pubs.acs.

org/Biomac Article

Extraction of High-Purity Lignins via Catalyst-free Organosolv

Pulping from Low-Input Crops
Jessica Rumpf, Xuan Tung Do, Rene ́ Burger, Yulia B. Monakhova, and Margit Schulze*
Cite This: Biomacromolecules 2020, 21, 1929−1942 Read Online

ACCESS Metrics & More Article Recommendations *

sı Supporting Information
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

ABSTRACT: A catalyst-free organosolv pulping process was applied

to cup plant (Silphium perfoliatum, S), Miscanthus grass (Miscanthus x
giganteus, M), and the Paulownia tree (Paulownia tomentosa, P),
resulting in high-purity lignins with no signals for cellulose,
Downloaded via UNIV OF EXETER on July 16, 2020 at 17:37:09 (UTC).

hemicellulose, or other impurities in two-dimensional heteronuclear

single quantum coherence (HSQC) nuclear magnetic resonance
(NMR) spectra. Different biomass particle sizes used for the
organosolv pulping (1.6−2.0 mm (1); 0.5−1.0 mm (2); <0.25 mm
(3)) influenced the molecular weight and chemical structure of the
isolated lignins. Principal component analysis (PCA) of 1H NMR data
revealed a high intergroup variance of Miscanthus and Paulownia
lignins, separating the small particle fraction from the larger ones.
Furthermore, monolignol ratios identified via HSQC NMR differ
significantly: Miscanthus lignins were composed of all three monolignols (guaiacyl (G), p-hydroxyphenyl (H), syringyl (S)), while for
Paulownia and Silphium lignins only G and S units were observed (except for P3).

1. INTRODUCTION perfoliatum) were recently listed as eligible ecological focus

In the context of the United Nations Sustainable Development areas (EFA).14 According to the European Common
Goals (SDGs), the Institute for Advanced Sustainability Agricultural Policy, farmers are requested to specify at least
Studies (IASS) specified the role of renewable resources in a 5% of their land as EFA. Originally, cup plants were cultivated
response to the current global economic and ecological in North America, whereas today they are bred worldwide.
challenges of the 20th century (climate change, limited Analogously to Miscanthus, Silphium crops require low
resources, rising energy demand).1 In addition, the European amounts of fertilizer and maintenance and show high yields
Commission released the European Bioeconomy Strategy in and fast growth.15 Thus, they are attractive candidates for
2018, including circular economy concepts and strategies to biogas and fuel production but can also as be used as
substitute fossil fuel with renewable resources for sustainable renewable source for particleboard production.16
energy, biofuel and chemical production.2,3 One promising This contribution focuses on lignins isolated from these low-
option to reach that goal is the utilization of lignocellulose input crops, namely Miscanthus, Silphium, and Paulownia. The
biomass, of which currently about 170 million metric tons are native three-dimensional macromolecular lignin structure is
annually produced.4 A variety of hard and soft wood species composed of three monolignols, namely p-coumaryl alcohol,
have been studied as potential biorefinery feedstock, among coniferyl alcohol, and sinapyl alcohol, that form the specific
them fast-growing trees (e.g., Bamboo, Poplar, Eucalyptus, Pine, residues p-hydroxyphenyl (H), guaiacyl (G), and syringyl (S)
and Paulownia).4,5 Paulownia trees show short growing cycles (Figure 1).
(silviculture rotations) of less than 15 years reaching heights These monolignols are randomly cross-linked via different
up to 20 m and diameters up to 40 cm and were originally linkages, mainly ether bonds (Figure 2) and carbon−carbon-
cultivated in tropical/subtropical regions, mainly China.6−8 In bonds (Figure 3) but also much more complex structures
addition, recent developments in biomass exploitation include where three or more monolignols are connected (Figure 4).
low-input crops, in particular, lignocellulose-rich grasses and
agroforestry residues.9,10 Miscanthus is such a low-input crop
and possesses a number of advantages; Miscanthus crops are Special Issue: Anselme Payen Award Special Issue
able to fix a higher amount of terrestrial carbon dioxide, show Received: January 26, 2020
high photosynthesis yields, require low amounts of fertilizer, Revised: March 17, 2020
and can be cultivated in arid conditions, compared to similar Published: March 18, 2020
species.11−13 Besides other perennial bioenergy crops, such as
short rotation coppice, Miscanthus and cup plants (Silphium

© 2020 American Chemical Society https://dx.doi.org/10.1021/acs.biomac.0c00123

1929 Biomacromolecules 2020, 21, 1929−1942
Biomacromolecules pubs.acs.org/Biomac Article

Figure 1. Lignin monolignol structures: p-coumaryl alcohol, coniferyl alcohol, and sinapyl alcohol forming the specific residues p-hydroxylphenyl
(H), guaiacyl (G), and syringyl (S). These units are randomly polymerized to form a complex three-dimensional macromolecule.

Figure 2. Ether linkages (from left to right: β-aryl-ether (β-O-4′), α-aryl-ether (α-O-4′), biphenyl ether (4-O-5′)); R = CH2OH, lignin.

Figure 3. Carbon−carbon linkages (from left to right: 1,2-diarylpropane (β-1′), biphenyl (5−5′), resinol (β−β′)); R = CH2OH, lignin.

Figure 4. More complex linkages (from left to right: dibenzodioxocin (α-O-4′/β-O-4′/5−5′), phenylcoumaran (α-O-4′/β-5′), spirodienone (β-1′/
β-O-4′)).

Because lignin biosynthesis is a complex process, monolignol Although the structure has been investigated for decades,
ratio and the corresponding number and nature of linkages many structural details remain unclear.19,20 The potential of
conventional univariate spectroscopic methods (e.g., FTIR,
strongly depend on the particular biosynthesis pathway.17 For
NMR) for structure elucidation and origin assignment of
example, the content of biphenyl fragments in wood-derived complex 3D molecules such as lignin is limited due to strong
lignins varies with the wood species.18 overlap of spectral bands even with 2D methods. Here,
1930 https://dx.doi.org/10.1021/acs.biomac.0c00123
Biomacromolecules 2020, 21, 1929−1942
Biomacromolecules pubs.acs.org/Biomac Article

