International Journal of Biological Macromolecules 97 (2017) 164172

Contents lists available at ScienceDirect

International Journal of Biological Macromolecules

journal homepage: www.elsevier.com/locate/ijbiomac

Isolation and further structural characterization of lignins from the

valonea of Quercus variabilis
Lina Yang, Dongmei Wang , Dan Zhou, Yawei Zhang, Tingting Yang
College of Forestry, Northwest A & F University, Yangling, Shaanxi, 712100, China

a r t i c l e i n f o a b s t r a c t

Article history: The isolation process of alkali lignin (AL) from the valonea of Quercus variabilis Blume was optimized
Received 28 October 2016 (liquid/solid ratio, 12.21; isolation time, 4.21 h; isolation temperature, 42.21 C; and alkali concentration,
Received in revised form 0.85 mol/L) using the response surface method (RSM), with the highest isolation rate obtained being
26 December 2016
22.67%. Then, the apparent structures of AL, enzymatic hydrolysis lignin (CEL) and milled wood lignin
Accepted 2 January 2017
(MWL) were studied by scanning electron microscopy (SEM), which indicated that the isolation processes
Available online 3 January 2017
of AL and CEL caused some damage to the apparent structure of lignin. The comprehensive structure
characteristics of lignin samples was studied using 1 H, 13 C and 2D-HSQC techniques based on former
Keywords:
Valonea of Quercus variabilis
studies. It was found that (1) three lignins were GSH-type; (2) the relative content of -O-4 linkages in
Lignin CEL (75.91%) was lower than those in AL (91.57%) and MWL (83.23%), suggesting that the -O-4 linkages
Structural characterization were largely cleaved during the CEL isolation process. In addition, the existence of phenylcoumarane,
Nuclear magnetic response (NMR) ferulic acid, p-coumarates and p-hydroxycinnamyl alcohol end groups can be found; (3) The S/G ratios
were estimated to be 8.72, 1.30 and 0.98 for AL, MWL and CEL, respectively, suggesting that the lignin
fragment rich in S-units was easily released under the alkali conditions.
2017 Elsevier B.V. All rights reserved.

1. Introduction
Lignin is a natural polymer with a three-dimensional structure
that widely exists in plant cells in higher plants, including ferns
[1]. It is found that lignin has a very complicated main structure,
which is mainly composed of guaiacyl (G), syringyl (S), and hydrox-
yphenyl (H) units [2]. The considerable variation in the structure of
lignin can be caused by the growing conditions, plant species, and
estimate the relative percentage of inter-unit linkages, S/G ratio and
acylation degree based on the signals at the separable contour level
presented in the HSQC spectrum, which has been used in some non-
woody plants [5]. Unfortunately, this method cannot reect the
quantity of inter-unit linkages truthfully via calculating the relative
percentage of inter-unit linkages [6]. Otherwise, previous literature
has shown that it is worth noting that the combination of HSQC and
13 C NMR provides more satisfactory quantitative results [7]. There-

isolation methods.
Meanwhile, as an aromatic phenolic polymer, the structure of
lignin dramatically affects the effective use of lignin, and thus, the
knowledge of the lignin structure is signicant to its ultimate util-
ity. For example, recently, the S/G ratio has been reported as the key
factor affecting the rate of pulping of hardwoods [3]. The reported
study showed that the S/G ratio appears to control the -O-4 con-
tent, E/T ratio, degree of condensation and methoxyl content [3].
13 C NMR and 2D-HSQC have been developed as the core of struc-
tural illustration of lignin on account of NMR having the advantages
of the analysis of the entire lignin structure and the direct detection
of lignin moieties [4]. A semi-quantitative 2D-NMR technique can
Corresponding author.
E-mail address: dmwli@163.com (D. Wang).
fore, the key technique in the quantication method adopted is to
choose a suitable internal standard reference signal for the lignin
sample with similar structural features. The signicant technique
is converting the relative integration values obtained from the cor-
responding 2-D spectrum to absolute values by selecting internal
standard references and coupling with the quantitative 13 C NMR
spectrum [6].
In addition, scanning electron microscopy (SEM) is used to
examine the microstructure, morphology, crystalline structure and
size of the lignin polymer [8]. However, SEM is limited in quanti-
tatively evaluating the purity of a lignin sample. More seriously,
there is no published algorithm that can convert such images into
numerical data in spite of it providing information on the particle
size and surface properties of lignin [8]. When the lignin samples
are isolated by different means of chemical or mechanical meth-
ods, the particle sizes might change, and the surface of lignin might

