Enhancing lignin depolymerizationviaa dithionite-assisted organosolv fractionation of birch sawdust

Article abstract of DOI:10.1039/d1gc00503k

Sodium dithionite is utilized as a reducing agent in the organosolv fractionation of lignocellulose to concomitantly produce cellulosic pulp and promote the reductive conversion of lignin into phenolic monomers. Reactions with model compounds highlight the role of sodium dithionite with respect to the reductive cleavage of β-O-4 bonds in lignin and the consequent formation of phenolic monomers.

Full text of DOI:10.1039/d1gc00503k

Green Chemistry
View Article Online
View Journal | View Issue
COMMUNICATION
Enhancing lignin depolymerization via a dithionite-
assisted organosolv fractionation of birch sawdust†
Cite this: Green Chem., 2021, 23,
3268
Filippo Brienza,a Korneel Van Aelst,b Karel Thielemans, b Bert F. Sels,
b
Received 8th February 2021,
Accepted 11th April 2021
Damien P. Debecker *c and Iwona Cybulska*a
DOI: 10.1039/d1gc00503k
rsc.li/greenchem
Sodium dithionite is utilized as a reducing agent in the organosolv stabilized by catalytic reduction reactions, which prevent re-
fractionation of lignocellulose to concomitantly produce cellulosic condensation towards the recalcitrant species usually found in
pulp and promote the reductive conversion of lignin into phenolic technical lignins. Several authors studied RCF of hardwoods
monomers. Reactions with model compounds highlight the role of and reported the achievement of outstanding delignification
sodium dithionite with respect to the reductive cleavage of β-O-4 yields (up to ∼90%) and phenolic monomer yields (up to
bonds in lignin and the consequent formation of phenolic monomers. 50–60%), as well as high yields of carbohydrate pulps with
nearly complete cellulose preservation, amenable toward enzy-
matic or chemical conversion.11–15
Despite providing a powerful means for full valorization of
lignocellulose, RCF processes suffer few limitations that com-
Introduction
In view of the pressing need for sustainable alternatives to
fossil feedstock, lignocellulosic biomass has gained consider-
able attention as a potential source of renewable energy and
materials.1 Recently, within the concept of integrated biorefi-
neries, promising one-step processes were proposed for the
simultaneous fractionation of lignocellulose and conversion of
lignin toward a handful of phenolics.2–7 These methods –
which target the formation of a lignin oil, rich in phenolic
monomers – are usually referred to as “lignin-first biorefi-
neries”, to highlight that the valorization of lignin to chemi-
cals is realized upfront.6–9
A popular strategy adopted within lignin-first processes is
Reductive Catalytic Fractionation (RCF),10 also referred to as
Early-stage Catalytic Conversion (ECCL)3 or Catalytic Upstream
Biorefining (CUB),11 consisting of an organosolv treatment of
lignocellulosic biomass in the presence of a heterogeneous
redox catalyst and of a source of hydrogen.6,10 Under such con-
ditions, labile β-O-4 bonds, the most abundant in lignin struc-
tures, are cleaved and the generated lignin fragments are
plicate their scalability and negatively impact their environ-
mental performance. Notably, the frequent use of costly pre-
cious metals and the adoption of high pressures of hydrogen
gas constitute primary issues that need to be addressed in
order to minimize processing costs as well as safety and equip-
ment requirements.7,16
Van den Bosch et al. conducted an in-depth study of the
role of the solvent, the catalyst and the hydrogen source in the
frame of RCF and pointed out that lignin depolymerization
occurs primarily via solvolysis.17 In particular, the authors pos-
tulated the necessity of a reductive step for the solvolytic clea-
vage of β-O-4 bonds in lignin structures, possibly involving the
solvent as a source of hydrogen. Metal-based catalysis is only
involved in the final stabilization of the generated reactive
lignin moieties, via the hydrogenation of alkylic side chains in
lignin units. More recently, Klinger et al. demonstrated that
reductive cleavage of β-O-4 bonds in pre-extracted, oxidized
lignin can be accomplished using nucleophilic thiols as redu-
cing agents, with the concomitant formation of the corres-
ponding disulfides.18,19 The authors reported a considerable
formation of phenolic monomers and short oligomers as well
as a minor incorporation of sulfur into the lignin matrix.
These results prompted us to investigate a modified organo-
solv process, which could lead to the simultaneous production
of cellulosic pulp and conversion of lignin into phenolic
monomers, but in the absence of hydrogen gas, heterogeneous
catalysts and nucleophilic thiols. Sodium dithionite (Na2S2O4),
an inexpensive and readily available sulfur-based compound,
aEarth and Life Institute (ELI), UCLouvain, Croix du Sud 2, 1348 Louvain-La-Neuve,
Belgium
bCentre for Sustainable Catalysis and Engineering, KU Leuven, Celestijnenlaan 200F,
3001 Leuven, Belgium
cInstitute of Condensed Matter and Nanoscience (IMCN), UCLouvain,
Place Louis Pasteur 1, 1348 Louvain-La-Neuve, Belgium.
