2.1. Kinetic Study of the Process of Birch Ethanol Lignin Sulfation
The scheme of birch ethanol lignin sulfation with the low-toxic sulfamic acid–urea mixture in the 1,4-dioxane medium is shown in
Figure 1. The isolation of sulfated lignin was isolated in the form of an ammonium salt.
The use of the low-toxic and non-corrosive mixture of sulfamic acid and urea for the sulfation of abies ethanol lignin was previously proposed [
16]. The yield of sulfated ethanol lignin of abies wood and sulfur content can be regulated by varying the sulfation process temperature, time, and the ratio of lignin to sulfating complex (sulfamic acid and urea mixture).
A high yield of sulfated lignin and sulfur content can be obtained at different combinations of the above-mentioned parameters of the sulfation process. Taking into account the data obtained when optimizing the process of abies ethanol lignin sulfation with a mixture of sulfamic acid and urea, the ratio of birch ethanol lignin to the sulfating complex was chosen in this study to be 1:3 mol/mol.
The sulfation process temperature ranged from 70 to 110 °C and the process time from 0.5 to 3.0 h. The data on the effect of the birch ethanol lignin sulfation conditions on the yield of sulfated lignin and sulfur content are given in
Table 1.
The high yield of sulfated birch ethanol lignin and the high sulfur content can be obtained at different combinations of the specified parameters of the sulfation process (temperature and time).
The sulfated ethanol lignin samples with a high yield (91.1–96.4%) and a sulfur content of 7.1–8.1 wt % were obtained at a process temperature of 90 °C and a time of 3 h or at a temperature of 100 °C and a time of 1.5–2.0 h. A further increase in the sulfation time and temperature did not significantly affect the sulfur content in the product but reduced the sulfated lignin yield. This may be due to the intensification of secondary condensation and destruction reactions under more severe conditions, which leads to the formation of products that are removed at the stage of sulfated lignin dialysis. It should be noted that the products of sulfation of birch ethanol lignin contain somewhat more sulfur than the products of sulfation of abies ethanol lignin under similar conditions [
16]. This is possibly related to the lower reactivity of coniferous lignins that contain more condensed structural units than hardwood lignins [
4].
The kinetics of the birch ethanol lignin sulfation with sulfamic acid–urea mixture in 1,4-dioxane medium was investigated in the temperature range of 70–100 °C (
Figure 2). The apparent rates of the birch ethanol lignin sulfation were calculated from the change in the sulfur content in sulfated ethanol lignin. The calculation was made using the first-order equation. The activation energy of the sulfation process was determined from the temperature dependence of the rate constants in the Arrhenius coordinates (
Figure 3).
The calculated apparent rate constants and activation energies of the birch ethanol lignin sulfation process are given in
Table 2.
The activation energies of the processes of sulfation of birch and abies ethanol lignins with sulfamic acid–urea mixture in 1,4-dioxane medium are similar: 10.7 kJ/mol for birch lignin (
Table 2) and 8.4 kJ/mol for abies lignin [
16]. It should be noted that, for the process of starch sulfation in a deep eutectic solvent (the sulfamic acid–urea mixture), the activation energy was 6.4 kJ/mol [
20] and, for the sulfation of arabinogalactan with sulfamic acid in DMSO it was 13.1 kJ/mol [
21]. It is known that the low activation energy of the process may indicate the presence of significant diffusion restrictions [
22,
23]. Taking this into account, we can conclude that, under the chosen conditions, the processes of biopolymer sulfation proceed under diffusion restrictions.
The solubility of the sulfated lignin in water increased with an increase in the content of sulfate groups. The maximum sulfur content in the sulfated ethanol lignin was estimated to be 10.6 wt %, taking into account the hypothetical structure of the Berkman spruce lignin [
4], in which one phenylpropane unit has 0.9 mol of free OH groups capable of sulfating. In order to find the conditions that ensure the production of sulfated birch ethanol lignin with maximum yield and sulfur content, a numerical optimization of the sulfation process was carried out.