chemometric methods became a versatile and efficient tool for the desired particle size (1.6−2.0 mm (1); 0.5−1.0 mm (2); <0.25
data analysis and interpretation, especially to classify natural mm (3)).
products such as LCF and lignin.21,22 Most often, FTIR and 2.3. Compositional Analysis of Biomass. The compositional
analysis (wt %) of all three biomasses (Miscanthus, Paulownia,
(HSQC) NMR spectroscopy are combined with multivariate
Silphium) was determined in quadruplicate according to Laboratory
data processing such as principle component analysis (PCA) to Analytical Procedures (LAP) published by the National Renewable
classify lignins according to their botanical origin and/or Energy Laboratory (NREL), following NREL/TP-510-42621 (Deter-
develop and explain structure−property-relationships (e.g., mination of Total Solids in Biomass and Total Dissolved Solids in
aliphatic and aromatic OH content, monolignol ratio, radical Liquid Process Samples),36 NREL/TP-510-42622 (Determination of
scavenging capacity)22 that are essential for the application of Ash in Biomass),37 and NREL/TP-510-42618 (Determination of
lignin as polyol or phenol substitute or antioxidant. Structural Carbohydrates and Lignin in Biomass).38
In previous studies of our group, a catalyst-free organosolv 2.4. Lignin Isolation. Organosolv pulping was performed for
process has been developed.23 In contrast to other extraction lignin isolation using the following biomass particle sizes: 0.5−1.0 mm
(M2; P2; S2); for Miscanthus and Paulownia, additional pulping was
processes using dioxan24 or ionic liquids,25 environmentally carried out with particle sizes of 1.6−2.0 mm (M1; P1) and <0.25
benign aqueous ethanol was used as solvent for lignin isolation mm (M3; P3). The biomass was mixed with an aqueous ethanol
from different biomasses (including M. sinensis, M. robustus, M. solution (80% ethanol) with a solid-to-liquid-ratio of 1:8 in a Parr
nagara, and different M. x giganteus samples, separated into leaf reactor with a 4848 Reactor controller and heated to 170 °C. After
and stem).26,27 Using chemometric processing of infrared data, holding this temperature for 90 min, the autoclave was cooled down,
stem versus leaf-derived lignins could be differentiated by their the biomass was filtered, and washed with aqueous ethanol solution
aromatic in-plane deformation signals at 1160 cm−1 corre- (80% ethanol). The organosolv lignin was then precipitated from the
sponding to the monolignol substitution pattern.26 Other filtrate by adding three volumes of acidified water and separated by
centrifugation. After washing the lignin with distilled water, samples
studies on Miscanthus genotypes specified the influence of were freeze-dried for 72 h. For each biomass and particle size, the
pulping conditions, process severity, or pretreatment con- pulping was carried out in triplicate to confirm the reproducibility of
ditions on lignin yield and structure.28−31 Comparing the process.
monolignol ratios (G/H/S) determined by pyrolysis gas Kraft lignin (KL) was precipitated from black liquor by slowly
chromatography/mass spectrometry (Py-GC/MS) and adding HCl (20 wt %) with a dropping funnel until pH 2 was reached.
HSQC NMR, deviations were obtained as investigated for a Afterward, the mixture was centrifuged and the solid residue was
number of different M. x giganteus samples.32 washed three times with acidified water (pH 2). Finally, the KL was
In contrast to a number of data reported for M. genotypes, freeze-dried for 72 h.
2.5. Chemometric Modeling. Matlab 2018b (The Math Works,
studies focusing on lignin extraction from Paulownia are very
Natick, MA, U.S.A.) and SAISIR package for MATLAB was used for
rare.7,33,34 Furthermore, data for Silphium crops are limited to statistical calculations.39 Data points from four spectral regions of the
plant composition16 and monolignol ratio (G, H, S) for the 1
H NMR (10.00−3.50 ppm, 3.24−2.65 ppm, 2.35−1.09 ppm, 1.03−
biomass (not the isolated lignins) obtained by OrganoCat 0.20 ppm) were used for multivariate modeling. Each data set was
pretreatment using Py-GC/MS.35 So far, no data is available preprocessed by bucketing with 0.01 ppm width. To normalize the
for monolignol linkages. intensities in different samples, buckets were scaled to the total
Here we present a comparative study of lignins isolated via intensity. PCA and factor discriminant analysis (FDA) were utilized as
organosolv pulping from Miscanthus x giganteus, Silphium chemometric approaches.
perfoliatum, and Paulownia tomentosa. Characterization was 2.6. Ultraviolet (UV) Spectroscopy. UV spectra were recorded
on a Hach Lange DR6000 from 200−400 nm using a solution of
performed using different spectroscopic methods (FTIR, UV,
lignin in sodium hydroxide (0.1 M) with a concentration of 50 μg
HSQC NMR) and size exclusion chromatography (SEC). mL−1.
Chemometric methods were used for exploratory analysis of 2.7. Fourier-Transform Infrared (FTIR) Spectroscopy. FTIR
acquired spectral information. spectra were recorded on a Jasco FT/IR 410 from 4000−700 cm−1
with a resolution of 4 cm−1 and 64 scans. The lignins were measured
2. EXPERIMENTAL SECTION as KBr disks containing 1% lignin.
2.8. Nuclear Magnetic Resonance (NMR) Spectroscopy.
2.1. Materials. Ethanol (absolute) was purchased from VWR Samples for NMR analysis were prepared by dissolving 50 mg of the
chemicals (Darmstadt, Germany); sodium hydroxide and potassium lignin in 900 μL DMSO-d6 + 0.03% (ν/ν) tetramethylsilane. NMR
bromide from Merck (Darmstadt, Germany); hydrochloric acid from measurements were performed on a Bruker Avance III 600 MHz
Carl Roth (Karlsruhe, Germany); DMSO-d6, degree of deuteration spectrometer (Bruker Biospin, Rheinstetten, Germany) with BBO
99.8%, with 0.03% (v/v) TMS from Eurisotop (Saarbrü cken, cryo probe equipped with a Bruker Automatic Sample Changer (B-
Germany); and Indulin AT from S3-chemical (Bad Oeynhausen, ACS 120) at 297 K. 1H NMR spectra were recorded with the
Germany). Black liquor obtained from kraft pulping (hardwood, standard pulse program (zg30 in Bruker language) using 16 scans
spruce/pine mixture) was obtained from Zellstoff- and Papierfabrik (NS) and 2 prior dummy scans (DS). The data of 132k points (TD)
Rosenthal GmbH (Blankenstein, Germany, MERCER group). For were acquired with a spectral width (SW) of 24.0155 ppm, and an
SEC calibration, polystyrenesulfonate sodium salt standards (891 Da acquisition time (AQ) of 4.5438 s. Receiver gain (RG) values were
to 151 kDa from Polymer Standards Service GmbH (Mainz, automatically adjusted. The 2D HSQC spectra were recorded with
Germany)) and benzenesulfonic acid sodium salt (180 Da from the following parameters: NS = 16, DS = 16, TD = 4096 (F2) and
Sigma-Aldrich (Steinheim, Germany)) were used. All chemicals were 256 (F1), RG = 2050, AQ = 0.2840 (F2) and 0.0044 (F1), SW =
used without further purification. 12.00 ppm (F2) and 190 ppm (F1). The transmitter offset was set to
2.2. Field Trials. The crops presented in this study are Miscanthus 5.0 for 1H and 80 ppm for 13C. The data were recorded automatically
x giganteus (M), Paulownia tomentosa (P) and Silphium perfoliatum under the control of ICON-NMR (Bruker Biospin, Rheinstetten,
(S); all crops were cultivated at the field lab of the Agricultural Germany). All NMR spectra were manually phased and baseline-
Faculty of the University of Bonn (Campus Klein-Altendorf, corrected using Topspin 4.0.7 (Bruker Biospin, Rheinstetten,
Rheinbach, Germany). Miscanthus samples were harvested in 2018, Germany).
Paulownia in 2016 and Silphium in 2018. All samples were milled and 2.9. Size Exclusion Chromatography (SEC). Samples for SEC
sieved using a model AS 200 basic from Retsch (Haan, Germany) to analysis were prepared in sodium hydroxide solution (0.1 M) with a

1931 https://dx.doi.org/10.1021/acs.biomac.0c00123
Biomacromolecules 2020, 21, 1929−1942
Biomacromolecules pubs.acs.org/Biomac Article

concentration of 5 mg mL−1 and filtered through 0.2 μm polyamide

(PA) filters prior to analysis. Weight-average (Mw) molecular weight
as well as polydispersity index (PDI) were determined on a PSS GPC
system based on Agilent 1100 Series (Mainz, Germany and
Waldbronn, Germany, respectively) with a PSS MCX 8 × 50 mm 5
μ precolumn and two PSS MCX 8 × 300 mm 5 μ 1000 Å columns
and refractive index detector (RID). Sodium hydroxide solution (0.1
M) as eluent at a flow rate of 1 mL min−1 and an injection volume of
10 μL were used. Calibration was performed with 11 molecular
weight standards (benzenesulfonic acid sodium salt and polystyr-
enesulfonate sodium salts) in a range from 180 Da to 151 kDa using a
fourth-degree polynomial function (R = 0.999).

3. RESULTS AND DISCUSSION

3.1. Compositional Analysis of Biomass. The chemical
composition of Miscanthus, Paulownia, and Silphium samples
has been determined according to the NREL protocol in
quadruplicates and is summarized in Table 1. Figure 5. Comparison of lignin yield from different biomasses
(Miscanthus (M), Paulownia (P), Silphium (S)) and particle size
Table 1. Chemical Composition of Miscanthus, Paulownia, fractions (1, 1.6−2.0 mm; 2, 0.5−1.0 mm; 3, <0.25 mm). Error bars
and Silphium, Determined According to NREL are obtained from standard deviation of the pulping process
Protocol36−38,a triplicates.