http://dx.doi.org/10.1016/j.ijbiomac.2017.01.008
0141-8130/ 2017 Elsevier B.V. All rights reserved.
 L. Yang et al. / International Journal of Biological Macromolecules 97 (2017) 164172
become smoother or rougher depending on the degree of intensity
of the isolation process [8].
Q. variabilis Bl., the Chinese cork oak, which is widely distributed
in northern and eastern China, is an oak species with a valonea
outer shell, and it is mainly used for timber, cork, and water-soil
conservation. Valonea contains approximately 28.45% tannin with
a high purity of 77.64% and has been used for tannin extract, which
has a signicant economic value for the leather industry. It has
long been commercially used to extract ellagitannins, a low added-
value product of approximately $500 per ton, as a tannin agent,
which convert hides/skins into leather. There are approximately
136.35 104 hm2 of Q. variabilis in China [9].
In our previous study, the lignin preparations from the valonea
of Q. variabilis Bl. were studied by TG, GPC, FTIR and UVvis tech-
niques. Additionally, the chemical compositions of valonea were
tested by HPLC. The results showed that there is approximately 32%
lignin in the valonea, and different isolation methods signicantly
affect the lignin structure [9]. However, the detailed structural
characteristics of lignin were not explained, which results in some
disadvantages in the utilization of lignin from Q. variabilis. In this
study, the isolation of alkali lignin (AL) was optimized by a response
surface method (RSM), and the effects of different isolation meth-
ods on the lignin structures were studied by scanning electron
microscopy (SEM) and nuclear magnetic resonance (NMR) tech-
niques. Concretely, the particle sizes and surface properties of lignin
preparations were demonstrated by SEM. The lignin structural
information, including the S/G ratio and types of major linkages
(inter-coupling bonds, -O-4 , -, -5 , etc.), was obtained from
the quantitative 13 C NMR and 2D-HSQC NMR spectroscopy.
2. Materials and methods
2.1. Materials
Q. variabilis Bl. was obtained from Yangling, Shanxi, China.
The valonea were dried in an oven at 50 C and then ground to
obtain particles with a size distribution between 450 and 900 m
(2040 mesh). The extracted material (toluene/ethanol, 2:1, v/v;
Soxhlet extractor; 12 h) was again dried at 60 C for 16 h and ball-
milled in a planetary ball mill for 5 h under N2 (room temperature),
followed by storage at 4 C in a refrigerator.
2.2. Isolation and purication of lignin
The lignin samples (MWL, AL and CEL) were extracted from the
valonea of Q. variabilis Bl. according to the method mentioned in a
previous study [9]. Additionally, RSM with a Box-Behnken design
was employed to design the experiments and investigate the effects
of the variables on the lignin isolation rate. A total of 29 experi-
mental runs with three replicates at the center points were carried
out for each method. The polynomial equation used for the four
variables is given below and designated as Eq. (1):
  