E-mail: damien.debecker@uclouvain.be
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
d1gc00503k
3268 | Green Chem., 2021, 23, 3268–3276
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Communication
was selected as a reducing agent. Na2S2O4 is already adopted The solid fraction
in the pulping industry for the bleaching of pulps20,21 and was
The light brown solid isolated after the fractionation was ana-
lyzed for dry matter (DM) and ash (Fig. S2†),27,28 and then sub-
jected to acid hydrolysis,29 which revealed that about 91.7% of
C6 polysaccharides initially contained in the biomass were
recovered in the solid fraction (Table 1). In addition, 81.9% of
C5 polysaccharides and 71.6% of the acid-insoluble lignin
(AIL) contents were removed, leading to a cellulosic pulp with
a high cellulose purity of 76.0%. The obtained pulp was ana-
lyzed via X-ray diffraction (Fig. S3†), and the cellulose crystalli-
nity index (CI) was determined by Segal’s method.30 Despite
the extensive removal of hemicellulose and lignin achieved
during the fractionation, which was reported to favor an
increased CI,31,32 no major difference was observed between
the CIs of the untreated birch sawdust and that of the pulp,
48.7% and 54.4%, respectively (Table 1). Partial cellulose depo-
lymerization27 or re-deposition of pseudolignin28 (also termed
“humins”) on cellulose fibers may constitute possible expla-
nations for the fact that the measured CI increase remained so
minor.
also recently employed in a patented process for the depoly-
merization of Kraft lignin.22 However, to our knowledge, there
is no record of the adoption of sodium dithionite in the frame
of organosolv fractionation of lignocellulose. Yet, the use of
Na2S2O4 for the depolymerization of lignin directly from native
biomass may constitute an interesting route for the valoriza-
tion of lignin during pulping.
Herein, we disclose the potential of the dithionite-assisted
organosolv fractionation (DAOF) process for the treatment of
lignocellulose and we rationalize the role of the reducing agent
in the formation of phenolic monomers via reactions with
model compounds. We argue that this strategy represents a
practical way to modify a pulping process while targeting the
reductive depolymerization of lignin for a sustainable pro-
duction of monoaromatics.
Results and discussion
The morphological features of pristine biomass and of the
obtained pulp were further inspected via scanning electron
microscopy (SEM). While the surface of birch sawdust
appeared rough and compact (Fig. S4a†), after treatment, the
pulp possessed a considerably smoother surface and bundles
of fibers with a diameter of about 10 µm could be observed
(Fig. S4b†). At higher magnification, thin fibers possessing
diameters of about 1 µm were apparent (Fig. S4c†), pointing at
the partial disassembly of fibrous bundles achieved during the
treatment, which is expected to favor further enzymatic conver-
sion of the pulp.33
Dithionite-assisted organosolv fractionation: proof of concept
In order to assess the potential of DAOF, batch experiments
were carried out in triplicate treating 3 g of birch sawdust
(Betula pendula, particle size ≤2 mm,23 biomass composition
reported in Table S1†) in 120 mL of an equivolumetric mixture
of n-butanol and water, in the presence of 1 g of sodium
dithionite, at a temperature of 200 °C for a duration of
3 hours. After the fractionation, a solid pulp and two liquid
fractions (an organic and an aqueous phase) were collected.
An outline of the process is presented in Fig. 1. As the solvent,
the n-butanol–water mixture facilitates the subsequent separ-
ation of the isolated fractions.24–26 Typically, the solid fraction
contains the retained polysaccharides (C5, C6), the organic
liquid fraction contains lignin-derived phenolics, and the
aqueous liquid fraction comprises solubilized non-condensed
C5 and C6 derivatives (e.g. mono- and oligosaccharides,
polyols and organic acids). The obtained product streams were
then extensively characterized by a series of analytical pro-
cedures (Fig. S1†).
In order to gain a better insight into the susceptibility of
the cellulosic pulp toward further processing, it was subjected
to enzymatic digestion (Table 1). A substantial enhancement
of both glucan and xylan convertibility to their corresponding
Table 1 Characterization of the solid and liquid fractions obtained for
the treatment of birch sawdust (3 g, particle size <2 mm) in a mixture of
n-butanol and water (120 mL, 50% v/v), at 200 °C, under 30 bar of N2
(introduced at ambient temperature), in the presence of Na2S2O4 (1 g),
for a duration of 3 hours
Solid fraction
Recovery (wt%)
C5
C6
18.1 3.2
91.7 1.3
28.4 2.7
91.4 4.6
92.8 4.0
54.4 3.5
93.4 6.8
18.1 1.7
0.1 0.0
3.2 0.4
0.5 0.1
7.5 0.4
Lignina
Glucan
Xylan
Conversion (%)
CI (%)
Yield (wt%)
Organic fraction
Aqueous fraction
Lignin oilb
Monophenolicsb
C5, C6 derivativesc
Lignin oilb
Yield (wt%)
Monophenolicsb
C5, C6 derivativesc
a Acid-insoluble lignin. b Expressed with respect to the weight of acid-
insoluble lignin contained in the initial biomass. c Non-condensed
carbohydrate derivatives, expressed with respect to the total weight of
polysaccharides contained in the initial biomass.
Fig. 1 Outline of the dithionite-assisted organosolv fractionation of
birch sawdust.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 3268–3276 | 3269

View Article Online
Communication
Green Chemistry
monosaccharides was observed for the solid pulp (91.4% and
92.8%, respectively) compared to untreated biomass (8.3% and
9.7%, respectively), confirming that the structural changes
occurring in the lignocellulose matrix during fractionation
ultimately favor the pulp processability. Interestingly, analo-
gous blank organosolv experiments carried out in the absence
of sodium dithionite produced a solid pulp with very similar
properties (Table S2†), indicating that the use of dithionite
does not affect the pulp composition, CI and enzymatic
digestibility. On the other hand, a relevant difference between
DAOF and a blank organosolv treatment was found in the
degree of polymerization (DP) of cellulose in the isolated
pulps, which was estimated via viscosimetric analysis, accord-
ing to the NF G 06-037 norm (Table S3†). Notably, DAOF pro-
duced a pulp with a DPv of about 780 AGU, considerably larger Fig. 2 Comparison of the GPC chromatograms of the lignin oil isolated
from the organic fraction obtained after treating birch sawdust (3 g, par-
ticle size <2 mm) in a mixture of n-butanol and water (120 mL, 50% v/v),
at 200 °C, under 30 bar of N2 (introduced at ambient temperature), for a
duration of 3 hours, in the absence (blank) and in the presence of
than that of the blank organosolv pulp (∼230 AGU). We reason
that this difference may be due to the higher acidity of the
medium in the blank organosolv experiment (see below),
which may favor cellulose depolymerization.