2.2. Numerical Optimization of the Process of Birch Ethanol Lignin Sulfation
As independent variables, we used two factors: process temperature
X1 (70, 80, 90, 100, 110 °C) and time
X2 (0.5, 0.75, 1.0, 1.5, 2.0, 3.0 h). The result of the sulfation process was characterized by two output parameters: sulfur content
Y1 (wt %) in the sulfated ethanol lignin and sulfated ethanol lignin yield
Y2 (wt %). The fixed parameter was the ratio L:SC = 1:3. A combined multilevel experiment plan (Users Design) was used in the calculations. The designations of the variables are listed in
Table 3.
The experimental results given in
Table 1 were used in the mathematical processing and optimization of the birch ethanol lignin sulfation process.
The dependences of the output parameters on the variable process factors were approximated by second-order regression equations. The results of the variance analysis are given in
Table 4.
The variance analysis showed that, within the limits of the experimental conditions used, the greatest contribution to the total variance of the output parameter was made by both factors: the temperature and time of the birch ethanol lignin sulfation process. This is indicated by the high variance ratios (
F) for the main effects, which are called also the influence efficiencies. The data in the columns of
Table 4 (
P) are interpreted similarly. The influence of the variance source on the output parameter is considered to be statistically significant if its significance level is lower than a specified critical value (in our case, 0.05).
The dependence of the sulfur content
Y1 in the sulfated birch ethanol lignin on the process variables is approximated by the regression equation
The predictive properties of Equation (1) are illustrated in
Figure 4, in which the experimental values of the output parameter
Y1 are compared with its values calculated using Equation (1). The straight line corresponds to the calculated
Y1 values, and the dots correspond to the observed values. The proximity of the experimental points to the straight line confirms the good predictive properties of Equation (1).
The approximation quality is characterized also by the determination coefficient R2adj. In the case under consideration, the value is R2adj = 86.1%, which indicates acceptable approximation quality. This confirms the adequacy of Equation (1) for the experiment and makes it possible to use this equation as a mathematical model of the process under study.
Using the mathematical model, the dependence of the output parameter
Y1 on the variables
X1 and
X2 was plotted in the form of a response surface (
Figure 5).
According to the calculation using mathematical model (1), the maximum predicted sulfur content (8.4 wt %) was obtained at the point corresponding to a process temperature of 107 °C and a process time of 2.3 h.
According to the results of the variance analysis within the limits of the chosen experimental conditions, the sulfation temperature contributes significantly to the total variance of the output parameter Y2 (sulfated lignin yield, wt %). This is indicated by the high variance relation (F) corresponding to this factor and the small P criterion.
The dependence of
Y2 on the variable process factors is approximated by the regression equation
The determination coefficient is fairly high,
R2adj = 92.8%, which evidences the good approximation quality. The latter is also confirmed by the good agreement between the output parameters calculated using Equation (2) and those obtained in the experimental measurements. This confirms the adequacy of Equation (2) for the experiment and its use as a mathematical model of the process under study (
Figure 6).
Figure 7 shows the graphical representation of the dependence of the sulfated birch ethanol lignin yield on the variable factors
X1 and
X2.
According to the above-described model, the optimum conditions for the sulfation of birch ethanol lignin that ensure the maximum yield of sulfated birch ethanol lignin (96.1 wt %) correspond to a process temperature of 78 °C and a time of 2.9 h.
2.3. Characterization of the Sulfated Birch Ethanol Lignin
The substitution of hydroxyl groups for sulfate groups during the sulfation of the birch ethanol lignin with the sulfamic acid–urea mixture was confirmed by FTIR and NMR spectroscopy.
The FTIR spectrum of birch ethanol lignin (
Figure 8) contains absorption bands characteristic of hardwood lignins (GS) [
24]. The band at 1123 cm
−1 corresponds to planar bending vibrations of the C–H syringyl aromatic rings, and the C–O stretching vibrations in secondary alcohols are dominant in the spectrum. The pronounced band with a maximum at 1327 cm
−1 belongs to the skeletal vibrations of the syringyl ring with the C–O stretching vibrations. In addition, the spectrum includes medium-intensity absorption bands around 1271 and 1034 cm
−1, characteristic of the vibrations of the guaiacyl units of lignin [
4].
In contrast to the spectrum of the initial ethanol lignin, the FTIR spectrum of the ammonium salts of the ethanol lignin sulfates contained a new absorption band at 798 cm
−1 (
Figure 8), which corresponds to the stretching vibrations of the C–O–S bond of the sulfate group and a broad absorption band with the maximum at 1218 cm
−1 corresponding to the asymmetric stretching vibrations υ
as(O=S=O).
