Reactions of dehydrodiferulates with ammonia

Article abstract of DOI:10.1039/c1ob05677h

Lignocellulosic materials derived from forages and agricultural residues are potential sustainable resources for production of bioethanol or other liquid biofuels. However, the natural recalcitrance of such materials to enzymatic hydrolysis is a major obstacle in their efficient utilization. In grasses, much of the recalcitrance is associated with ferulate cross-linking in the cell wall, i.e., with polysaccharide-polysaccharide cross-linking that results from ferulate dehydrodimerization or with lignin-polysaccharide cross-linking that results from the incorporation of (polysaccharide-bound) ferulates or diferulates into lignin, mainly via free-radical coupling reactions. Many pretreatment methods have been developed to address recalcitrance, with ammonia pretreatments in general, and the AFEX (Ammonia Fiber Expansion) process in particular, among the more promising methods. In order to understand the polysaccharide liberating reactions involved in the cleavage of diferulate cell wall cross-links during AFEX pretreatment, reaction products from five esters modeling the major diferulates in grass cell walls treated under AFEX-like conditions were separated and characterized by NMR and HR-MS. Results from this study indicate that, beyond the anticipated amide products, a range of degradation products derive from an array of cleavage and substitution reactions, and reveal various pathways for incorporating ammonia-based nitrogen into biomass.

Full text of DOI:10.1039/c1ob05677h

View Article Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
Dynamic Article Links
Cite this: Org. Biomol. Chem., 2011, 9, 6779
www.rsc.org/obc
PAPER
Reactions of dehydrodiferulates with ammonia†
Ali Azarpira,*a Fachuang Lua,b and John Ralpha,b
Received 29th April 2011, Accepted 23rd June 2011
DOI: 10.1039/c1ob05677h
Lignocellulosic materials derived from forages and agricultural residues are potential sustainable
resources for production of bioethanol or other liquid biofuels. However, the natural recalcitrance of
such materials to enzymatic hydrolysis is a major obstacle in their efficient utilization. In grasses, much
of the recalcitrance is associated with ferulate cross-linking in the cell wall, i.e., with
polysaccharide–polysaccharide cross-linking that results from ferulate dehydrodimerization or with
lignin–polysaccharide cross-linking that results from the incorporation of (polysaccharide-bound)
ferulates or diferulates into lignin, mainly via free-radical coupling reactions. Many pretreatment
methods have been developed to address recalcitrance, with ammonia pretreatments in general, and the
AFEX (Ammonia Fiber Expansion) process in particular, among the more promising methods. In
order to understand the polysaccharide liberating reactions involved in the cleavage of diferulate cell
wall cross-links during AFEX pretreatment, reaction products from five esters modeling the major
diferulates in grass cell walls treated under AFEX-like conditions were separated and characterized by
NMR and HR-MS. Results from this study indicate that, beyond the anticipated amide products, a
range of degradation products derive from an array of cleavage and substitution reactions, and reveal
various pathways for incorporating ammonia-based nitrogen into biomass.
Lignin, one of the most abundant polymers in nature, is
Introduction
present in the cell walls of all plants. It is an amorphous
and structurally complex polymer produced by combinatorial
oxidative coupling of mainly three monolignols, i.e., p-coumaryl
(normally only as a minor component), coniferyl and sinapyl
alcohols (4-hydroxycinnamyl alcohols with different degrees of
methoxylation ortho to the phenol). These monolignols generate
p-hydroxyphenyl (H), guaiacyl (G) and syringyl (S) units in the
polymer (Fig. 1). These units are asymmetrically joined to each
other and, more importantly, to the growing polymer, or between
polymer units, mostly via b–O–4-, b–5-, b–b-, 5–5-, and 5–O–
4-linkages.9,10 In grasses, ferulates and diferulates may serve as
nucleation or initiation sites for lignification.11
All three hydroxycinnamates, p-coumarate, ferulate and sina-
pate (Fig. 1) are associated with cell wall polysaccharides; p-
coumarate is found mainly acylating lignins, in a form that
does not contribute to the recalcitrance.12 Grasses, in particular,
have relatively high levels of ferulate and low levels of p-
coumarate acylating the primary hemicellulosic polymers, the
(glucurono)arabinoxylans.13–15 In addition to the problem lignin
poses to efficient saccharification, much of the recalcitrance in
grasses is due to cell wall cross-linking that is mediated by ferulates.
Polysaccharide–polysaccharide cross-linking results from ferulate
dehydrodimerization in the wall, and lignin–polysaccharide cross-
linking results from the incorporation of ferulates and diferulates
into the radical cross-coupling polymerization reactions of ligni-
fication, with which they are fully compatible.9,10,16,17 Both cross-
linking mechanisms have been shown to affect the rate and/or
Plants provide substantial renewable sources of polysaccharides
that can be used as feedstocks for liquid biofuel production.
However, in order to make these natural complexes efficiently
accessible to microbes and enzymes it is necessary to use some
kind of pretreatment to overcome their natural resistance to
deconstruction. A number of different pretreatment methods have
been developed. They operate at different pH conditions and
temperatures and can address varying degrees of recalcitrance.1–3
Ammonia-based pretreatments are gaining favor, particularly for
grasses. Ammonia Fiber Expansion (AFEX), in particular, is
one of the more promising physiochemical pretreatments for
lignocellulosic materials having the advantage of retaining the
whole biomass without producing the usual array of compounds
that inhibit saccharification enzymes or are toxic to fermentation
microbes. In the AFEX process, liquid ammonia to biomass
loading is (0.3–2) : 1 (w/w) and water to biomass is (0.6–2) : 1
(w/w) at 60–130 ◦C under pressure (200–400 psi) for 5–30 min.4–8
aDOE Great Lakes Bioenergy Research Center, University of Wisconsin-
Madison, Madison WI 53726, USA. E-mail: azarpira@wisc.edu; Fax: +1
608 265 2904; Tel: +1 608 262 1629
bDepartment of Biochemistry and Wisconsin Bioenergy Initiative, University
of Wisconsin-Madison, Madison WI 53726, USA. E-mail: fachuanglu@
wisc.edu, jralph@wisc.edu; Fax: +1 608 265 2904; Tel: +1 608 890 2429
† Electronic supplementary information (ESI) available: HPLC chro-
matograms of crude reaction mixtures, proposed mechanisms for some of
the reactions, photograph of the pressure vessel, NMR spectra for isolated
compounds and ESI spectra for 2d and 2e. See DOI: 10.1039/c1ob05677h
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 6779–6787 | 6779
©

View Article Online
discuss the chemical reactions occurring during the treatment of
lignin with ammonia, but the information in this field is limited due
to the complex nature of lignin and the possible diverse reactions
occurring under different treatment conditions.22
To obtain a better understanding of the chemical reactions and
structural changes occurring when grass cell walls are treated with
ammonia, we studied milligram-scale ‘AFEX’ pretreatment of
models for the important polysaccharide–polysaccharide cross-
linking structures, the series of five dehydrodiferulates. Along
with anticipated reactions, some rather striking bond-cleavage
reactions were observed, which also hint at mechanisms by which
macromolecular lignin might be depolymerizing. The reactions of
the various dehydrodiferulates are described in this paper.