Miscanthus Paulownia Silphium

total lignin [%] 26.0 ± 0.5 32.8 ± 0.2 24.8 ± 0.2 For Miscanthus, no significant differences were observed
cellulose [%] 40.6 ± 1.1 36.9 ± 1.0 41.8 ± 0.3 depending on the particle sizes, as the yield varies between
hemicellulose [%] 24.2 ± 0.8 20.5 ± 0.5 22.4 ± 0.4 14.2−15.3%, which is in good accordance with literature
ash [%] 3.2 ± 0.4 1.3 ± 0.4 5.0 ± 0.1 results; Obama et al. extracted lignin from Miscanthus x
a
Errors are given as standard deviation from the fourfold giganteus using ethanol organosolv pulping at similar
determination of each sample. conditions (170 °C, 60 min, 1% sulfuric acid as catalyst) and
achieved 14.4%.44 Bauer et al. reported higher lignin yields for
For Miscanthus, the obtained data were in accordance with Miscanthus x giganteus, ranging from 29−32% when using a
literature results for lignin (21−26%) and ash contents (3− 75−95% aqueous ethanol solution.28 In contrast to our studies,
7%).26,29,40 The chemical composition, in particular lignin the biomass was extracted three times with ethanol prior to the
content, of Silphium resembles Miscanthus crops, confirming pulping process carried out under atmospheric pressure for 4 h
data reported by Damm et al., who conducted cell wall in the presence of hydrochloric acid as the catalyst, meaning
compositional analysis of both plants and found lignin contents major efforts and additional steps were required to achieve
(determined as acetyl bromide soluble lignin) between 18− these higher yields. Interestingly, the catalyst-free treatment
22%.35 For cup plant stalks, lignin contents of 21.4% were used in this study resulted in lignin yields comparable with
published.16 Even if these values are slightly lower than our catalyzed ethanol organosolv treatments and did not decrease
data (obtained according to NREL protocols), they are in significantly, as one could have expected.
good accordance with each other and deviations might be due For Silphium crops, the lignin yield (15.7%) is equal to that
to varying analysis methods. For Paulownia, a significantly of Miscanthus. Literature data for Silphium so far are limited to
higher lignin amount was found compared to the values stated lignin contents and chemical composition in whole plant cell
in literature: Gong et al. measured 23.8% total lignin in walls.35
Paulownia tomentosa,33 whereas Garciá et al. reported 22.1% Paulownia possesses the highest lignin content according to
lignin6 and Ye et al. 20.5% lignin.41 In contrast to our studies, the NREL data (see Table 1) and therefore seems to be
these values were not obtained following the NREL protocol, promising for lignin recovery. However, considering particle
which might be one reason for the deviation. Nevertheless, the sizes of 1.6−2.0 mm and 0.5−1.0 mm, Paulownia provides
measured values are generally higher than lignin contents in significantly lower yields compared to Miscanthus and Silphium.
other hardwood biomasses (18−24%).42 Another aspect to be Recently, Rashid et al. reported similar results when comparing
considered is the amount of extractives usually included in the lignin content and extraction efficiency of other biomass
biomasses. If a sample contains a high portion of extractives, species: a decrease in lignin extraction efficiency was observed
this might result in deviations of the lignin content, especially with increasing lignin content.45 Only for the smallest fraction
for the acid insoluble lignin, as observed for the Paulownia (P3), higher yields up to 18.4% were observed, making it the
samples. Nevertheless, woody biomass is expected to possess only biomass where the particle size makes a difference
higher amounts of lignin compared to grasses or herbaceous concerning the lignin yield in this study. In very general terms,
plants,43 which could be confirmed by our results. it was reported that a decreased particle size of biomass leads
3.2. Lignin Yield. The lignin yields given in percent to an increase of total accessible specific surface area and thus
referred to the total lignin content of respective biomass (see to a shorter mass transfer distance.45,46 Rencoret et al. obtained
Table 1) are shown in Figure 5. All pulping processes were milled wood lignin (MWL) from Paulownia fortunei (<0.15
conducted in triplicate resulting in comparable lignin yields mm) with a yield of approximately 20% of original Klason
with absolute errors between 0.3% (M3) and 1.7% (P3), which lignin content,7 which is in accordance with our results despite
is a first indicator for the reproducibility of the pulping process. the different genotype and recovery process.
1932 https://dx.doi.org/10.1021/acs.biomac.0c00123
Biomacromolecules 2020, 21, 1929−1942
Biomacromolecules pubs.acs.org/Biomac Article

3.3. Intergroup Variance of the Isolated Lignins. First,

to investigate possible differences between Miscanthus (M),
Paulownia (P), and Silphium (S) lignins, a PCA of 1H NMR
data was performed. The results of this PCA are shown in
Figure 6. It can be clearly seen that Miscanthus, Silphium, and

Figure 7. Loadings plot of the FDA.

For both F1 and F2, the signal at 5.99 ppm is strongly

negative. That means a strong signal at 5.99 ppm is typical for
the Paulownia lignin. This peak can actually only be found in
all Paulownia samples, neither in the other organosolv lignins
Figure 6. Principal component analysis of 1H NMR data of lignin nor in the reference Kraft lignins. Lundquist et al. have
from different biomasses (Miscanthus (M), Paulownia (P), Silphium assigned a similar-looking signal at 6.01 ppm to the Hα in β-O-
(S)) and particle size fractions (1, 1.6−2.0 mm; 2, 0.5−1.0 mm; 3, 4 and β-1 linkages of acetylated birch and spruce lignin,47,48 yet
<0.25 mm). no acetylation or other treatment of the Paulownia lignins in
comparison to the Miscanthus or Silphium lignins has been
Paulownia lignins differ from each other but also that pulping performed here. In HSQC NMR spectra, this signal can be
under the same conditions and with the same particle size leads found at around 100/6 ppm (δC/δH) but usually the
to comparable results. This confirms that next to the lignin assignment of Hα of the common β-O-4 linkage in organosolv
yield the structure of isolated lignins also can be reproduced lignins is at 71.6/4.9 ppm.28 Because of this contradiction,
with the applied catalyst-free ethanol organosolv process. Yet, a further investigations on the origin of this signal is required.
high intergroup variance from Miscanthus and Paulownia Signals leading to high F1 values, which are typical for
lignins is visible; the smaller fractions M3 and P3 are separated Silphium samples, were found at 2.01, 2.18, 2.73, and especially
from the larger core groups of M1/M2 and P1/P2, at 5.32 ppm. A signal region leading to high F2 values, which
respectively. Because only one size fraction of Silphium was are typical for Miscanthus samples, is that around 7.5 ppm,
used for lignin extraction, no separation could be investigated which corresponds to the aromatic protons of the H-type
here. In the Supporting Information, a PCA also including the monolignol (see Monolignol Ratio and Lignin Linkages),
KL and Indulin AT as reference is shown (Figure S1), and mainly present in Miscanthus lignins. Another signal leading to
both appear separate from the other clusters but next to each high F2 values and thus is typical for Miscanthus samples was
other. This was expected because Indulin AT is also a type of observed at 6.79 ppm, whereas the signal at 6.85 ppm
Kraft lignin. corresponds to a low F2 value, which was found in the
To investigate the spectral differences between the Paulownia samples. Both of those signals are in the region of
Miscanthus, Silphium, and Paulownia lignins, an FDA with aromatic protons of the G unit. These variations in monolignol
these three groups was performed. To check the discriminating ratio of the different biomasses were also found in the analysis
accuracy of the FDA approach, a model with a randomly of HSQC NMR and will be discussed there in further detail
chosen portion of 75% of the samples was created, using the (see Monolignol Ratio and Lignin Linkages).
3.4. Qualitative Structure Analysis and Functional
remaining 25% as an independent test set. It was shown that all
Groups. Lignin is a highly branched irregular phenolic
samples of the independent test set were correctly assigned. In
macromolecule made up of different amounts of the three
Figure 7, the loadings plots of the FDA are depicted. It shows monolignols with various functional groups (e.g., hydroxy,
how the variables are weighted to calculate the position in the methoxy, carboxyl, and carbonyl groups), depending on the
F1/F2 plane (see Supporting Information, Figure S2). A biomass as well as the pulping process. Thus, structure analysis
sample is assigned to one of the groups by the smallest distance is of great importance for the classification and assignment of
between the sample and the barycenters of the groups. different lignins according to their origin or other criteria (e.g.,
Knowing that Paulownia samples show both low F1 and F2 pulping process parameters, solvents used for isolation). UV
values and that Miscanthus samples show low F1 values with and FTIR spectroscopy were chosen as fast, simple, and easily
high F2 values and Silphium samples show high F1 values with performable techniques to get a brief overview of lignin
F2 around zero, spectral differences between the lignin groups structure including monolignols and functional groups. For
have been estimated using the loadings plot. future exploitation of lignin as an industrial raw material, fast
1933 https://dx.doi.org/10.1021/acs.biomac.0c00123
Biomacromolecules 2020, 21, 1929−1942
Biomacromolecules pubs.acs.org/Biomac Article

