*3.4. <sup>1</sup>H and <sup>13</sup>C NMR*

NMR analysis is frequently used to predict lignin's structural details concerning its molecular characteristics, reactivity, and composition. The acetylation of lignin before NMR analysis aims to decrease the impurities in lignin that may interfere with the spectrum [25]. Due to the complex structure of polymer lignin, typically simple proton <sup>1</sup>H NMR resulted in overlapping spectra which is difficult to justify the structure. Hence, <sup>13</sup>C NMR is needed to support the hypothesis of <sup>1</sup>H NMR. The presence of condensed and uncondensed aliphatic and aromatic carbon and aryl ethers can be detected by natural <sup>13</sup>C isotope NMR. However, longer scanning and acquisition times are required to improve signal sensitivity due to the low abundance of carbon isotope in the lignin molecule [4]. Still, quantitative <sup>13</sup>C-NMR can be a useful technique for lignin structural investigation, particularly in determining molecular alterations caused by different isolation procedures and biomass sources [8,25,35]. *Polymers* **2022**, *14*, x FOR PEER REVIEW 10 of 20

> The <sup>1</sup>H NMR spectrum in Figure 4 shows similar peaks among the three lignin samples. The small peak at 0.8 and 1.23 ppm occurred because of saturated aliphatic lignin protons in the methyl and methylene chain. The intense signal at 1.98 ppm indicates the presence of an aliphatic acetate group. The strong signal at 2.5 ppm and 3.3 is because of protons in water and DMSO. A pronounced peak at 3.76 ppm corresponds to methoxyl protons (-OCH3). A sharper signal at 6.7–6.9 in isolated lignin spectrum suggests more syringyl units than reference lignin. The 1H NMR spectrum in Figure 4 shows similar peaks among the three lignin samples. The small peak at 0.8 and 1.23 ppm occurred because of saturated aliphatic lignin protons in the methyl and methylene chain. The intense signal at 1.98 ppm indicates the presence of an aliphatic acetate group. The strong signal at 2.5 ppm and 3.3 is because of protons in water and DMSO. A pronounced peak at 3.76 ppm corresponds to methoxyl protons (-OCH3). A sharper signal at 6.7–6.9 in isolated lignin spectrum suggests more syringyl units than reference lignin.

**Figure 4.** 1H NMR signal of standard lignin (**a**), single-step lignin (**b**), and fractionated-step lignin **Figure 4.** <sup>1</sup>H NMR signal of standard lignin (**a**), single-step lignin (**b**), and fractionated-step lignin (**c**).

(**c**). Conversely, a more intensive peak at 7 ppm was found in reference lignin due to the higher guaicyl content of reference lignin than isolated lignin. This trend agrees with the FTIR and Pyr-GC/MS results, where isolated lignin has more syringyl units than reference lignin. The other strong signal in the range of 7.5–8.5 ppm reveals an aromatic presence in p-hydroxyphenyl proton positions 2 and 6 [38]. The obtained signal in the NMR spectra of isolated lignin was similar to lignin from sweet sorghum stem (SST) [34] and lignin kraft [8]. Nevertheless, the peak spectrum was slightly different with lignin from Ginkgo shells, where the Hα signals at 5.5–5.9 ppm related to linkages of β-O-4′ and β-5′ were produced [39]. The results may reveal that different sources of lignin generated different structures. Conversely, a more intensive peak at 7 ppm was found in reference lignin due to the higher guaicyl content of reference lignin than isolated lignin. This trend agrees with the FTIR and Pyr-GC/MS results, where isolated lignin has more syringyl units than reference lignin. The other strong signal in the range of 7.5–8.5 ppm reveals an aromatic presence in p-hydroxyphenyl proton positions 2 and 6 [38]. The obtained signal in the NMR spectra of isolated lignin was similar to lignin from sweet sorghum stem (SST) [34] and lignin kraft [8]. Nevertheless, the peak spectrum was slightly different with lignin from Ginkgo shells, where the Hα signals at 5.5–5.9 ppm related to linkages of β-O-40 and β-50 were produced [39]. The results may reveal that different sources of lignin generated different structures.