Y = 0 + i Xi + ij Xi Xj + ii Xi2
(Y is the response value; 0 is a constant; i is the linear effect; ii
is the squared effect; ij is the interaction effect; and Xi and Xj are
independent variables).
RSM experiment design and analysis were conducted using the
Design Expert software program (Version 8). The software was also
used to establish the model equation, graph the 3-D plot, graph the
2-D contour of the response and predict the optimum values for
the four response variables.
165
2.3. Characterization of lignin
SEM images of three lignin samples (AL, MWL and CEL) were
taken in the FESEM S-4800. The lignin samples were dissolved in
1,4-dioxane and then formed a lm on the 18 18 glass slide. The
surfaces of the lms were sputtered with conductive glue and then
observed and photographed [10].
The NMR spectra of three lignin samples (AL, MWL and CEL)
were obtained on a Bruker AVIII 500 MHz spectrometer at 25 C
in DMSO-d6 . A 50-mg lignin sample was dissolved in 0.5 mL of
DMSO-d6 (99.8%) for the 1 H NMR, and 140 mg of lignin sample was
dissolved in 0.5 mL of DMSO-d6 for the quantitative 13 C NMR exper-
iment, with 128 scans taken for the 1 H NMR and 30000 scans taken
for the 13 C NMR. The chemical shift was determined by the solvent
peak of DMSO (H = 2.49 ppm; C = 39.5 ppm) as an internal refer-
ence. For the quantitative 2D-HSQC spectra, a 25-mg lignin sample
was dissolved in 0.5 mL of DMSO-d6 . The widths were set at 5000
and 20000 Hz for the 1 H NMR and 13 C NMR dimensions, respec-
tively. The central solvent (DMSO) peak was used as an internal
reference [6].
3. Results and discussion
3.1. RSM analysis for alkali isolation
The effects of isolation factors, such as X1 (liquid/solid ratio),
X2 (concentration), X3 (time), and X4 (temperature), on the isola-
tion rate for the alkali isolation method were carefully studied. The
isolation rate is shown in Table 1 and could be expressed by the
following second order polynomial equation:
Y = 22.01 + 0.39X1 + 3.67X2 + 0.37X3 + 1.57X4 + 0.13X1 X2
+0.012X1 X3 0.014X1 X4 + 1.00X2 X3 + 0.43X2 X4
+0.046X3 X4 2.04X12 8.33X22 1.49X32 3.82X24 (2)
where Y is the isolation rate, and X1 , X2 , X3 and X4 are the coded val-
ues for the liquid/solid ratio, concentration, time, and temperature,
respectively.
The signicance of each coefcient measured using a p-value
and a F-value is listed in Table 2. The p-value of the model was
less than 0.0001, which indicated that the model was signicant
and could be used to optimize the isolated variables. The two inde-
pendent variables (X2 , X4 ) and three quadratic terms (X1 2 , X2 2 and
X4 2 ) signicantly affected the isolation rate, and the quadratic term
X3 2 was signicant (p < 0.01). X1 , X3 , X1 X2 , X1 X3 , X1 X4 , X2 X3 , X2 X4
and X3 X4 had less of an effect on the isolation rate due to a higher
p-value (p > 0.05). Meanwhile, concentration was the most signi-
cant factor affecting the isolation rate, followed by the temperature,
liquid/solid ratio and time.
Fig. 1 shows the response surfaces for the effects of the indepen-
dent variables on the isolation rate for the alkali isolation method.
It was found that all six response surfaces were convex in shape,
which indicated that the ranges of variables were chosen prop-
erly. In the case of the isolation rate, all the liquid/solid ratios
(1) (X1 ), concentrations (X2 ), times (X3 ) and temperatures (X4 ) used
had quadratic effects on the isolation rate. When the isolation
parameter was kept at one level, the isolation rate increased when
increasing the isolation parameter within a certain range and then
decreased with extending it.
The suitability of the model equation for predicting the optimum
response values was tested using the selected optimum conditions.
The optimal conditions were a liquid/solid ratio (X1 ) of 12.21, a
concentration (X2 ) of 0.85 mol/L, a time (X3 ) of 4.21 h, and a tem-
perature (X4 ) of 42.21 C, under which the predicted isolation rate
was 22.67%. Considering the operability in actual production, the
166 L. Yang et al. / International Journal of Biological Macromolecules 97 (2017) 164172

Table 1
Total isolation rate of lignin under different conditions for alkali lignin isolation.

Run Liquid/Solid (X1 ) Isolation Concentration (mol/L) (X2 ) Isolation Time (h) (X3 ) Isolation Temperature ( C) (X4 ) Isolation Rate (%)
(Y)

1 10 0.6 4 40 7.3002
2 14 0.6 4 40 7.6385
3 10 1 4 40 13.9697
4 14 1 4 40 14.8327
5 12 0.8 3 30 14.3241
6 12 0.8 5 30 14.3605
7 12 0.8 3 50 17.5334
8 12 0.8 5 50 17.7529
9 10 0.8 4 30 14.7671
10 14 0.8 4 30 15.3733
11 10 0.8 4 50 17.2954
12 14 0.8 4 50 17.8442
13 12 0.6 3 40 8.2174
14 12 1 3 40 14.4912
15 12 0.6 5 40 8.2331
16 12 1 5 40 18.5224
17 10 0.8 3 40 18.4291
18 14 0.8 3 40 19.5413
19 10 0.8 5 40 18.4437
20 14 0.8 5 40 19.6049
21 12 0.6 4 30 5.60841
22 12 1 4 30 11.5307
23 12 0.6 4 50 8.3625
24 12 1 4 50 16.0116
25 12 0.8 4 40 21.48
26 12 0.8 4 40 22.03
27 12 0.8 4 40 21.47
28 12 0.8 4 40 22.19
29 12 0.8 4 40 22.86

All results are the means SD (n = 3).

Table 2
ANOVA for response surface quadratic model for alkali lignin isolation.