and reliable procedures are essential for quality control and cm−1, the CH in-plane deformation at 1123−1128 cm−1 and
product assurance.4,49 the CH out-of-plane deformation at 830−833 cm−1, which
The UV spectra of lignin typically show broad and indistinct were exclusively observed for the three organosolv-derived
absorption bands due to the overlap of numerous electronic lignins and not for KL and Indulin AT. Absorption bands
transition bands derived from the different light-absorbing indicating G units were observed for all lignins, namely the
units in lignin structure.50 Nevertheless, it provides a fast, CC stretch of the aromatic skeleton at 1509−1515 cm−1 and
cheap and simple technique to get a brief overview of the lignin the CO stretch of the aromatic skeleton at 1266−1268
macromolecule. As given in Figure 8, all lignins show a sharp cm−1. In addition, two characteristic peaks appeared in the
fingerprint region at 852−854 cm−1 and 815−823 cm−1 that
belong to lignins mainly composed of G units, such as KL and
Indulin AT. Grass lignins like Miscanthus are the only ones
with a considerable amount of H units, which was indicated by
the typical peak for HGS lignins at 1164 cm−1, exclusively for
the Miscanthus-derived lignins.
The signal at 1266−1268 cm−1 not only indicates the CO
stretching vibration of G units but could also be assigned to the
COC stretching in ethers. Additionally, the CO
deformation of ethers is located at 1080−1086 cm−1. Ether
linkages are the predominant interunit linkages in lignin
(approximately 50%),55 depending on biomass source and
isolation process. Comparing the FTIR spectra of different
lignins, both KL and Indulin AT show more pronounced
signals for ether linkages, suggesting a higher portion of these
bonds, which could also be verified by HSQC NMR (see
Monolignol Ratio and Lignin Linkages).
Furthermore, all lignins show a broad signal at 3411−3439
cm−1 indicating the OH stretch typical for lignins, as they
Figure 8. UV spectra of Miscanthus lignin (M2), Paulownia lignin possess many phenolic and aliphatic hydroxyl groups. Another
(P2), Silphium lignin (S2), Kraft lignin (KL), and Indulin AT; typical absorption band is the CH stretch in alkyl groups
measured in 0.1 M NaOH from 200−400 nm. (2923−2937 cm−1), whereby differences in the ratio of OH
and CH stretch could be specified; for Miscanthus and KL as
absorption band between 200−210 nm originating from well as Indulin AT, the OH stretch is predominant, for
nonconjugated phenolic groups (π−π*) and a second Paulownia both are approximately equal, whereas Silphium
maximum around 290−300 nm arising from conjugated possesses by far the highest absorbance for the CH stretch.
phenolic groups (n−π*), as well as a shoulder at 230 Moreover, carbonyl groups are also expected, giving rise to
nm.51,52 In general, the spectra of hardwood lignins (Paulownia signals at 1704−1710 cm−1 for the nonconjugated CO and
lignin, blue) consisting of G and S units exhibit bands at 1656−1661 cm−1 for the conjugated CO stretch. Both were
shorter wavelengths in contrast to softwood lignins (KL, observed to different extents in all spectra, depending on the
orange and Indulin AT, violet)53 which are mainly made up of chemical environment of these carbonyls.
G units. For grass lignins (Miscanthus lignin, red), an additional Overall, differences between the lignins from varying
band is observed around 380 nm due to the occurrence of H biomasses and pulping processes could be specified, even
units to a greater extent in this type of lignin. The Silphium with rather simple techniques such as UV and FTIR
lignin closely resembles the Paulownia sample, indicating a spectroscopy. According to the obtained results, the catalyst-
lignin composed of G as well as S units. free pulping process guarantees a reproducible and reliable
These differences revealed by UV spectroscopy regarding quality of the isolated lignins for all three biomasses
the monolignol composition of the lignins were confirmed by (Miscanthus, Paulownia, Silphie). However, structural differ-
FTIR results. The FTIR spectra of the three organosolv lignins, ences due to particle sizes could not be distinguished. For that
the technical KL obtained from pulp mill black liquor, and purpose, HSQC NMR as more a specific and sensitive tool was
Indulin AT as a commercially available Alkali lignin are shown used, as well as for the quantitative determination of the
in Figure 9. Among others, differences in FTIR signals can be differences in monolignol and interunit linkage ratio,
attributed to varying monolignol ratios of the studied crops respectively.
(hardwood, softwood, and grass lignin). The signals corre- 3.5. Monolignol Ratio and Lignin Linkages. The 2D
sponding to different monolignol units are shown as colored HSQC NMR is one of the most specific methods to determine
lines in the FTIR spectra, labeled according to the structures of monolignol ratios and assign bonding patterns, as the 13C−1H
same color shown in Figure 10. The assignment of character- correlation combines the sensitivity of 1H NMR with the
istic FTIR signals according to literature data23,26,29,50,54 is higher resolution of 13C NMR.56 This means, that each carbon
available in the Supporting Information in Table S1. atom (y-axis) is directly correlated with its attached protons (x-
KL and Indulin AT only show a slight shoulder at 2843− axis), allowing the elucidation of overlapping peaks in 1D
2850 cm−1, while the organosolv lignins have a sharp band spectra.57 According to literature, the spectra can be divided
arising from the higher amount of S units in grasses and into two regions, namely the aliphatic side chain region (δC/δH
hardwoods compared to softwoods, which is accompanied by a 50−90/2.8−5.8 ppm) and the aromatic core region (δC/δH
higher amount of methoxy groups. Other signals indicating S 100−150/6.0−8.0 ppm). Both regions are exemplarily shown
units are the CO stretch of the aromatic skeleton at 1329 in Figure 11 for M2, P2, and S2, respectively, and a detailed
1934 https://dx.doi.org/10.1021/acs.biomac.0c00123
Biomacromolecules 2020, 21, 1929−1942
Biomacromolecules pubs.acs.org/Biomac Article

Figure 9. FTIR spectra of Miscanthus lignin (M2), Paulownia lignin (P2), Silphium lignin (S2), Kraft lignin (KL), and Indulin AT from 4000−2500
cm−1 and 1850−700 cm−1. The most important signals showing differences between lignins are labeled with the corresponding functional group;
signals indicating different monolignols are labeled according to the colored structures in Figure 10.

assignment of lignin signals according to literature The prevalent interunit linkages present in all types of lignins
data7,26,28,29,58 can be found in the Supporting Information are β-aryl-ether (A), resinol (B), and phenylcoumaran (C).
(Table S2 and Table S3), as well as the HSQC spectra of M1, Unsaturated esters (D) were exclusively observed for the
M3, P1, P3, KL, and Indulin AT (Figure S3). The signals were Miscanthus-derived lignins at 113.5/6.3 ppm for Dβ and 144.5/
integrated using TopSpin 3.6.2 and the S/G ratio was 7.4 ppm for Dα, whereas spirodienone structures (E) only
determined based on the C2, 6−H2,6 correlations of S units appeared in Paulownia lignin at 60.6/2.8 ppm for Eβ and 84.8/
and the sum of C2−H2 and C6−H6 correlations of G units. For 4.8 ppm for Eα′. For Aβ there are two signals, depending on
the determination of interunit linkages, different approaches whether it is connected to a G (83.5/4.3 ppm) or S (85.7/4.1
exist; here we integrated all interunit linkages assuming that ppm) unit.28,29 Next to Aγ (59.4/3.4 ppm), a second signal
100% of them have been detected to calculate the relative appeared at 63.2/4.3 ppm in Miscanthus and Silphium lignin,
portion of each bonding type.59 corresponding to γ-acetylated or γ-p-coumaroylated (A′γ)
The biggest peak in all HSQC NMR spectra at δC/δH = substructures.29,60 Bauer et al. reported an additional signal
55.3/4.0 ppm belongs to the methoxy groups that are present for ethanol organosolv lignins from α-ethoxylated A sub-
in the G and S units and thus in the dominating monolignols. structures28 that could also be detected in all organosolv
1935 https://dx.doi.org/10.1021/acs.biomac.0c00123
Biomacromolecules 2020, 21, 1929−1942
Biomacromolecules pubs.acs.org/Biomac Article

Figure 10. Substructures identified in the different lignins: (G) guaiacyl unit, (S) syringyl unit, (S′) oxidized syringyl unit, (H) p-hydroxyphenyl
unit, (I) cinnamyl alcohol end group, (A) β-O-4 ether linkage, (B) resinol, (C) phenylcoumaran, (D) unsaturated ester, and (E) spirodienone.

lignins in this study at 63.8/3.3 ppm. In addition, next to the For Paulownia, only G and S units were observed in P1 and
S2,6 signal originally located at 103.5/6.7 ppm a second peak P2, whereas P3 also possesses small amounts of H units and S/
appears at 103.9/7.3 ppm due to α-oxidized S units (Cα- G ratios of 0.57 (P1) and 0.61 (P2). A difference between P1/
ketone).60 For the H units, only signals belonging to the H2,6 P2 and P3 could also be observed in the PCA of the 1H NMR
atoms were found at 127.0/7.2 ppm and 129.8/7.4 ppm, most (see Intergroup Variance of the Isolated Lignins). Literature
likely due to the overlap of H3,5 with signals from G5 (114.9/ data for Paulownia lignin are rare, but in accordance with our
6.8 ppm) and G6 (119.8/6.8 ppm), which display considerably results. Rencoret et al. investigated MWL from Paulownia
higher intensities.61 fortunei and found mainly S and G units and only a small
Figure 12 shows the monolignol ratio for each biomass and portion of H units, whereas the S/G ratio was determined to
particle size: Miscanthus lignin, independent of the particle size be 0.66.7 Gong et al. examined Paulownia elongata (S/G =
of the biomass, is composed of all three monolignols, 0.88) and Paulownia tomentosa (S/G = 0.98) and measured
confirming the results of UV and FTIR spectroscopy. The higher ratios, which might be due to the application of a
S/G ratio varies between 0.93 (M2) and 0.99 (M3). In modified nitrobenzene oxidation for the determination instead
literature, different S/G ratios can be found for Miscanthus of HSQC NMR.33
lignin, depending on the pulping process: Bauer et al. observed In conclusion, regarding the lignin content in the plant itself,
S/G ratios between 0.52−0.68 for ethanol organosolv pulping Silphium resembles Miscanthus, but regarding the structural
at atmospheric pressure using hydrochloric acid as the analysis this resemblance cannot be confirmed, as Miscanthus is
catalyst,28 while Wang et al. studied higher ratios of 1.1 for made up of G, S, and H units in approximately equal
basic ethanol organosolv pulping with sodium hydroxide as proportions, whereas Silphium possesses mainly S units, which
catalyst.29 In our studies, a catalyst-free treatment was used for is usually expected for hardwoods.42 Taking also UV and FTIR
pulping mainly to avoid corrosion problems and still the S/G spectroscopy into account, one can conclude that all methods
ratio is comparable to literature data. are in good accordance with each other. Miscanthus could be
Silphium has by far the highest S/G ratio with 3.25 and no H identified as HGS lignin, whereas for Silphium and Paulownia
units were detected. So far, there is no literature data available G and S units could be detected. KL and Indulin AT, both
for this kind of isolated lignin, but monolignol profiles in whole softwood lignins, are only composed of G units as expected,62
plant cell walls of Silphium and also Miscanthus after which could also be verified by all three spectroscopic methods
OrganoCat pretreatment have been determined using Py- (FTIR, UV, NMR). In addition, the pulping processes
GC/MS.35 The authors found that both plants possess a performed in triplicate revealed comparable results, as can be
comparable amount of G units that make up over 40%, while seen from the standard deviations in Table S4 and Table S5 in
Silphium has more S than G units and only minor amounts of the Supporting Information, thus conforming its reproduci-
H units in contrast to Miscanthus that is made up of more H bility.
than S units. However, a study by Groenewold et al. showed Next to the monolignol content, also the interunit linkages
that quantitative pyrolysis two-dimensional gas chromatog- were determined from the HSQC NMR spectra (Figure 13).
raphy mass spectrometry (Py/GCxGC/MS) of Miscanthus In all lignins, A is the dominating linkage with 63% (M1) to
biomass is useful to describe thermal processing but not 65% (M3) for Miscanthus, 57% (P1) to 61% (P3) for
appropriate to discover quantitative biomass composition and Paulownia and 84% for Silphium. For MWL, from Miscanthus x
monolignol ratios,32 which might be a reason for the giganteus even higher portions up to 93% of A are reported in
deviations. literature.63 MWL is used to study native lignins, as it
1936 https://dx.doi.org/10.1021/acs.biomac.0c00123
Biomacromolecules 2020, 21, 1929−1942
Biomacromolecules pubs.acs.org/Biomac Article