The 13C NMR spectra (Figure 5) shows a unique signal related to the lignins in this study. All the lignin samples show a similar trend of peaks where five typical lignin sig-

in lignin where isolated lignin (b–c) has a stronger signal than commercial lignin (a), which corresponds with FTIR spectra (2918 and 2854 cm−1) and 1H NMR (0.8 and 1.23 ppm). Meanwhile, the strong peak at 40 ppm belongs to DMSO as a solvent. The intensive signal at 55.9 ppm is attributed to the methoxy group in the G and S units. A signal related to esterified syringyl unit in C3/C5 is observed at 152 ppm in isolated lignin (b–c) but is not detected in reference lignin (a). Repeatedly, this result agrees with FTIR, Py-GC/MS, and 1H NMR. A strong signal at a 170–160 ppm range implies ester linkage (-COO−) at γ position [34].

The <sup>13</sup>C NMR spectra (Figure 5) shows a unique signal related to the lignins in this study. All the lignin samples show a similar trend of peaks where five typical lignin signals show strong resonance such as aliphatic chains, solvent (DMSO), methoxy, C3/C5, and ester. The signal in region between 20–30 ppm represents an aliphatic chain structure in lignin where isolated lignin (b–c) has a stronger signal than commercial lignin (a), which corresponds with FTIR spectra (2918 and 2854 cm−<sup>1</sup> ) and <sup>1</sup>H NMR (0.8 and 1.23 ppm). Meanwhile, the strong peak at 40 ppm belongs to DMSO as a solvent. The intensive signal at 55.9 ppm is attributed to the methoxy group in the G and S units. A signal related to esterified syringyl unit in C3/C5 is observed at 152 ppm in isolated lignin (b–c) but is not detected in reference lignin (a). Repeatedly, this result agrees with FTIR, Py-GC/MS, and <sup>1</sup>H NMR. A strong signal at a 170–160 ppm range implies ester linkage (-COO−) at γ position [34]. *Polymers* **2022**, *14*, x FOR PEER REVIEW 11 of 20

**Figure 5.** 13C NMR spectra of standard lignin (**a**), single-step lignin (**b**), and fractionated-step lignin (**c**). **Figure 5.** <sup>13</sup>C NMR spectra of standard lignin (**a**), single-step lignin (**b**), and fractionated-step lignin (**c**).

In general, the obtained signal in this study was also detected in lignin from sweet sorghum stem SST [34] and lignin kraft from the industrial residue [32]. However, the assigned peak correlated to the G unit is not detected in this current spectrum. it is likely that this is caused by either the lignin concentration being too dilute or the presence of the Nuclear Overhauser Effect (NOE). According to a report by Wen et al. [40], an important aspect of lignin characterization fulfills three criteria. First, lignin should be free from impurities; for this case, acetylation improved the purity of lignin, which was proved by residual carbohydrate's absence signal at 62, 73–75, and 100–102 ppm [25]. Second, lignin solution must be concentrated to minimize baselined phasing distortion and increase the signal-to-noise ratio, yet this requirement negatively affects the LC column. Third, to avoid the NOE, the inverse-gated decoupling sequence (i.e., C13IG pulse) should be utilized, which entails turning off the proton decoupling during the recovery between pulses In general, the obtained signal in this study was also detected in lignin from sweet sorghum stem SST [34] and lignin kraft from the industrial residue [32]. However, the assigned peak correlated to the G unit is not detected in this current spectrum. it is likely that this is caused by either the lignin concentration being too dilute or the presence of the Nuclear Overhauser Effect (NOE). According to a report by Wen et al. [40], an important aspect of lignin characterization fulfills three criteria. First, lignin should be free from impurities; for this case, acetylation improved the purity of lignin, which was proved by residual carbohydrate's absence signal at 62, 73–75, and 100–102 ppm [25]. Second, lignin solution must be concentrated to minimize baselined phasing distortion and increase the signal-to-noise ratio, yet this requirement negatively affects the LC column. Third, to avoid the NOE, the inverse-gated decoupling sequence (i.e., C13IG pulse) should be utilized, which entails turning off the proton decoupling during the recovery between pulses [40].