The FTIR spectra of birch and abies ethanol lignin sulfates [
16] obtained in a similar way are almost identical, except for the presence in the spectrum of the sulfated birch ethanol lignin of the adsorption band at 1331 cm
−1 characteristic of the syringyl structures.
The 2D HSQC NMR spectra of the initial and sulfated birch ethanol lignins are shown in
Figure 9 and
Figure 10, respectively. The main
1H–
13C peaks in the HSQC spectra identified using the literature data [
25,
26,
27] are given in
Table 5, together with the chemical shifts of some low-intensity peaks (not shown in
Figure 9 and
Figure 10). The main structural units and fragments of the initial and sulfated birch ethanol lignins are presented in
Figure 11.
The HSQC spectra of the initial and sulfated ethanol lignin samples were compared in the regions of the chemical shifts of atoms from the lignin side chains (δC/δH 50–90/2.9–5.7 ppm) and aromatic rings (δC/δH 100–150/5.5–8.0 ppm).
Considering the region of the
1H–
13C side chain signals in the HSQC spectrum of the birch ethanol lignin (
Figure 9), we can see that it contains the intense correlation peaks of β-aryl ethers (A), pinoresinol (B), and phenylcoumaran fragments (C) (see
Figure 11). A part of β-aryl ethers is ethoxylated in the α-position, judging by the fact that the spectra contain signals of the methylene group in the α-ethoxylated β-O-4′ bonds (δ
C/δ
H 64.4/3.33) and the α-position of the α-acylated β-O-4′ bonds (δ
C/δ
H 81.2/4.56). This assumption is confirmed by the presence of a correlation signal of the methyl group at δ
C/δ
H 14.3/1.00 ppm.
In the spectrum of the sulfated ethanol lignin (
Figure 10), the group of peaks assigned to the phenylcoumaran fragments (C) is characterized by a change in the position of the correlation signals Cγ-H
γ and Cβ-H
β (
Figure 10), which is related to the effect of sulfation of OH groups in the γ position (Cγ-H
γ: a shift of Δδ
C ~ 4 ppm toward weak fields; Cβ-H
β: a shift of Δδ
C ~ 3 ppm toward strong fields).
A similar change in the peak positions along the carbon atom axis in the sulfated ethanol lignin spectrum is observed for β-aryl ethers β-O-4′ (A) and α-ethoxylated β-aryl ethers (A′). The Cγ-Hγ correlation signals of the sulfated lignin are shifted relative to the initial lignin signals toward weak fields by ~5 ppm and the Cβ-Hβ signals toward strong fields by ~4 ppm. Such shifts of the signals in the spectra are most likely due to the sulfation of OH groups of β-aryl ethers of the lignin macromolecule in the γ position. In addition, free OH groups in the α position of β-aryl ethers (A) are probably subjected to the sulfation, since the Cα-Hα peak at δC/δH 72.5/4.88 ppm is missing in the sulfated lignin spectrum.
The shift of signals in the aliphatic region of the spectrum of the sulfated lignin sample as compared with the spectrum of the initial lignin was also found for the Cγ–Hγ correlations of the cinnamyl alcohol end groups (I). This is also indicative of the sulfation of OH groups bonded with carbon atoms Cγ of this lignin fragment.
Despite the presence of the fairly intense Cα-Hα, Cβ-Hβ, and Cγ-Hγ signals of pinoresinol (β–β′) fragments (B) in the spectrum of the initial ethanol lignin, the intensity of these signals in the spectrum of the sulfated sample drops dramatically. The peaks assigned to the Cγ-Hγ correlations (δC/δH 71.6/3.80 and 4.18 ppm) disappear almost completely.
It is important to note the appearance of a peak at δC/δH 53.6/5.00 ppm assigned to the CH3 or CH groups in the aliphatic region of the sulfated ethanol lignin spectrum. We failed to establish an unambiguous correspondence of this peak to any structural fragment.