Figure 11. HSQC NMR spectra of Miscanthus lignin (M2), Paulownia lignin (P2), and Silphium lignin (S2) of (left) aliphatic side chain region
(δC/δH 50−90/2.8−5.8 ppm) and (right) aromatic core region (δC/δH 100−150/6.0−8.0 ppm). Colored peaks are assigned to the respective
structures of same color in Figure 10.

undergoes only minor structural changes during its extraction. for alkaline ethanol organosolv lignins from Miscanthus a
During the organosolv pulping used in this study, the ether relatively high amount of A (83%) is reported29 as well as 82−
linkages are partially cleaved, explaining the lower amount of A 84% for acidic ethanol organosolv lignins.28 Bergs et al. used an
found in organosolv-derived lignins. In addition, the higher ethanol organosolv process without any catalyst like we did in
process intensity of organosolv pulping leads to the enhanced our studies and found 55−65% A, which is in good accordance
cleavage of lignin-carbohydrate linkages, so that lower amounts with our results.26 For MWL, from Paulownia fortunei 62% of A
of those are detected in the spectra compared to MWL. Also, are reported,7 which seems to be rather low compared to
1937 https://dx.doi.org/10.1021/acs.biomac.0c00123
Biomacromolecules 2020, 21, 1929−1942
Biomacromolecules pubs.acs.org/Biomac Article

MWL from Miscanthus but is in good accordance with our

findings. Moreover, Paulownia has the highest content of
condensed substructures of the lignins investigated with a total
of 43% (P1) to 39% (P3). The biggest part of it is resinol (B)
which is mainly formed via dimerization reactions during lignin
synthesis and not during lignin modification during the pulping
process. Phenylcoumaran (C) on the other hand could be the
product of such condensation reactions, but as α-ethoxylation
also could be observed in all OSL neither intra- nor
intermolecular condensation reactions are very likely due to
the substitution of the α-carbon.28 Silphium possesses the
smallest fractions linked by carbon−carbon bonds, which
might be due to the high amount of S units with methoxyl-
substituted ortho-positions. These fragments cannot form 5-5
or β-5 carbon−carbon bonds and therefore possess a higher
amount of ether linkages.62
Spirodienone structures have already been reported in
Figure 12. Ratio of monolignols (G, S, H units, see Figure 10) in Miscanthus lignins64 but only in trace amounts, and there are
Miscanthus, Paulownia, and Silphium lignins derived from biomasses
with different particle sizes in comparison with KL and Indulin AT.
also several other studies that found no correlations belonging
to this subunit.25,26,58 Here, we only found this structure in
Paulownia lignin in small contents of 2% (P3) to 3% (P1, P2),
which is in accordance with literature data.7
In contrast to most studies found in literature,28,29 there are
no hints for cellulose, hemicellulose, or other impurities in the
HSQC spectra, which confirms the high purity of lignins
obtained via catalyst-free organosolv pulping. El Hage et al.
detected minor amounts of carbohydrates in ethanol organo-
solv lignin from Miscanthus,30,65 determined by 13C NMR. In
contrast, Vanderghem et al. found up to 14% residual
carbohydrates in lignin isolated from Miscanthus via ammonia
pretreatment.66
Future studies should be aimed at developing a more
detailed understanding of the molecular structure of lignin,
which is necessary for the application of lignin in polymer
systems or as antioxidant agent; one option here is to apply
complex 2D NMR experiments such as HSQC/HMQC,
HSQC-TOCSY, and HMBC to evaluate long-distance
Figure 13. Ratio of most abundant linkages in Miscanthus, Paulownia, coupling patterns. Because of some disadvantages to this
and Silphium lignins derived from biomasses with different particle approach (long experiment time, semiquantitative character of
sizes in comparison with KL and Indulin AT. Abbreviations (A, B, C,
D, E) belong to interunit linkages shown in Figure 10.
2D NMR), the application of chemometric techniques to 1D
NMR profiles of lignin is a promising way to obtain
quantitative properties in a short time. Now, such models

Figure 14. (Left) Weight-average molar mass Mw and (right) polydispersity index (PDI) of different lignins: Miscanthus (M1, M2, M3), Paulownia
(P1, P2, P3) and Silphium (S2). Boxes indicate 25/75th percentiles, whiskers indicate outmost data points which are not outliers (indicated by
rhomb).

1938 https://dx.doi.org/10.1021/acs.biomac.0c00123
Biomacromolecules 2020, 21, 1929−1942
Biomacromolecules pubs.acs.org/Biomac Article

can be transferred to other high-resolution devices without

repeating the multivariate calibration procedure, but still large
calibration data sets are needed.67
3.6. Molar Masses and Polydispersities. Figure 14
shows weight-average molar masses (Mw) and polydispersity
indexes (PDI) for the different lignins. Comparing samples
with the same biomass but with different particle sizes, a trend
that smaller particle sizes result in higher molar masses can be
observed. It is reported that the accessibility for the pulping
process is influenced by particle size.46 Moreover, higher molar
masses could be caused by repolymerization processes in finely
ground particles, as observed in various studies.68,69
When comparing the same particle size but different
biomasses, it is obvious that lignins from Silphium have higher
molar masses than from Paulownia or from Miscanthus.
Weight-average molar masses of the Miscanthus derived lignins
(M1, 2.21 kDa; M2, 2.26 kDa; M3, 2.78 kDa) are in
accordance to literature data, where values between 1.76−2.39
kDa26 and 2.25−2.72 kDa28,70 are reported, although it should Figure 15. Size exclusion chromatograms of three different lignins:
be noted that those values are relative to polystyrene and were Miscanthus (M2), Paulownia (P2), and Silphium (S2).
obtained using THF as eluent. Lignins extracted from
Paulownia have weight-average molar masses of 3.19 kDa,
3.50 kDa, and 5.64 kDa for P1, P2, and P3, respectively. No
data were reported for molar masses of Paulownia-derived Miscanthus. In contrast, the lignins isolated from Silphium
lignins. As Paulownia is considered a hardwood,7 it should be perfoliatum resemble Paulownia and other hardwood-based
legitimate to compare those results to other hardwoods. lignins (exclusively composed of G and S units) and therefore
Zijlstra et al. extracted lignins from beech and cedar wood clearly differ from grass-derived lignins. Deviations in
using a similar isolation method (80:20 ethanol/water mixture, monolignol ratio (and corresponding linkages) have an impact
120 °C for 5 h but with HCl) and reported weight-average on properties such as antioxidant activity and antimicrobial
molar masses of 4.30 kDa, 3.29 kDa, and 2.93 kDa, capacity. So, the lignins isolated in this study are promising
respectively.71 Considering the particle size, Silphium by far candidates and are currently investigated as environmentally
benign antioxidants.