#### [40]. *3.5. Thermal Behavior of Lignin*

nation of humidity [42].

*3.5. Thermal Behavior of Lignin*  Mass loss (TG) and mass loss rate (DTG) curves of *A. mangium* lignin are shown in Figure 6a to indicate the similar thermal characteristics between isolated lignin and reference lignin. The reaction region of all stages shifts toward a higher temperature by increasing the heating rate for isolated and reference lignin. The primary loss stage of two isolated lignins and their reference was located in a broad temperature range (between Mass loss (TG) and mass loss rate (DTG) curves of *A. mangium* lignin are shown in Figure 6a to indicate the similar thermal characteristics between isolated lignin and reference lignin. The reaction region of all stages shifts toward a higher temperature by increasing the heating rate for isolated and reference lignin. The primary loss stage of two isolated lignins and their reference was located in a broad temperature range (between 100 ◦C and 700 ◦C), representing a complex structure consisting of phenolic hydroxyl, carbonyl benzylic hydroxyl functionalities [41]. The decomposition of lignin by

100 °C and 700 °C), representing a complex structure consisting of phenolic hydroxyl, carbonyl benzylic hydroxyl functionalities [41]. The decomposition of lignin by temperature can be divided into three stages. The initial pyrolysis stage at around 100 °C with a higher

mainly attributed to the moisture evaporation in lignin and releasing of volatile products such as carbon dioxide and carbon monoxide [30]. A similar degradation study of lignin reported that an endothermic peak ranges from 100–180 °C, corresponding to the elimi-

The second pyrolysis stage, between 120 °C and 270 °C, indicated the decomposition of lignin into some possible degradation products and the removal of carbohydrates from lignin. The peak of the stage was around 200 °C, and below this peak, lignin is thermally stable. The losses in this stage were derived by aromatic decomposition as a phenolic com-

temperature can be divided into three stages. The initial pyrolysis stage at around 100 ◦C with a higher mass loss was represented by the fractionated-step lignin. This first stage, up to 200 ◦C, is mainly attributed to the moisture evaporation in lignin and releasing of volatile products such as carbon dioxide and carbon monoxide [30]. A similar degradation study of lignin reported that an endothermic peak ranges from 100–180 ◦C, corresponding to the elimination of humidity [42]. applications [35]. The Tg value varies widely depending on the method of lignin isolation, adsorbed water, molecular weight, and thermal history [49]. Since single-step lignins with higher Tg values are more stable at high temperatures, the process requires higher temperature operation.

nin, attributed to volatiles' intensive evolution in the third stage [43].

*Polymers* **2022**, *14*, x FOR PEER REVIEW 12 of 20

degradation of the propanoic side chains of lignin [41].

pounds, such as the cleavage of ether linkages among the C9 units [43]. Thermal degradation of lignin is followed by condensation processes, leading to unsaturated C=C bonds occurring in the temperature range of 160 to 270 °C. Afterward, the production of vinyl guaiacol, ethyl, and methyl byproducts is usually obtained at 230 and 260 °C with the

The third pyrolysis stage had a temperature range of 270 °C to 700 °C, with the prominent peak being around 350 °C and 380 °C for the fractionated-step lignin and singlestep, respectively. The reference lignin reached the highest peak, indicating it as the most stable. The lignin structure is decomposed majorly at a temperature of 260–478 °C. At temperatures below 310 °C, aryl ether links tend to cleave, caused by low thermal stability [44]. At higher temperatures (>500 °C), aromatic structures rearranged and condensed the lead into char [44,45] and released volatile products. The high capacity to produce char by lignin makes it an efficient alternative to improve the flame retardancy of polymers [7]. Similar tendencies were also observed in other studies concerning the mass loss and the evolution of the volatiles against the origin and pyrolysis temperature [46]. Hu et al. [47] studied the isolated lignins extracted by different solvents and reported that CH4, CO, and phenols are lignin's main mass loss stage. Based on TGA analysis in the third stage, the fractionated-step lignin was decomposed at a lower temperature with moderate mass loss than the single-step lignin due to the extraction method used. This result indicates that fractionated-step lignin yields better purity than single-step lignin. The mass loss in this third stage was remarkable over in the second stage for isolated lignin and reference lig-

