Ultrasonics Sonochemistry 9 (2002) 85±93

www.elsevier.com/locate/ultsonch

Comparative study of lignins isolated by alkali and

ultrasound-assisted alkali extractions from wheat straw
RunCang Sun a,b,*, Jeremy Tomkinson b
a
State Key Laboratory of Pulp and Paper Engineering, College of Paper and Environment Engineering, South China University of Technology,
Guangzhou 510640, People's Republic of China
b
The Biocomposites Centre, University of Wales, Bangor Gwynedd LL57 2UW, UK
Received 19 March 2001; accepted 21 May 2001

Abstract
Treatment of the dewaxed wheat straw with 0.5 M KOH at 35°C for 2.5 h without ultrasonic irradiation and with ultrasound
assistance for 5, 10, 15, 20, 25, 30, and 35 min resulted in a dissolution of 43.9%, 43.9%, 43.9%, 44.4%, 44.4%, 45.6%, 46.8%, and
49.1% of the original lignin, respectively. Much better results were achieved under the ultrasonic irradiation time for 35 min where
nearly 50% of the original lignin was solubilized at 35°C for 2.5 h. The purity of the lignin preparations isolated by alkali with
ultrasound assistance was higher than that of the lignin fraction obtained by alkali without ultrasonic irradiation, and their purity
increased with an increment of irradiation time between 5 and 35 min, in which the content of associated polysaccharides in the
former lignin preparations (0.87±1.06%) was lower than that of the latter lignin fraction (1.16%). In addition, the lignins isolated by
alkali with ultrasonic irradiation time between 5 and 30 min showed a slightly higher molecular weight and thermal stability than the
lignin obtained by alkali without ultrasound assistance. No substantial di€erences in the main structure features between the lignin
preparations isolated by alkali and ultrasound-assisted alkali extractions were found. Ó 2002 Elsevier Science B.V. All rights reserved.

Keywords: Ultrasound; Irradiation time; Lignin; Phenolics; Wheat straw

1. Introduction Lignin in wood or straw can be described as a three-

dimensional macromolecule with high molecular weight
In the past decades, an increase in the world popu- in the range of 100 KD. It originates from phenyl-
lation and economy has resulted in a raising demand for propanoid precursors such as coumaryl, coniferyl, and
various raw materials for industries such as raw mate- sinapyl alcohol C6 ±C3 and is present in vascular plants
rials for pulp and papermaking. At an average popula- [4]. Also it acts as an adhensive that binds together the
tion annual rate of 3.2%, the global demand for pulp cellulosic ®bre structure and imparts sti€ness to the
and paper is expected to reach 420 million metric tons wood matrix [5]. However, unlike hardwood lignin,
by year 2010 [1]. To reach this demand, various re- straw lignin additionally contains small but signi®cant
sources of the pulp and paper need to be concerned. amounts of structural elements derived from p-coumaryl
Among these, agricultural residues such as cereal straws alcohol. Straw lignin also contains p-coumaric and
may represent a potential material for pulp and paper ferulic acids attached to the core lignin through ester
industry since they contain comparable amounts of the linkages [6]. In spite of technology's progress at the end
major constituents common to wood species. In parti- of last century, it has not been possible to determine
cular, wheat straw is extensively used as raw material for exactly the inter- and intra-molecular bonds involving
pulp and paper. More than 9 million tons of straw pulp lignin and other polymers in the cell wall. In addition, its
are produced annual in China, which account for about heterogeneity has been a struggle for all wood re-
90% of the world's total straw pulp [2,3]. searchers worldwide. Deligni®cation in pulping pro-
cesses consists of the degradation and dissolution of
*
Corresponding author. Tel.: +44-1248-370-588; fax: +44-1248-370-
lignin macromolecule into smaller fragments. In the case
594. of alkaline pulping, the important reactions include
E-mail address: bcs00a@bangor.ac.uk (R.C. Sun). the cleavage of phenolic a-O-4 linkages, cleavage of
1350-4177/02/$ - see front matter Ó 2002 Elsevier Science B.V. All rights reserved.
PII: S 1 3 5 0 - 4 1 7 7 ( 0 1 ) 0 0 1 0 6 - 7
86 R.C. Sun, J. Tomkinson / Ultrasonics Sonochemistry 9 (2002) 85±93