Source Sum of Squares df Mean Square F Value Prob > F

Model 684.71 14 48.91 54.92 <0.0001

X1 1.79 1 1.79 2.01 0.1786
X2 161.32 1 161.32 181.15 <0.0001
X3 1.6 1 1.6 1.8 0.2015
X4 29.57 1 29.57 33.2 <0.0001
X1 X2 0.069 1 0.069 0.077 0.7851
X1 X3 6.00E-04 1.00E+00 6.00E-04 6.74E-04 0.9797
X1 X4 8.24E-04 1.00E+00 8.24E-04 9.25E-04 0.9762
X2 X3 4.03 1 4.03 4.53 0.0516
X2 X4 0.75 1 0.75 0.84 0.3757
X3 X4 8.38E-03 1.00E+00 8.38E-03 9.41E-03 0.9241
X1 2 26.97 1 26.97 30.29 <0.0001
X2 2 450.02 1 450.02 505.34 <0.0001
X3 2 14.35 1 14.35 16.12 0.0013
X4 2 94.84 1 94.84 106.49 <0.0001
Residual 12.47 14 0.89
Lack of Fit 11.14 10 1.11 3.36 0.1273
Pure Error 1.33 4 0.33
Cor. Total 697.18 28

R2 = 0.9821; Adj. R2 = 0.9642.

All results are the means SD (n = 3).

optimum conditions were modied as follows: liquid/solid ratio
(X1 ) of 12.2, concentration (X2 ) of 0.85 mol/L, time (X3 ) of 4.2 h,
and temperature (X4 ) of 42.2 C, under which the experimental
isolation rate was 22.70% (n = 3), which agreed closely with the pre-
dicted isolation rate, consequently indicating that the RSM model
was satisfactory and accurate.
3.2. Characterization of lignin
3.2.1. SEM
The results of SEM are shown in Fig. 2. The isolation conditions
of lignin samples had signicant effects on the microstructures of
the lignin samples. It can be observed that the shape of the MWL
was irregular polygonal with a smooth surface. After alkaline treat-
ment, the surface gradually became smoother, which may be due
to the strongly basic isolation conditions, which polished the sur-
face more sleekly. Additionally, the extreme isolation conditions
resulted in some breakage of the surface. The lignin particles also
presented a better dispersion under alkaline treatment. Compared
to MWL, the surface of CEL became rough gradually, and many pores
were formed due to the release of volatiles [11]. It was speculated
that the delignication and hemicellulose removal destroyed the
matrix composed of cellulose, hemicellulose, and lignin [12]. The
surface of CEL had distinct sunken areas, which strongly proved
the effects of cellulase and hemicellulase in the hydrolysis process.
 L. Yang et al. / International Journal of Biological Macromolecules 97 (2017) 164172 167

Fig. 1. Response surface plots showing the effect of variables on the isolation rate for alkali isolation method.

Fig. 2. SEM images of lignin preparations obtained from Q. variabilis Bl.

Therefore, the AL and CEL isolation processes caused some damage 3.2.2. 1 H NMR analysis
to the apparent lignin structure. 1 H NMR is probably the fastest and cheapest NMR technique

to investigate the chemical structure of lignin [13]. The 1 H NMR

168 L. Yang et al. / International Journal of Biological Macromolecules 97 (2017) 164172

1
Fig. 3. H NMR spectra of AL, MWL and CEL obtained from Q. variabilis Bl.

spectra of AL, MWL and CEL are shown in Fig. 3. The hydrogen sig-
nal integrations were subdivided into different structural regions
[14]. In the 1 H NMR spectra, the signals between 6.0-8.0 ppm were
attributed to the aromatic nucleus of G and S units. Concretely, the
signals at 6.7 ppm and 6.6 ppm stand for the G and S units, respec-
tively. The weak signal at 5.3 ppm is attributed to the H- of -5
structures. The signal at 4.9 ppm originated from the H- of -O-4
structures. The signals between 4.6-4.7 ppm and 4.0-4.5 ppm stand
for H- and H-, respectively. The signal at 2.50 ppm arose from
DMSO-d6 . The signals at 2.0, 2.1, and 2.2 ppm are assigned for the H
atoms of methyl or carbonyl groups. Signals between 0.8-1.5 ppm
arose from the H atoms of aliphatic functional groups. For example,
the signal at 1.2 ppm corresponded to the methylene groups of the
aliphatic chain.
3.2.3. 13 C NMR analysis
Further information on the chemical structures of extracted
lignin samples was obtained by 13 C NMR. The 13 C NMR spectrums
of AL, MWL and CEL are shown in Fig. 4. It was observed that
the spectra of the lignin preparations were very similar except
for several peaks. The absence of signals between 90 and 102 ppm
indicated low concentrations of residual sugars in these lignin sam-
ples [15]. The portion between 102 and 160 ppm in the 13 C NMR
spectra represents the aromatic moiety of lignin, and it is set as
the reference. Signal assignment for 13 C NMR of the three lignin
samples is shown in Table 3. The signals at 174.43, 174.61 and
174.76 ppm in AL, MWL and CEL were assigned to the aliphatic
carboxyl groups. However, the signal assigned for the aliphatic car-
boxyl groups was less pronounced in the AL spectrum than in the
 L. Yang et al. / International Journal of Biological Macromolecules 97 (2017) 164172 169

