Tandem catalytic depolymerization of lignin by water-tolerant Lewis acids and rhodium complexes

Article abstract of DOI:10.1002/cssc.201600683

Lignin is an attractive renewable feedstock for aromatic bulk and fine chemicals production, provided that suitable depolymerization procedures are developed. Here, we describe a tandem catalysis strategy for ether linkage cleavage within lignin, involving ether hydrolysis by water-tolerant Lewis acids followed by aldehyde decarbonylation by a Rh complex. In situ decarbonylation of the reactive aldehydes limits loss of monomers by recondensation, a major issue in acid-catalyzed lignin depolymerization. Rate of hydrolysis and decarbonylation were matched using lignin model compounds, allowing the method to be successfully applied to softwood, hardwood, and herbaceous dioxasolv lignins, as well as poplar sawdust, to give the anticipated decarbonylation products and, rather surprisingly, 4-(1-propenyl)phenols. Promisingly, product selectivity can be tuned by variation of the Lewis-acid strength and lignin source.

Full text of DOI:10.1002/cssc.201600683

DOI: 10.1002/cssc.201600683
Communications
Tandem Catalytic Depolymerization of Lignin by Water-
Tolerant Lewis Acids and Rhodium Complexes
[a]
[a]
[b]
[b]
Robin Jastrzebski, Sandra Constant, Christopher S. Lancefield, Nicholas J. Westwood,
[a]
[a]
Bert M. Weckhuysen, and Pieter C. A. Bruijnincx*
[
2a]
Lignin is an attractive renewable feedstock for aromatic bulk
and fine chemicals production, provided that suitable depoly-
merization procedures are developed. Here, we describe
a tandem catalysis strategy for ether linkage cleavage within
lignin, involving ether hydrolysis by water-tolerant Lewis acids
followed by aldehyde decarbonylation by a Rh complex. In situ
decarbonylation of the reactive aldehydes limits loss of mono-
mers by recondensation, a major issue in acid-catalyzed lignin
depolymerization. Rate of hydrolysis and decarbonylation were
matched using lignin model compounds, allowing the method
to be successfully applied to softwood, hardwood, and herba-
ceous dioxasolv lignins, as well as poplar sawdust, to give the
anticipated decarbonylation products and, rather surprisingly,
nins more recalcitrant to further upgrading. Exactly this pro-
pensity to structural change, compounded by the structural
complexity of lignin, hampers targeted (catalytic) lignin depoly-
merization, making its efficient conversion to a small number
of valuable products extremely challenging. Some selective de-
[
4]
polymerization strategies have nonetheless been reported,
methods which typically rely on (excess) stoichiometric re-
agents to form stable end products before significant recon-
densation can occur. In addition, such strategies are often de-
veloped on lignin model compounds and translating the
chemistry to actual lignins can prove particularly difficult.
Illustrative of the challenges listed above is lignin ether
cleavage by acidolysis, one of the oldest catalytic methods for
[5]
4-(1-propenyl)phenols. Promisingly, product selectivity can be
lignin depolymerization, typically giving low yields of mono-
meric species. Indeed, acid-induced deformylation and hydroly-
sis of a b-O-4 ether result in styryl ether and aldehyde forma-
tion, both of which are highly susceptible to aldol condensa-
tuned by variation of the Lewis-acid strength and lignin
source.
[6]
tion, affording higher molecular weight products (Figure 1).
Lignin is the only aromatic biomass component of substantial
abundance and an ideal candidate for renewable aromatic
Only the fraction of the ethers that does not undergo deformy-
lation gives rise to a mixture of the relatively stable Hibbert’s
[
1]
[7]
bulk and fine chemicals production. Native lignins consist pri-
ketones. Deuss et al. recently demonstrated that loss of mon-
marily of syringyl (S), guaiacyl (G), and p-hydroxyphenyl (H)-
omers to recondensation could be limited by stoichiometrically
trapping intermediate aldehydes formed in triflic acid-catalyzed
[
2]
containing units, with structure varying with plant species
and isolation method: grasses and softwoods typically are rich
in H and/or G and hardwoods are rich in S. Different linkages
are found in the macromolecule, of which up to 60% are b-O-
[8]
lignin depolymerization, for instance by acetalization with
1,2-ethanediol. Preferably, a trapping strategy does not use
sacrificial trapping reagents, is catalytic, and yields valuable ar-
omatic monomers. Catalytic aldehyde decarbonylation is par-
ticularly promising in this respect, but also highly challenging
[
3]
4
ethers. Isolation invariably changes the structure, with new
carbon–carbon bonds forming by recondensation at the ex-
[9]
pense of the ether bonds (b-O-4 in particular), resulting in lig-
with only two known systems showing reasonable efficiency.