non-phenolic b-O-4 linkages, and removal of residual the straw is cellulose 38.8%, hemicelluloses 39.5%, lignin
lignin fractions, either by cleavage of carbon±carbon 17.1%, ash 1.8%, and wax 2.2% on a dry weight basis.
linkages or by carbohydrate degradation, releasing lig- After being dried at 60°C in an oven for 16 h, the straw
nin-carbohydrate fractions, which are mainly oxidised was ground to pass through a 0.7 mm screen and stored
into aliphatic carboxylic acids [7]. These degraded lig- at 5°C until use.
nins in principle can be used either as fuel or as feed-
stock for chemical conversions. 2.2. Ultrasound-assisted extraction and isolation of lignins
Our previous studies have shown that high yield of
lignin from herbaceous plants such as straw and grass A scheme for alkali and ultrasound-assisted alkali
can be obtained using mild alkaline treatments, even at extractions and isolation of lignins is shown in Fig. 1.
room temperature, in which over 60% of the lignin from The dried straw was ®rst extracted with toluene±ethanol
wheat straw was extracted with 1.5% NaOH at 40°C for (2:1, v/v) in a Soxhlet extractor for 6 h. To 9.78 g de-
6 h [8]. It was found that the considerable proportion of waxed sample in a 500 ml beaker, 300 ml 0.5 M KOH
free phenolic groups present in the guaiacyl units seems aqueous solution was added. The irradiation was carried
to play an important role in the solubility properties of using the Sonic system SOMERSET (England, 20 kHz)
Gramineae lignins [9]. Meanwhile, the participation of provided with a horn at sonic power of 100 W and so-
ester- or ether-linked cinnamic acids, particularly ferulic nication time for 0, 5, 10, 15, 20, 25, 30, and 35 min in
acid, as linked bridges between lignin and hemicelluloses 0.5 M KOH aqueous solution at 35°C, respectively. The
in the cell walls of herbaceous plants is considered to play mixture was then successively mixed at 35°C for a total
a role in the high extraction yields of alkaline lignin [10± period of 2.5 h. After ®ltration on a nylon cloth, the
12]. This high solubility of the lignins from Gramineae hemicelluloses were isolated from the hydrolysates by
in alkali can be considered as a suitable method for precipitation of the acidi®ed hydrolysate (pH 5.5 ad-
fractional isolation and structural study of the lig- justed with 6 M acetic acid) with three volumes of 95%
nins [13]. ethanol. The pellets rich in the hemicelluloses were ®l-
More recently, applications of ultrasound tech- tered, washed with 70% ethanol and air-dried. After
niques in food industry result in more attentions, to evaporation of ethanol, the alkali soluble lignins were
depolymerise macromolecules, make emulsions, disrupt obtained by precipitation at pH 1.5 adjusted by 6 M
biological cells, and de¯occulate droplets [14,15]. Soni- HCl from the corresponding supernatants. The isolated
cation has been reported to improve pectin technology acid-insoluble lignin preparations were washed with acidi-
from apple pressings [16] and isolation of pharmaceu- ®ed water (pH 1.5±2.0) and freeze-dried. The residues
tically active compounds from Salvia ocinalis [17], and rich in cellulose were washed with water and ethanol,
increase of the yield of xylans from corn hulls [18] and
corn cobs [19]. In the case of polysaccharide depoly-
merisation, studies by ultrasound have been preliminary
investigated on various polysaccharides [14,20,21]. It has
been found that the extraction of organic compounds
contained within the body of plants and seeds by a
solvent is signi®cantly improved by the use of power
ultrasound. The mechanical e€ects of ultrasound pro-
vide a greater penetration of solvent into cellular ma-
terials and improves mass transfer [15]. However, the
use of ultrasound for direct extraction of lignin polymer
from straw or wood has not been reported. The aim of
this study was to compare the yield, composition,
physico-chemical properties, and structural features of
lignins extractable from wheat straw, isolated with 0.5
M KOH aqueous solution by conventional and ultra-
sound-assisted alkali extractions.

2. Experimental

2.1. Materials

Wheat straw (Variety Riband) was kindly supplied by Fig. 1. Scheme for isolation of acid-insoluble lignins from wheat straw
B Lloyd Co., Llangefni. The composition (%, w/w) of at di€erent ultrasonic irradiation times.
 R.C. Sun, J. Tomkinson / Ultrasonics Sonochemistry 9 (2002) 85±93 87

then dried at 60°C for 16 h. All experiments were per- 3. Results and discussion
formed at least in duplicate. Yield of lignin fractions is
given on a dry weight basis related to the wheat straw. 3.1. Yield, purity, and molecular properties