Results and discussion
Dehydrodiferulates, from here on termed simply diferulates, in
the cell wall result from radical coupling of ferulate monomers
that acylate wall (glucurono)arabinoxylans. Like lignification
reactions, ferulate dehydrodimerization is a combinatorial process,
with ferulates coupling with various regiochemistries, giving rise
to a series of 5 predominant diferulates (Fig. 2).9,10,23,24 Although
the synthesis of all of these diferulates has been described,10,23,25
we needed improved methods to generate larger quantities for
this and other studies. These dehydrodimers are now available
more conveniently from a simpler set of coupling reactions,
requiring little in the way of complex organic synthesis. All of the
diferulate model compounds were synthesized via radical coupling
reactions of ethyl or methyl ferulate initiated by the CuCl(OH)–
tetramethylethylenediamine complex in acetonitrile or hydrogen
peroxide/peroxidase in phosphate buffer as reported elsewhere.26
All compounds were identical to those previously reported.23
From the reaction mixture of ammonia with the acyclic 8–8-
diferulate 1 several compounds, in addition to some unreacted
compound 1, were obtained out of which five major products were
determined, Scheme 1A. Vanillin 1a was the simplest product.
Compound 1b, after purification by HPLC, was identified as an
oxopyrrolidine derivative. The high resolution electrospray ion-
ization mass spectrometry (HRESIMS) peak for 1b corresponded
with the molecular formula C12H14N2NaO4. In the 13C NMR
spectrum, signals from the 3-methoxy-4-hydroxyphenyl moiety at
dC 147.6 (C-3), 146.2 (C-4), 132.9 (C-1), 118.4 (C-6), 115.3 (C-
5), and 109.9 (C-2) indicated the presence of only one aromatic
ring. On the other hand, 13C NMR signals at 60.0 (C-7), 49.3
(C-8), 34.9 (C-8¢) and 174.7 (C-9¢) along with COSY correlations
between dH 2.83 (H-8) and 4.60 (H-7) as well as dH 2.83 (H-
8) with 2.30 (H-8¢a) and 2.44 (H-8¢b) and HMBC correlations
between dH 4.60 (H-7), 2.83 (H-8), 2.30 (H-8¢a) and 2.44 (H-8¢b)
with dC 174.7 (C-9¢) were consistent with a structure containing
an oxopyrrolidine ring. HMBC correlations between dH 4.60 (H-
7) with carbon signals at dC 49.3 (C-8), 118.4 (C-6), 109.9 (C-2)
and 132.9 (C-1) established the connectivity between the two rings.
Compound 1c, isolated from the reaction mixture of 1, showed a
HRESIMS (M + Na)+ signal corresponding with the molecular
formula C20H19NNaO7. 13C NMR signals at dC 47.2 (C-8), 179.7
(C-9), 53.8 (C-8¢), and 174.6 (C-9¢) combined with the HMBC
correlation between the doublet proton at dH 4.88 (H-8¢) with
carbon signals at dC 47.2 (C-8), 179.7 (C-9) and 174.6 (C-9¢) as
well as the COSY correlation between proton signals at dH 3.46
Fig. 1 (a) Primary lignin monomers, (b) generic lignin units, and (c) the
hydroxycinnamates.
extent of enzymatic cell wall saccharification.18 In other words, all
of these cross-linking mechanisms between polysaccharides, with
each other and with lignin, have a negative impact on the enzymatic
bioconversion of cellulose and other wall polysaccharides into
fermentable sugars. One of the reasons that AFEX pretreatment
works so effectively to improve the digestibility of grasses is that
these ferulate (and diferulate) ester bonds are readily cleaved by
ammonia, effectively resulting in an efficient destruction of such
cross-links, and rendering the polysaccharides more accessible.
Cleavage of ferulate ester linkages therefore contributes to the
unusually high extractability of grass lignins and the improved
enzymatic degradability of grass cell walls after mild alkaline
pretreatments.19 A particular advantage of ammonia-based pre-
treatments, and AFEX in particular (which retains the entire
biomass, without requiring its fractionation), is that nutrients for
fermentation are preserved (or even enhanced, due to the increase
of N-content) in the biomass.20
Utilization of plant derived lignins (or lignin derivatives), which
may be generated in significant quantities during pulp and paper
making as well as biomass treatment for biofuel production, has
been the subject of many studies. Lignin is one of the important
substances comprising the organic fraction of soil. Moreover,
introduction of nitrogen into lignin using different chemical reac-
tions has drawn special attention due to the successful application
of such products as humus-like soil fertilizers.21,22 Several articles
6780 | Org. Biomol. Chem., 2011, 9, 6779–6787
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Fig. 3 Proposed structures for compounds 2d and 2e and important
HMBC correlations.
spectrum of 2c, the presence of two (broad) singlets at dH 8.25 (H-
7) and dH 7.63 (H-8¢) and their correlations with carbonyl signal
at dC 168.2 (C-9), in the HMBC spectrum, confirmed the presence
of the amide carbonyl at C-9, i.e., attached to C8.