Na2S2O4 (1 g). The total surface of each chromatogram was normalized
with respect to the yield of lignin oil obtained for the corresponding
experiment.
The organic fraction
The organic liquid fraction obtained from the process was ana-
lyzed for dry matter and ash (Fig. S2†),27,28 and then a lignin
oil was isolated from the organic fraction by drying a sample mers was much smaller, and, instead, the presence of peaks at
and subjecting it to liquid–liquid extraction with dichloro- higher molecular weights was substantially larger than in the
methane and water, following
a
procedure reported DAOF process. Also, the tail extending to even higher mole-
elsewhere.12,17 Drying the dichloromethane extract yielded a cular weight products was more pronounced in the case of the
viscous brown lignin oil, corresponding to about 93.4% of the blank experiment. Thus, overall, the analysis of the MWD of
weight of AIL contained in the initial biomass (Table 1). lignin oil unambiguously demonstrates that the introduction
Interestingly, this value is higher than 71.6% measured for of sodium dithionite within the organosolv process leads to
lignin removal in the pulp, suggesting the possible redeposi- enhanced depolymerization of the isolated lignin, with an
tion of pseudolignin on the pulp or the incorporation of increased production of phenolic monomers.
(hemi)cellulose derivatives into lignin oil. Indeed, humic pro-
The monomer composition of the lignin oil was analyzed
ducts, often generated during hydrothermal processing of the by gas chromatography (GC-MS and GC-FID). The chromato-
lignocellulosic material via condensation of dehydration pro- gram reported in Fig. S5a† and the data summarized in
ducts from polysaccharides,34 are known to be partially soluble Table S4† show that several monophenolic compounds are
in alcohols and dichloromethane.35 Thus, we reason that present in the lignin oil, accounting for a total monomer yield
humins are likely responsible for the apparent generation of of 18.1% (based on the content of AIL in the initial biomass).
lignin during the fractionation.
The most abundant phenolic monomers that were detected
The molecular weight distribution (MWD) of the lignin oil are: 4-ethoxy-3,5-dimethoxybenzaldehyde (4.3%), 4-propenyl
was evaluated via gel permeation chromatography (GPC), and syringol (4.3%), desaspidinol (3.2%) and acetosyringone
compared to that obtained for a blank organosolv reaction, (2.7%). The formation of phenolic compounds possessing
carried out in the absence of sodium dithionite (Fig. 2). A aldehyde or ketonic groups within the alkylic side chains is
broad but small peak was detected at a molecular weight typical of acid-catalyzed organosolv depolymerization of
below 100 g mol−1 for the DAOF experiment, which may be lignin,3 while the formation of monomers with unsaturated
related to degradation products from the solvent and the redu- alkylic side chains was postulated or demonstrated to occur
cing agent. Two major peaks were found at about 160 g mol−1 under RCF conditions.14,17,36 Therefore, we reason that lignin
and 240 g mol−1, corresponding to phenolic monomers. These depolymerization during DAOF may follow a combination of
constitute the major fraction of the chromatogram, indicating acid-catalyzed and reductive pathways for the cleavage of β-O-4
the formation of monophenolics during the fractionation bonds in lignin and for the generation of phenolic monomers
process. In addition, two smaller peaks were observed at 370 g (Scheme 1). However, the formation of structures with an
mol−1 and 610 g mol−1, likely corresponding to phenolic alkylic side chain containing four carbon atoms, such as
dimers and trimers. These were followed by a tail extending to desaspidinol, is somewhat surprising (see below).
higher molecular weights, which reveals the presence of some
For comparison, the analysis of phenolic monomers in the
condensation products (lignin oligomers and humins). For the lignin oil isolated from blank organosolv reactions performed
blank experiment, the peak corresponding to phenolic mono- in the absence of dithionite revealed a considerably lower for-
3270 | Green Chem., 2021, 23, 3268–3276
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Communication
Scheme 1 Possible pathways leading to the formation of phenolic monomers during organosolv fractionation in the presence of sodium dithionite.
Pathways (A) and (B) rely on the acid-catalyzed cleavage of β-O-4 bonds,3,4 while the pathway (C) consists of a reductive cleavage of ether bonds.17
The combination of acid-catalyzed and reductive lignin depolymerization pathways may explain the different structures observed for the phenolic
monomers produced during the process (enclosed in the green oval). The generated monomers can subsequently undergo repolymerization to an
extent that depends on reaction conditions.
mation of monophenolics (total monomer yield of 3.6%, see during hydrothermal treatment, via hydrolysis, retro-aldol and
Table S5†), corroborating the deductions made based on GPC. dehydration pathways.37–39 The presence of a large amount of
Notably, marginal formation of sinapyl alcohol occurred for formic acid in the aqueous phase is remarkable. In addition,
the blank organosolv experiments, while such monophenolic acetic acid (6.5%) was detected in the aqueous phase, likely
with an unsaturated alkyl side chain possessing a terminal generated from the deacetylation of hemicellulose.40,41 The
hydroxyl group was not observed in DAOF.
formation of organic acids during the process is known to con-
To inspect the presence of organic acids and dehydration tribute to lowering the pH of the medium and to enhancing
products from carbohydrates, a sample of the organic phase acid-catalyzed cleavage of ether and ester bonds in the ligno-
was subjected to HPLC analysis. The chromatogram in cellulose structure.42
Fig. S6a† and the data in Table S6† show that 0.1% of formic
Blank organosolv experiments, carried out in the absence of
acid and 0.9% of acetic acid were detected (on a DM basis), dithionite, led to the formation of lower amounts of formic
indicating a minimal presence of non-condensed carbohydrate and acetic acid (0.5% and 3.1%, respectively), and a larger
derivatives in the organic phase. A similar outcome was yield of xylose (1.2%), indicating that the presence of Na2S2O4
observed also for the blank organosolv experiment.
affects (hemi)cellulose decomposition. Despite the larger pro-
duction of formic and acetic acid observed during DAOF, the
pH value of the aqueous phase was larger in this case, com-
pared to that measured after a blank organosolv treatment
(about 5.9 and 3.3, respectively), highlighting how the pres-
ence of dithionite leads to a lower acidity of the medium.