Decarbonylation was previously also attempted in the triflic
[a] Dr. R. Jastrzebski, Dr. S. Constant, Prof. Dr. B. M. Weckhuysen,
Dr. P. C. A. Bruijnincx
Inorganic Chemistry and Catalysis
Debye Institute for Nanomaterials Science
Utrecht University
Universiteitsweg 99, 3584 CG Utrecht (The Netherlands)
E-mail: p.c.a.bruijnincx@uu.nl
[b] Dr. C. S. Lancefield, Prof. Dr. N. J. Westwood
School of Chemistry and Biomedical Science Research Complex
University of St. Andrews and EaStCHEM
North Haugh, St. Andrews, Fife KY16 9ST (United Kingdom)
Supporting Information and the ORCID identification number(s) for the
author(s) of this article can be found under http://dx.doi.org/10.1002/
cssc.201600683.
ꢀ
2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA.
This is an open access article under the terms of the Creative Commons
Attribution-NonCommercial-NoDerivs License, which permits use and
distribution in any medium, provided the original work is properly cited,
the use is non-commercial and no modifications or adaptations are
made.
Figure 1. Lewis acid-catalyzed cleavage of lignin, here represented as b-O-4
fragments, combined with Rh-catalyzed decarbonylation prevents loss of
monomers to recondensation.
ChemSusChem 2016, 9, 1 – 7
1
ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Communications
acid-catalyzed depolymerization, but only with limited success:
whereas the decarbonylation worked on model compounds,
the decarbonylation rate could not compete effectively with
the condensation side-reactions during actual lignin depolyme-
Table 1. Results of the tandem catalytic cleavage of b-O-4 model com-
pounds 1a–1d and major products observed.
[a]
[
8]
rization. Another recent example of a trapping strategy made
use of a combination of a Lewis acid metal triflate and Pd/C,
with the latter catalyst serving to hydrogenate reactive inter-
mediates formed by Lewis acid-catalyzed elimination reac-
[
10]
tions. However, this method was not applied to actual lignin.
We here report an innovative and highly efficient tandem
catalytic approach involving lignin hydrolysis/decarbonylation
(
Figure 1), using water-tolerant Lewis acids [M(OTf) , M=Sc, Y,
3
[
11]
In, Ga, and Yb]. Rh-catalyzed decarbonylation then effective-
ly scavenges the aldehyde hydrolysis products to afford the
desired decarbonylated 4-methylphenols in reactions with
model compounds, various dioxasolv lignins, as well as poplar
sawdust. Remarkably, a second alternative trapping pathway is
observed with the actual lignins, leading instead to the forma-
tion of 4-(1-propenyl)phenols (Figure 1), which are attractive
high-value/low-volume lignin products. Crucially, we show that
the trapping pathway can be controlled by variation of the
Lewis-acid strength. The tandem catalytic reaction was first de-
[
b]
[c]
Entry
b-O-4
model
LA
T
[8C]
Conv.
[%]
Product yield [%]
[mol%]
2
3
4
5
7
1
2
3
4
5
6
7
8
9
1a
1a
1b
1b
1b
1b
1b
1b
1c
1c
1d
10
200
200
200
175
175
175
175
175
175
175
175
57
100
96
99
99
92
99
98
97
24
-
47
63
77
81
85
41
84
61
71
12
88
–
–
–
–
–
3
–
–
3
8
–
2
–
–
–
–
44
–
25
4
16
–
–
9
1
17
4
–
31
–
–
–
3
[
d]
10
10
10
5
65
37
66
36
5
45
67
7
veloped on b-O-4 model compounds with Sc(OTf) as the
3
[
12]
2.5
Lewis acid
in combination with an in situ prepared
[
e]
5
[
Rh(cod)Cl] /dppp decarbonylation catalyst (cod=1,5-cyclooc-
[e]
2
2.5
5
2.5
5
[9a]
tadiene, dppp=bis-1,3-(diphenylphosphino) propane). A 9:1
mixture of 1,4-dioxane and water as the solvent ensured good
hydrolysis activity, without the Rh catalyst being negatively im-
pacted by the excess water. Cleavage of the simple model
compound 1a (Table 1, entry 1) gave significant amounts of
the expected products phenol and toluene at 2008C. Phenol is
liberated by hydrolysis of the intermediate enol ether and tolu-
ene is subsequently obtained by decarbonylation of 2-phenyla-
cetaldehyde (6a). Although the latter intermediate was not ob-
served in detectable amounts, 2-phenylethanol was found
10
11
39
100
51
[a] Conditions: 2.4 mmol substrate, 2.5 mol% [Rh(cod)Cl] , 2:1 dppp/Rh,
2
Sc(OTf)
water; yields determined by GC. [b] Lewis acid. [c] Conversion (%).