■
yields lignins with the highest molar mass of 5.87 kDa. So far,
no data about SEC of lignins derived from Silphium have been
ASSOCIATED CONTENT
published yet, but to put those values into perspective the well-
known Indulin AT and our KL show molar masses of 4.57 kDa *
sı Supporting Information

and 4.92 kDa, respectively. Rohde et al. measured molar The Supporting Information is available free of charge at
masses of 4.50 kDa for acetylated Indulin AT in THF relative https://pubs.acs.org/doi/10.1021/acs.biomac.0c00123.
to polystyrene, which is in good accordance to our results, PCA of 1H NMR data, including lignins isolated from
despite the differing measuring parameters.72 Miscanthus (M1, M2, M3), Paulownia (P1, P2, P3), and
Except for M1−M2 and P3−S2, all lignin molar masses Silphium (S2) of different particle size (1, 1.6−2.0 mm;
differ from each other significantly (95%). A similar trend can 2, 0.5−1.0 mm; 3: <0.25 mm) and one Kraft lignin (KL)
be observed for the polydispersity indexes, although not and Indulin AT (AT) as reference; FDA of 1H NMR
significantly. All three repetitions of the pulping process show data of lignins isolated from Miscanthus (M), Paulownia
comparable molar masses, once more emphasizing that the (P), and Silphium (S) of different particle size (1, 1.6−
isolation procedure results in reproducible and reliable data. 2.0 mm; 2, 0.5−1.0 mm; 3: <0.25 mm); assignment of
Exemplarily, three size exclusion chromatograms of M2, P2, FTIR signals; assignment of main 1H−13C correlations
and S2 are depicted in Figure 15. M2 and S2 have a peak mass in HSQC-NMR spectra of lignins in the aliphatic
share at 1.42 kDa, but M2 additionally has a relatively high oxygenated region; assignment of main 1 H−13 C
mass share at 274 Da corresponding to a high content of small correlations in HSQC-NMR spectra of lignins in the
molecules such as dimers. S2 instead has one at 36.3 kDa. P2 aromatic core region; HSQC NMR spectra of
has a slightly higher peak mass at 1.55 kDa with no distinct Miscanthus lignin (M1, M3), Paulownia lignin (P1,
second peak. P3), Kraft lignin (KL) and Indulin AT of aliphatic side
chain region and aromatic core region; ratio of
4. CONCLUSION monolignols (S, G, H) in lignins from different
According to the presented results, the catalyst-free ethanol biomasses; ratio of most abundant linkages in lignins
organosolv treatment applied for lignin isolation serves as an from different biomasses (PDF)
appropriate method to produce lignins of high purity, a
fundamental aspect for further lignin valorization and
exploitation. For the first time, data are reported for a new
■ AUTHOR INFORMATION
Corresponding Author
low-input crop (Silphium perfoliatum), recently listed as an Margit Schulze − Department of Natural Sciences, Bonn-Rhein-
eligible EFA by the Common Agricultural Policy of the Sieg University of Applied Sciences, D-53359 Rheinbach,
European Union. As shown, the chemical biomass composition Germany; orcid.org/0000-0002-8975-1753;
of Silphium perfoliatum is similar to that of grasses such as Email: [email protected]
1939 https://dx.doi.org/10.1021/acs.biomac.0c00123
Biomacromolecules 2020, 21, 1929−1942
Biomacromolecules

■
pubs.acs.org/Biomac Article

Authors REFERENCES
Jessica Rumpf − Department of Natural Sciences, Bonn-Rhein- (1) Müller, A.; Weigelt, J.; Götz, A.; Schmidt, O.; Lobos Alva, I.;
Sieg University of Applied Sciences, D-53359 Rheinbach, Matuschke, I.; Ehling, U.; Beringer, T. The Role of Biomass in the
Germany Sustainable Development Goals: A Reality Check and Governance
Xuan Tung Do − Department of Natural Sciences, Bonn-Rhein- Implications: IASS Working Paper, 2015.
Sieg University of Applied Sciences, D-53359 Rheinbach, (2) European Commission. A sustainable Bioeconomy for Europe:
Germany Strengthening the connection between economy, society and the environ-
René Burger − Department of Natural Sciences, Bonn-Rhein- ment: Communication from the Commission to the European Parliament,
Sieg University of Applied Sciences, D-53359 Rheinbach, the Council, the European Economic and Social Committee and the
Germany Committee of the Regions, 2018.
Yulia B. Monakhova − Spectral Service AG, D-50996 Köln, (3) van Meerbeek, K.; Muys, B.; Hermy, M. Lignocellulosic biomass
Germany; Institute of Chemistry, Saratov State University, for bioenergy beyond intensive cropland and forests. Renewable
410012 Saratov, Russia Sustainable Energy Rev. 2019, 102, 139−149.
(4) Hassan, S. S.; Williams, G. A.; Jaiswal, A. K. Moving towards the
Complete contact information is available at: second generation of lignocellulosic biorefineries in the EU: Drivers,
https://pubs.acs.org/10.1021/acs.biomac.0c00123 challenges, and opportunities. Renewable Sustainable Energy Rev. 2019,
101, 590−599.
Author Contributions (5) Renders, T.; van den Bossche, G.; Vangeel, T.; van Aelst, K.;
The manuscript was written through contributions of all Sels, B. Reductive catalytic fractionation: state of the art of the lignin-
authors. All authors have given approval to the final version of first biorefinery. Curr. Opin. Biotechnol. 2019, 56, 193−201.
the manuscript. (6) García, J. C.; Zamudio, M. A. M.; Pérez, A.; Feria, M. J.; Gomide,
J. L.; Colodette, J. L.; López, F. Sod-AQ Pulping of Paulownia wood
Funding after Hydrolysis Treatment. Bioresources 2011, 6, 971−986.
This study was supported by the BMBF (Grant 13FH102PX8) (7) Rencoret, J.; Marques, G.; Gutiérrez, A.; Nieto, L.; Jiménez-
and EFRE/NRW “Biobasierte Produkte” (Grant EFRE Barbero, J.; Martínez, Á . T.; del Río, J. C. Isolation and structural
0500035). J.R. gratefully acknowledges a scholarship given by characterization of the milled-wood lignin from Paulownia fortunei
the Graduate Institute of the Bonn-Rhein-Sieg University of wood. Ind. Crops Prod. 2009, 30, 137−143.
Applied Sciences. Y.M. acknowledges support of the Russian (8) Ye, X.-k.; Chen, Y. Kinetics study of enzymatic hydrolysis of
Science Foundation (Project 18-73-10009). Paulownia by dilute acid, alkali, and ultrasonic-assisted alkali
pretreatments. Biotechnol. Bioprocess Eng. 2015, 20, 242−248.
Notes (9) Clifton-Brown, J.; Harfouche, A.; Casler, M. D.; Dylan Jones, H.;
The authors declare no competing financial interest. Macalpine, W. J.; Murphy-Bokern, D.; Smart, L. B.; Adler, A.;

■ ACKNOWLEDGMENTS
The authors thank Ralf Pude, Georg Völkering and Jan Niklas
Ashman, C.; Awty-Carroll, D.; et al. Breeding progress and
preparedness for mass-scale deployment of perennial lignocellulosic
biomass crops switchgrass, miscanthus, willow and poplar. GCB
Bioenergy 2019, 11, 118−151.
Frase (INRES, University of Bonn, Germany) for providing (10) Von Cossel, M.; Lewandowski, I.; Elbersen, B.; Staritsky, I.;
the biomasses, Zellstoff- und Papierfabrik Rosenthal GmbH Van Eupen, M.; Iqbal, Y.; Mantel, S.; Scordia, D.; Testa, G.;
(Blankenstein, Germany, MERCER group) for providing the Cosentino, S. L.; Maliarenko, O.; Eleftheriadis, I.; Zanetti, F.; Monti,
black liquor, Friederike Armbruster (Bonn-Rhein-Sieg Uni- A.; Lazdina, D.; Neimane, S.; Lamy, I.; Ciadamidaro, L.; Sanz, M.;
versity of Applied Sciences, Germany) for assistance in lignin Esteban Carrasco, J.; Ciria, P.; McCallum, I.; Trindade, L. M.; Van
recovery from Paulownia tomentosa; and Michael Larkins Loo, E. N.; Elbersen, W.; Fernando, A. L.; Papazoglou, E. G.;
(North Carolina State University, Raleigh, U.S.A.) for English Alexopoulou, E.; et al. Marginal Agricultural Land Low-Input Systems
language revision. for Biomass Production. Energies 2019, 12, 3123−3147.