The glass transition temperature (Tg) of the lignin fractionated-step and single-step was higher than that of reference lignin, with Tg temperatures of 184, 167, and 154 °C, respectively (Figure 6b). Furthermore, the first peak at 50 °C indicated an endothermic process in which the lignin absorbed heat energy to evaporate water and other volatile substances [48]. Tg represents the end of an endothermic process in which the lignin structure changes from a glassy state into a rubbery (plasticized) state. The wide range of Tg values indicated the flexibility and stiffness at higher temperatures, beneficial in industrial

**Figure 6.** TGA (**a**) and DSC (**b**) thermogram of lignin standard, single-step lignin, and fractionatedstep lignin. **Figure 6.** TGA (**a**) and DSC (**b**) thermogram of lignin standard, single-step lignin, and fractionatedstep lignin.

The second pyrolysis stage, between 120 ◦C and 270 ◦C, indicated the decomposition of lignin into some possible degradation products and the removal of carbohydrates from lignin. The peak of the stage was around 200 ◦C, and below this peak, lignin is thermally stable. The losses in this stage were derived by aromatic decomposition as a phenolic compounds, such as the cleavage of ether linkages among the C9 units [43]. Thermal degradation of lignin is followed by condensation processes, leading to unsaturated C=C bonds occurring in the temperature range of 160 to 270 ◦C. Afterward, the production of vinyl guaiacol, ethyl, and methyl byproducts is usually obtained at 230 and 260 ◦C with the degradation of the propanoic side chains of lignin [41].

The third pyrolysis stage had a temperature range of 270 ◦C to 700 ◦C, with the prominent peak being around 350 ◦C and 380 ◦C for the fractionated-step lignin and single-step, respectively. The reference lignin reached the highest peak, indicating it as the most stable. The lignin structure is decomposed majorly at a temperature of 260–478 ◦C. At temperatures below 310 ◦C, aryl ether links tend to cleave, caused by low thermal stability [44]. At higher temperatures (>500 ◦C), aromatic structures rearranged and condensed the lead into char [44,45] and released volatile products. The high capacity to produce char by lignin makes it an efficient alternative to improve the flame retardancy of polymers [7]. Similar tendencies were also observed in other studies concerning the mass loss and the evolution of the volatiles against the origin and pyrolysis temperature [46]. Hu et al. [47] studied the isolated lignins extracted by different solvents and reported that CH4, CO, and phenols are lignin's main mass loss stage. Based on TGA analysis in the third stage, the fractionated-step lignin was decomposed at a lower temperature with moderate mass loss than the single-step lignin due to the extraction method used. This result indicates that fractionated-step lignin yields better purity than single-step lignin. The mass loss in this third stage was remarkable over in the second stage for isolated lignin and reference lignin, attributed to volatiles' intensive evolution in the third stage [43].

The glass transition temperature (Tg) of the lignin fractionated-step and single-step was higher than that of reference lignin, with T<sup>g</sup> temperatures of 184, 167, and 154 ◦C, respectively (Figure 6b). Furthermore, the first peak at 50 ◦C indicated an endothermic process in which the lignin absorbed heat energy to evaporate water and other volatile substances [48]. T<sup>g</sup> represents the end of an endothermic process in which the lignin structure changes from a glassy state into a rubbery (plasticized) state. The wide range of T<sup>g</sup> values indicated the flexibility and stiffness at higher temperatures, beneficial in industrial applications [35]. The T<sup>g</sup> value varies widely depending on the method of lignin isolation, adsorbed water, molecular weight, and thermal history [49]. Since single-step lignins with higher T<sup>g</sup> values are more stable at high temperatures, the process requires higher temperature operation.