d] Only Sc(OTf) , no [Rh(cod)Cl] . [e] In(OTf)
3 3
[or In(OTf) ] as listed; 175–2008C, 2 h, 22 mL 9:1 1,4-dioxane/
[
3
2
3
.
observed in about 1% yield. With decarbonylation being too
slow, higher molecular weight species were formed, such as
(12%), as well as dehydrogenated (a-ketone) starting material,
[15]
suggesting that catalytic transfer hydrogenation is also possi-
homoveratryl aldehyde-derived 7b (Scheme 1). Indeed, it is
imperative that the kinetics of hydrolysis and decarbonylation
are carefully matched. Lowering the amount of Sc(OTf)3
(Table 1, entries 4–6), as expected, resulted in an increase in
[
13]
ble if an appropriate hydrogen donor is available. With only
the Lewis acid, full conversion was observed with a reasonable
phenol yield, but neither toluene nor 6a could be detected. In-
stead, higher molecular weight (M ) stilbenes and 2-phenyl-
w
naphthalene (7a) were found instead (Table 1, entry 2). With
[14]
Sc(OTf) also being an excellent aldol condensation catalyst,
3
7
a is likely formed by self-condensation of 6a, followed by in-
tramolecular cyclization. These results clearly demonstrate the
value of the tandem approach, with rapid removal of the reac-
tive aldehydes by decarbonylation preventing undesired side
reactions.
Model compounds 1b–1d allowed the influence of addi-
tional functional groups to be assessed. Cleavage of G–G
model 1b thus afforded both 4-methylveratrole (2b) and
guaiacol (3b) in good yield (Table 1, entry 3; Figure 2). Interest-
ingly, 1b could also be readily cleaved at 1758C; the high
guaiacol yield (81%) at this temperature indicated efficient hy-
drolysis, but the much lower yield of 4-methylveratrole (2b,
Figure 2. The reaction profile of 1b as function of time [determined through
individual reactions; 2.5 mol% Sc(OTf) ] shows the accumulation and subse-
3
38%) showed that the decarbonylation catalyst cannot keep
up with aldehyde formation. Homoveratryl aldehyde (6b) was
quent decrease of 5b as an intermediate. C=conversion.
&
ChemSusChem 2016, 9, 1 – 7
www.chemsuschem.org
2
ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communications
Model compound 1d (erythro/threo mixture) mimics the full
glyceryl b-O-4 backbone and was also cleaved effectively
(
entry 11), albeit with slightly lower yields than with 1b. Defor-
mylation, as observed before during lignin model compound
[6]
acidolysis, is thought to precede hydrolysis and decarbonyla-
tion. The mechanistic insights provided by the model com-
pounds are summarized in Scheme 1. Initial Lewis acid-
catalyzed dehydration of the a-hydroxyl functionality in 1b
forms an isomeric mixture of enol ethers 5b, followed by
a slower Lewis acid-catalyzed hydrolysis step to the corre-
sponding acetaldehyde 6b. A sufficiently active decarbonyla-
tion catalyst then gives the corresponding methyl-substituted
product 2b, whilst in the absence thereof aldol condensation
takes place instead to give 7b. For the first time, an alternative
pathway is also observed, tentatively originating from the enol
ether 5b, leading to the formation of a styrene derivative 4b
by insertion of a formal hydrogen equivalent.
The tandem strategy was then translated to actual lignin.
With maximum monomer yield being approximately quadratic
Scheme 1. Proposed mechanism for the cleavage of b-O-4 models and pos-
sible pathways to styrene formation.
[4c,d]
with the fraction of cleavable bonds,
a poplar sawdust diox-
asolv lignin was isolated to ensure a relatively high abundance
[
4d]
4
-methylveratrole (2b) yield to 66%, with a concomitant drop
of the targeted b-O-4 linkage. Dioxasolv pulping furthermore
ensures that 1) the a-OH remains available for initial dehydra-
tion and 2) the isolated lignin will actually be soluble in (wet)
in condensation products. Such an effect was also observed in
the tandem catalytic system employed by Marks and cowork-
ers, where using a weaker Lewis acid enabled the hydrogena-
tion catalyst to remove reactive intermediates more effectively,
1,4-dioxane, the solvent for the tandem reaction. The M was
w
3
determined to be 3.0ꢁ10 Da by gel permeation chromatogra-
[
10]
leading to higher yields. Further lowering the Sc(OTf) load-
phy (GPC). HSQC NMR analysis (Figure S1) showed 39 b-O-4, 3
b-5, and 5 b-b linkages per 100 aromatic units, with a S/G/H
ratio of 2.1:1.0:0.0. In addition, 16 per 100 aromatic units of
p-hydroxybenzoate units were observed. This value is typical
3
ing (entry 6) proved very revealing mechanistically, with cis
and trans enol ether intermediates (5b) being detected, the
latter being predominant.