13
Fig. 4. C NMR spectra of AL, MWL and CEL obtained from Q. variabilis Bl.

MWL and CEL spectra, suggesting that both the aliphatic esters
and acetyl groups in AL were mostly saponied during the alka-
line treatment [16]. The three lignin preparations contained mostly
non-etheried S units, which was a symbol of a low degree of con-
densation [16]. The S units were detected by some signals in the
spectra of AL, MWL and CEL including the following: 152.63, 152.71,
and 152.72 ppm (C-3/C-5, S etheried); and 148.70/148.54/148.65,
148.65, and 148.64 ppm (C-3/C-5, S nonetheried). Meanwhile, the
G units were detected by several signals in the AL, MWL and CEL
samples as follows: 147.52, 147.54, and 147.53 ppm (C-4 etheri-
ed); 144.98, 145.12, and 145.14 ppm (C-4 none etheried); 111.46,
112.07, and 112.05 ppm (C-2); and 119.73, 119.71, and 119.63 ppm
(C-6). The P-hydroxyphenyl group (H) was indicated by small C-
2/C-6 signals at 128.23, 127.25, and 128.22 ppm for AL, MWL and
CEL, respectively, conrming that the lignin samples can be referred
to as GSH lignin, which was consistent with the results of the 1 H
NMR of the nitrobenzene oxidation products of lignin in a previous
study [9].
In addition, there was some evidence of the existence of p-
coumaric esters (p-CA) in AL, MWL and CEL (159.62, 159.62,
and 159.62 ppm (C-4); 130.10, 130.11, and 130.10 ppm (C-2/C-6)).
The existence of -O-4 substructures was proven by the fol-
lowing signals: 86.79, 86.79, and 85.56 ppm (C-); 72.81, 72.19,
and 72.45 ppm (C-); and 60.11, 60.42, and 59.65 ppm (C-).
170 L. Yang et al. / International Journal of Biological Macromolecules 97 (2017) 164172