A polar fraction obtained from the reaction mixture of cyclic
8–8-diferulate 2 by flash chromatography was purified on HPLC
(Luna phenyl-hexyl column) and yielded structural isomers ten-
tatively assigned as 2d and 2e (Fig. 3 and 4). The HRESIMS
spectrum for 2e showed a (M + Na)+ peak corresponding with
the molecular formula C19H15N3NaO5. Both compounds showed
same ESI (M + H)+ and (M - H)- peak masses (ESI spectra on
pages S45 and S46 in the ESI†). Their NMR data are presented in
Table 4. In 2d, the three-bond HMBC correlation between the
proton signal at dH 8.13 (H-7) and the carbonyl signal at dC
168.8 (C-9) showed the presence of a carbonyl group connected
at C-8. Although another HMBC correlation between the proton
signal at dH 8.13 (H-7) with a carbon signal at dC 122.1 (C-8¢)
was observed, the signal for the C-8 quaternary carbon did not
appear in the NMR spectra. The presence of three nitrogen atoms
(realized from the MS spectrum) along with the presence of one
carbonyl and two quaternary carbons at C-8 and C-8¢ (deduced
from the NMR spectra) suggested the presence of the ring C. As
is observed in Table 4, the main differences in NMR data of 2d
and 2e are limited to C-2, C-3, C-4 and C-5 (of ring A), whereas
the data for rings B, C and D are quite similar. The 2D NMR
observations suggested that the methoxyl and hydroxyl group in
ring A of 2d are switched from those in 2e (Fig. 3). This was
deduced from three-bond HMBC correlations in 2d between the
proton signal at dH 7.49 (H-2) with carbon signals at dC 120.9
(C-7), 130.9 (C-6), and 150.8 (C-4) which showed the presence of
the methoxyl group at C-4 of ring A. Also, three-bond HMBC
correlations in 2d between the proton signal at dH 7.12 (H-5) with
Fig. 2 Diferulates from oxidative coupling of ferulate esters.
(H-8) and 4.88 (H-8¢) suggested the presence of the pyrrolidine-
dione moiety. The HMBC correlations between proton signals at
dH 2.74 (H-7a) and 3.05 (H-7b) with carbons at dC 179.7 (C-9)
and 53.8 (C-8¢) established the connectivity of the vanillyl moiety
and the pyrrolidinedione ring at C-8 while the correlation between
dH 3.46 (H-8) and 4.88 (H-8¢) with dC 192.3 (C-7¢) established the
connection of the second aromatic ring at C-8¢. Compound 1d,
obtained as a colorless gum following HPLC purification, was also
identified as a pyrrolidinedione derivative. Its HRESIMS peak (M
+ Na)+ corresponded with the molecular formula C20H20N2NaO6.
The differences between the NMR spectra of 1c and 1d were due
to the presence of an olefinic double bond substituted with an
amino group in 1d. Simultaneous hydrolysis and aminolysis of
compound 1 yielded compound 1e.
Compounds 2a–2e (Scheme 1B, Fig. 3) were isolated from the
reaction mixture of the cyclic 8–8-diferulate 2. Compound 2a
was the result of complete aminolysis of the ester 2. HRESIMS
confirmed a molecular formula of C19H17NNaO5 for both 2b and
2c. Their 13C NMR spectra indicated the presence of only one
1
carbonyl in the molecule. In the H NMR spectrum of 2b, the
presence of two doublets at dH 7.70 (H-7) and dH 7.27 (H-8) and
their correlation in COSY spectra suggested that the carbonyl
moiety is attached to C-8¢. This assignment was supported by the
3-bond HMBC correlation between the proton signal at dH 7.27
(H-8) and the carbonyl signal at dC 171.5 (C-9¢). In the 1H NMR
Fig. 4 HPLC profile of the fraction containing 2d and 2e (see Experi-
mental section for details).
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 6779–6787 | 6781
©

View Article Online
Scheme 1 Reactions of diferulates with liquid ammonia in the presence of water. Isolated yields of the identified products are given.
carbon signals at dC 137.8 (C-7¢), 131.7 (C-1), and 149.6 (C-3)
indicated that the C-3 of ring A was hydroxylated (Fig. 3). In 2e,
on the other hand, three-bond HMBC correlations between the
proton signal at dH 7.68 (H-2) with carbon signals at dC 121.5
(C-7), 130.6 (C-6), and 149.6 (C-4) indicated that the methoxyl
group was at C-3 of ring A. Also, three-bond HMBC correlations
between the proton signal at dH 7.12 (H-5) with carbon signals
at dC 137.0 (C-7¢), 131.5 (C-1), and 150.7 (C-3) further indicated
that the hydroxyl group was at C-4 of ring A in 2e. Currently, we
don’t have an easy way of explaining the formation of isomer 2d
mechanistically, and the structural assignments for both 2d and 2e
remain somewhat tentative.
Feruloyl amide (3a) was obtained via cleavage of the ether bond
in the 8–O–4-linked diferulate 3, Scheme 1C, whereas compounds
3b and 3c were more trivially obtained by aminolysis/hydrolysis.