[
16]
The time profile of 1b with
[17]
2
.5 mol% Sc(OTf) indeed shows that 5b is an intermediate en
for poplar,
but the actual content is likely lower as such
3
route to decarbonylation (Figure 2). That homoveratryl alde-
hyde (6b) and condensation product 7b are hardly detected,
demonstrates the efficient coupling of the tandem reaction.
The subtle interplay between the Lewis acid and Rh catalyst is
units are systematically overestimated by 2D NMR methods.
Interestingly, Hibbert’s ketone structures could be detected in
the starting lignin, in line with previous organosolv lignin ana-
[2b,18]
lyses.
evidenced by the stronger acid In(OTf) (Table 1, entries 7 and
Generally, no char formation was observed and a very limit-
ed number of volatile products detected (Figure 3). Two of the
3
8), which shows faster hydrolysis, but also accelerated alde-
hyde condensation at the expense of decarbonylation. Notably,
small amounts (ca. 3%) of 3,4-dimethoxystyrene (4b) were also
observed on treatment of 1b with 2.5 mol% Sc(OTf) (entry 6,
3
see below for further discussion). The S–G model 1c also af-
forded excellent conversion and good guaiacol (3b) and 3,4,5-
trimethoxytoluene (2c) yields with 5 mol% Sc(OTf) at 1758C
3
(
entry 9). At only 2.5 mol% Sc(OTf) (entry 10), conversion was
3
only 39%, with only 7% yield of 3,4,5-trimethoxytoluene (2c)
and a 16% yield of the enol ethers 5c. Again, the formation of
8
% of a styrene derivative (3,4,5-trimethoxystyrene, 4c) was
observed.
The formation of the styrene derivatives 4 is remarkable and
has, to the best of our knowledge, not yet been reported.
Whether the styrenes 4b and 4c are formed directly from the
b-O-4 models or from the enol ethers is unclear at this point;
nonetheless there must be a formal hydrogenolysis step in-
volved in the mechanism. The alternative of styrene formation
by transfer hydrogenation of the aldehyde (6b/6c), followed
by acid-catalyzed alcohol dehydration would not be in line
with lower acid concentrations favoring styrene formation.
Figure 3. GC ꢁ GC–MS chromatogram of the volatile components after reac-
tion {300 mg poplar dioxasolv lignin, 0.060 mmol [Rh(cod)Cl]
2
, 0.240 mmol
dppp, 0.120 mmol (ScOTf) ; 1758C, 2 h, 22 mL 1,4-dioxane/water}. Lignin-de-
3
rived products in white; non-lignin derived compounds in yellow.
ChemSusChem 2016, 9, 1 – 7
www.chemsuschem.org
3
ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Communications
major products were indeed the anticipated decarbonylation
products, 4-methylguaiacol (8a) and 4-methylsyringol (9a), de-
rived from S and G units in the lignin, showing that the model
compound chemistry translates well to actual lignin. Surpris-
ingly, significant quantities of iso-eugenol (8b) and 4-(1-prope-
nyl)syringol (9b) were also found, proposed to form analo-
gously to the styrene derivatives 4 seen above for the b-O-4
model compounds. Formation of such compounds has been
observed previously in reactive organosolv pulping, although
rather similar, a comparison of the monomer yield revealed
some interesting trends (Table 2, entries 2–6). Yb(OTf)3 gave
a relatively low amount of decarbonylation products (8a, 9a),
but a relatively large amount of 4-(1-propenyl)phenols (8b,
9b), with a high total amount of monomers. Conversely, with
In(OTf) or Ga(OTf) , the selectivity was found to be the oppo-
3
3
site. Stronger Lewis acids, as expressed by their hydrolysis con-
[21]
stants (pKh),
thus give more decarbonylation products,
whereas the weaker acids favor 4-(1-propenyl)phenol forma-
tion (Figure 4a). The differences seen with the triflic acid run,
showing an even greater preference for the formation of 4-(1-
propenyl)phenols, demonstrate that the metal triflates do not
simply act as precursors to Brønsted acidity by metal triflate
hydrolysis.
[
19]
they could not be obtained from isolated lignin. Compounds
identified in smaller amounts included 4-ethylguaiacol (8d), 4-
ethylsyringol (9d), (4-guaiacyl)acetone (8c), and (4-syringyl)a-
cetone (9c). A potential mechanism of formation for the latter
two may involve (transfer) hydrogenolysis of the terminal g-hy-
droxyl of the original Hibbert’s ketone. Combined,
all identified monomers comprised 9.5 wt% or ap-
proximately 12 mol% of the original lignin intake.