The higher Tg value of isolated lignin compared to reference lignin was due to the higher amount of phOH content in the isolated lignin (Table 1). Intramolecular hydrogen bonds between phOH groups in the main back bonds of lignin contributed to the higher Tg. The bonds created a physically cross-linked structure [50]. The T<sup>g</sup> value is influenced by the solubility of the organic solvents, where higher solubility is obtained with a lower T<sup>g</sup> value [51,52]. The reference lignin had a higher absorbance in dioxane, the organic solvent, than isolated lignin, as presented in Figure 1b (solubility lignin in dioxane). This finding is supported by the lower T<sup>g</sup> value of reference lignin than the isolated lignin and agrees with the results reported by Dastpak et al. [52] where Kraft lignin had higher T<sup>g</sup> and lower solubility in the organic solvent than organosolv lignin. The T<sup>g</sup> value corresponds positively with the molecular mass of lignin [53–55]. The T<sup>g</sup> shifts to higher temperatures by increasing the average molar mass [56]. Based on the T<sup>g</sup> value, the fractionated-step lignin had a lower molecular mass than the single-step lignin. This might be caused by the more extended process obtained by using acid precipitation to condense lignin, continued by ethanol addition to adsorbing the carbohydrate attached to lignin during the kraft pulping process. This suggestion was supported by GPC analysis in the next section.

Impurities influenced the T<sup>g</sup> value in the lignin sample represented by ash content. The fractionated-step lignin and reference lignin had higher ash content (1.96% and 2.58%, respectively) compared to single-step lignin (0.48%), resulting in a lower T<sup>g</sup> value. Sameni et al. (2013) also reported a higher percentage of impurities obtained with a lower T<sup>g</sup> value. The abundance of aromatic rings in the main backbone of lignin can also contribute to the higher T<sup>g</sup> value of isolated lignin. The varied T<sup>g</sup> values were due to the heterogeneous structures and the broad molecular weight of isolated lignin samples [57]. These factors also affect interchain hydrogen bonding, cross-linking density, and rigid phenyl groups [58]. Although several studies reported an increase of char residue with higher Tg values, the results may be inconsistent due to the plant sources and extraction conditions [35]. The two isolated lignins from *A. mangium* in this study had higher T<sup>g</sup> values in comparison to the others hardwood lignin, for instance T<sup>g</sup> value from *Eucalyptus Grandis* was 161 ◦C [59], while eucalyptus kraft lignin was 133 ◦C [60], and other hardwood kraft lignin showed values such as 108 ◦C [61] and 138 ◦C [62]. pH solution conditions also influenced the T<sup>g</sup> during lignin precipitation that varied from 106.12–131.81 ◦C at pH 1–5 [55].

#### *3.6. Chemical Elucidation by Mass Spectrometry*

Py-GCMS helps determine the lignin degradation and the existence of carbohydrates and other additives [63]. The PyGCMS method can help unravel the nature of lignins, elemental composition, number of formed products, and the isolation method [64]. The monomer unit in hardwood lignin consists of syringyl (S) and guaiacyl (G) units with varying proportions, while softwood is dominated by a high proportion of G units with less p-hydroxyl phenyl (H) units [65]. In native wood, higher S units are easier to delignify than lower S units contributed by the erythro-rich and S unit rich in β-O-4 structure [66]. Some kinds of pyrolysis products can be found in pyrogram (Figure 7)-derived S unit, G unit, and H unit with different relative abundance between lignin samples. The higher retention time indicated that the compound was degraded at a higher temperature. At a

pyrolysis temperature of 350 ◦C, phenolic compounds such as eugenol (G6), aldehydes, ketones, or alcohol group from G- and S-unit were released. Other compounds such as vanillin (G7) and acetoguaiacone from G-unit were degraded at 200–400 ◦C, in which the β-ether was separated [67]. relative abundance of S unit higher than the single-step lignin, which showed that the fractionated-step lignin had more β-O-4 ether linkages and was more reactive than the single-step lignin.