In this study, eight alkaline lignin preparations were

2.3. Physico-chemical and thermal analysis of the isolated isolated with 0.5 M KOH at 35°C for 2.5 h with and
acid-insoluble lignins without application of ultrasound and their yields are
summarized in Table 1. As expected, treatment of the
Both qualitative identi®cation and quantitative de- dewaxed wheat straw with 0.5 M KOH without appli-
termination of phenolic compounds released by alkaline cation of ultrasound at 35°C for 2.5 h and with ultra-
nitrobenzene oxidation of the non-condensed mono- sonic irradiation for 5, 10, 15, 20, 25, 30, and 35 min
meric units of the acid-insoluble lignin preparations [7] resulted in dissolution of 43.9%, 43.9%, 43.9%, 44.4%,
were carried out on a Hichrom H5ODS HPLC column 44.4%, 45.6%, 46.8%, and 49.1% of the original lignin
of dimensions 250 mm  4.6 mm. All nitrobenzene oxi- and 62.8%, 63.0%, 62.8%, 62.8%, 63.5%, 63.5%, 64.3%,
dation results represent the mean of at least triplicate and 64.5% of the original hemicelluloses (data not
samples and each oxidation mixture was chromato- shown in Table 1), respectively. Obviously, between the
graphed twice. The carbohydrate moieties associated irradiation time 15 and 35 min the release of the lignin
with acid-insoluble lignin preparations were determined components from the cell walls was increased from 7.6%
by hydrolysis of the lignin samples with 2 M tri¯uoro- to 8.4%. Approximately half of the total lignin in wheat
acetic acid for 2 h at 120°C. Liberated neutral sugars straw was released during the alkali extraction at soni-
were analyzed as their alditol acetate derivatives by gas cation time of 35 min. Thus, application of sonication
chromatography (GC) [22]. Methods for recording UV for 35 min resulted in raising solubility of lignin by 0.9%
spectra and determination of molecular-average weights in comparison to the alkaline extraction procedure
of the acid-insoluble lignin fractions are described in without ultrasound assistance. However, the di€erences
previous papers [23,24]. between ultrasound-assisted extractions for 5±10 min
FT-IR spectra were obtained on an FT-IR spectro- and the alkali treatment without ultrasonic irradiation
photometer (Nicolet 750) using a KBr disc containing were found to be negligible. This implied that ultrasonic
1% ®nely ground lignin samples. The 13 C-NMR spec- irradiation time was a main parameter a€ecting the
trum was obtained on a Bruker 250 AC spectrometer lignin yields under the conditions used. The higher e-
operating in the FT mode at 62.4 MHz under total ciency of the ultrasound-assisted extractions can be ex-
proton decoupled conditions. It was recorded at 25°C plained by the mechanical action of ultrasound on the
from 180 mg of sample dissolved in 1.0 ml DMSO-d6 cell walls resulting in an increased accessibility and ex-
after 20,000 scans. A 60° pulse ¯ipping angle, a 3.9 ls tractibility of the lignin component. Such e€ects of so-
pulse width and a 0.85 s acquisition time were used. nication were known to improve the isolation of
Thermal analysis of acid-insoluble lignin prepara- extractives from various plant materials [15] and xylans
tions was performed using thermogravimetric analysis from corn cob [19]. More importantly, the yield of acid-
(TGA) and di€erential scanning calorimetry (DSC) on insoluble lignin preparation representing the major
a simultaneous thermal analyzer (STA 625). The appa- fraction (Table 1), comprised 65.3±72.6% of the total
ratus was continually ¯ushed with nitrogen. The sam- solubilized lignins, while the yield of the lignin frac-
ple weighed between 8 and 12 mg. Each sample was tion associated in the solubilized hemicelluloses, ac-
heated from room temperature to 600°C at a rate of counted only for 13.1±19.0% of the total released lignin.
10°C min 1 . The last was particularly lower when the extraction was

Table 1
The yield of lignin fractions (% dry matter) obtained by alkaline and ultrasonic-assisted alkaline extractions of wheat straw with 0.5 M KOH at 35°C
for 2.5 h under di€erent ultrasonic times
Lignin fractions Ultrasonic time (min)
0 5 10 15 20 25 30 35
Total solubilized lignins 7.5 7.5 7.5 7.6 7.6 7.8 8.0 8.4
Acid-insoluble ligninsa 4.9 5.0 5.1 5.1 5.3 5.2 5.8 6.1
Acid-soluble ligninsb 1.0 0.9 1.0 1.1 0.9 1.2 1.1 1.2
Lignin associated in isolated 1.6 1.4 1.4 1.4 1.4 1.4 1.1 1.1
hemicelluloses
a
Represent the lignin fractions obtained by precipitation of the supernatant solution at pH 1.5 after isolation of the solubilized hemicelluloses.
b
Represent the lignin fractions which are still solubilized in the pH 1.5 supernatant after precipitation of the acid-insoluble lignin fractions and
obtained by di€erence (total solubilized lignin acid-insoluble lignin lignin associated in the solubilized hemicelluloses).
88 R.C. Sun, J. Tomkinson / Ultrasonics Sonochemistry 9 (2002) 85±93