The
1H–
13C aromatic region in the HSQC spectrum of the birch ethanol lignin (
Figure 9) contains characteristic correlation peaks of the syringyl (S) and guaiacyl (G) units, p-coumarates (pCA), and cinnamyl aldehyde end groups (J). The syringyl units are of several types. In particular, using the assignments made in [
28,
29,
30], we found S units with a substituent in the 4-position (δ
C/δ
H 104.5/6.70), S units with a free hydroxyl group in the 4-position (δ
C/δ
H 106.2/6.51), S units with a carbonyl group in the α-position (δ
C/δ
H 107.1/7.35), and S units with a carboxyl group in the α position (δ
C/δ
H 107.2/7.23).
In addition, there are high-intensity signals at δ
C/δ
H = 129.1/7.73 and 131.9/7.67 ppm corresponding to the CH or CH
3 groups. Some researchers believe that the signals located at these chemical shifts can be attributed to both C
α,β−H
α,β in stilbenes [
31] and C2,6-H
2,6 in p-benzoates [
32].
In the aromatic region of the sulfated lignin spectrum containing the main part of the peaks of the syringyl (S
2,6eth, S′
2,6, S″
2,6) and guaiacyl units (G
2, G
5, G
6), p-coumarates (pCA
3,5), and cinnamyl aldehyde end groups (J
2,6) (see
Table 6), the signal of the C2,6-H
2,6 syringyl units with a free hydroxyl group in the 4-position disappears almost completely. This is apparently due to the replacement of this hydroxyl group by the sulfate one. In addition, in this region of the spectrum, a new peak at δ
C/δ
H = 120.8/7.38 and 7.32 ppm appears, which is most likely a signal of the C5-H
5 guaiacyl (Gs) and C3,5-H
3,5 p-coumarate (pCAs) structural units with sulfate groups attached to the 4-position. These changes in the chemical shifts of adjacent positions 3 and 5 of the aromatic ring caused by esterification correspond to the expected change of the substituent in phenol [
33]. The possible substitution of the sulfate group of the phenolic hydroxyls for the guaiacyl (G) units in the 4-position is evidenced also by the almost complete disappearance of the peak at δ
C/δ
H = 115.7/6.97 ppm characteristic of the C5-H
5 guaiacyl units (G
5) with unsubstituted hydroxyl in the 4-position [
34].
Based on the data obtained, it can be concluded that sulfation affects the acceptable aliphatic hydroxyl groups of lignin at the γ-positions of β-aryl ethers, α-ethoxylated β-aryl ethers, phenylcoumaran substructures, and cinnamyl alcohol end groups, as well as the unsubstituted hydroxyl groups in the α-position of β-aryl ethers. In addition, free phenolic hydroxyl groups in the 4-position of syringyl and guaiacyl units and p-coumarates can be subjected to sulfation.
The comparison of the 2D NMR spectroscopy data on birch and abies ethanol lignin sulfates [
16] revealed higher structural diversity of the sulfated birch ethanol lignin. A significant difference between the HSQC spectra of birch ethanol lignin and abies ethanol lignin in the area of correlations of the aromatic fragments is the presence of several types of syringyl units in the former. However, both samples are built from the phenylpropane structural units linked by simple ether (β-O-4′) and C–C (β–β′, β–5′) bonds, while the sulfate groups are localized mainly in the γ and α positions of the side chains and, probably, in the 4-position of aromatic rings.
Data on the molecular weight distribution of the initial and sulfated birch ethanol lignin were obtained by the GPC method. The molecular weight distribution curves for the initial and sulfated birch ethanol lignin samples are shown in
Figure 12. The average molecular weights and polydispersity of the initial and sulfated ethanol lignins are indicated in
Table 6.
The birch ethanol lignin obtained in [
35] had a low molecular weight (
Mw = 1800 Da) and a monomodal distribution (PD = 2.02), which evidences higher homogeneity of the sample as compared to Alcell birch lignin (
Mw = 3470 and PD = 4.1) [
17].