Table 3 CEL, such as -O-4 arylether (A), resinol (, B) and phenyl-

Signal assignment for 13 C NMR spectra of AL, MWL and CEL.
coumaran (-5, C), were identied by the cross peaks in the
Assignments (ppm) 2D-HSQC spectra (Fig. 5). The detailed assignments of the typi-
cal substructures could be found in recent publications [13,17,18].
AL MWL CEL
The detailed assignments are listed in Table 4. Herein, the differ-
COOH 174.43 174.61 174.76
ences in the spectra of AL, MWL and CEL are primarily discussed in
C-4, p-CA 159.62 159.62 159.62
C-3/C-5, S etheried 152.63 152.71 152.72 this study. For example, the phenylcoumaran lignin units (C) were
C-3/C-5, S nonetheried 148.70/148.65/148.54 148.65 148.64 observed at C /H 85.79/5.34 (C ), 54.40/3.48 (C ) and 63.18/3.19
C-4, G etheried 147.52 147.54 147.53 (C ) in MWL as well as C /H 86.56/5.31 (C ), 54.25/3.44 (C ) and
C-4, G nonetheried 144.98 145.12 145.14 63.32/3.70 (C ) in CEL, but the corresponding signals disappeared
C-2/C-6, p-CA 130.10 130.11 130.10
C-2/C-6, H units 128.23 128.25 127.22
in AL at the current contour level, suggesting that most phenyl-
C-6, G units 119.73 119.71 119.63 coumaran groups have been cleaved after the alkaline extraction.
C-5, G units 115.23 115.88 115.79 The aromatic region in the HSQC spectra (C /H 95160/5.5-
C-2, G units 111.46 112.07 112.05 8.5 ppm) of AL, MWL, and CEL revealed a large number of
C-, -O-4 86.79 85.79 86.56
cross-signals from G, S, and H lignin units. This region allowed the
C-, resinol substructure 85.93 85.79 85.62
C-, -O-4 72.81 72.19 72.45 information about the S/G ratio of the lignin preparations to be
C-, resinol substructure N. D. 71.65 70.04 estimated and gave the oxidation of side chains of the S and G units
C-, -O-4 60.11 59.42 60.65 [13]. The S units exhibited the signals for the C2,6 H2,6 correlation
Methoxyls 56.45 56.44 56.46 at 104.88/6.71, 104.85/6.79, and 104.94/6.80 ppm in AL, MWL and
C-, resinol substructure N. D. 54.40 54.25
CEL spectra, respectively. Meanwhile, the G units were represented
by several lower intensity correlation signals. The C2 H2 correla-
tion was represented by the signals at 111.46/6.94, 112.07/6.90,
and 112.05/6.91 ppm in AL, MWL and CEL spectra, respectively.
Additionally, we found a trace of the resinol substructure in AL,
The C6 H6 linkages were proven by the signals at 119.73/6.71,
MWL and CEL, which can be proven by the following signals: 85.93,
119.71/6.79, and 119.63/6.73 ppm for AL, MWL and CEL, respec-
85.79, and 85.62 ppm (C-); 70.65 and 71.04 ppm for MWL and CEL
tively.
(C-); and 54.40 and 54.25 ppm (C-).
In addition to the above-mentioned signals, the existence
Meanwhile, xylan exhibited some absorption in the three spec-
of conifery alcohol was proven by the signals at 129.71/5.33,
tra, including the signals at 102.85, 102.88 and 102.28 ppm (C-3,
129.60/5.34, and 130.17/5.36 ppm in AL, MWL and CEL, respec-
xylan); 74.91, 74.02, and 74.67 ppm (C-1, xylan); 72.81, 72.22,
tively, which could be attributed to the C2,6 H2,6 in conifery
and 73.99 ppm (C-1, xylan); and 62.57, 63.18, and 63.32 ppm (C-
alcohol. Furthermore, the C H in p-coumarlylated substruc-
5, xylan), which proved that xylan was exhibited in AL, MWL and
tures was certied by the signals at 115.24/6.71, 115.38/6.79,
CEL, respectively. This conrmed that xylan may be the main carbo-
and 114.45/6.80 ppm in AL, MWL and CEL, respectively. The cor-
hydrate contamination in CEL. This phenomenon proved that the
relations of C5 H5 in ferulic acid were proven by the cross
process of hydrolysis was incomplete. As reported in a previous
peaks at 110.89/7.51, 110.83/7.50, and 110.86/7.51 ppm in AL,
study [17], this fact conrmed that xylan was covalently bonded
MWL, and CEL, respectively. The xylan exhibited strong signals
to lignin. Additionally, the trace of methoxyls can be proven by
at 102.28/4.30 ppm (C-1), 73.11/3.09 ppm (C-2), 74.57/3.27 ppm
the signals at 56.45, 56.44, and 56.46 ppm in AL, MWL and CEL,
(C-3), 75.86/3.55 ppm (C-4), and 63.41/3.25 ppm (C-5) in the CEL
respectively.
spectrum, which proved that xylan was the main carbohydrate
contamination in CEL. This was consistent with the results of the
3.2.4. 2D-HSQC analysis 13 C NMR and those found in the literature [19]. The reported
To obtain a further comprehensive structural characterization study [17] suggested that xylose acted as a protector against
of the extracted lignin samples, including the substructures (inter- cooking in the pulping and paper industry. C H correlations in p-
coupling bonds) and S/G ratio, all the lignin preparations were hydroxycinnamyl alcohol end groups appeared at 67.57/3.95 ppm,
subjected to 2D-HSQC analysis. The substructures in AL, MWL and

Table 4
Assignments of 13 C-1 H cross signals in the HSQC spectra of AL, MWL and CEL.

