated wood flour
Table 3 indicates a decrease in WF thermal stability following alkali treatment across all samples, evidenced by reductions in both peak (Tpeak = 332 °C) and onset degradation temperatures (Tonset = 287 °C), depicted in Fig. 1 ESM (Appendix, Electronic Supplementary Material, ESM). Contrary to the anticipated rise in Tonset due to the removal of less thermally stable extractives (Markets and Markets 2020; Ohtani et al. 2001), alkali treatment resulted in a decline. This suggests that while extractives were removed during alkalization, it coincided with the elimination or alteration of the lignocellulosic material, as evidenced in the FTIR spectra in Fig. 2. The decreased thermal stability can be attributed to several factors: an increase in the less thermally stable amorphous cellulose content, cellulose depolymerization (Mwaikambo and Ansell 2002), and potential changes in the lignin structure (Chen et al. 2014; Ohtani et al. 2001). Furthermore, due to lignin’s wider range of thermal degradation temperatures (Tserki et al. 2005), the degradation or removal of lignin has an influence on the degradation of hemicelluloses as it leaves the hemicelluloses exposed to degradation (Chen et al. 2014; Ohtani et al. 2001). Notably, alterations in Tonset primarily relate to changes in the hemicellulose structure (Y Chen et al. 2005; Ohtani et al. 2001), whereas shifts in Tpeak are predominantly associated with modifications in lignin (Chen et al. 2014; Ohtani et al. 2001) and cellulose structures (Mwaikambo and Ansell 2002). The thermal stability variations among the treated WF samples align with sequences 1 and 2, indicating sample 13 (at 38 °C, 60 min, and 16% NaOH) exhibited the lowest thermal stability, while sample 5 (at 50 °C, 90 min, and 5% NaOH) demonstrated the most thermally stable treatment.
Table 3
WF treatment conditions and corresponding thermal stabilities (Tonset & Tpeak) and lignocellulosic contents (LCM) obtained after alkalization
Comparison of the different WF samples to get a clear picture of the individual or combined effects of alkalization factors on WF modification was attempted using multiple regression analyses.
Regression analyses of WF treatment conditions on Tonset and Tpeak
The sample treatment conditions followed a 2-level central composite design at temperatures relatively close to ambient temperatures (22–55 °C), time range 24–96 min and concentration of 4–16% NaOH (Tables 1, 2). Multiple regression analyses were done at 95% confidence interval (CI) and null hypothesis, Ho = 0, in an attempt to find a model that can best explain the effects of treatment conditions on Tonset and Tpeak (see Tables 4, 5).
Table 4
Multiple linear regression with %∆Tonset as response variable
Coefficients
Standard error
t Stat
P value
Lower 95%
Upper 95%
Intercept
24.2226
11.7142
2.0678
0.0725
− 2.7903
51.2355
Temperature
− 0.5456
0.2756
− 1.9797
0.0831
− 1.1811
0.0900
Time
− 0.1229
0.1440
− 0.8535
0.4182
− 0.4550
0.2093
Concentration
− 0.8114
0.8640
− 0.9391
0.3752
− 2.8039
1.1811
Time*Conc
− 0.0013
0.0078
− 0.1636
0.8741
− 0.0192
0.0167
Temp*Time
0.0036
0.0031
1.1638
0.2780
− 0.0036
0.0108
Temp*Conc
0.0301
0.0187
1.6150
0.1450
− 0.0129
0.0731
Table 5
Multiple linear regression with %∆Tpeak as response variable
Coefficients
Standard Error
t Stat
P value
Lower 95%
Upper 95%
Intercept
22.1730
11.7521
1.8867
0.0960
− 4.9275
49.2734
Temperature
− 0.4893
0.2765
− 1.7696
0.1148
− 1.1268
0.1483
Time
− 0.1508
0.1445
− 1.0438
0.3271
− 0.4840
0.1824
Concentration
− 0.4016
0.8668
− 0.4633
0.6555
− 2.4001
1.5973
Time*Conc
− 0.0024
0.0078
− 0.3038
0.7690
− 0.0204
0.0156
Temp.*Time
0.0045
0.0031
1.4452
0.1860
− 0.0027
0.0117
Temp.*Conc
0.0201
0.0187
1.0712
0.3152
− 0.0231
0.0632
Tables 4 and 5 show that no model according to multiple linear regression analyses can be used to explain the effect of treatment conditions on the changes in thermal stability of the WF. The P values for the individual as well as combined factors are all greater than 5% within 95% confidence interval (P value > 0.05, 95% CI); therefore, the null hypothesis is accepted. This means that the varied factors show no defined effect on thermal stability and therefore cannot be modelled through multiple linear regression to explain or predict WF thermal stability changes upon alkalization. This is emphasized by R-squared values of 0.41 and 0.36 for changes in thermal stability (%∆Tonset and %∆Tpeak), respectively. The inexplicability of the obtained data through regression is likely due to the dynamic nature of plant fibres as well as their dynamic structural changes during alkalization; as a result, single-factor analyses were done to show individual effects of process factors on WF thermal stability.