The HRESIMS spectrum for 3d showed an (M + Na)+ peak
6782 | Org. Biomol. Chem., 2011, 9, 6779–6787
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Table 1 13C NMR chemical shift data for compoundsa
Compounds
Carbon
number
1a
1b
1c
1d
1e
2a
2b
2c
3a
3b
3c
3d
4b
4c
5a
5b
5c
1
127.9 132.9 128.6 127.7 127.5 131.2 128.7 128.2 126.6 123.3 123.5 123.5 123.9 120.9 122.8 118.3 124.7
110.4 109.9 112.9 113.1 113.1 112.1 106.5 107.7 111.0 112.7 112.5 112.7 108.9 109.1 108.7 107.6 109.1
148.4 147.6 147.3 146.6 147.1 147.2 149.2 149.2 148.3 147.5 147.3 148.1 148.1 148.9 149.0 151.1 147.8
154.4 146.2 144.9 144.7 146.7 147.7 147.5 148.7 149.8 148.7 147.7 147.2 148.6 153.3 150.5 159.4 146.3
115.5 115.3 115.3 114.8 114.9 116.2 108.6 108.0 116.0 115.5 115.2 115.3 123.5 123.4 126.5 128.3 124.7
126.5 118.4 121.1 121.5 123.6 126.4 128.0 128.8 122.1 124.3 124.0 124.1 123.4 125.8 125.8 126.8 123.4
2
3
4
5
6
7
190.7 60.0
34.5
47.2
33.6
45.7
131.9 132.4 125.3 125.7 140.2 122.7 122.4 122.3 139.9 145.7 145.6 146.7 139.6
136.5 135.0 121.6 128.5 119.3 140.4 140.3 141.1 117.9 111.4 113.2 109.2 118.7
8
—
—
—
—
55.5
—
—
—
—
—
—
—
—
—
—
—
—
49.3
9
173.8 179.7 178.8 170.8 169.6
—
—
—
55.5
168.2 167.6 164.4 164.3 164.6 167.3 166.9 166.8 167.1 167.1
10
—
—
—
—
—
—
—
—
—
—
—
—
—
55.8
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
59.4
14.3
55.6
—
59.5
14.3
55.8
59.1
14.2
55.2
—
—
11
—
55.4
3-OMe
55.5
—
—
—
—
—
—
55.5
55.8
55.2
55.7
55.0
54.8
54.8
55.6
—
—
—
—
—
—
—
37.4
55.5
1¢
127.4 126.9 127.9 123.1 129.2 131.4
111.7 115.5 112.2 111.7 114.3 113.8
147.4 148.5 147.1 146.2 146.9 147.4
153.5 147.6 146.3 144.9 145.7 146.0
114.8 111.5 115.0 114.9 115.0 115.4
124.9 120.8 122.9 119.5 122.4 122.1
129.5 128.9 140.4
111.0 111.2 110.7
148.5 148.7 147.3
146.4 146.9 143.6
113.2 112.9 112.6
120.8 121.8 118.0
138.9 143.5 52.0
120.9 117.5 44.7
124.0 124.4 124.7
108.8 109.4 109.1
148.7 149.6 147.8
150.5 150.5 146.3
126.4 128.9 124.7
123.8 120.7 123.4
2¢
—
—
—
—
—
—
38.1
3¢
4¢
5¢
6¢
7¢
—
34.9
192.3 156.4 131.2 44.8
53.8 90.8 133.8 47.4
134.1 137.5
133.2 123.0
140.1 51.5
118.1 39.6
139.6
118.7
8¢
9¢
174.7 174.6 178.3 171.8 173.2 171.5
—
—
—
55.6
166.8 167.7 172.7 173.2 173.4 167.3 171.1 167.1
10¢
—
—
—
—
—
55.2
—
—
55.2
—
—
55.3
—
—
55.7
—
—
55.4
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
11¢
—
55.6
3¢-OMe
55.5
55.3
55.7
55.4
55.5
a 125 MHz in DMSO-d6.
Table 2 1H NMR data for compounds 1a–1e and 2a–2ca
Compound
Carbon
number
1a
1b
6.79 (s)
1c
1d
1e
2a
6.84 (s)
2b
2c
2
5
6
7
7.34 (s)
6.93 (d, 8.1) 6.74 (d, 8.1)
7.38 (d, 8.1) 6.63 (d, 8.1)
6.47 (s)
6.24 (s)
7.46 (s)
6.62 (d, 8.3) 6.52 (s)
6.91 (d, 8.3)
7.03 (s)
7.32 (s)
6.91 (s)
—
7.24 (s)
7.39 (s)
—
6.53 (d, 8.0)
6.57 (d, 8.0)
6.52 (d, 7.9)
6.19 (d, 7.9)
H-a: 2.64 (d, 11.2)
—
7.32 (s)
9.72 (s)
4.60 (d, 6.8)
H-a: 2.74 (t, 12.5)
H-b: 3.05 (dd, 13.8, 4.2) H-b: 2.02 (dd, 13.3, 5.1)
7.70 (d, 8.2) 8.25 (s)
8
10
11
—
—
—
2.83 (m)
—
—
3.74 (s)
—
—
3.46 (m)
—
—
3.76 (s)
7.27 (s)
6.77 (d, 8.3)
7.35 (d, 8.3)
—
3.86 (s)
—
—
3.81 (s)
7.15 (s)
6.91 (d, 8.1)
7.10 (d, 8.1)
—
—
—
—
3.64 (s)
7.22 (s)
—
—
—
3.77 (s)
6.74 (s)
7.27 (d, 8.2)
—
—
3.73 (s)
6.87 (s)
—
—
—
3-OMe 3.81 (s)
3.90 (s)
1.98 (d, 1.6)
2¢
5¢
6¢
7¢
8¢
—
—
—
—
—
6.60 (d, 8.3) 6.53 (d, 8.0) 6.84 (d, 7.9) 6.91 (d, 7.9)
6.79 (d, 8.3) 6.21 (d, 8.0) 6.67 (d, 7.9) 6.85 (dd, 7.9, 1.6)
7.05 (s)
—
—
—
4.27 (s)
3.60 (s)
—
—
—
7.63 (s)
H-a: 2.30 (dd, 4.88 (d, 4.6)
16.4, 8.5)
—
H-b: 2.44 (dd,
16.4, 8.5)
—
—
—
10¢
—
—
—
—
—
3.55 (s)
—
—
3.66 (s)
—
—
3.68 (s)
—
—
3.68 (s)
—
—
3.86 (s)
—
—
3.81 (s)
11¢
3¢-OMe
a 500 MHz in DMSO-d6, (multiplicity, J in Hz).
corresponding with the molecular formula C20H23N3NaO6. In 3b,
the NMR spectra showed the presence of an acrylamide side-chain
on the B-ring whereas new signals at dC 104.4 (C-1¢), 52.0 (C-7¢),
44.7 (C-8¢), and 172.7 (C-9¢) in 3d suggested further ammination
of the side-chain. The position of the amino group at C-7¢ was
confirmed by HMBC correlations between proton signals at dH
2.31 (H-8¢), 6.77 (H-6¢) and 7.12 (H-2¢) with the carbon signal at
dC 52.0 (C-7¢) as well as the COSY correlation between protons at
dH 4.13 (H-7¢) with 2.31 (H-8¢).