Labels Chemical shift (13 C/1 H,ppm) Assignment

AL MWL CEL

MeO 56.45/3.74 56.44/3.75 56.46/3.77 C-H in methoxyls

A 72.81/4.82 72.19/4.79 72.45/4.83 C -H in -O-4 substructures (A)
A 86.79/4.25 85.79/4.21 86.56/4.29 C -H in -O-4 substructures (A)
A 60.11/3.40 59.42/3.48 60.65/3.47 C -H in -O-4 substructures (A)
B 85.93/4.63 85.79/4.62 85.62/4.67 C -H in resinol substructures (B)
B 85.79/3.06 54.40/3.10 54.25/3.10 C -H in resinol substructures (B)
B 85.62/3.74 71.65/3.78 70.04/3.84 C -H in resinol substructures (B)
C N. D 85.79/5.34 86.56/5.31 C -H in phenylcoumaran substructures (C)
C N. D 54.40/3.48 54.25/3.44 C -H in phenylcoumaran substructures (C)
C N. D 63.18/3.79 63.32/3.70 C -H in phenylcoumaran substructures (C)
FA7 110.89/7.51 110.83/7.50 110.86/7.51 C7 -H7 in ferulate (FA)
S2,6 104.88/6.71 104.85/6.79 104.94/6.80 C2,6 -H2,6 in S units (S)
G2 111.46/6.94 112.07/6.90 112.05/6.91 C2 -H2 in G units (G)
G6 119.73/6.71 119.71/6.79 119.63/6.73 C6 -H6 in G units (G)
F 67.57/3.95 66.85/3.93 66.86/3.97 C -H in p-hydroxycinnamyl alcohol end groups (F)
C2,6 129.71/5.33 129.60/5.34 130.17/5.36 C2,6 -H2,6 in conifery alcohol
PCE3,5 115.24/6.71 115.38/6.79 114.45/6.80 C3,5 -H3,5 in p-coumarate ester (PCE)
a
Not detected.
 L. Yang et al. / International Journal of Biological Macromolecules 97 (2017) 164172
Fig. 5. 2D-HSQC spectra of the aliphatic-oxygenated (C/H 5090/2.56.0) and the aromatic (C/H 90150/6.08.0) regions for AL, MWL and CEL obtained from Q. variabilis
Bl.
Table 5
Structural characteristics (relative abundance of main inter-unit linkages) from the
integration of 13 C1 H correlation signals in the HSQC spectra of AL, MWL and CEL.
The linkage structures Relative abundance (%)
AL MWL
-O-4 91.57 83.23
- 8.43 12.86
5 N. Da 3.91
S/G ratio 8.72b 1.30
a
Not detected.
b
S/G ratio obtained by the this equation: S/G = IS2,6 /2IG2 .
66.85/3.93 ppm and 66.86/3.97 ppm for AL, MWL and CEL, respec-
tively. This structure has been reported to probably correspond to
C4 -etheried sinapyl alcohol [20], and only the C H correlation
was observed due to its higher sensitivity in the HSQC spectrum
[21,22]. In summary, the 2D-HSQC spectra and the identied sub-
structures are both plotted and depicted in Fig. 6
There were several semi-quantitative parameters that were cal-
culated by means of spectral integration in order to compare the AL,
MWL and CEL based on the documented calculation methodologies
[7]. Thus, the relative contents of the main inter-unit linkages and
terminal structures as well as the S/G ratio were calculated from
the HSQC spectra (Table 5). It was found that the relative content of
-O-4 linkages in CEL (75.91%) was lower than those in AL (91.57%)
and MWL (83.23%), suggesting that the -O-4 linkage was cleaved
greatly during the CEL isolation process [16]. Meanwhile, the rela-
tive contents of - were 8.43%, 12.86%, and 12.48% for AL, MWL
and CEL, respectively. The relative contents of -5 were 3.91% and
11.61% for MWL and CEL, respectively. The S/G ratio of the lignin
samples was considered to be an important parameter in the delig-
nication process of pulping [23]. The S/G ratios were estimated to
171
be 1.30 and 0.98 for MWL and CEL, respectively, while the ratio was
8.72 for AL. Clearly, AL had a higher S/G ratio than MWL and CEL.
This may be caused by fractionation during different handing pro-
cesses [6]. This potentially suggested that the lignin fragment rich
CEL in S-units could be easily released under the alkali conditions. The
environment was so erce that it produced more active points in
75.91
12.48 the G units, which resulted in the G units reacting with other struc-
11.61 tures more easily. Thus, the G units were reduced, and the ratio of
0.98 S/G was increased relatively. On the other hand, a high S/G ratio
always led to a high content of -O-4 linkages in lignin macro-