Given a b-O-4 content of 39%, which puts the actual
maximum monomer yield at 20 mol%, the monomer
yield thus corresponds to 50% of the theoretical
yield. GPC and HSQC NMR analysis of the reaction
mixture (Figure S1–2, Table S1) showed the M drop
w
2
to 7.7ꢁ10 Da and all b-O-4 bonds to be successfully
cleaved. Signals for the b-5 fragment also disap-
peared and the amount of b-b linkages was also sig-
nificantly reduced, suggesting that at least the ether
linkages in these fragments are susceptible to cleav-
age as well. No b-b linkage epimerization was ob-
[
20]
served. As a control experiment, lignin depolyme-
rization without the Rh catalyst yielded no mono-
mers, but a large amount of dark precipitate. With
only the Rh complex, some monomers were ob-
served, albeit in very small amounts (Table 2,
[
21]
Figure 4. (a) Lewis acid strength correlates with 4-methylsyringol (9a) yields; (b) yields
of 4-methylphenols versus 4-(1-propenyl)phenols as a function of Sc(OTf)
onstrating the ability to tune product distribution.
3
loading, dem-
entry 1). Possibly the Rh complex is also able to act as (weak)
Likewise, variation of the amount of Lewis acid also allowed
[
21]
Lewis acid providing some hydrolysis activity.
The Lewis acids Yb(OTf) , In(OTf) and Ga(OTf) were also in-
control over selectivity (Table 2, entries 4 and 7–9). With larger
amounts of Sc(OTf) , the relative yield of decarbonylation prod-
3
3
3
3
vestigated, as was triflic acid (3 equiv. compared to the metal
triflates). Whereas the GPC traces of these reactions proved
ucts was highest, while with lower Sc(OTf) concentrations the
3
4-(1-propenyl)phenols were again formed as the major prod-
ucts (Figure 4b). This is in line with the model compound stud-
ies (Table 1) and suggests that for both lignin and the model
compounds the formation of the decarbonylation products
proceeds through an identical mechanism. Increased decar-
bonylation activity with more Lewis acid or a stronger one is
then the result of faster enol ether hydrolysis. In turn, a weak
acid or low loading favors a competitive (reductive) mecha-
nism that leads to formation of the 4-(1-propenyl)phenols. The
selectivity differences between the guaiacyl and syringyl-
derived products are also in line with the model compound
studies: the former tend to form decarbonylation products
more rapidly, whilst the latter form more 4-(1-propenyl)phe-
nols. Maximum yields of decarbonylation products (5.1 wt%)
are obtained with the strongest Lewis acid gallium triflate. On
the other hand, using only 0.030 mmol of scandium triflate the
4-(1-propenyl)phenol yield is the highest, with iso-eugenol
Table 2. Lignin depolymerization with different (amounts of) Lewis acids.
[
b]
Entry Lewis acid
type
Monomers 8 and 9 yield [wt%]
Sum
amt [mmol]
a
b
c
d
[wt%]
1
2
3
4
5
6
7
8
9
0
–
–
0.2
2.5
3.2
4.2
5.0
5.1
4.2
2.2
1.2
0.9
1.2
1.4
5.8
5.6
2.7
0.6
0.3
0.8
7.3
9.0
8.6
9.2
0.0
1.1
1.3
1.8
2.7
3.3
2.2
0.8
0.4
0.4
0.4
0.1
0.6
0.7
0.5
1.0
0.3
0.4
0.7
0.5
0.6
0.9
2.6
10.6
11.4
9.5
9.7
9.4
HOTf
Yb(OTf)
Sc(OTf)
In(OTf)
Ga(OTf)
Sc(OTf)
Sc(OTf)
Sc(OTf)
Yb(OTf)
Yb(OTf)
0.360
0.120
0.120
0.120
0.120
0.240
0.060
0.030
0.030
0.030
3
3
3
3
3
7.9
3
11.8
12.1
11.2
12.4
3
1
3
3
[c]
11
[
0
a] Conditions: 300 mg poplar dioxasolv lignin, 0.060 mmol [Rh(cod)Cl]
.240 mmol dppp, acid; 1758C, 2 h, 22 mL 1,4-dioxane/water (9:1); yields
determined by GC. [b] Sum of all identified monomers (Table S5–13).
c] 4 h reaction.