Opening of the tetrahydrofuran ring in the 8–5-linked diferulate
4 yielded vanillin 1a, with 4b and 4c also isolated from the
reaction mixture, Scheme 1D. The HRESIMS spectrum for 4b
showed a (M + Na)+ peak corresponding with the molecular
formula C12H14N2NaO4. The NMR spectra of 4b showed the
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 6779–6787 | 6783
©

View Article Online
Table 3 1H NMR data for compounds 3a–3d, 4b, 4c, 5a–5ca
Compound
Carbon
number
3a
3b
3c
3d
4b
4c
5a
5b
5c
2
7.11 (s)
6.78 (d, 7.7)
6.97 (d, 7.7)
7.30 (d, 15.7) 7.16 (s)
6.41 (d, 15.7)
—
—
3.79 (s)
—
—
—
—
—
—
—
—
7.27 (d, 1.7)
6.74 (d, 8.3)
7.01 (d, 8.3)
7.27 (s)
6.71 (d, 8.2)
7.02 (d, 8.2)
7.15 (s)
—
—
—
3.61 (s)
7.45 (s)
6.65 (d, 8.1)
7.11 (d, 8.1)
7.28 (s)
6.73 (d, 8.0)
7.03 (d, 8.0)
7.10 (s)
—
—
—
3.59 (s)
7.12 (s)
6.57 (d, 8.1)
6.77 (d, 8.1)
6.99 (d, 1.7)
—
6.89 (d, 1.7)
7.12 (s)
—
6.97 (s)
7.27 (s)
—
7.21 (s)
7.03 (s)
—
7.16 (s)
7.12 (s)
—
6.96 (s)
5
6
7
7.26 (d, 15.7) 7.46 (d, 15.3) 7.56 (d, 15.8) 7.52 (d, 15.5) 7.32 (d, 15.8)
6.37 (d, 15.7) 6.25 (d, 15.3) 6.43 (d, 15.8) 6.20 (d, 15.5) 6.44 (d, 15.8)
—
—
3.79 (s)
—
—
—
—
3.35 (s)
—
8
—
—
—
10
4.12, (q, 7.1) 4.15 (q, 7.1)
1.23, (t, 7.1)
55.6, (s)
—
—
—
—
3.31, (s)
—
—
4.13 (q, 7.1)
1.23 (t, 7.1)
3.71 (s)
6.81 (s)
—
—
—
11
1.24 (t, 7.1)
3.82 (s)
7.07 (s)
—
7.01 (s)
7.33 (d, 15.7) 4.43 (t, 6.8)
6.42 (d, 15.7) 2.68 (d, 6.8)
—
—
3.38 (s)
3-OMe
3.60 (s)
7.29 (d, 1.6)
6.67 (d, 8.3)
7.01 (d, 8.3)
7.32 (d, 15.8) 7.48 (d, 15.8) 4.13 (s, br)
6.51 (d, 15.8) 6.46 (d, 15.8) 2.31 (d, 4.8)
—
—
3.91 (s)
3.86 (s)
7.12 (s)
—
6.96 (s)
7.32 (d, 15.8)
6.44 (d, 15.8)
—
2¢
5¢
6¢
6.99 (s)
7¢
8¢
10¢
—
—
3.92 (s)
—
—
—
—
—
3.71 (s)
11¢
—
—
—
3.86 (s)
3¢-OMe
—
a 500 MHz in DMSO-d6, (multiplicity, J in Hz).
Table 4 1H NMRa and 13C NMRb data for compounds 2d and 2e
2d
2e
Carbon number
dC
dH (multiplicity, J in Hz)
dC
dH (multiplicity, J in Hz)
1
131.7
112.3
149.6
150.8
106.7
130.9
120.9
Xc
—
7.49 (s)
—
—
7.12 (s)
—
8.13 (s)
—
—
—
—
3.69 (s)
—
6.97 (s)
—
—
6.89 (d, 8.0)
6.80 (d, 8.0)
—
—
—
—
—
3.74 (s)
131.5
109.3
150.7
149.6
110.2
130.6
121.5
Xc
—
7.68 (s)
—
—
7.12 (s)
—
8.18 (s)
—
—
—
—
3.93 (s)
—
6.89 (s)
—
—
6.90 (d, 8.0)
6.73 (d, 8.0)
—
—
—
—
—
3.74 (s)
2
3
4
5
6
7
8
9
168.8
—
—
168.9
—
—
10
11
Ring A-OMe
55.1
125.3
114.2
146.9
146.3
115.0
122.6
137.8
122.1
—
55.6
125.6
114.1
146.9
146.2
114.9
122.4
137.0
122.7
—
1¢
2¢
3¢
4¢
5¢
6¢
7¢
8¢
9¢
10¢
—
—
55.5
—
—
55.5
11¢
3¢-OMe
a 500 MHz in DMSO-d6, (multiplicity, J in Hz). b 125 MHz in DMSO-d6. c Missing carbon signal.
presence of a 5-substituted 4-hydroxy-3-methoxyphenyl moiety.
The HMBC correlation between dH 7.26 (H-7) with dC 108.9 (C-
2) and 123.4 (C-6) showed the acrylamide connection at C-1, and
correlation between dH 3.35 (H-8¢) and dC 123.4 (C-6), 123.5 (C-5),
and 148.6 (C-4) showed the connection of an acetamide moiety
at C-5.
Aminolysis/hydrolysis of 5–5-coupled diferulate 5, Scheme 1E,
yielded relatively few compounds, from which 5a, 5b and 5c were
isolated and characterized. In 5b, similarly to in 3d, ammination
of the acrylamide side-chain was observed.
From the outset, we suspected that AFEX pretreatments (or
AFEX-like reactions, e.g., ammoniation) are not quite as simple as
just cleaving esters and making the corresponding amides.27 Some
esters apparently don’t cleave readily, others hydrolyze (rather
than being attacked by the nucleophilic ammonia), and some
structures undergo more extensive degradation than might be
anticipated. Tentative mechanisms have been proposed to show
possible pathways leading to the formed products (ESI, Scheme
S1†). Phenolic units in lignin are converted into quinone methides
◦
at above ~150 C during the alkaline wood pulping process.24
6784 | Org. Biomol. Chem., 2011, 9, 6779–6787
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Similarly, all of the diferulate model compounds 1–5, containing
phenolic hydroxyl groups para to a conjugated side-chain, can
form quinone methide intermediates in basic media; such reactions
might be the key steps in the degradation processes (see the ESI†).