■
(11) da Costa, R. M. F.; Lee, S. J.; Allison, G. G.; Hazen, S. P.;
ABBREVIATIONS Winters, A.; Bosch, M. Genotype, development and tissue-derived
variation of cell-wall properties in the lignocellulosic energy crop
EFA Ecological focus area Miscanthus. Ann. Bot. 2014, 114, 1265−1277.
FDA Factor discriminant analysis (12) Scagline-Mellor, S.; Griggs, T.; Skousen, J.; Wolfrum, E.;
HSQC Heteronuclear single quantum coherence Holásková, I. Switchgrass and Giant Miscanthus Biomass and
HMBC Heteronuclear multiple bond correlation Theoretical Ethanol Production from Reclaimed Mine Lands.
HMQC Heteronuclear multiple-quantum correlation BioEnergy Res. 2018, 11, 562−573.
KL Kraft lignin (13) Schäfer, J.; Sattler, M.; Iqbal, Y.; Lewandowski, I.; Bunzel, M.
LAP Laboratory analytical procedures Characterization of Miscanthus cell wall polymers. GCB Bioenergy
LCF Lignocellulosic feedstock 2019, 11, 191−205.
M Miscanthus (14) Emmerling, C.; Pude, R. Introducing Miscanthus to the
Mn Number-average molecular weight greening measures of the EU Common Agricultural Policy. GCB
Bioenergy 2017, 9, 274−279.
Mw Weight-average molecular weight
(15) Gansberger, M.; Montgomery, L. F.R.; Liebhard, P. Botanical
MWL Milled wood lignin characteristics, crop management and potential of Silphium
NREL National renewable energy laboratory perfoliatum L. as a renewable resource for biogas production: A
P Paulownia review. Ind. Crops Prod. 2015, 63, 362−372.
PCA Principle component analysis (16) Klímek, P.; Meinlschmidt, P.; Wimmer, R.; Plinke, B.; Schirp,
PDI Polydispersity index A. Using sunflower (Helianthus annuus L.), topinambour (Helianthus
S Silphium tuberosus L.) and cup-plant (Silphium perfoliatum L.) stalks as
SDG Sustainable development goals alternative raw materials for particleboards. Ind. Crops Prod. 2016, 92,
TOCSY Total correlation spectroscopy 157−164.

1940 https://dx.doi.org/10.1021/acs.biomac.0c00123
Biomacromolecules 2020, 21, 1929−1942
Biomacromolecules pubs.acs.org/Biomac Article

(17) Vanholme, R.; De Meester, B.; Ralph, J.; Boerjan, W. Lignin Characterization and application of lignin. J. Chem. Technol.
biosynthesis and its integration into metabolism. Curr. Opin. Biotechnol. 2015, 90, 534−542.
Biotechnol. 2019, 56, 230−239. (35) Damm, T.; Grande, P. M.; Jablonowski, N. D.; Thiele, B.;
(18) Hirayama, H.; Akiyama, T.; Tamai, A.; Nawawi, D. S.; Syafii, Disko, U.; Mann, U.; Schurr, U.; Leitner, W.; Usadel, B.; Domínguez
W.; Yokoyama, T.; Matsumoto, Y. Variation of the contents of de María, P.; et al. OrganoCat pretreatment of perennial plants:
biphenyl structures in lignins among wood species. Holzforschung Synergies between a biogenic fractionation and valuable feedstocks.
2019, 73, 569−578. Bioresour. Technol. 2017, 244, 889−896.
(19) Collins, M. N.; Nechifor, M.; Tanasă, F.; Zănoagă, M.; (36) Sluiter, A.; Hames, B.; Hyman, D.; Payne, C.; Ruiz, R.; Scarlata,
McLoughlin, A.; Stróżyk, M. A.; Culebras, M.; Teacă, C.-A. C.; Sluiter, J.; Templeton, D.; Wolfe, J. Determination of Total Solids in
Valorization of lignin in polymer and composite systems for advanced Biomass and Total Dissolved Solids in Liquid Process Samples:
engineering applications - A review. Int. J. Biol. Macromol. 2019, 131, Laboratory Analytical Procedure (LAP); Golden, CO, 2008.
828−849. (37) Sluiter, A.; Hames, B.; Ruiz, R.; Scarlata, C.; Sluiter, J.;
(20) Giummarella, N.; Pu, Y.; Ragauskas, A. J.; Lawoko, M. A critical Templeton, D. Determination of Ash in Biomass: Laboratory Analytical
review on the analysis of lignin carbohydrate bonds. Green Chem. Procedure (LAP); Golden, CO, 2008.
2019, 21, 1573−1595. (38) Sluiter, A.; Hames, B.; Ruiz, R.; Scarlata, C.; Sluiter, J.;
(21) Mancini, M.; Duca, D.; Toscano, G. Laboratory customized Templeton, D.; Crocker, D. Determination of Structural Carbohydrates
online measurements for the prediction of the key-parameters of and Lignin in Biomass: Laboratory Analytical Procedure (LAP) (Revised
biomass quality control. J. Near Infrared Spectrosc. 2019, 27, 15−25. July 2011): Laboratory Analytical Procedure (LAP); Golden, CO, 2012.
(22) Lancefield, C. S.; Constant, S.; de Peinder, P.; Bruijnincx, P. C. (39) Cordella, C. B.Y.; Bertrand, D. SAISIR: A new general
A. Linkage Abundance and Molecular Weight Characteristics of chemometric toolbox. TrAC, Trends Anal. Chem. 2014, 54, 75−82.
Technical Lignins by Attenuated Total Reflection-FTIR Spectroscopy (40) Villaverde, J. J.; Ligero, P.; de Vega, A. Miscanthus x giganteus
Combined with Multivariate Analysis. ChemSusChem 2019, 12, as a Source Of Biobased Products Through Organosolv Fractionation:
1139−1146. A Mini Review. Open Agric. J. 2010, 4, 102−110.
(23) Hansen, B.; Kusch, P.; Schulze, M.; Kamm, B. Qualitative and (41) Ye, X.; Zhang, Z.; Chen, Y.; Cheng, J.; Tang, Z.; Hu, Y.
Quantitative Analysis of Lignin Produced from Beech Wood by Physico-chemical pretreatment technologies of bioconversion effi-
Different Conditions of the Organosolv Process. J. Polym. Environ. ciency of Paulownia tomentosa (Thunb.) Steud. Ind. Crops Prod.
2016, 24, 85−97. 2016, 87, 280−286.
(24) Saha, M.; Saynik, P. B.; Borah, A.; Malani, R. S.; Arya, P.; (42) Wang, S.; Dai, G.; Yang, H.; Luo, Z. Lignocellulosic biomass
Shivangi; Moholkar, V. S. Dioxane-based extraction process for pyrolysis mechanism: A state-of-the-art review. Prog. Energy Combust.
Sci. 2017, 62, 33−86.
production of high quality lignin. Bioresour. Technol. Rep. 2019, 5,
(43) Haghdan, S.; Renneckar, S.; Smith, G. D. Sources of Lignin. In
206−211.
Lignin in Polymer Composites; Sain, M., Faruk, O., Eds.; Elsevier:
(25) Brandt, A.; Chen, L.; van Dongen, B. E.; Welton, T.; Hallett, J.
Kidlington, Oxford, U.K., 2016; pp 1−11.
P. Structural changes in lignins isolated using an acidic ionic liquid
(44) Obama, P.; Ricochon, G.; Muniglia, L.; Brosse, N.
water mixture. Green Chem. 2015, 17, 5019−5034.
Combination of enzymatic hydrolysis and ethanol organosolv
(26) Bergs, M.; Völkering, G.; Kraska, T.; Pude, R.; Do, X. T.;
pretreatments: effect on lignin structures, delignification yields and
Kusch, P.; Monakhova, Y.; Konow, C.; Schulze, M. Miscanthus x
cellulose-to-glucose conversion. Bioresour. Technol. 2012, 112, 156−
giganteus Stem Versus Leaf-Derived Lignins Differing in Monolignol
163.
Ratio and Linkage. Int. J. Mol. Sci. 2019, 20, 1200. (45) Rashid, T.; Gnanasundaram, N.; Appusamy, A.; Kait, C. F.;
(27) Bergs, M.; Do, X. T.; Rumpf, J.; Kusch, P.; Monakhova, Y.; Thanabalan, M. Enhanced lignin extraction from different species of
Konow, C.; Völkering, G.; Pude, R.; Schulze, M. Comparing chemical oil palm biomass: Kinetics and optimization of extraction conditions.
composition and lignin structure of Miscanthus x giganteus and Ind. Crops Prod. 2018, 116, 122−136.
Miscanthus nagara harvested in autumn and spring and separated into (46) Barakat, A.; Mayer-Laigle, C.; Solhy, A.; Arancon, R. A. D.; de
stems and leaves. RSC Adv. 2020, 10, 10740−10751. Vries, H.; Luque, R. Mechanical pretreatments of lignocellulosic
(28) Bauer, S.; Sorek, H.; Mitchell, V. D.; Ibáñez, A. B.; Wemmer, D. biomass: towards facile and environmentally sound technologies for
E. Characterization of Miscanthus giganteus lignin isolated by ethanol biofuels production. RSC Adv. 2014, 4, 48109−48127.
organosolv process under reflux condition. J. Agric. Food Chem. 2012, (47) Lundquist, K.; Sjöholm, R.; Teien, G.; Pakkanen, T.; Servin, R.;
60, 8203−8212. Sternerup, H.; Wistrand, L.-G.; Nørskov, L.; Schroll, G. NMR Studies
(29) Wang, K.; Bauer, S.; Sun, R.-c. Structural transformation of of Lignins. 2. Interpretation of the 1H NMR Spectrum of Acetylated
Miscanthus × giganteus lignin fractionated under mild formosolv, Birch Lignin. Acta Chem. Scand. 1979, 33b, 27−30.
basic organosolv, and cellulolytic enzyme conditions. J. Agric. Food (48) Lundquist, K.; Aasen, A. J.; Daasvatn, K.; Forsgren, B.;
Chem. 2012, 60, 144−152. Gustafsson, J.-Å.; Högberg, B.; Becher, J. NMR Studies of Lignins. 4.
(30) El Hage, R.; Brosse, N.; Sannigrahi, P.; Ragauskas, A. Effects of Investigation of Spruce Lignin by 1H NMR Spectroscopy. Acta Chem.
process severity on the chemical structure of Miscanthus ethanol Scand. 1980, 34b, 21−26.
organosolv lignin. Polym. Degrad. Stab. 2010, 95, 997−1003. (49) Alzagameem, A.; Bergs, M.; Do, X. T.; Klein, S. E.; Rumpf, J.;
(31) El Hage, R.; Chrusciel, L.; Desharnais, L.; Brosse, N. Effect of Larkins, M.; Monakhova, Y.; Pude, R.; Schulze, M. Low-Input Crops
autohydrolysis of Miscanthus x giganteus on lignin structure and as Lignocellulosic Feedstock for Second-Generation Biorefineries and
organosolv delignification. Bioresour. Technol. 2010, 101, 9321−9329. the Potential of Chemometrics in Biomass Quality Control. Appl. Sci.
(32) Groenewold, G. S.; Johnson, K. M.; Fox, S. C.; Rae, C.; 2019, 9, 2252−2278.
Zarzana, C. A.; Kersten, B. R.; Rowe, S. M.; Westover, T. L.; (50) Lupoi, J. S.; Singh, S.; Parthasarathi, R.; Simmons, B. A.; Henry,
Gresham, G. L.; Emerson, R. M.; et al. Pyrolysis Two-Dimensional R. J. Recent innovations in analytical methods for the qualitative and
GC-MS of Miscanthus Biomass: Quantitative Measurement Using an quantitative assessment of lignin. Renewable Sustainable Energy Rev.
Internal Standard Method. Energy Fuels 2017, 31, 1620−1630. 2015, 49, 871−906.
(33) Gong, C.; Bujanovic, B. Impact of Hot-Water Extraction on (51) Alzagameem, A.; Khaldi-Hansen, B. E.; Büchner, D.; Larkins,
Acetone-Water Oxygen Delignification of Paulownia Spp. and Lignin M.; Kamm, B.; Witzleben, S.; Schulze, M. Lignocellulosic Biomass as
Recovery. Energies 2014, 7, 857−873. Source for Lignin-Based Environmentally Benign Antioxidants.
(34) Zamudio, M. A.M.; Alfaro, A.; de Alva, H. E.; Garcia, J. C.; Molecules 2018, 23, 2664−2288.
Garcia-Morales, M.; Lopez, F. Biorefinery of paulownia by (52) Ahuja, D.; Kaushik, A.; Chauhan, G. S. Fractionation and
autohydrolysis and soda-anthraquinone delignification process. physicochemical characterization of lignin from waste jute bags: Effect