performed with ultrasound assistance for 30 and 35 min. Table 2

These results indicated that alkaline extractions with The content of neutral sugars (% of lignin sample, w/w) in isolated
acid-insoluble lignin preparations obtained at di€erent ultrasonic times
ultrasonic irradiation under the conditions used had a from wheat straw
more e€ect on the cleavage of the ether bonds between
Neutral Ultrasonic time (min)
lignin and hemicelluloses from the cell walls of wheat sugars 0 5 10 15 20 25 30 35
straw than the alkaline treatment without ultrasound
assistance. The current observations were consistent Arabinose 0.25 0.22 0.22 0.22 0.20 0.18 0.16 0.16
Xylose 0.53 0.50 0.49 0.46 0.47 0.47 0.48 0.45
with the studies of the cleavage of interunitary bonds in
Glucose 0.23 0.21 0.21 0.21 0.22 0.20 0.18 0.17
lignin induced by ultrasonic irradiation of milled wood Galactose 0.13 0.13 0.12 0.11 0.12 0.10 0.11 0.09
lignin by Yoshioka et al. [25]. The authors stated that
Total 1.14 1.06 1.04 1.00 1.03 0.95 0.93 0.87
the major ether linkages, i.e., b-O-4 bonds between lignin
interunitary linkages and a-O-4 ether linkages between
lignin and hemicelluloses can be homolytically rup-
the lignin fraction obtained by alkali without ultrasonic
tured or cleaved to some extent by the ultrasonic irra-
irradiation. Obviously, an increase in sonication time
diation.
from 5 to 35 min resulted in a decrease in the level of
The isolated acid-insoluble lignin fractions were
associated hemicelluloses from 1.06% to 0.87% in the
characterized by UV spectroscopy at k 250±390 nm.
acid-insoluble lignin fractions. These data suggested that
Fig. 2 illustrates the four lignin preparations isolated by
treatment of the straw with alkali under the ultrasound-
0.5 M KOH at 35°C for 2.5 h without ultrasonic irra-
assisted conditions improved the cleavage of a-ether
diation (spectrum a) and with ultrasound assistance for
bonds between lignin and hemicelluloses from the cell
5 (spectrum b), 10 (spectrum c), and 15 min (spectrum
walls of wheat straw. Xylose (0.45±0.53%) was identi®ed
d). Evidently, the four lignin fractions showed the basic
as the major sugar component. Arabinose (0.16±0.25%),
UV spectrum typical of lignins with two maxima at 280
glucose (0.17±0.23%), and galactose (0.09±0.13%) were
and 320 nm. The ®rst absorption maximum corresponds
present in noticeable amounts.
to nonconjugated phenolic groups in the lignin [11] and
the second one is assigned to bound hydroxycinnamic
acid, such as p-coumaric and ferulic acids. A lowest 3.2. Lignin composition
absorption coecient of the lignin preparation, isolated
by 0.5 M KOH at 35°C for 2.5 h without ultrasonic To verify the lignin composition, the eight acid-
irradiation (spectrum a), was undoubtedly due to the insoluble lignin fractions were oxidized by alkaline ni-
more co-precipitated non-lignin materials such as car- trobenzene at 170°C for 2.5 h. In this case, the three
bohydrate degradation products, ash and salts. constitutive monomeric lignin units, i.e. p-hydroxyphe-
Comparison of the sugar content between the eight nyl, guaiacyl, and syringyl, produce the corresponding
acid-insoluble lignin fractions given in Table 2 indicated p-hydroxybenzaldehyde, vanillin, and syringaldehyde,
that all the lignin preparations obtained with ultrasound respectively. Table 3 summarizes the yield of phenolic
assistance contained lower amounts of bound hemicel- acids and aldehydes and the molar proportions of S
luloses as shown by 0.87±1.06% in comparison to that of (relative total moles of syringaldehyde and syringic

Fig. 2. UV spectra of acid-insoluble lignin preparations isolated with 0.5 M KOH at 55°C for 2.5 h without ultrasonic irradiation (spectrum a) and
with sonication time for 5 min (spectrum b), 10 min (spectrum c), and 15 min (spectrum d).
 R.C. Sun, J. Tomkinson / Ultrasonics Sonochemistry 9 (2002) 85±93 89