The aqueous fraction
The aqueous liquid fraction obtained after DAOF was analyzed
for dry matter and ash (Fig. S2†);27,28 then a lignin oil was iso-
lated, corresponding to about 3.2% of the weight of AIL con-
tained in the initial biomass (Table 1). GC analyses (Fig. S5b
and Table S4†) showed that the total yield of monophenols
The origin of desaspidinol
was as low as 0.5%. These findings demonstrate that efficient The generation of phenolic compounds with alkylic side
isolation of lignin derivatives in the organic phase was chains containing four carbon atoms within DAOF is rather
achieved. In order to inspect the presence of non-condensed unexpected since these structures are not present in native
derivatives that may arise from the decomposition of C5 and lignin.3 On the other hand, desaspidinol was reported to form
C6 polysaccharides (mono- and oligosaccharides, polyols and in the frame of the catalytic hydrogenolysis of lignin, and a
organic acids), a sample of the aqueous phase was subjected correlation between its production and the use of formic acid
to HPLC analysis. Xylose, 1,2-propanediol and formic acid are as a solvent for lignin depolymerization was recognized.43–47
the major carbohydrate derivatives that were identified, Therefore, we surmise that, in DAOF, carbohydrate derivatives,
respectively, accounting for 0.1%, 0.7% and 3.7% of the initial such as formic acid, produced by the decomposition of (hemi)
biomass DM (Fig. S6b and Table S6†). These components were cellulose may lead to the subsequent formation of
reported to arise from the decomposition of (hemi)cellulose desaspidinol.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 3268–3276 | 3271

View Article Online
Communication
Green Chemistry
To shed light on the role of carbohydrate derivatives with components. On the one hand, a relevant amount of C5
respect to the formation of desaspidinol during the fraction- derivatives was found in all the isolated fractions, also substan-
ation, milled wood lignin (MWL) was extracted from birch tially contributing to material losses. On the other hand, C6
sawdust and subjected to an analogous process. GC analysis of polysaccharides and lignin appeared to be mostly recovered in
the isolated lignin oil (Table S7†) showed that the total yield of the solid and organic fraction, respectively, showing that
phenolic monomers was more than two times lower than that efficient separation of cellulose and aromatics was achieved,
observed for DAOF of lignocellulosic biomass, possibly due to which conveniently facilitates subsequent downstream
partial modifications of the native lignin structure during operations.
MWL extraction. A significant finding is that the yield of desas-
Incorporation of dithionite derivatives into the output
pidinol was about 25 times lower for the MWL, which unam- streams obtained from the fractionation was assessed by
biguously confirms that the formation of this monophenol is quantification of the sulfur content in each fraction via
related to the presence of carbohydrate derivatives in the reac- ICP-AES (Table S8†). About 50.5% of the sulfur introduced in
tion mixture. A similar experiment was performed treating the process was recovered in the organic fraction, which points
MWL in the presence of formic acid. In this case, the yield of out the formation of organosulfur compounds during the
desaspidinol increased slightly, even if it remained much process, such as (poly)sulfides, sulfonates or thiols. An
lower than in DAOF of lignocellulosic biomass. Thus, while additional 27.4% of sulfur was found in the aqueous fraction,
formic acid appears to play a role in the formation of desaspi- corresponding to inorganic derivatives of dithionite, such as
dinol during DAOF, other (hemi)cellulose derivatives produced sulfate, sulfite and thiosulfate. The solid fraction accounted
during the fractionation may also influence this phenomenon. for 4.0% of the sulfur introduced, possibly owing to the pro-
duction of elementary sulfur or to the incorporation of sulfur
Overview of the fate of biomass components, dithionite and
butanol
into the re-deposited pseudolignin. Notably, nearly 18% of the
sulfur introduced was not recovered in the product streams,
The compositional investigation of the solid and the organic suggesting the formation of volatile sulfur species during
and the aqueous fractions obtained from DAOF gave insight DAOF.
into the fate of biomass components during processing. An
The stability of n-butanol during the process was
overall characterization of biomass derivatives in the product assessed by analyzing its recovery in the liquid fractions via
streams is presented in Table S6,† and, based on that, an HPLC, which revealed that about 97% of the butanol co-
overall mass balance for lignocellulose components is illus- solvent was recovered in the output streams (Table S9†),
trated in Fig. 3. About 89% of the DM initially introduced into with 93% of it being recovered in the organic phase, indi-
the reactor was recovered in the output streams; losses are pre- cating that no relevant decomposition of n-butanol occurs
sumably due to the partial conversion of biomass to volatile during the process.
Fig. 3 Schematic view of the mass balance (based on DM of the biomass introduced) for biomass components, obtained for the treatment of birch
sawdust (3 g, particle size <2 mm) in a mixture of n-butanol and water (120 mL, 50% v/v), at 200 °C, under 30 bar of N2 (introduced at ambient temp-
erature), in the presence of Na2S2O4 (1 g), for a duration of 3 hours. The reaction was carried out in triplicate. The ash fraction was not taken into
consideration in the mass balance, since it only accounts for a marginal fraction of DM in biomass (0.5%), while most of the ash content in the reac-
tion products derives from Na2S2O4. Lignin oil, besides phenolic monomers, was assumed to be composed only of oligomers.