Single-factor analyses of WF treatment conditions vs. thermal stability
Effect of treatment temperature on thermal stability
Figure 4 shows that at given controlled conditions of time and alkali concentration, increases in temperature from 25 to 50 °C result in corresponding decreases in Tpeak and Tonset except at 30 min and 5% NaOH, which were controlled conditions. The effects of increasing temperature are clearly depicted on the SEM images in Fig. 3. Increasing temperature from 25 to 50 °C at 30 min/15% NaOH, (i.e. sample 2 to 6) resulted in WF surface modification from a surface showing small traces of extraction to a dry surface associated with rough edges and ruptured fibre surfaces. Similarly, increasing temperature for samples 3–7 (at 5%/90 min) and 4–8 (15%/90 min) also resulted in intense morphological changes according to SEM images. Faint, shallow pits on sample 3 changed to holes on sample 7, indicating intensive extractives and lignocellulosic material removal. The observed holes indicate removal of lignin or cellulose as suggested by the significant change in Tpeak for sample 7 (Fig. 4b) compared to Tonset (Fig. 4a). So even though the alkali concentration was low (5%), it seems the longer treatment duration (i.e. 90 min) coupled with doubling temperature effected the degree of modification at the Tpeak region. Increasing temperature from sample 4 to 8 showed a formation of rough, bumpy surface features associated with the removal of surface extractives and hemicelluloses at Tonset. Extraction pits and holes which have been previously associated with the removal of lignocellulosic material at the Tpeak region were also observed. At mild conditions (5%/30 min), increasing temperature from samples 1 to 5 resulted in an increase in thermal stability of the WF samples rather than a decrease. This observation corroborates the work by Mwaikambo and Ansell (2002), where alkalization was reported to result in an initial increase in crystallinity at mild conditions. This was, however, followed by a decline in WF crystallinity and thermal stability as the less thermally stable amorphous cellulose was formed due to WF structural damage from the prolonged alkali treatment. The increase in thermal stability at mild conditions can therefore be explained by the removal of the less thermally stable extractives, hemicelluloses and an increase in cellulose crystallinity through cell wall thickening without the detrimental rupture of cell walls (Aboul-Fadl et al. 1985; Bisanda and Ansell 1991).
Fig. 4
Effect of treatment temperature on aTonset and bTpeak
×
Figure 5 shows a general decrease in the amount of lignocellulosic material with an increase in temperature except for treatment at 5%/30 min. The decrease in lignocellulosic material with temperature shows how increases in temperature facilitate the removal of lignocellulosic material and extractives. The lignocellulosic material left after WF treatment contained extractives except for samples 11 (53 °C/60 min/10%), 13 (38 °C/96 min/10%), 14 (38 °C/60 min/4%) & 15 (38 °C/60 min/16%) (Fig. 1), which may indicate intense alkalization of these samples.
Fig. 5
Effect of treatment temperature on % lignocellulosic material
×
Effect of NaOH concentration on thermal stability
Tripling the alkali concentration from 5 to 15% NaOH resulted in a decrease in both the Tonset and Tpeak for all controlled conditions, except for the mild conditions observed at 25 °C/30 min. These findings suggest that cell wall damage occurs due to prolonged exposure to high alkali concentrations, leading to the formation of less thermally stable amorphous cellulose and shorter crystallites (Aboul-Fadl et al. 1985). This explanation is supported by the significant decline in thermal stability observed at approximately 50 °C/30 min (Fig. 6). The increase in alkali concentration from 5 to 15%, as depicted in Fig. 7, led to a higher loss of lignocellulosic material and extractives. Specifically, a maximum loss of 9% in lignocellulosic material was recorded with the increase in alkali concentration at around 50 °C/30 min. Conversely, the rise in the amount of remaining lignocellulosic material after extraction at approximately 25 °C/30 min indicates a greater removal of extractives than lignocellulosic material.