As a result of birch ethanol lignin sulfation, the weight average molecular weight
Mw of the samples increased from ~1800 to ~7600 Da. Such a significant growth is related to an increase in the weight of lignin macromolecules due to the introduction of sulfate groups and the removal of the low molecular weight fraction of the sulfated lignin along with inorganic impurities at the dialysis stage. A feature distinguishing sulfated birch ethanol lignin from abies ethanol lignin sulfated under similar conditions [
16] is the bimodal molecular weight distribution (
Figure 12). The molecular weight distribution curve of birch ethanol lignin has two pronounced peaks with molecular weights of ~5000 and ~12,000 Da. These peaks can be attributed to the heterogeneity of the initial ethanol lignin molecules, which enter into the sulfation reaction in different ways. The low molecular weight lignin fraction is possibly less sulfated than the high molecular weight fraction, which is reflected in the separation of the peaks in the molecular weight distribution curve. Sulfated birch ethanol lignin has a higher polydispersity and a higher average molecular weight than sulfated abies ethanol lignin (
Mw~5300 Da, PD = 1.63) [
16].
2.4. Thermochemical Properties of the Birch Ethanol Lignin
The thermochemical properties of the birch ethanol lignin were studied using the non-isothermal TG/DTG analysis in an argon medium in the temperature range of 30–900 °C.
The thermal decomposition of the ethanol lignin occurred over a wide temperature range, since its structure contains various functional groups with different thermal stabilities (
Figure 13).
The sample weight loss at temperatures of 30–180 °C was found to be less than 1%. This is explained by the loss of moisture and adsorbed gases. The main thermal decomposition of the ethanol lignin started after 200 °C and practically ended at 600 °C. The solid residue yield gradually decreased with an increase in temperature to 700 °C and then remained constant invariable. At a pyrolysis temperature of 900 °C, the solid residue yield was 34.6 wt %, which is somewhat less than in the case of pyrolysis of the abies ethanol lignin under similar conditions (36.2 wt %) [
36]. As is known [
37], coniferous lignins consist mainly of the guaiacyl structures, while in hardwood lignins, the syringyl structures dominate. The high yield of the carbon residue during the thermal decomposition of abies ethanol lignin is probably due to the tendency of the guaiacyl propane units to condensation reactions [
38].
The DTG curve has a broad peak corresponding to the main thermal decomposition of ethanol lignin and an implicit peak. The maximum rate of thermal degradation of the birch ethanol lignin (4.1%/min) was reached at 372 °C. In the temperature range of 350–400 °C, the main lignin structural moieties (guaiacyl and syringyl) underwent cracking with the formation of phenol-type compounds of different molecular weights, the yield of which increased with temperature [
39].
At this temperature range, the pyrolysis products represent a complex mixture of organic compounds containing the aromatic, hydroxyl, and alkyl groups and reflecting the composition and structural features of the initial lignin [
40]. During the thermal decomposition of the lignin, the competing depolymerization reactions with the formation of lower molecular weight aromatic products and cross-linking reactions of aromatic compounds and their carbonization occurred [
40]. In the temperature range of 450–600 °C, the birch ethanol lignin weight loss rate significantly decreased and the thermal decomposition was mainly completed at 600 °C. In this case, some of the aromatic rings in the lignin probably decomposed and condensed into carbon products [
41].
The sulfation of the birch ethanol lignin noticeably changed the nature of its thermal transformation (
Figure 14).
According to the data presented in
Figure 14a, the sulfation of the birch ethanol lignin reduced its thermal stability. At temperature 300 °C, the sulfated ethanol lignin lost 26.4% of its initial weight, while the initial lignin lost only 15.7%. This tendency continued until the completion of the pyrolysis process.
The sulfation of the birch ethanol lignin also changed its thermal transformation profile (
Figure 14b). In the temperature range of 100–150 °C, the sulfated ethanol lignin weight loss rate was much higher than in the case of initial ethanol lignin. As was shown in [
42], in this temperature range, the aliphatic hydroxyl groups, carbonyl groups, and C–C bonds in the lignin side chains are broken.
In the temperature range of 200–350 °C, an intense narrow peak appeared in the DTG curve of the sulfated birch ethanol lignin, with a maximum weight loss rate of 6.7%/min at 317 °C, which is attributed to the thermal decomposition of sulfate groups [
43].
Thus, the TG/DTG study showed that the syringyl structure of hardwood (birch) ethanol lignin was thermally less stable than the guaiacyl structure dominating in softwood (abies) ethanol lignin. Additionally, the introduction of sulfate groups into the structure of birch ethanol lignin reduced its thermal stability.