3272 | Green Chem., 2021, 23, 3268–3276
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Communication
The role of sodium dithionite explained with model
compounds
In order to gain a better insight into the role of dithionite
within lignin depolymerization, 2-phenoxy-1-phenylethanol (1)
was selected as a model compound for β-O-4 bonds in
lignin,48,49 and subjected to similar process conditions.
Scheme 2 illustrates the two-step reaction, involving an initial
cleavage of 1 (Step I) to generate styrene (2) and phenol (3), fol-
lowed by the hydrogenation of styrene to form ethylbenzene
(4) (Step II).
In the absence of dithionite, the conversion of 1 was low
and mostly high molecular weight products were formed, as
suggested by the long tail after the peak at about 190 g mol−1
,
corresponding to the dimer (Fig. 4a). In contrast, when the
reducing agent was introduced, a greater conversion of 1 was
observed and another peak at about 110 g mol−1 appeared,
which became larger as the loading of dithionite was
increased, confirming that an excess of sodium dithionite sub-
stantially boosted the cleavage of β-O-4 bonds and the for-
mation of monomeric products. The tail extending to high
molecular weights appeared less pronounced in the presence
of dithionite. However, the small peak at 290 g mol−1 indicates
that repolymerization reactions were not totally suppressed in
the presence of dithionite.
The identification and quantification of the monomeric
reaction products (Fig. 4b) corroborate the deductions made
based on GPC. The addition of sodium dithionite boosted con-
version and triggered the formation of 2 and 3. As the loading
of the reducing agent was increased, both the dimer conver-
sion and the monomer yields increased. The lack of selectivity
toward monomers constitutes additional indication that a
portion of 1 underwent repolymerization. Furthermore, at high
loadings of dithionite, 4 was found as well, showing that the
reducing agent can trigger the hydrogenation of alkylic side
chains in the monomeric species generated via reductive clea-
vage of β-O-4 bonds.
A similar set of experiments was performed using 4-prope-
nyl guaiacol (5) as a model substrate, a phenolic monomer
that was actually generated during DAOF of birch sawdust. In
this case also, the conversion increased considerably with the
introduction of the reducing agent (Fig. S7†). Similarly, a
small amount of 4-propyl guaiacol (6) was formed, pointing
again to the ability of dithionite to saturate the alkylic side
chains in monophenolics.
Fig. 4 Effect of Na2S2O4 on the conversion of 2-phenoxy-1-pheny-
lethanol. (a) Molecular weight distribution of the products determined
by GPC and (b) conversion and yield of monomers determined by
GC-MS/FID. The reactions were performed in duplicate, treating 0.02 g
of 2-phenoxy-1-phenylethanol in a mixture of n-butanol and water
(120 mL, 50% v/v), at 250 °C, under 30 bar of N2 (introduced at ambient
temperature), for a duration of 3 hours. The arrows in the figure indicate
the reference axis.
these are normally found in native lignin structures,3 and were
also observed in marginal amounts in blank organosolv experi-
ments. Thus, to inspect whether the reducing agent is respon-
sible for the cleavage of the terminal hydroxyl group in mono-
lignols’ alkylic side chains, an analogous set of experiments
was carried out using 3-phenyl-1-propenol (7) as a model sub-
strate. When no dithionite was added to the reaction mixture,
7 reacted with n-butanol to form 3-butoxy-propenyl benzene
(Fig. S8†). In contrast, in the presence of the reducing agent,
the ether was not formed and the conversion of 7 to 3-phenyl-
1-propanol (8) was observed, highlighting again the reducing
effect of dithionite on the alkyl chain. Interestingly, no other
monomers were found, implying that the presence of dithio-
No phenolic monomers with propanol-side chains were
detected in the reaction products from DAOF, even though
Scheme 2 Reductive cleavage of 2-phenoxy-1-phenylethanol in the presence of Na2S2O4 (Step I) and subsequent hydrogenation of styrene to
ethylbenzene (Step II).
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 3268–3276 | 3273

View Article Online
Communication
Green Chemistry
nite does not determine per se the cleavage of the hydroxyl to well-established protocols developed by the National
group in the side chains of phenolic monomers. Otherwise, it Renewable Energy Laboratory (NREL).27–29,50 Namely, dry
would be expected to be depleted by reaction with the alcohol matter and ash contents were determined by drying and
co-solvent (n-butanol); the latter was in fact almost completely ashing biomass. The contents of water- and ethanol-soluble
recovered in the output liquid streams. We surmise that the extractives were measured by subjecting biomass to two con-
loss of terminal hydroxyl groups in monophenolic products secutive extractions with water and ethanol, using a Soxhlet
could occur through dehydration reactions in the conditions extractor. Quantification of structural carbohydrates and lignin
of DAOF. The resulting unsaturated side chains may eventually was performed by subjecting the extracted biomass to a two-
be further reduced by dithionite.
step acid hydrolysis in sulfuric acid.
DAOF experiments were carried out in a 300 mL Parr batch
reactor (Parr Instrument Company, Moline, IL, U.S.): 3 g of
biomass were introduced into the reactor, together with 60 mL
of n-butanol, 60 mL of water and 1 g of sodium dithionite. Air
Conclusions
A dithionite-assisted organosolv fractionation process is pre- in the reactor was displaced by purging with N2, and then the
sented, for the concomitant production of cellulosic pulp and reactor was pressurized with 30 bar of N2. The impeller speed
the reductive conversion of lignin into phenolic monomers was set to 750 rpm and the temperature was increased at a rate
and short oligomers. The process relies on the treatment of of about 10 °C min−1, up to 200 °C. Once the temperature set-
lignocellulosic biomass in an equivolumetric mixture of point was reached, the mixture was left to react for a duration
n-butanol and water, in the presence of sodium dithionite as a of 3 hours. Subsequently, the reactor was cooled down to
reducing agent. The immiscibility of n-butanol and water ambient temperature by letting water flow through the cooling
under ambient conditions facilitates product separation. coil and depressurized, and its content was collected.