Fig. 6
Effect of NaOH concentration on thermal stability
Fig. 7
Effect of alkali concentration on % lignocellulosic material
×
×
Effect of treatment duration on thermal stability
The impact of treatment duration on thermal stability appears to be influenced by the controlled conditions of alkali concentration and temperature. Although no consistent trend is observed, under mild conditions (25 °C & 5% NaOH), extending the treatment duration from 30 to 90 min results in a 16 °C increase in thermal stability. This observed enhancement in thermal stability at mild conditions mirrors the behaviour noted for the effect of temperature and alkali concentration on WF. Reviewing Figs. 8 and 9 at 25 °C/5%, WF modification over 90 min appears to involve the extraction of less thermally stable extractives and lignocellulosic material, potentially accompanied by an improvement in the order of crystallites through cell wall thickening (Mwaikambo and Ansell 2002). These structural alterations manifest as increases in thermal stability when the treatment time is extended from 30 to 90 min at 25 °C/5%. However, continued WF treatment or heightened severity of conditions ultimately damages the cell wall structure (Bisanda and Ansell 1991; Nguyen et al. 1981; Tserki et al. 2005). This deterioration is evident in Fig. 9, where longer treatment times (90 min) of the thermally stable WF at 50 °C/5% result in a significant decline in thermal stability. Consequently, the thermal stability of WF treated under severe conditions, such as 90 min/15%/50 °C, is relatively low.
Fig. 8
Effect of treatment duration on thermal stability
Fig. 9
Effect of alkali concentration on % lignocellulosic material
×
×
WF crystallinity changes due to alkalization
Figure 10 shows X-ray patterns observed for the WF samples. The crystalline phase diffraction intensities (I002) were determined from the maximum height of the (002) lattice diffraction intensity at 2θ angle between 21° and 23°, whereas the amorphous diffraction intensities (Iam) were determined from the intensity minimum of the (002) and (111) lattice peaks (i.e. between 2θ = 18° and 23°). The results obtained show little differences in percentage crystallinity for most but a few samples. This might be due to the narrow experimental domain used in the investigation. To attain marked differences in crystallinity, one may therefore have to increase the temperature, concentration or time range. Nonetheless, it is evident from Fig. 10 that alkalization resulted in changes in the WF crystallinity, which corroborates existing work (Aboul-Fadl et al. 1985; Mwaikambo and Ansell 2002). For brevity, only samples that show significant differences in crystallinity from the neat sample are discussed.
Fig. 10
Overlay of neat (untreated WF) and treated WF XRD patterns at 2θ = 5°–70° range. Online version in colour
×
The observed changes in crystallinity upon alkalization in Fig. 10 are likely a result of the removal of non-cellulosic amorphous material (Xie et al. 2016), swelling of cell walls, depolymerization of cellulose (Mwaikambo and Ansell 2002), or a combination of any of these factors. Changes in crystallinity can be used to explain some of the changes in thermal stability and lignocellulosic ratios observed under single-factor analyses. In an attempt to explain WF behavioural changes, Fig. 11 was used to relate observed crystallinity changes to WF behaviour.
Fig. 11
Overlay of neat (sample 0) and treated WF % crystallinity, crystallite size and changes in peak degradation temperature
×
Figures 8 and 9 show that sample 1 (at 25 °C/5%/30 min) has lower lignocellulosic content and thermal stability than sample 5 (at 50 °C/5%/30 min). Looking at Fig. 11, at 25 °C/5%/30 min (i.e. sample 1) the WF shows high % crystallinity which is associated with relatively smaller sized crystallites compared to sample 5. Figure 11 reveals that the relatively smaller crystallites are due to depolymerization, hence the observed increase in %crystallinity and low thermal stability. In other words, the increase in %crystallinity results from the shorter crystallites which increase crystallite order rather than crystallinity (Aboul-Fadl et al. 1985; Li et al. 2007; Mwaikambo and Ansell 2002). The loss of amorphous material in the process, as shown by the relatively low lignocellulosic content, further adds to the observed high % crystallinity (Aboul-Fadl et al. 1985; Xie et al. 2016).
Sample 6 (at 15%/30 min/50 °C) shows relatively bigger %crystallinity, while crystallite size appears to be similar to that of the neat sample. This can be explained by the removal of amorphous material and or swelling of the cell wall. The former is supported by the observed decline in the lignocellulosic content shown in Fig. 5 on effect of temperature.
Samples 5 and 9 showed the lowest changes in thermal stability compared to the neat sample. Evidently, they exhibit % crystallinity and crystallite sizes similar to the untreated WF sample as well as high lignocellulose content (69% and 67.5%, respectively). On the other hand, sample 13 showed the lowest crystallite size, associated with low thermal stability and increased %crystallinity. This implies that sample 13 had the highest degree of depolymerization compared to all the treatment conditions and this can be explained by the longest treatment duration of 96 min at 10% NaOH for sample 13. Reductions in crystallite size due to high NaOH concentration over long treatment durations, i.e. 10–15% NaOH over 60–96 min (Tables 1, 2), were also observed for samples 4, 8, 10, 11, 12, and 15 (Fig. 12).