1941 https://dx.doi.org/10.1021/acs.biomac.0c00123
Biomacromolecules 2020, 21, 1929−1942
Biomacromolecules pubs.acs.org/Biomac Article

of process parameters on yield and thermal degradation. Int. J. Biol. Ethanol Extraction and Its Effect on the Depolymerization Yield. J.
Macromol. 2017, 97, 403−410. Visualized Exp. 2019, e58575.
(53) Lin, S. Y. Ultraviolet Spectrophotometry. In Methods in Lignin (72) Rohde, V.; Böringer, S.; Tübke, B.; Adam, C.; Dahmen, N.;
Chemistry; Lin, S. Y., Dence, C. W., Eds.; Springer Series in Wood Schmiedl, D. Fractionation of three different lignins by thermal
Science; Springer: Berlin, 1992; pp 217−232. separation techniques-A comparative study. GCB Bioenergy 2019, 11,
(54) Vivekanand, V.; Chawade, A.; Larsson, M.; Larsson, A.; Olsson, 206−217.
O. Identification and qualitative characterization of high and low
lignin lines from an oat TILLING population. Ind. Crops Prod. 2014,
59, 1−8.
(55) Tribot, A.; Amer, G.; Abdou Alio, M.; de Baynast, H.; Delattre,
C.; Pons, A.; Mathias, J.-D.; Callois, J.-M.; Vial, C.; Michaud, P.;
Dussap, C.-G.; et al. Wood-lignin: Supply, extraction processes and
use as bio-based material. Eur. Polym. J. 2019, 112, 228−240.
(56) Crestini, C.; Melone, F.; Sette, M.; Saladino, R. Milled wood
lignin: A linear oligomer. Biomacromolecules 2011, 12, 3928−3935.
(57) Ralph, J.; Landucci, L. L. NMR of Lignins. In Lignin and
Lignans: Advances in Chemistry; Heitner, C., Dimmel, D., Schmidt, J.
A., Eds.; Taylor & Francis: Boca Raton, 2010; 137−243.
(58) Cheng, K.; Sorek, H.; Zimmermann, H.; Wemmer, D. E.; Pauly,
M. Solution-state 2D NMR spectroscopy of plant cell walls enabled by
a dimethylsulfoxide-d6/1-ethyl-3-methylimidazolium acetate solvent.
Anal. Chem. 2013, 85, 3213−3221.
(59) Hu, Z.; Du, X.; Liu, J.; Chang, H.-m.; Jameel, H. Structural
Characterization of Pine Kraft Lignin: BioChoice Lignin vs Indulin
AT. J. Wood Chem. Technol. 2016, 36, 432−446.
(60) Patil, N. D.; Tanguy, N. R.; Yan, N. Lignin Interunit Linkages
and Model Compounds. In Lignin in Polymer Composites; Sain, M.,
Faruk, O., Eds.; Elsevier: Kidlington, Oxford, U.K., 2016; pp 27−47.
(61) Calvo-Flores, F. G.; Dobado Jiménez, J. A.; Garcia, J. I.; Martín-
Martínez, F. J. Lignin and Lignans as Renewable Raw Materials:
Chemistry, Technology and Applications; Wiley Series in Renewable
Resources; John Wiley and Sons Inc: Chichester West Sussex, U.K.,
2015.
(62) Schutyser, W.; Renders, T.; van den Bosch, S.; Koelewijn, S.-F.;
Beckham, G. T.; Sels, B. F. Chemicals from lignin: an interplay of
lignocellulose fractionation, depolymerisation, and upgrading. Chem.
Soc. Rev. 2018, 47, 852−908.
(63) Villaverde, J. J.; Li, J.; Ek, M.; Ligero, P.; de Vega, A. Native
lignin structure of Miscanthus x giganteus and its changes during
acetic and formic acid fractionation. J. Agric. Food Chem. 2009, 57,
6262−6270.
(64) Villaverde, J. J.; Li, J.; Ligero, P.; Ek, M.; de Vega, A. Mild
peroxyformic acid fractionation of Miscanthus x giganteus bark.
Behaviour and structural characterization of lignin. Ind. Crops Prod.
2012, 35, 261−268.
(65) El Hage, R.; Brosse, N.; Chrusciel, L.; Sanchez, C.; Sannigrahi,
P.; Ragauskas, A. Characterization of milled wood lignin and ethanol
organosolv lignin from miscanthus. Polym. Degrad. Stab. 2009, 94,
1632−1638.
(66) Vanderghem, C.; Richel, A.; Jacquet, N.; Blecker, C.; Paquot,
M. Impact of formic/acetic acid and ammonia pre-treatments on
chemical structure and physico-chemical properties of Miscanthus x
giganteus lignins. Polym. Degrad. Stab. 2011, 96, 1761−1770.
(67) Monakhova, Y. B.; Diehl, B. W. K. Transfer of multivariate
regression models between high-resolution NMR instruments:
application to authenticity control of sunflower lecithin. Magn.
Reson. Chem. 2016, 54, 712−717.
(68) Sun, Z.; Fridrich, B.; de Santi, A.; Elangovan, S.; Barta, K.
Bright Side of Lignin Depolymerization: Toward New Platform
Chemicals. Chem. Rev. 2018, 118, 614−678.
(69) Li, J.; Henriksson, G.; Gellerstedt, G. Lignin depolymerization/
repolymerization and its critical role for delignification of aspen wood
by steam explosion. Bioresour. Technol. 2007, 98, 3061−3068.
(70) Tolbert, A.; Akinosho, H.; Khunsupat, R.; Naskar, A. K.;
Ragauskas, A. J. Characterization and analysis of the molecular weight
of lignin for biorefining studies. Biofuels, Bioprod. Biorefin. 2014, 8,
836−856.
(71) Zijlstra, D. S.; Santi, A. de; Oldenburger, B.; Vries, J. de; Barta,
K.; Deuss, P. J. Extraction of Lignin with High β-O-4 Content by Mild

1942 https://dx.doi.org/10.1021/acs.biomac.0c00123
Biomacromolecules 2020, 21, 1929−1942