result indicated that the primarily pyrolysis composition of *A. mangium* kraft lignin from kraft black liquor of industrial pulp and paper was an S unit. This result is similar to wheat straw and pine sawdust lignin, mainly G unit. However, a different finding was reported in palm kernel shell (PKS) lignin, which was dominated by the H unit [69]. The classification of reference lignin was G-lignin, while single-step and fractionated-step lignin are classified as SGH-lignin. The total relative abundance of H and G units of single-step lignin was higher than that of fractionated-step. Inversely, the total relative abundance of the S unit of the single-step lignin was lower than fractionated-step lignin. H and G units are easier formed in terms of the possibility of forming condensation in the acid precipitation process. Based on this classification, the biphenyl bond of single-step lignin is relatively higher than fractionated-step lignin. The H unit does not have a methoxy group, and the G unit only has one methoxy group, leading to the formation of C-C biphenyl linkages. Thus, there is a higher yield of single-step lignin compared to fractionated-step lignin. The biphenyl linkages are included as covalent linkages, which are relatively more stable than β-O-4 ether linkages [68], so single-step lignin is relatively more stable and resistant to thermal treatment and biodegradation. The fractionated-step lignin had a total

*Polymers* **2022**, *14*, x FOR PEER REVIEW 14 of 20

**Figure 7.** Pyrogram of single-step lignin and fractionated-step lignin compared to reference lignin. **Figure 7.** Pyrogram of single-step lignin and fractionated-step lignin compared to reference lignin.

**Table 3.** The list of pyrolysis product reference lignin, single-step lignin, and fractionated-step lignin. **Unit Pyrolysis Product Relative Abundance (%) Reference Fragmentation (***m/z***) Lignin Single-Step Lignin Fractionated-Step Lignin H1** Phenol 1.02 2.76 3.58 94, 66, 45 **H2** Phenol, 2-methyl- 1.18 1.39 1.32 108, 90, 79 **H3** Phenol, 3 + 4-methyll 2.24 3.25 2.26 107, 90, 79 **H4** Phenol, 2,4-dimethyl- 1.00 0.40 0.00 122, 107, 77 **H5** Phenol, 4-vinyl 0.20 0.48 0.66 120, 91, 65, 40 **H6** Catechol, 3-methyl 3.63 9.70 9.66 124, 78 **H7** Catechol, 4-methyl 3.90 3.81 1.99 124, 78 Table 3 shows pyrolysis products of reference, single-step, and fractionated-step lignin. The compounds detected by Py-GCMS are classified into five categories: aliphatic oxygen compounds and hydrocarbons, aromatic hydrocarbons, furan and phenol derivatives [68]. Based on pyrolysis products, single-step lignin had the highest total relative abundance of H unit followed by fractionated-step lignin and reference lignin. G unit presented the highest portion compared to the S and H unit in both isolated and reference lignin. The result indicated that the primarily pyrolysis composition of *A. mangium* kraft lignin from kraft black liquor of industrial pulp and paper was an S unit. This result is similar to wheat straw and pine sawdust lignin, mainly G unit. However, a different finding was reported in palm kernel shell (PKS) lignin, which was dominated by the H unit [69]. The classification of reference lignin was G-lignin, while single-step and fractionated-step lignin are classified as SGH-lignin. The total relative abundance of H and G units of single-step lignin was higher than that of fractionated-step. Inversely, the total relative abundance of the S unit of the single-step lignin was lower than fractionated-step lignin. H and G units are easier formed in terms of the possibility of forming condensation in the acid precipitation process. Based on this classification, the biphenyl bond of single-step lignin is relatively higher than fractionated-step lignin. The H unit does not have a methoxy group, and the G unit only has one methoxy group, leading to the formation of C-C biphenyl linkages. Thus, there is a higher yield of single-step lignin compared to fractionated-step lignin. The biphenyl linkages are included as covalent linkages, which are relatively more stable than β-O-4 ether linkages [68], so single-step lignin is relatively more stable and resistant to thermal treatment and biodegradation. The fractionated-step lignin had a total relative abundance of S unit higher than the single-step lignin, which showed that the fractionated-step lignin had more β-O-4 ether linkages and was more reactive than the single-step lignin.


**Table 3.** The list of pyrolysis product reference lignin, single-step lignin, and fractionated-step lignin.