Fig. 12
FTIR analyses for untreated and alkali-treated WF samples. The colour of arrows show grey (fatty acids, esters, extractives), red (lignin), black (hemicellulose and cellulose). Online version in colour
×
FTIR analyses of treated WF samples
FTIR analyses were done to track changes in the WF chemical structure and possibly find explanations for some WF behavioural changes. Due to the large data size only FTIR spectra for the neat untreated WF, most thermally stable treatment (sample 5, 25 °C/3 0 min/5%), least thermally stable (sample 13; 38 °C/96 min/15%) and median (sample 4; 25 °C/90 min/15%) were shown. The peaks at the range 1736 and 1744 cm−1 represent ester groups present in the extractives in the neat WF as well as remnant extractives after alkalization (Zhang et al. 2009). These absorption peaks tend to disappear due to de-esterification that occurs during alkalization, as seen in Fig. 12. During de-esterification, the ester bond is severed via hydrolysis reactions that result in a reduction of ester groups. Consequently, the peak intensity at this region for alkalized WF decreases depending on the degree of hydrolysis, while the neat sample shows a relatively intense absorption peak (Zhang et al. 2009). Furthermore, at 1596 cm−1, the fatty acids peak is observed more for untreated WF than alkalized WF.
The enhanced peak intensities in absorption range 1000 to 1500 cm−1 are indicative of hemicelluloses (Zhang et al. 2009). Peak enhancement arises due to destruction of native hydrogen bonds by reacting with the alkali, thereby exposing hydroxyl groups (Mwaikambo and Ansell 2002). However, during alkalization some hemicelluloses are removed, resulting in the observed low intensity peaks in the range 1245–1259 cm−1 (Maheswari et al. 2012). Between 1028 and 1055 cm−1 are peaks ch”aracteristic of both hemicelluloses and cellulose (Zhang et al. 2009). Unique peaks characteristic of cellulose are seen at 898 cm−1, which are indicative of the ꞵ-glycosidic bonds joining glucose moieties (Mwaikambo and Ansell 2002). At 1509–27 cm−1 are C–C double bonds stretching vibrations characteristic of aromatic rings found in lignin (Maheswari et al. 2012). It is important to note the reduced intensities for the alkalized WF samples. Not all the samples were shown; however, from the few samples analysed it is apparent that during alkalization a combination of physical and chemical processes happens to the WF and these changes are responsible for the final WF behaviour.
Conclusion
Alkalization results in WF physico-chemical changes that, if not monitored accurately, may result in serious loss of essential WF properties. Careful optimization of alkalization can result in WF with improved thermal stability, crystallinity and surface features without compromising the WF structural integrity. According to the experimental domain, alkalization of meranti WF resulted in thermal stabilities lower than untreated WF. However, good thermal stability of alkali treated WF was found at mild conditions (e.g. 50 °C/30 min/5%), while severe treatment conditions lowered WF thermal stability (e.g. at 38 °C/96 min/15%). No model could be established to predict and explain alkalization through multiple linear regression analyses due to the co-dependence of treatment factors. It was revealed in the investigation that temperature, alkali concentration and time exhibit individual as well as combined effects on WF treatment. As such, an increase in temperature or alkali concentration under mild conditions (e.g. 30 min/5% or 30 min/25 °C, respectively) increased thermal stability, whereas increases in temperature or alkali concentration under harsher conditions (e.g. 96 min/15% or 90 min/50 °C, respectively) resulted in reduced thermal stability. Furthermore, high alkali concentrations and longer treatment durations show a combined effect and seem to be detrimental to WF thermal stability. Based on the findings from the investigation, it therefore can be recommended that alkalization of meranti WF be done under mild conditions to improve WF thermal stability and surface features without compromising the WF chemical and structural integrity. Also, due to the use and generation of high temperatures (⁓180–260 °C) during internal mixing, extrusion or injection moulding, the use of alkalized WF with compromised thermal stability would likely result in exacerbated degradation of the WF during composite processing and, consequently, poor mechanical properties of composites. Therefore, careful optimization of WF alkalization processes is imperative prior composite processing.
Acknowledgements
This work was financially supported by the Council for Scientific and Industrial Research (CSIR) and the Nelson Mandela University (NMU) Research Capacity and Development.
Declarations
Conflict of interest
The authors declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