Kinetex C18 (150 ¥ 4.6 mm, 2.6 mm, Phenomenex, USA) columns
at 35 ◦C. The injection volume was 30–50 mL. Aqueous 0.1%
formic acid (v/v, solvent system A) and methanol or acetonitrile
(0.1% formic acid, solvent system B) served as the mobile phase in
a gradient mode with a flow rate of 2 mL min-1. Detection was at
280 nm via a SPD-M20A photodiode array detector. For analysis
of the compounds, the following gradient was mostly applied with
some modifications for each sample: B from 5 to 30% in 20 min,
from 30 to 50% in 5 min, from 50 to 80% in 10 min, and from
80 to 100% in 4 min. Liquid anhydrous ammonia was 99.99%
pure. A 16 ml pressure vessel with a multi-port lid (stainless steel
grade 316) from HEL Inc. USA was used to perform the AFEX
reactions (see the ESI, Fig. S2†).
Conclusions
There are reports on the analysis of lignocellulosic materials
treated with ammonia (in the form of anhydrous ammonia,
ammonium hydroxide, or via ammoxidation, etc.) but their major
interest focused on the analysis of nitrogen content in crude
product mixtures which enabled these products to be considered
as nitrogenous fertilizers.22,28,29 Comprehensive reports on the
isolation and structural elucidation of products from these types of
complex reactions are not evident. Here, under conditions similar
to those in the AFEX process, diferulates were converted to the
expected amides, although acids, amines, and aldehydes were also
produced. The survival of some ferulate esters indicated that the
AFEX conditions used here might not be optimal for cleaving
ferulate esters. Finding nitrogen-containing products other than
the expected amides revealed interesting reaction pathways for
nitrogen incorporation into the degradation products from grass
cell walls and enhance our knowledge of the mechanisms involved
in AFEX pretreatment. The carbon–carbon bond cleavage prod-
ucts resulting from 8–5-diferulate, in particular, imply that more
complicated reactions than just ester cleavage are involved during
the AFEX process. Moreover, results from this study are providing
a basis for understanding lignin AFEX reactions that will be
further investigated.
Typical procedure for AFEX reactions, and purification of products
Anhydrous ammonia gas was liquefied in a liquid nitrogen-cooled
conical flask. Note: ammonia gas/liquid is extremely dangerous
to health and is chemically corrosive. Caution should be exercised
at all times.. To a cooled steel vessel containing a mixture of frozen
deionized water (1 mL) and the diferulate model compound (100
mg) was added 6 mL liquid ammonia via a cooled graduated
glass pipette. The steel vessel was then sealed and heated up to
100 ◦C (at which temperature the pressure reached ~400 psi)
for 18 h in a heating block. After cooling down the vessel in
liquid nitrogen, it was opened under the hood, the ammonia
allowed to evaporate as it warmed to room temperature, and
the products washed out with methanol. The collected methanol
solution was evaporated under vacuum. Fractionation by silica-gel
flash chromatography (or via TLC) of the obtained dark residue
using a gradient of hexane/EtOAc 3 : 1 and then EtOAc/MeOH
10 : 1 yielded several fractions which were purified over preparative
TLC or reverse-phase HPLC to obtain the compounds shown in
Scheme 1. We had difficulty isolating and elucidating all of the
compounds from the reaction mixtures—the yields are isolated
yields and the mass balance is not complete. See the ESI for HPLC
chromatograms of the crude reaction mixtures showing, in some
cases, the considerable complexity of the mixtures of products
produced, but also identifying the compounds that were isolated
and characterized.†
Experimental section
General experimental procedures
NMR spectra were acquired on Bruker Biospin (Billerica, MA)
AVANCE 500 (500 MHz) spectrometer fitted with a cryogenically
cooled 5 mm probe with inverse geometry (proton coils closer
to the sample). The central solvent peak was used as reference
(DMSO-d6 dH 2.49, dC 39.50 ppm). The usual array of 1H, 13C and
2D NMR experiments (gradient-selected COSY, 1H-detected adia-
batic 2D-HSQC, and HMBC) were used for structural elucidation
of compounds. J-values are given in Hz. Conventional lignin
numbering was used for carbon numbering. NMR data for all of
the compounds are given in Tables 1–4. Ultraviolet-visible (UV-
Vis) absorption spectra were recorded on a Shimadzu BioSpec-
nano (Shimadzu, Kyoto, Japan) spectrophotometer equipped with
a quartz cell adapter. All solvents were HPLC grade and were
used as supplied. Flash chromatography was performed on an
Isolera One system using SNAP KP-Sil 10 g silica-gel cartridges
(Biotage, Dyax Corp., Charlottesville, VA) equipped with a UA-
6 UV-vis detector (ISCO, Lincoln, NE). Preparative TLC plates
(Analtech TLC Uniplates, 20 ¥ 20 cm coated with 1.5 mm thick
silica gel GF with UV 254) were from Newark, DE, USA. Solvents
were removed first by rotary evaporator (<50 ◦C), then further
under higher vacuum (100–200 mTorr). HPLC purification and
analysis was conducted on a Shimadzu LCMS-2020 (Shimadzu,
USA) system comprised of LC-20AD pumps, a SIL-20AC HT
autosampler, a CTO-20AC column oven, and using Luna phenyl-
hexyl or Luna C18(2) (250 ¥ 10 mm, 5 mm, Phenomenex, USA) or
4-Hydroxy-3-methoxybenzaldehyde (vanillin) 1a. 10% isolated
yield from diferulate 1 and 12% from diferulate 4. Colorless gum,
UV (l max MeOH) 229, 277, 307; ESIMS m/z (M - H)- 151;
HRESIMS calc. for C8H8O3 (M - H)- 151.0400, found 151.0401.
2-(4-Hydroxy-3-methoxyphenyl)-5-oxopyrrolidine-3-carbox-
amide 1b. 5% isolated yield from diferulate 1. Yellow gum, UV
(l max MeOH) 230, 280; ESIMS m/z (M + Na)+ 273; HRESIMS
calc. for C12H14N2NaO4 (M + Na)+ 273.0846, found 273.0840.
3-(4-Hydroxy-3-methoxybenzoyl)-4-(4-hydroxy-3-methoxyben-
zyl)pyrrolidine-2,5-dione 1c. 3% isolated yield from diferulate 1.