molecules [19]. This is consistent with the results of the S/G ratio
value.
4. Conclusion
In the present study, RSM was used to optimize the alkali lignin
(AL) isolation process from the valonea of Q. variabilis Bl., and SEM
and NMR techniques were used to characterize the AL, MWL and
CEL. From the results, it was found that the optimum conditions
were a liquid/solid ratio of 12.21, an isolation concentration of
0.85 mol/L, an isolation time of 4.21 h, and an isolation tempera-
ture of 42.21 C, under which the isolation rate was 22.67%. There
was some damage to the apparent structures of AL and CEL during
the isolation process. The three lignin samples were all GSH types,
and the relative contents of main -O-4 linkages in the three lignin
samples were 91.57%, 83.23% and 75.91%, respectively. The relative
contents of - in AL, MWL and CEL were 8.43%, 12.86% and 12.48%,
respectively. In addition, the existence of phenylcoumarane, ferulic
acid, p-coumarates and p-hydroxycinnamyl alcohol end groups can
be found. The S/G ratios calculated from 2D-HSQC were 8.27, 1.30
and 0.98 in AL, MWL and CEL, respectively, which provide some
theoretical basis for further studies.
172 L. Yang et al. / International Journal of Biological Macromolecules 97 (2017) 164172
Fig. 6. Main classical structures in the lignin preparations (A) -O-4 linkages; (B) resinol substructures; (C) phenylcoumarin structures; (F) p-hydroxycinnamyl alcohol end
groups; (FA) ferulate; (PCE) p-coumarcoumarate ester; (G) guaiaicyl; (S) syringyl unit.
Acknowledgment
This work was supported by the program from the Special
Fund for Forestry Scientic Research in the Public Interest of China
(201504320).
References
[1] T.K. Kirk, R.L. Farrell, Ann. Rev. Microbiol. 41 (1987) 465505.
[2] H. Kim, J. Ralph, T. Akiyama, BioEnergy Res. 1 (2008) 5666.
[3] R.B. Santos, E.A. Capanema, M.Y. Balakshin, H.-M. Chang, H. Jameel,
BioResources 6 (2011) 36233637.
[4] E.A. Capanema, M.Y. Balakshin, J.F. Kadla, J. Agric. Food Chem. 53 (2005)
96399649.
[5] .T. Martnez, J. Rencoret, G. Marques, A. Gutirrez, D. Ibarra, J.
Jimnez-Barbero, J.C.d. Rio, Phytochemistry 69 (2008) 28312843.
[6] J.L. Wen, B.L. Xue, F. Xu, R.C. Sun, BioEnergy Res. 5 (2012) 886903.
[7] L. Zhang, G. Gellerstedt, Magn. Reson. Chem. 45 (2007) 3745.
[8] S.H. Ghaffar, M. Fan, Biomass Bioenergy 57 (2013) 264279.
[9] L.N. Yang, D.M. Wang, D. Zhou, Y.W. Zhang, J. Biol. Macromol. 85 (2016)
417424.
[10] H. Deka, A. Mohanty, M. Misra, Macromol. Mater. Eng. 299 (2014) 552559.
[11] F. Cotana, G. Cavalaglio, A. Nicolini, M. Gelosia, V. Coccia, A. Petrozzi, L.
Brinchi, Energy Procedia 45 (2014) 5260.
[12] M.F. Li, Y.M. Fan, F. Xu, R.C. Sun, X.L. Zhang, Ind. Crop. Prod. 32 (2010) 551559.
[13] C. Fernndez-Costas, S. Gouveia, M.A. Sanromn, D. Moldes, Biomass
Bioenergy 63 (2014) 156166.
[14] A. Guerra, R. Mendonca, A. Ferraz, F. Lu, J. Ralph, Appl. Environ. Microbiol. 70
(2004) 40734078.
[15] R. El Hage, N. Brosse, L. Chrusciel, C. Sanchez, P. Sannigrahi, A. Ragauskas,
Polym. Degrad. Stab. 94 (2009) 16321638.
[16] J.L. Wen, B.L. Xue, F. Xu, R.C. Sun, A. Pinkert, Ind. Crop. Prod. 42 (2013)
332343.
[17] A.P. Duarte, D. Robert, D. Lachenal, Holzforschung 55 (2001) 645651.
[18] K. Wang, S. Bauer, R.C. Sun, J. Agric. Food Chem. 60 (2012) 144152.
[19] Z.J. Shi, L.P. Xiao, J. Deng, R.C. Sun, BioEnergy Res. 6 (2013) 12121222.
[20] D. Ibarra, M.I. Chvez, J. Rencoret, Jor Carlos del Ro, A. Gutirrez, J. Romero,
S. Camarero, M.J. Martnez, J. Jimnez-Barbero, A.T. Martnez, J. Agric. Food
Chem. 55 (2007) 34773490.
[21] Y. Liu, S. Wei, M. Liao, Ind. Crop. Prod. 49 (2013) 837843.
[22] M. Yu, Balakshin, E.A. Capanema, C.-L. Chen, H.S. Gracz, J. Agric. Food Chem.,
51 (2003) 61166127.
[23] H.Z. Li, Z.J. Zhang, T.Y. Hou, X.J. Li, T. Chen, Ind. Crop. Prod. 76 (2015) 1824.