Table 3
The content (% of lignin sample, w/w) of phenolic acids and aldehydes from nitrobenzene oxidation of the acid-insoluble lignin preparations obtained
at di€erent ultrasonic times from wheat straw
Phenolic acids and aldehydes Ultrasonic time (min)
0 5 10 15 20 25 30 35
p-Hydroxybenzoic acid 1.84 1.84 1.64 1.32 1.20 0.96 0.88 0.52
p-Hydroxybenzaldehyde 2.92 2.85 2.48 1.88 1.44 1.56 1.54 1.44
Vanillic acid 1.87 1.38 1.34 0.76 0.63 0.58 0.58 0.40
Syringic acid 2.53 2.95 2.30 2.42 2.05 1.84 1.86 1.84
Vanillin 20.32 19.56 20.30 18.26 15.90 15.29 15.78 15.52
Syringaldehyde 18.04 18.01 18.20 15.00 10.65 10.44 10.37 10.30
p-Coumaric acid 1.08 1.00 0.92 0.73 0.78 0.74 0.50 0.48
Ferulic acid 1.67 1.56 1.48 1.39 1.48 1.44 0.98 0.96
Total 50.27 49.15 48.66 41.76 34.13 32.85 32.49 31.46
a
Molar ratio (S:V:H) 3:4:1 3:4:1 4:5:1 4:5:1 3:5:1 3:5:1 3:5:1 3:5:1
a
S represents the sum of total moles of syringaldehyde and syringic acid; V represents the sum of total moles of vanillin and vanillic acid; and H
represents the sum of total moles of p-hydroxybenzaldehyde and p-hydroxybenzoic acid.

acid):V (relative total moles of vanillin and vanillic densation [26]. It is higher in the lignin fraction obtained
acid):H (relative total moles of p-hydroxybenzaldehyde by alkali without ultrasonic irradiation than in the lignin
and p-hydroxybenzoic acid) from nitrobenzene oxida- preparations isolated by alkali with ultrasound assis-
tion of the acid-insoluble lignin fractions. As can be tance, and it decreased substantially from 49.15% to
seen, the predominant oxidation products were vanillin 31.46% with an increase of ultrasonic irradiation time
and syringaldehyde, which together represented 76.3%, from 5 to 35 min. This result demonstrated that ultra-
76.4%, 79.1%, 79.6%, 77.8%, 78.3%, 80.5%, and 82.1% sound-assisted extractions under the alkaline conditions
of the total phenolic compounds in the lignin fractions used resulted in signi®cant condensation of the solubi-
isolated by alkali without ultrasonic irradiation and with lized lignin molecules.
ultrasound assistance for 5, 10, 15, 20, 25, 30, and 35 The values of the weight-average (M w ) and number-
min, respectively. The yield of vanillin was higher than average (M n ) molecular weights and the polydispersity
that of the syringaldehyde in all the lignin preparations. (M w =M n ) of the eight acid-insoluble lignin fractions are
The relatively higher amounts of vanillin was particu- given in Table 4. It should be noted that the data shown
larly signi®cant in the lignin fractions obtained by alkali in this table can be used only in a comparative study
with ultrasonic irradiation time between 20 and 35 min since the calibration was carried out using polystyrene
as shown by 3:5:1 molar ratios of S:V:H. Extension of standards. As shown in Table 4, the six lignin prepara-
ultrasonic irradiation time between 20 and 35 min en- tions isolated by alkali with ultrasound assistance for 5±
hanced the release of guaiacyl lignin. The high ratio of 30 min showed a slightly higher molecular-average
V/S units in the lignin fractions revealed that these lig- weights, ranging M w from 3010 to 3570 g mol 1 than
nins were released mainly from the primary cell wall that of the lignin fraction obtained by alkali without
since a larger amount of guaiacyl lignin is formed in ultrasonic irradiation (M w ˆ 2890 g mol 1 ). An increase
the early stage of xylem di€erentiation than in the in sonication time from 5 to 20 min led to an increment
later stages, where the lignin is rich in syringyl units. of M w from 3010 to 3570 g mol 1 , indicating an in-
The presence of fewer p-hydroxybenzaldehyde and p- creasing in solubilization of large molecular size lignins
hydroxybenzoic acid was considered most probably to under the ultrasound conditions used. The reason for
be indicative of non-condensed p-hydroxyphenyl units, this increase in M w is probably due to the condensation
indicating the incorporation of p-hydroxycinnamoyl al- reactions between the lignin structures under ultrasound
cohol in the wheat straw alkali lignins. In addition,
except of partial conversion of p-coumaric acid to p-
hydroxybenzaldehyde and ferulic acid to vanillin during Table 4
the nitrobenzene oxidation process, the remaining Weight-average (M w ) and number-average (M n ) molecular weights and
occurrence of noticeable amounts of ferulic acid (0.96± polydispersity (M w =M n ) of the isolated acid-insoluble lignin prepara-
tions obtained at di€erent ultrasonic times from wheat straw
1.67%) and small quantities of p-coumaric acid (0.48±
1.08%) in all the lignin fractions indicated that these two Ultrasonic time (min)
phenolic acids are closely linked with lignins in the cell 0 5 10 15 20 25 30 35
walls of wheat straw, in accord with our previous study Mw 2890 3010 3130 3470 3570 3300 3070 2760
[18]. Furthermore, the yield of nitrobenzene oxidation is Mn 1490 1540 1550 1620 1620 1340 1150 1040
M w =M n 1.94 1.95 2.02 2.14 2.20 2.46 2.66 2.65
considered to be related to the degree of lignin con-
90 R.C. Sun, J. Tomkinson / Ultrasonics Sonochemistry 9 (2002) 85±93