2
,
(
8b) and 4-(1-propenyl)syringol (9b) being obtained instead in
an overall yield of 9.0 wt%. Using a lower amount of the weak-
[
est Lewis acid [Yb(OTf) ; Table 2, entry 11], led both the highest
3
&
ChemSusChem 2016, 9, 1 – 7
www.chemsuschem.org
4
ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communications
yield of 4-(1-propenyl)phenols (8b and 9b, total yield of
.2 wt%) as well as the highest overall monomer yield
9
(12.4 wt%) after 4 h of reaction.
The tandem approach was also successfully applied to diox-
asolv lignins isolated from brewer’s spent grain (BSP) and pine
sawdust, representative for grasses and softwoods, respective-
ly. HSQC NMR analysis of the BSP lignin revealed a b-O-4 con-
tent of 42 per 100 aromatic units, an S/G/H ratio of 0.5:1.0:0.1
as well as p-coumarate, ferulate, and tricin units, common in
[
17]
grasses (Figure S4, Table S2).
Incomplete solubility in THF
precluded GPC analysis for this lignin. The amount of mono-
mers formed with BSP lignin was lower than with poplar lignin
and some dark precipitate was detected. In line with its higher
G content, a larger amount of guaiacyl-substituted products
were observed (Table 3, entry 1). HSQC analysis of the pine
Figure 5. Three protypical reactions demonstrate the ability to tune the
product distribution of lignin depolymerization. Lignins with an appropriate
S/G ratio affords either S or G (9 or 8) products, whereas the choice of Lewis
acid adjusts the selectivity between methylphenols (a) or 1-propenylphenols
(b).
[
a]
Table 3. Results of the depolymerization of other lignins and sawdust.
Entry Substrate Lewis acid
type
Monomer yield [wt%]
amt [mmol] 8a 9a 8b 9b Sum
[
b]
1
2
3
4
5
BSP
pine
sawdust
sawdust
sawdust
Sc(OTf)
Sc(OTf)
Sc(OTf)
Ga(OTf)
Yb(OTf)
3
3
3
0.120
0.120
0.120
0.120
0.030
1.7 0.7 0.4 0.8 5.2
5.1 0.0 0.6 0.0 6.8
2.1 1.4 0.7 2.9 9.8
2.4 1.6 0.3 1.4 7.1
0.0 0.1 1.6 3.0 7.6
lar monomer yield might be a result of condensation reactions
of the aromatics with (liberated) sugar compounds. Likewise,
reducing sugars may be competitive substrates, inhibiting the
decarbonylation reaction. Nonetheless, in line with the results
on isolated lignin, changing the Lewis acid co-catalyst again al-
lowed the product distribution to be tuned (Table 3, entries 3–
5): when using Ga(OTf)3 predominantly the decarbonylation
products 4-methylguaiacol (8a) and 4-methylsyringol (9a)
3
[c]
3
[a] Conditions: 300 mg lignin or 2.0 g poplar sawdust, 0.060 mmol
[Rh(cod)Cl] , 0.240 mmol dppp, acid; 1758C, 2 h, 22 mL 1,4-dioxane/water.
[b] Sum of all identified monomers (see Tables S14–18 for complete list-
2
ing). [c] 4 h reaction.
were observed; whereas Yb(OTf) gave a larger amount of the
3
lignin revealed it to consist exclusively of G units, with 29 b-O-
4-(1-propenyl)phenols.
4
, 12 b-5, and 4 b-b linkages per 100 aromatic units (Figure S6,
In conclusion, an efficient tandem catalysis approach for
cleavage of b-O-4 fragments in lignin models and (isolated) lig-
nins has been developed. This strategy addresses the key chal-
lenge of loss of yield of mono-aromatics owing to lignin recon-
densation. Building on the mechanistic insight provided by the
model compound studies, the anticipated decarbonylation
products could be obtained in good yields when the rate of
both reactions was appropriately matched. Importantly, the se-
lectivity towards either 4-methylphenols or 4-(1-propenyl)phe-
nols can be controlled by varying the amount and strength of
the Lewis acid catalyst. By using a lignin with an appropriate
S/G ratio and selection of the right Lewis acid, each of the four
major products can then be specifically targeted. Trapping of
reactive intermediates in catalytic lignin depolymerization is
thus shown to be a very versatile approach. Further detailed
mechanistic studies on the remarkable cleavage of the b-O-4
aryl alkyl ethers to 4-(1-propenyl)phenols are currently under
way.
Table S3). Correspondingly, after reaction only guaiacyl-derived
products are observed, further decreasing the complexity of
the product mixture. The lower total monomer yield is in line
with the lower b-O-4 content. On the other hand, at 5.1 wt%
of 4-methylguaiacol (8a), the highest single product yield for
a decarbonylation product is observed with this lignin. Conse-
quently, lignin depolymerization selectivity can be tuned in
two dimensions (Figure 5): by selection of a biomass source
with an appropriate S/G ratio and by selection of the appropri-
ate (amount of) Lewis acid catalyst.