Highly processable cellulose is recovered in the solid, while Centrifugation of the product mixture led to the recovery of a
lignin and humins migrate to the organic phase and non-con- solid and a liquid fraction. The solid pulp obtained from the
densed carbohydrate derivatives are collected in the aqueous process was washed twice with n-butanol (40 mL) and water
phase. Importantly, as compared to classical organosolv con- (40 mL) to remove polar and apolar components weakly
ditions, the dithionite-assisted process reaches enhanced adsorbed to the solid matrix, before being dried at 60 °C to a
lignin depolymerization efficiency and much higher phenolic constant weight. Thereafter, the washing solvents were added
monomer yields, potentially offering an economically viable to the liquid fraction, and the mixture was filtered to eliminate
alternative to conventional pulping processes (see the ESI† for residual solid particles. The filtrate was transferred to a separ-
a simplified economic assessment). Apart from the monomers ating funnel, where it separated into an organic phase and an
classically found in similar processes, the formation of desas- aqueous liquid phase, which were subsequently collected.
pidinol was highlighted and explained by the action of (hemi)
cellulose derivatives formed during processing, such as formic were determined according to NREL protocols.27,28 The solid
acid. fraction was analyzed for carbohydrates and lignin, following
The dry matter and ash contents of all the isolated fractions
The study of reactions with model compounds revealed that the same two-step acid hydrolysis procedure used for biomass
sodium dithionite favors the reductive cleavage of β-O-4 bonds characterization. In addition, the pulp was characterized via
and the hydrogenation of alkylic side chains in the produced X-ray diffraction analysis, to determine cellulose crystallinity
monomers, without, however, suppressing totally the occur- (measured according to Segal’s method),30 via scanning elec-
rence of repolymerization reactions.
tron microscopy analysis, via viscosimetric analysis to deter-
While the operating conditions should be further optimized mine the degree of polymerization of cellulose (according to
to exploit the full potential of the method, this first study the NF G 06-037 norm), and via enzymatic saccharification,
shows that dithionite-assisted organosolv fractionation (DAOF) performed following NREL procedures.51
is an attractive process for implementing lignin valorization
Portions of the organic and aqueous liquid fractions were
during pulping, in the perspective of developing an integrated dried under nitrogen flow and subjected to a three-fold liquid–
biorefinery approach that does not require the use of a catalyst liquid extraction with dichloromethane and water, to isolate
or hydrogen gas to achieve partial depolymerization of lignin.
lignin derivatives from more polar products. The three di-
chloromethane fractions were mixed and then dried under
vacuum to yield a viscous brown lignin oil that was sub-
sequently analyzed via GPC to inspect its MWD. In order to
assess the phenolic monomer composition of lignin oil, a
Experimental section
A complete list of the materials used throughout the work and weighed amount of 2-isopropylphenol (internal standard) was
a detailed description of the experimental procedures are avail- added to the dichloromethane extract, which was then ana-
able in the ESI.†
lyzed via GC-MS and GC-FID.
Silver birch (Betula pendula) wood was collected in Belgium,
Furthermore HPLC-RID analysis of the organic and
milled (particle size ≤2 mm) and dried in air under ambient aqueous fractions was performed to quantify their content of
conditions. The biomass composition was assessed according mono- and oligosaccharides, polyols and organic acids, as well
3274 | Green Chem., 2021, 23, 3268–3276
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Communication
as to determine the recovery of the butanol co-solvent in the GC-MS/FID methods. D. Magnin is acknowledged for perform-
liquid fractions. ing SEM analyses. E. Devos is acknowledged for performing
The presence of derivatives of sodium dithionite in the ICP-AES analyses.
output streams obtained after fractionation was assessed by
determination of the sulfur content in each fraction via
ICP-AES analysis.
Notes and references
Milled wood lignin was prepared according to a well-estab-
lished procedure.52 Birch sawdust underwent Soxhlet extrac-
tion as described above and then was milled (800 rpm,
12 hours) using an Emax ball mill (Retsch, Haan, Germany) to
yield a fine powder. Milled wood was dispersed in aqueous
dioxane (96% v/v) for 48 hours, recycling the wood into fresh
solvent after 24 hours. The vacuum-dried extract was dissolved
in acetic acid (90% v/v), precipitated in water and air dried,
before being dissolved again in 1,2-dichloroethane–ethanol
(2 : 1 v/v), precipitated in ethyl ether and dried to yield MWL.
The isolated lignin was then subjected to DAOF as previously
illustrated for lignocellulosic biomass. For an optimal reprodu-
cibility of the reaction conditions between experiments with
biomass and MWL, the amount of MWL employed equaled
that of AIL contained in the biomass (about 0.63 g). The treat-
ment of MWL in the presence of formic acid was carried out
with an amount of formic acid equal to that obtained after
DAOF of raw biomass (about 0.11 g).
1 A. K. Chandel, V. K. Garlapati, A. K. Singh, F. A. F. Antunes
and S. S. da Silva, Bioresour. Technol., 2018, 264, 370–381.
2 M. V. Galkin and J. S. M. Samec, ChemSusChem, 2016, 9,
1544–1558.
3 R. Rinaldi, R. Jastrzebski, M. T. Clough, J. Ralph,
M. Kennema, P. C. A. Bruijnincx and B. M. Weckhuysen,
Angew. Chem., Int. Ed., 2016, 55, 8164–8215.