irradiation conditions given. In contrast, as the irradi- skeleton vibrations in the lignin preparations are as-
ation time was further increased from 20 to 25, to 30, signed at 1600, 1510, and 1424 cm 1 . Absorption at 1464
and to 35 min, the M w decreased from 3570 to 3300, to cm 1 relates to C±H deformations and aromatic ring
3070, and to 2760 g mol 1 , respectively. This decrease vibrations. The aliphatic C±H stretching in CH3 gives a
in M w is assigned to the cleavage of the b-O-4 linkages small band at 1362 cm 1 . The bands at 1331, 1265, and
between the lignin precursors under a relatively longer 1230 cm 1 are indicative of ring breathing with C±O
sonication period. Similar phenomenon was observed by stretching. The 1126 and 1032 cm 1 bands arise from
Yoshioka et al. [25] in the study of homolytic scission of the aromatic CH in-plain deformation for syringyl type
interunitary bond in wood lignins induced by ultrasonic and guaiacyl type, respectively. Aromatic C±H out of
irradiation. The authors stated that the alkyl phenyl bending appears at 845 cm 1 . The spectral pro®les and
ether bonds, CH±O±phenyl, known as interunitary the relative intensities of the absorption bands were
bonds in lignins were homolytically cleaved by the ul- rather similar, indicating that application of ultrasound
trasonic irradiation. On the other hand, the polydis- under the alkaline extraction conditions given did not
persity of the molecular distribution of seven lignin a€ect the structure of alkali lignin from wheat straw.
preparations isolated by alkali under ultrasonic irradi- To further investigate the structural changes of the
ation was higher than that of the lignin fraction ob- acid-insoluble lignin individually by ultrasonic irradia-
tained by alkali without ultrasound assistance, and it tion, the lignin fraction isolated with 0.5 M KOH under
increased with increasing the irradiation time from 5 to sonication time for 35 min was investigated by 13 C-
30 min. This implied that an increase in ultrasonic ir- NMR spectroscopy (Fig. 4). Most of the observed sig-
radiation time between 5 and 30 min under the alkaline nals have been previously assigned in wood and straw
condition used also led to lignin fractions solubilized lignin spectra [7,11,28,29]. As shown in Fig. 4, the lignin
having a more broad molecular weight distribution. fraction is the almost absence of typical polysaccharide
signals between 57 and 103 ppm. The spectrum does
3.3. Spectroscopic characterization show two signals at 79.6 and 62.5 ppm, which are at-
tributed to C-a etheri®ed to carbohydrates and C-5 in
FT-IR spectra of some the acid-insoluble lignin Xyl internal unit, respectively. However, the peak in-
preparations are illustrated in Fig. 3. All the lignins tensities are rather weak, indicating a trace amount
showed the spectral features of GSH type lignins, of associated carbohydrates in the lignin fraction. This
namely, the bands at 1126 and 845 cm 1 and a shoulder result corresponded to the lower carbohydrate content.
at 1158 cm 1 [27]. The band at 1702 cm 1 corresponds Similar observation of C-a etheri®ed to polysaccharides
to the unconjugated ketone and carboxyl group in lignin side chain has been reported by Xie and co-
stretching, while the band at 1657 cm 1 is attribute to workers [30] from the study of ginkgo lignin±carbohy-
conjugated carbonyl stretching in lignin. Aromatic drate complexes (LCC). The authors demonstrated that

Fig. 3. FT-IR spectra of wheat straw acid-insoluble lignin preparations obtained by treatment with 0.5 M KOH (35°C, 2.5 h) without ultrasonic-
assistance (spectrum 1) and with ultrasonic assistance for 5 min (spectrum 2), 15 min (spectrum 3), and 35 min (spectrum 4).
 R.C. Sun, J. Tomkinson / Ultrasonics Sonochemistry 9 (2002) 85±93 91

Fig. 4. 13 C-NMR spectrum of acid-insoluble lignin fraction isolated with 0.5 M KOH with ultrasonic irradiation time for 35 min from wheat straw.