Reactive lignin upgrading directly on the biomass has re-
[
18,22]
cently gained interest,
more recalcitrant lignins by recondensation.
yields of small aromatics have been obtained with this ‘lignin-
as this avoids the generation of
[
22a]
Indeed, high
[
22]
first’ approach. Our reaction was therefore also performed
directly on ball-milled poplar sawdust (Klason lignin 22.6 wt%).
Gratifyingly, under standard conditions using 2.0 g sawdust,
lignin depolymerization products identical to those from the
isolated lignins were obtained, as well as sugar-derived furfural
and hydroxymethylfurfural. The monomer yield obtained with
the whole biomass (accounting for the lignin content) is com-
parable to that observed for the isolated lignin (Table 3, en-
tries 3–5). This is somewhat surprising as, based on the higher
amount of b-O-4 linkages in the unprocessed biomass, one
would expect larger amounts of liberated monomers. The simi-
Acknowledgements
P.C.A.B. thanks the Netherlands Organization for Scientific Re-
search (NWO) for a Vernieuwingsimpuls Veni Grant. The CatchBio
program is acknowledged for financial support. N.J.W. thanks
EPSRC for funding by the EP/J018139/1, EP/K00445X/1 grants and
ChemSusChem 2016, 9, 1 – 7
www.chemsuschem.org
5
ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Communications
an EPSRC Doctoral Prize Fellowship to C.S.L. Saumya Dabral and
Prof. Carsten Bolm (RWTH Aachen) are gratefully acknowledged
for providing an authentic sample of 5b. Dr. Wouter Huijgen
[8] P. J. Deuss, M. Scott, F. Tran, N. J. Westwood, J. G. de Vries, K. Barta, J.
Am. Chem. Soc. 2015, 137, 7456–7467.
[
9] a) M. Kreis, A. Palmelund, L. Bunch, R. Madsen, Adv. Synth. Catal. 2006,
48, 2148–2154; b) E. P. K. Olsen, R. Madsen, Chem. Eur. J. 2012, 18,
16023–16029.
[10] T. L. Lohr, Z. Li, T. J. Marks, ACS Catal. 2015, 5, 7004–7007.
3
(ECN) is acknowledged for Klason lignin determination. SuBiCat
Initial Training Network, Call FP7-PEOPLE-2013-ITN, #607044) is
gratefully acknowledged.
[
11] a) S. Kobayashi, Synlett 1994, 689–701; b) G. K. S. Prakash, T. Mathew,
G. A. Olah, Acc. Chem. Res. 2012, 45, 565–577; c) A. C. Atesin, N. A. Ray,
P. C. Stair, T. J. Marks, J. Am. Chem. Soc. 2012, 134, 14682–14685; d) L.
Yang, Y. Li, P. E. Savage, Ind. Eng. Chem. Res. 2014, 53, 2633–2639.
12] S. Constant, C. Basset, C. Dumas, F. Di Renzo, M. Robitzer, A. Barakat, F.
Quignard, Ind. Crops Prod. 2015, 65, 180–189.
13] a) F. Hutschka, A. Dedieu, M. Eichberger, R. Fornika, W. Leitner, J. Am.
Chem. Soc. 1997, 119, 4432–4443; b) O. Pꢃmies, J.-E. Bꢄckvall, Chem.
Eur. J. 2001, 7, 5052–5058.
Keywords: biomass · decarbonylation · lewis acid · lignin ·
tandem catalysis
[
[
[
1] a) J. Zakzeski, P. C. A. Bruijnincx, A. L. Jongerius, B. M. Weckhuysen,
Chem. Rev. 2010, 110, 3552–3599; b) A. J. Ragauskas, G. T. Beckham,
M. J. Biddy, R. Chandra, F. Chen, M. F. Davis, B. H. Davison, R. A. Dixon, P.
Gilna, M. Keller, P. Langan, A. K. Naskar, J. N. Saddler, T. J. Tschaplinski,
G. A. Tuskan, C. E. Wyman, Science 2014, 344, 1246843; c) R. Rinaldi, R.
Jastrzebski, M. T. Clough, J. Ralph, M. Kennema, P. C. A. Bruijnincx, B. M.
Weckhuysen, Angew. Chem. Int. Ed. 2016, DOI: 10.1002/anie.201510351.
2] a) F. S. Chakar, A. J. Ragauskas, Ind. Crops Prod. 2004, 20, 131–141; b) S.