4 W. Schutyser, T. Renders, S. Van den Bosch,
S.-F. Koelewijn, G. T. Beckham and B. F. Sels, Chem. Soc.
Rev., 2018, 47, 852–908.
5 T. I. Korányi, B. Fridrich, A. Pineda and K. Barta, Molecules,
2020, 25, 2815.
6 T. Renders, S. Van den Bosch, S.-F. Koelewijn, W. Schutyser
and B. F. Sels, Energy Environ. Sci., 2017, 10, 1551–1557.
7 M. M. Abu-Omar, K. Barta, G. T. Beckham,
J. S. Luterbacher, J. Ralph, R. Rinaldi, Y. Román-Leshkov,
J. S. M. Samec, B. F. Sels and F. Wang, Energy Environ. Sci.,
2021, 262–292.
8 Y. M. Questell-Santiago, M. V. Galkin, K. Barta and
J. S. Luterbacher, Nat. Rev. Chem., 2020, 4, 311–330.
9 X. Luo, Y. Li, N. K. Gupta, B. Sels, J. Ralph and L. Shuai,
Angew. Chem., 2020, 132, 11800–11812.
The reactions with model compounds were carried out
using an identical set-up to that adopted for DAOF experi-
ments with biomass. A weighed amount of the model com-
pound (0.02 g for 2-phenoxy-1-phenylethanol, 0.05 g for 4-pro-
penyl guaiacol and 3-phenyl-1-propenol) was introduced into
the reactor, together with 60 mL of n-butanol, 60 mL of water 10 T. Renders, G. Van den Bossche, T. Vangeel, K. Van Aelst
and a chosen amount of sodium dithionite (between 0 and and B. Sels, Curr. Opin. Biotechnol., 2019, 56, 193–201.
12 molar equivalents). Subsequently, the reaction and the ana- 11 P. Ferrini, C. A. Rezende and R. Rinaldi, ChemSusChem,
lysis of the products were performed as described above. For 2016, 9, 3171–3180.
the treatment of 2-phenoxy-1-phenylethanol, a higher reaction 12 S. Van den Bosch, W. Schutyser, R. Vanholme, T. Driessen,
temperature of 250 °C was adopted, to promote the cleavage of
the dimer.48
S.-F. Koelewijn, T. Renders, B. De Meester, W. J. J. Huijgen,
W. Dehaen, C. M. Courtin, B. Lagrain, W. Boerjan and
B. F. Sels, Energy Environ. Sci., 2015, 8, 1748–1763.
13 T. Parsell, S. Yohe, J. Degenstein, T. Jarrell, I. Klein,
E. Gencer, B. Hewetson, M. Hurt, J. I. Kim, H. Choudhari,
B. Saha, R. Meilan, N. Mosier, F. Ribeiro, W. N. Delgass,
C. Chapple, H. I. Kenttämaa, R. Agrawal and M. M. Abu-
Omar, Green Chem., 2015, 17, 1492–1499.
Conflicts of interest
There are no conflicts to declare.
14 M. V. Galkin, A. T. Smit, E. Subbotina, K. A. Artemenko,
J. Bergquist, W. J. J. Huijgen and J. S. M. Samec,
ChemSusChem, 2016, 9, 3280–3287.
Acknowledgements
This work was funded by an FSR grant (Fonds spéciaux de 15 A. Deneyer, E. Peeters, T. Renders, S. Van den Bosch,
recherche) from Université catholique de Louvain. D. P. D.
thanks the Francqui Foundation for his Francqui Research
Professor Chair. K. V. A. acknowledges funding through
N. Van Oeckel, T. Ennaert, T. Szarvas, T. I. Korányi,
M. Dusselier and B. F. Sels, Nat. Energy, 2018, 3, 969–
977.
FWO-SBO project BioWood and Catalisti-SBO project 16 E. Cooreman, T. Vangeel, K. Van Aelst, J. Van Aelst,
NIBCON. B. S. acknowledges funding from internal funds of
J. Lauwaert, J. W. Thybaut, S. Van den Bosch and B. F. Sels,
KU Leuven (IDN project – FFASSD) and EoS project BIOFACT
Ind. Eng. Chem. Res., 2020, 59, 17035–17045.
of financial support for biorefinery research. F. Devred is 17 S. Van den Bosch, T. Renders, S. Kennis, S.-F. Koelewijn,
acknowledged for performing the XRD analyses. M. Leclercq
and T. Nicolay are acknowledged for their help with assem-
bling the experimental setup and with the development of
G. Van den Bossche, T. Vangeel, A. Deneyer, D. Depuydt,
C. M. Courtin, J. M. Thevelein, W. Schutyser and B. F. Sels,
Green Chem., 2017, 19, 3313–3326.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 3268–3276 | 3275

View Article Online
Communication
Green Chemistry
18 G. E. Klinger, Y. Zhou, P. Hao, J. Robbins, J. M. Aquilina, 33 Y. Liao, B. O. de Beeck, K. Thielemans, T. Ennaert,
J. E. Jackson and E. L. Hegg, ChemSusChem, 2019, 12,
4775–4779.
J. Snelders, M. Dusselier, C. M. Courtin and B. F. Sels, Mol.
Catal., 2020, 487, 110883.
19 G. E. Klinger, Y. Zhou, J. A. Foote, A. M. Wester, Y. Cui, 34 S. D. Shinde, X. Meng, R. Kumar and A. J. Ragauskas,
M. Alherech, S. S. Stahl, J. E. Jackson and E. L. Hegg,
ChemSusChem, 2020, 13, 4394–4399.
20 E. Sjöström, Wood chemistry: fundamentals and applications,
Academic Press, San Diego, 2nd edn, 1993.
Green Chem., 2018, 20, 2192–2205.