Colorless gum, UV (l max MeOH) 230, 281, 312; ESIMS m/z
(M + Na)+ 408; HRESIMS calc. for C20H19NNaO7 (M + Na)+
408.1059, found 408.1057.
(Z)-3-[Amino(4-hydroxy-3-methoxyphenyl)methylene]-4-(4-hy-
droxy-3-methoxybenzyl)pyrrolidine-2,5-dione 1d. 3% isolated
yield from diferulate 1. Colorless gum, UV (l max MeOH)
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 6779–6787 | 6785
©

View Article Online
288, 318; ESIMS m/z (M + Na)+ 407; HRESIMS calc. for
(E)-3-[3-(2-Amino-2-oxoethyl)-4-hydroxy-5-methoxyphenyl]-
acrylamide 4b. 20% isolated yield from diferulate 4. Colorless
gum, UV (l max MeOH) 317; ESIMS m/z (M + Na)+ 273;
HRESIMS calc. for C12H14N2NaO4 (M + Na)+ 273.0853, found
273.0846.
C20H20N2NaO6 (M + Na)+ 407.1214, found 407.1217.
(2E,3E)-3-Carbamoyl-2-(4-hydroxy-3-methoxybenzylidene)-4-
(4-hydroxy-3-methoxyphenyl)but-3-enoic acid 1e. 2% isolated
yield from diferulate 1. Colorless gum, UV (l max MeOH)
285, 313; ESIMS m/z (M + Na)+ 408; HRESIMS calc. for
C20H19NNaO7 (M + Na)+ 408.1059, found 408.1055.
(E)-Ethyl 3-[3-(2-Amino-2-oxoethyl)-4-hydroxy-5-methoxyphe-
nyl]acrylate 4c. 15% isolated yield from diferulate 4. Colorless
gum, UV (l max MeOH) 305, 325; ESIMS m/z (M + Na)+ 302;
HRESIMS calc. for C14H17NNaO5 (M + Na)+ 302.0999, found
302.1004.
7-Hydroxy-1-(4-hydroxy-3-methoxyphenyl)-6-methoxy-1,2-di-
hydronaphthalene-2,3-dicarboxamide 2a. 30% isolated yield from
diferulate 2. Colorless gum, UV (l max MeOH) 248, 288; ESIMS
m/z (M + Na)+ 407; HRESIMS calc. for C20H20N2NaO6 (M +
Na)+ 407.1214, found 407.1214.
(E)-Ethyl
3-{5¢-[(E)-3-amino-3-oxoprop-1-en-1-yl]-2¢,6-dihy-
5a. 30%
droxy-3¢,5-dimethoxy-[1,1¢-biphenyl]-3-yl}acrylate
isolated yield from diferulate 5. Colorless gum, UV (l max
MeOH) 241, 323; ESIMS m/z (M + Na)+ 436; HRESIMS calc.
for C22H23NNaO7 (M + Na)+ 436.1372, found 436.1373.
7-Hydroxy-1-(4-hydroxy-3-methoxyphenyl)-6-methoxy-2-naph-
thamide 2b. 5% isolated yield from diferulate 2. Colorless gum,
UV (l max MeOH) 245, 289; ESIMS m/z (M + Na)+ 362;
HRESIMS calc. for C19H17NNaO5 (M + Na)+ 362.1004, found
362.1012.
(E)-Ethyl 3-[5¢-(1,3-diamino-3-oxopropyl)-2¢,6-dihydroxy-3¢,5-
dimethoxy-(1,1¢-biphenyl)-3-yl]acrylate 5b. 10% isolated yield
from diferulate 5. Colorless gum, UV (l max MeOH) 326; ESIMS
m/z (M + Na)+ 453; HRESIMS calc. for C22H26N2NaO7 (M +
Na)+ 453.1633, found 453.1653.
6-Hydroxy-4-(4-hydroxy-3-methoxyphenyl)-7-methoxy-2-naph-
thamide 2c. 6% isolated yield from diferulate 2. Colorless gum,
UV (l max MeOH) 251, 310; ESIMS m/z (M + Na)+ 362;
HRESIMS calc. for C19H17NNaO5 (M + Na)+ 362.1004, found
362.1010.
(2E,2¢E)-3,3¢-[6,6¢-Dihydroxy-5,5¢-dimethoxy-(1,1¢-biphenyl)-
3,3¢-diyl]diacrylamide 5c. 6% isolated yield from diferulate 5.
Colorless gum, UV (l max MeOH) 239, 320; ESIMS m/z (M
+ Na)+ 407; HRESIMS calc. for C20H20N2NaO6 (M + Na)+
407.1219, found 407.1205.
7-Hydroxy-10-(4-hydroxy-3-methoxyphenyl)-8-methoxy-10,
10a-dihydronaphtho[2,3-d][1,2,3]triazin-4(3H)-one 2d. (tentative
structural assignment). 2% isolated yield from diferulate 2. Color-
less gum, UV (l max MeOH) 283; ESIMS for C19H15N3O5 (M +
H)+ found 366 and (M-H)- found 364.
Acknowledgements
This work was funded by the DOE Great Lakes Bioenergy
Research Center (DOE BER Office of Science DE-FC02-
07ER64494). We thank Hoon Kim (Dept. of Biochemistry,
U. Wisconsin-Madison) for discussions regarding the NMR of 2d
and 2e, and Bruce E. Dale, Shishir P. S. Chundawat, and Venkatesh
Balan of the Michigan State University branch of the Great Lakes
Bioenergy Research Center for discussions on AFEX and for their
collaboration on related projects.
8-Hydroxy-10-(4-hydroxy-3-methoxyphenyl)-7-methoxy-10,
10a-dihydronaphtho[2,3-d][1,2,3]triazin-4(3H)-one 2e. (tentative
structural assignment). 2.2% isolated yield from diferulate 2.
Colorless gum, UV (l max MeOH) 283; ESIMS m/z (M+H)+
366; (M - H)- found 364; HRESIMS calc. for C19H15N3NaO5 (M
+ Na)+ 388.0904, found 388.0910.
(E)-3-(4-Hydroxy-3-methoxyphenyl)acrylamide 3a. 3% iso-
lated yield from diferulate 3. Colorless gum, UV (l max MeOH)
317, 232; ESIMS m/z (M + H)+ 194; HRESIMS calc. for
C10H12NO3 (M + H)+ 194.0812, found 194.0818.