Constant, H. L. J. Wienk, A. E. Frissen, P. de Peinder, R. Boelens, D. S.
van Es, R. J. H. Grisel, B. M. Weckhuysen, W. J. J. Huijgen, R. J. A. Gosse-
link, P. C. A. Bruijnincx, Green. Chem. 2016, 18, 2651–2665.
[
14] S. Ishikawa, T. Hamada, S. Kobayashi, J. Am. Chem. Soc. 2004, 126,
12236–12237.
[
[
15] G. Brunow, K. Lundquist, Acta Chem. Scand. Ser. B 1984, 38, 323–325.
16] S. Dabral, J. Mottweiler, T. Rinesch, C. Bolm, Green Chem. 2015, 17,
4908–4912.
[
[
[
17] H. Kim, J. Ralph, Org. Biomol. Chem. 2010, 8, 576–591.
[
18] C. S. Lancefield, N. J. Westwood, Green Chem. 2015, 17, 4980–4990.
19] a) M. V. Galkin, S. Sawadjoon, V. Rohde, M. Dawange, J. S. M. Samec,
ChemCatChem 2014, 6, 179–184; b) M. V. Galkin, J. S. M. Samec, Chem-
SusChem 2014, 7, 2154–2158.
20] F. Tran, C. S. Lancefield, P. C. J. Kamer, T. Lebl, N. J. Westwood, Green
Chem. 2015, 17, 244–249.
21] S. Kobayashi, K. Manabe, Acc. Chem. Res. 2002, 35, 209–217.
22] a) P. Ferrini, R. Rinaldi, Angew. Chem. Int. Ed. 2014, 53, 8634–8639;
Angew. Chem. 2014, 126, 8778–8783; b) S. Van den Bosch, W. Schutyser,
R. Vanholme, T. Driessen, S. F. Koelewijn, T. Renders, B. De Meester,
W. J. J. Huijgen, W. Dehaen, C. M. Courtin, B. Largrain, W. Boerjan, B. F.
Sels, Energy Environ. Sci. 2015, 8, 1748–1763; c) T. Parsell, S. Yohe, T. Jar-
rell, I. Klein, E. Gencer, B. Hewetson, H. Choudhari, B. Saha, R. Meilan, N.
Mosier, F. Ribeiro, W. N. Delgass, C. Chapple, H. I. Kenttꢄmaa, R. Agrawal,
M. M. Abu-Omar, Green Chem. 2015, 17, 1492–1499.
[
[
3] J. Ralph, K. Lundquist, G. Brunow, F. Lu, H. Kim, P. F. Schatz, J. M. Marita,
R. D. Hatfield, S. A. Ralph, J. H. Christensen, W. Boerjan, Phytochem. Rev.
[
2004, 3, 29–60.
[
[
4] a) S. Son, F. D. Toste, Angew. Chem. Int. Ed. 2010, 49, 3791–3794; Angew.
Chem. 2010, 122, 3879–3882; b) A. Rahimi, A. Ulbrich, J. J. Coon, S. S.
Stahl, Nature 2014, 515, 249–252; c) E. Feghali, G. Carrot, P. Thuꢂry, C.
Genre, T. Cantat, Energy Environ. Sci. 2015, 8, 2734–2743; d) C. S. Lance-
field, O. S. Ojo, F. Tran, N. J. Westwood, Angew. Chem. Int. Ed. 2015, 54,
258–262; Angew. Chem. 2015, 127, 260–264; e) P. J. Deuss, K. Barta,
Coord. Chem. Rev. 2016, 306, 510–532.
[
[
5] E. Adler, Wood Sci. Technol. 1977, 11, 169–218.
6] a) L. H. Hoo, K. V. Sarkanen, C. D. Anderson, J. Wood Chem. Technol.
1
983, 3, 223–243; b) T. Yokoyama, J. Wood Chem. Technol. 2014, 35,
7–42.
7] M. Kulka, H. Hibbert, J. Am. Chem. Soc. 1943, 65, 1180–1185.
2
Received: May 23, 2016
Published online on && &&, 0000
[
&
ChemSusChem 2016, 9, 1 – 7
www.chemsuschem.org
6
ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

COMMUNICATIONS
Catch and release: A tandem acidoly-
sis/decarbonylation strategy allows for
efficient catalytic scavenging of reactive
lignin fragments avoiding monomer
loss by recondensation. Product selec-
tivity can be tuned by variation of the
Lewis acid and lignin source.
R. Jastrzebski, S. Constant,
C. S. Lancefield, N. J. Westwood,
B. M. Weckhuysen, P. C. A. Bruijnincx*
&& – &&
Tandem Catalytic Depolymerization of
Lignin by Water-Tolerant Lewis Acids
and Rhodium Complexes
ChemSusChem 2016, 9, 1 – 7
www.chemsuschem.org
7
ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