three lignin±carbohydrate linkages (i.e., ether type, ester In Fig. 4 three resonances at 86.0, 72.1, and 60.1 ppm
type, ketal type) were found at the C-a position of the are assigned to C-b in b-O-4, C-a in b-O-4, and C-c in
side chain of phenylpropane units in ginkgo LCC, and b-O-4, respectively. These signals revealed that the treat-
no lignin±carbohydrate bond at the C-b or C-c position ment with 0.5 M KOH at 35°C under ultrasound as-
of the lignin side chain was observed in the 13 C-NMR sistance for 35 min did not attack the b-aryl ether
spectra of the 13 C-enriched LCC. structure to a signi®cant extent. A very strong signal at
The signals for aromatic part of the lignin units can 56.0 ppm is due to the OCH3 in syringyl and guaiacyl
be observed in the region between 104.3 ppm and 152.2 units. The signals for the c-methyl, a- and b-methylene
ppm. The syringyl residues were identi®ed by signals at groups in n-propyl side chains of the lignin fraction
152.2 (C-3/C-5 in syringyl units), 147.1 (C-3/C-5 in non- occurred in the spectrum between 14.0 and 33.8 ppm.
etheri®ed syringyl units), 138.2 (C-4 in etheri®ed syringyl
units, data not shown), 134.2 (C-1 in etheri®ed syringyl 3.4. Thermal analysis
units, data not shown), 133.0 (C-1 in non-etheri®ed sy-
ringyl units, data not shown), and 104.3 ppm (C-2/C-6 The thermal stability of the lignin preparations was
in syringyl units). The guaiacyl units give signals at 147.1 determined by TGA and DSC. Fig. 5 illustrates the
(C-4 in etheri®ed guaiacyl units), 145.4 (C-4 in non- thermograms of the lignin preparation obtained by 0.5
etheri®ed guaiacyl units), 134.2 (C-1 in etheri®ed gua- M KOH without ultrasound assistance (Fig. 5a) and the
iacyl units), 133.0 (C-1 in non-etheri®ed guaiacyl units), two lignin fractions isolated by alkali with sonication
119.0 (C-6 in guaiacyl units, data not shown), 114.8 ppm time for 10 (Fig. 5b) and 30 min (Fig. 5c). As can be seen
(C-5 in guaiacyl units). The signals at 159.8, (C-4 in from the Fig. 5, the decomposition temperature of the
esteri®ed p-coumaric acid), 130.2 ppm (C-2/C-6 in este- three lignin fractions in Fig. 5a±c started at 176°C,
ri®ed p-coumaric acid), 125.4 (C-1 in esteri®ed p- 208°C, and 205°C, respectively. Similarly, at 10% weight
coumaric acid), and 115.8 ppm (C-3/C-5 in esteri®ed loss the decomposition temperature of the three lignins
p-coumaric acid) represent the esteri®ed p-coumaric acid. appeared at 238°C in Fig. 5a, 254°C in Fig. 5b, and
Etheri®ed ferulic acid gives signals at 168.0 ppm (C-c in 248°C in Fig. 5c. This trend of the initial weight loss
etheri®ed ferulic acid), 144.3 (C-a in etheri®ed ferulic correlated well with the lignin molecular weights given
acid), and 122.5 ppm (C-6 in etheri®ed ferulic acid), in Table 4, and showed that the thermal stability of the
while the esteri®ed ferulic acid exhibits a signal at 122.9 lignins increased with the molecular weight even though
ppm (C-6 in esteri®ed ferulic acid). These observations they showed a similar maximum decomposition tem-
strongly supported our previous ®ndings that p-cou- perature ranged between 300°C and 500°C and gave
maric acid is linked to lignin by ester bonds, whereas similar thermal stability when the temperature raised to
ferulic acid is linked to lignin by both ether and ester 500°C. At this temperature the weight loss amounted to
bonds [24]. 50% for all the three lignin preparations. The current
92 R.C. Sun, J. Tomkinson / Ultrasonics Sonochemistry 9 (2002) 85±93

4. Conclusions

The above results indicated that ultrasonic irradiation

in alkaline medium promoted the extraction of lignins
from wheat straw, and their yield and purity increased
with sonication time from 5 to 35 min under the con-
ditions given. Except for the lignin preparation isolated
with 0.5 M KOH with ultrasonic irradiation time for 35
min, all other lignin fractions obtained by the alkali
under ultrasound-assisted extractions showed a higher
molecular weight than that of the lignin sample ex-
tracted with alkali without ultrasonic irradiation. The
content of associated carbohydrates in the lignin prepa-
rations obtained by alkali with ultrasonic irradiation
was lower than that of the lignin fraction isolated by
alkali without ultrasound assistance, and it decreased
with an increase in sonication time between 5 and 35
min. However, ultrasonication under the alkaline con-
ditions used did not cause signi®cant changes in lignin
composition and its structure. This is of major impor-
tance from industrial point of view and makes the ul-
trasound-assisted extraction process very advantageous.

Acknowledgements

We are grateful for the ®nancial support of this re-

search from the LINK program (Fractionation of wheat
straw for industrial utilizations) of the UK Ministry of
Agriculture, Fisheries, and Food.

References

[1] S. Agrupis, E. Maekawa, K. Suzuki, J. Wood Sci. 46 (2000) 222.