References
1 N. Mosier, C. Wyman, B. Dale, R. Elander, Y. Y. Lee, M. Holtzapple
and M. Ladisch, Bioresour. Technol., 2005, 96, 673–686.
2 A. T. W. M. Hendriks and G. Zeeman, Bioresour. Technol., 2009, 100,
10–18.
3 D. K. Johnson and R. T. Elander, Biomass Recalcitrance, 2008, 436–
453.
4 E. Sendich, M. Laser, S. Kim, H. Alizadeh, L. Laureano-Perez, B. Dale
and L. Lynd, Bioresour. Technol., 2008, 99, 8429–8435.
5 F. Teymouri, L. Laureano-Perez, H. Alizadeh and B. E. Dale, Bioresour.
Technol., 2005, 96, 2014–2018.
6 B. Bals, H. Murnen, M. Allen and B. Dale, Anim. Feed Sci. Technol.,
2010, 155, 147–155.
(Z)-2-(4-((E)-3-Amino-3-oxoprop-1-en-1-yl)-2-methoxypheno-
xy)-3-(4-hydroxy-3-methoxyphenyl)acrylamide 3b. 23% isolated
yield from diferulate 3. Colorless gum, UV (l max MeOH)
294, 316; ESIMS m/z (M + Na)+ 407; HRESIMS calc. for
C20H20N2NaO6 (M + Na)+ 407.1214, found 407.1207.
(Z)-2-{4-[(E)-3-Amino-3-oxoprop-1-en-1-yl]-2-methoxypheno-
xy}-3-(4-hydroxy-3-methoxyphenyl)acrylic acid 3c. 15% isolated
yield from diferulate 3. Colorless gum, UV (l max MeOH)
292, 315; ESIMS m/z (M + Na)+ 408; HRESIMS calc. for
C20H19NNaO7 (M + Na)+ 408.1054, found 408.1060.
7 V. Balan, L. D. Sousa, S. P. S. Chundawat, D. Marshall, L. N. Sharma,
C. K. Chambliss and B. E. Dale, Biotechnol. Prog., 2009, 25, 365–375.
8 S. P. S. Chundawat, B. S. Donohoe, L. D. Sousa, T. Elder, U. P. Agarwal,
F. C. Lu, J. Ralph, M. E. Himmel, V. Balan and B. E. Dale, Energy
Environ. Sci., 2011, 4, 973–984.
(Z)-2-(4-(1,3-Diamino-3-oxopropyl)-2-methoxyphenoxy)-3-(4-
hydroxy-3-methoxyphenyl)acrylamide 3d. 5% isolated yield from
diferulate 3. Colorless gum, UV (l max MeOH) 295, 317; ESIMS
m/z (M + Na)+ 424; HRESIMS calc. for C20H23N3NaO6 (M +
Na)+ 424.1485, found 424.1483.
9 J. Ralph, Phytochem. Rev., 2010, 9, 65–83.
10 J. Ralph, M. Bunzel, J. M. Marita, R. D. Hatfield, F. Lu, H. Kim,
P. F. Schatz, J. H. Grabber and H. Steinhart, Phytochem. Rev., 2004, 3,
79–96.
11 J. Ralph, J. H. Grabber and R. D. Hatfield, Carbohydr. Res., 1995, 275,
167–178.
6786 | Org. Biomol. Chem., 2011, 9, 6779–6787
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
12 J. Ralph, R. D. Hatfield, S. Quideau, R. F. Helm, J. H.
Grabber and H. J. G. Jung, J. Am. Chem. Soc., 1994, 116, 9448–
9456.
21 F. Ramirez, V. Gonzalez, M. Crespo, D. Meier, O. Faix and V. Zuniga,
Bioresour. Technol., 1997, 61, 43–46.
22 K. Fischer and R. Schiene, Chemical Modification, Properties, and
Usage of Lignin, 2002, 167–198.
23 J. Ralph, S. Quideau, J. H. Grabber and R. D. Hatfield, J. Chem. Soc.,
Perkin Trans. 1, 1994, 3485–3498.
24 J. Ralph, G. Brunow, P. J. Harris, R. A. Dixon, P. F. Schatz and
W. Boerjan, in Recent Advances in Polyphenol Research, vol 1, ed.
F. Daayf, A. El Hadrami, L. Adam and G. M. Ballance, Wiley-
Blackwell Publishing, Oxford, UK, 2008, vol. 1, pp. 36-66.
25 J. Ralph, M. T. Garcia-Conesa and G. Williamson, J. Agric. Food
Chem., 1998, 46, 2531–2532.
26 F. Lu, L. Wei, A. Azarpira and J. Ralph, submitted, 2011.
27 A. U. Buranov and G. Mazza, Ind. Crops Prod., 2008, 28, 237–259.
28 D. Meier, V. Zuniga-Partida, F. Ramirez-Cano, N.-C. Hahn and O.
Faix, Bioresour. Technol., 1994, 49, 121–128.
13 E. Sjo¨estrom, Wood Chemistry: Fundamentals and Applications,, 2nd
edn, 1993.
14 M. M. D. Buanafina, Mol. Plant, 2009, 2, 861–872.
15 T. Koshijima and T. Watanabe, Association Between Lignin and
Carbohydrates in Wood and Other Plant Tissues, 2003.
16 J. Ralph, R. D. Hatfield, J. H. Grabber, H.-J. G. Jung, S. Quideau and
R. F. Helm, ACS Symp. Ser., 1998, 697, 209–236.
17 J. Ralph and R. F. Helm, Forage Cell Wall Struct. Dig., Int. Symp.,
1993, 201–246.
18 J. H. Grabber, R. D. Hatfield, F. C. Lu and J. Ralph, Biomacromolecules,
2008, 9, 2510–2516.
19 M. Baucher, C. Halpin, M. Petit-Conil and W. Boerjan, Crit. Rev.
Biochem. Mol. Biol., 2003, 38, 305–350.
20 M. W. Lau and B. E. Dale, Proc. Natl. Acad. Sci. U. S. A., 2009, 106,
1368–1373.
29 A. Potthast, R. Schiene and K. Fischer, Holzforschung, 1996, 50, 554–
562.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 6779–6787 | 6787
©