35 Z. Cheng, J. L. Everhart, G. Tsilomelekis, V. Nikolakis,
B. Saha and D. G. Vlachos, Green Chem., 2018, 20, 997–
1006.
21 H. U. Suess, Pulp bleaching today, De Gruyter, Berlin; 36 M. V. Galkin and J. S. M. Samec, ChemSusChem, 2014, 7,
New York, 2010. 2154–2158.
22 A. Orebom, D. Christian, J. Löfstedtl and J. Samec, 37 N. Paksung and Y. Matsumura, Ind. Eng. Chem. Res., 2015,
Depolymerisation of lignin, WO 2015199608, 2015. 54, 7604–7613.
23 N. E. Thornburg, M. B. Pecha, D. G. Brandner, M. L. Reed, 38 C. Li, M. Zheng, A. Wang and T. Zhang, Energy Environ.
J. V. Vermaas, W. E. Michener, R. Katahira, T. B. Vinzant, Sci., 2012, 5, 6383–6390.
T. D. Foust, B. S. Donohoe, Y. Román-Leshkov, 39 Q. Jing and X. Lü, Chin. J. Chem. Eng., 2007, 15, 666–
P. N. Ciesielski and G. T. Beckham, ChemSusChem, 2020,
13, 4495–4509.
24 Y. Kawamata, T. Yoshikawa, Y. Nakasaka, Y. Koyama,
669.
40 G. Garrote, H. Domínguez and J. C. Parajó, Holz Roh-
Werkst., 2001, 59, 53–59.
E. Fumoto, S. Sato, T. Tago and T. Masuda, J. Jpn. Pet. Inst., 41 N. Akiya and P. E. Savage, Chem. Rev., 2002, 102, 2725–
2019, 62, 37–44. 2750.
25 Q. Schmetz, H. Teramura, K. Morita, T. Oshima, A. Richel, 42 P. Ferrini and R. Rinaldi, Angew. Chem., Int. Ed., 2014, 53,
C. Ogino and A. Kondo, ACS Sustainable Chem. Eng., 2019,
7, 11069–11079.
26 T. Renders, E. Cooreman, S. Van den Bosch, W. Schutyser,
8634–8639.
43 A. Toledano, L. Serrano, A. Pineda, A. A. Romero, R. Luque
and J. Labidi, Appl. Catal., B, 2014, 145, 43–55.
S.-F. Koelewijn, T. Vangeel, A. Deneyer, G. Van den 44 A. Toledano, L. Serrano, A. M. Balu, R. Luque, A. Pineda
Bossche, C. M. Courtin and B. F. Sels, Green Chem., 2018,
20, 4607–4619.
27 A. Sluiter, B. Hames, D. Hyman, C. Payne, R. Ruiz,
and J. Labidi, ChemSusChem, 2013, 6, 529–536.
45 A. Toledano, L. Serrano, J. Labidi, A. Pineda, A. M. Balu
and R. Luque, ChemCatChem, 2013, 5, 977–985.
C. Scarlata, J. Sluiter, D. Templeton and J. Wolfe, 46 S. Liang and C. Wan, Waste Biomass Valorization, 2017, 8,
Determination of Total Solids in Biomass and Total Dissolved 393–400.
Solids in Liquid Process Samples, National Renewable 47 C. Xu, R. A. D. Arancon, J. Labidi and R. Luque, Chem. Soc.
Energy Laboratory, Golden, CO, USA, 2008. Rev., 2014, 43, 7485–7500.
28 A. Sluiter, B. Hames, R. Ruiz, C. Scarlata, J. Sluiter and 48 M. R. Sturgeon, S. Kim, K. Lawrence, R. S. Paton,
D. Templeton, Determination of Ash in Biomass, National
S. C. Chmely, M. Nimlos, T. D. Foust and G. T. Beckham,
Renewable Energy Laboratory, Golden, CO, USA, 2005.
ACS Sustainable Chem. Eng., 2014, 2, 472–485.
29 D. Crocker, D. Templeton, J. Sluiter, C. Scarlata, R. Ruiz, 49 M. R. Sturgeon, M. H. O’Brien, P. N. Ciesielski, R. Katahira,
B. Hames and A. Sluiter, Determination of Structural
Carbohydrates and Lignin in Biomass, National Renewable
Energy Laboratory, Golden, CO, USA, 2012.
J. S. Kruger, S. C. Chmely, J. Hamlin, K. Lawrence,
G. B. Hunsinger, T. D. Foust, R. M. Baldwin, M. J. Biddy
and G. T. Beckham, Green Chem., 2014, 16, 824–835.
30 L. Segal, J. J. Creely, A. E. Martin and C. M. Conrad, Text. 50 A. Sluiter, R. Ruiz, C. Scarlata, J. Sluiter and D. Templeton,
Res. J., 1959, 29, 786–794.
Determination of Extractives in Biomass, National Renewable
31 Q. Sun, M. Foston, X. Meng, D. Sawada, S. V. Pingali,
Energy Laboratory, Golden, CO, USA, 2005.
H. M. O’Neill, H. Li, C. E. Wyman, P. Langan, 51 M. Selig, N. Weiss and Y. Ji, Enzymatic Saccharification
A. J. Ragauskas and R. Kumar, Biotechnol. Biofuels, 2014, 7,
of Lignocellulosic Biomass, National Renewable Energy
150.
Laboratory, Golden, CO, USA, 2008.
32 W.-C. Tu, L. Weigand, M. Hummel, H. Sixta, A. Brandt- 52 J. R. Obst and T. K. Kirk, in Methods in Enzymology,
Talbot and J. P. Hallett, Cellulose, 2020, 27, 4745–4761. Elsevier, 1988, vol. 161, pp. 3–12.
3276 | Green Chem., 2021, 23, 3268–3276
This journal is © The Royal Society of Chemistry 2021