[2] J.E. Atchison, Tappi J. 79 (1996) 87.
[3] X.J. Pan, Y. Sano, Holzforschung 54 (2000) 61.
[4] H.D.S. Abreu, A.M.D. Nascimento, M.A. Maria, Wood Fibre
Sci. 31 (1999) 426.
[5] M.A. Gilarranz, F. Rodriguez, M. Oliet, Holzforschung 54 (2000)
373.
Fig. 5. TGA/DSC curves of acid-insoluble lignin fractions isolated [6] S.Y. Lin, C.W. Dence, Methods in Lignin Chemistry, Springer,
with 0.5 M KOH (35°C, 2.5 h) (a) without ultrasonic-assistance, (b) Berlin, 1992.
with ultrasonic assistance for 10 and (c) 30 min. [7] R.C. Sun, J.M. Fang, J. Tomkinson, Ind. Crops Prod. 12 (2000)
71.
[8] R.C. Sun, J.M. Lawther, W.B. Banks, Ind. Crops Prod. 4 (1995)
results were in good agreement with the thermal stability 127.
of lignins from wood samples, in which the thermal [9] C. Lapierre, D. Jouin, B. Monties, Phytochem. 28 (1989) 1401.
stability increased with the molecular weight [31,32]. In [10] T. Higuchi, I. Kawamura, Holzforschung 20 (1967) 16.
[11] A. Scalbert, B. Monties, E. Guitter, J.Y. Lallemand, Holzfors-
addition, the DSC curve of the lignin fraction, isolated chung 40 (1986) 119.
with 0.5 M KOH without ultrasound assistance (Fig. [12] K. Iiyama, T.B.T. Lam, B.A. Stone, Phytochem. 29 (1990) 733.
5a), showed a big exothermic peak between 200°C and [13] M.C. Terron, M.L. Fidalgo, G. Almendros, A.E. Gonzalez,
600°C, whereas in the curves of the two lignin fractions, Rapid Commun. Mass Spectrom. 10 (1996) 413.
isolated by alkali with sonication time for 10 (Fig. 5b) [14] T.J. Mason, J.P. Lorimer, Sonochemistry: Theory, Applications
and Uses of Ultrasound in Chemistry, Ellis Horwood, Chichester,
and 30 min (Fig. 5c), this exothermic peak shifted to UK, 1988.
a higher temperature of 400±600°C, due to exothermic [15] T.J. Mason, L. Paniwnyk, J.P. Lorimer, Ultrason. Sonochem. 3
reactions of the lignin [33]. (1996) S253.
 R.C. Sun, J. Tomkinson / Ultrasonics Sonochemistry 9 (2002) 85±93 93

[16] I.N. Panchev, N.A. Kirtchev, C.G. Kratchanov, Food Hydrocoll. [25] A. Yoshioka, T. Seino, M. Tabata, M. Takai, Holzforschung 54
8 (1994) 9. (2000) 357.
[17] M. Salisova, S. Toma, T.J. Mason, Ultrason. Sonochem. 4 (1997) 131. [26] T. Lam, K. Iiyama, B. Stone, J. Sci. Food Agri. 51 (1990) 493.
[18] A. Ebringerova, Z. Hromadkova, Ultrason. Sonochem. 4 (1997) 305. [27] O. Fax, Holzforschung 45 (1991) 21.
[19] Z. Hromadkova, J. Kovacikova, A. Ebringerova, Ind. Crops [28] H.H. Nimz, D. Robert, O. Faix, M. Nemr, Holzforschung 35
Prod. 9 (1999) 101. (1981) 16.
[20] M. Aliyu, M.J. Hepher, Ultrason. Sonochem. 7 (2000) 265. [29] C. Lapierre, B. Monties, E. Guittet, J.Y. Lallemand, Holzfors-
[21] F. Kath, S. Lange, W.M. Kulicke, Die Angew. Makromol. Chem. chung 38 (1984) 333.
271 (1999) 28. [30] Y.M. Xie, S. Yasuda, H. Wu, H.B. Liu, J. Wood Sci. 46 (2000)
[22] A.B. Blakeney, P.J. Harris, R.J. Henry, B.A. Stone, Carbohydr. 130.
Res. 113 (1983) 291. [31] F.C. Becall, Wood Fibre 1 (1969) 215.
[23] R.C. Sun, T. Tomkinson, W. Zhu, S.Q. Wang, J. Agri. Food [32] H. Yoshida, R. Morck, K.P. Kringstad, Holzforschung 41 (1987)
Chem. 48 (2000) 1253. 171.
[24] R.C. Sun, B. Xiao, J.M. Lawther, J. Appl. Polym. Sci. 68 (1998) [33] C.A. Wotttz, C.E.L. Pasquali, H.A. Herrera, Latin Amer. Appl.
1633. Res. 31 (2001) 37.