Rapid syntheses of dehydrodiferulates via biomimetic radical coupling reactions of ethyl ferulate

Article abstract of DOI:10.1021/jf302140k

Dehydrodimerization of ferulates in grass cell walls provides a pathway toward cross-linking polysaccharide chains limiting the digestibility of carbohydrates by ruminant bacteria and in general affecting the utilization of grass as a renewable bioresource. Analysis of dehydrodiferulates (henceforth termed diferulates) in plant cell walls is useful in the evaluation of the quality of dairy forages as animal feeds. Therefore, there has been considerable demand for quantities of diferulates as standards for such analyses. Described here are syntheses of diferulates from ethyl ferulate via biomimetic radical coupling reactions using the copper(II)-tetramethylethylenediamine [CuCl(OH)-TMEDA] complex as oxidant or catalyst. Although CuCl(OH)-TMEDA oxidation of ethyl ferulate in acetonitrile produced mixtures composed of 8-O-4-, 8-5-, 8-8- (cyclic and noncyclic), and 5-5-coupled diferulates, a catalyzed oxidation using CuCl(OH)-TMEDA as catalyst and oxygen as an oxidant resulted in better overall yields of such diferulates. Flash chromatographic fractionation allowed isolation of 8-8- and 5-5-coupled diferulates. 8-5-Diferulate coeluted with 8-O-4-diferulate but was separated from it via crystallization; the 8-O-4 diferulate left in the mother solution was isolated by rechromatography following a simple tetrabutylammonium fluoride treatment that converted 8-5-diferulate to another useful diferulate, 8-5-(noncyclic) diferulate. Therefore, six of the nine (5-5, 8-O-4, 8-5-c, 8-5-nc, 8-5-dc, 8-8-c, 8-8-nc, 8-8-THF, 4-O-5) diferulic acids that have to date been found in the alkaline hydrolysates of plant cell walls can be readily synthesized by the CuCl(OH)-TMEDA catalyzed aerobic oxidative coupling reaction and subsequent saponification described here.

Full text of DOI:10.1021/jf302140k

Article
pubs.acs.org/JAFC
Rapid Syntheses of Dehydrodiferulates via Biomimetic Radical
Coupling Reactions of Ethyl Ferulate
Fachuang Lu,*,†,‡ Liping Wei,‡,§ Ali Azarpira,‡ and John Ralph†,‡,§
‡
†Department of Biochemistry and the Wisconsin Bioenergy Initiative, The DOE Great Lakes Bioenergy Research Center, and
§Department of Biological Systems Engineering, University of WisconsinMadison, Madison, Wisconsin 53706, United States
ABSTRACT: Dehydrodimerization of ferulates in grass cell walls provides a pathway toward cross-linking polysaccharide chains
limiting the digestibility of carbohydrates by ruminant bacteria and in general affecting the utilization of grass as a renewable
bioresource. Analysis of dehydrodiferulates (henceforth termed diferulates) in plant cell walls is useful in the evaluation of the
quality of dairy forages as animal feeds. Therefore, there has been considerable demand for quantities of diferulates as standards
for such analyses. Described here are syntheses of diferulates from ethyl ferulate via biomimetic radical coupling reactions using
the copper(II)−tetramethylethylenediamine [CuCl(OH)−TMEDA] complex as oxidant or catalyst. Although CuCl(OH)−
TMEDA oxidation of ethyl ferulate in acetonitrile produced mixtures composed of 8−O−4-, 8−5-, 8−8- (cyclic and noncyclic),
and 5−5-coupled diferulates, a catalyzed oxidation using CuCl(OH)−TMEDA as catalyst and oxygen as an oxidant resulted in
better overall yields of such diferulates. Flash chromatographic fractionation allowed isolation of 8−8- and 5−5-coupled
diferulates. 8−5-Diferulate coeluted with 8−O−4-diferulate but was separated from it via crystallization; the 8−O−4 diferulate
left in the mother solution was isolated by rechromatography following a simple tetrabutylammonium fluoride treatment that
converted 8−5-diferulate to another useful diferulate, 8−5-(noncyclic) diferulate. Therefore, six of the nine (5−5, 8−O−4, 8−5-
c, 8−5-nc, 8−5-dc, 8−8-c, 8−8-nc, 8−8-THF, 4−O−5) diferulic acids that have to date been found in the alkaline hydrolysates of
plant cell walls can be readily synthesized by the CuCl(OH)−TMEDA catalyzed aerobic oxidative coupling reaction and
subsequent saponification described here.
KEYWORDS: copper(II)−tetramethylethylenediamine, diferulates, oxidative coupling, plant cell walls, synthesis, NMR
proposed to exhibit a wide range of important biological and
therapeutic properties including anti-inflammatory, antibacte-
rial, antidiabetic, anticarcinogenic, antiaging, and neuroprotec-
tive effects.10 In recent years, with increased interest in
renewable bioresources as alternatives to fossil fuels for energy
and materials production, lignocellulosic biomass sources,
including grasses, are emerging as potential feedstocks for the
production of biofuels and chemicals.11
Pretreatment processes become essential before biological
conversion of lignocellulosic biomass to biofuels. Modification
of grasses via manipulation of genes responsible for
feruloylation of polysaccharides has been a focus in plant
research because such modified plants potentially have cell
walls with desirable properties, including better digestibility by
ruminants, and are more easily processed as feedstocks that are
saccharified and fermented for bioenergy and materials
production.12
Therefore, studying the effects of ferulates and especially
their dehydrodimers, the dehydrodiferulates (henceforth simply
termed diferulates or ferulate dimers), which are markers for
polysaccharide−polysaccharide cross-linking, on cell wall
properties has become crucial. Such studies require ready
accessibility to diferulates as reference compounds for
qualitative and quantitative analysis of diferulates,13 as well as
INTRODUCTION
■
Ferulates, particularly in the form of polysaccharide ferulate
esters, are widely found in the cell walls of forage plants,
including within monocotyledons of the Poaceae and within
dicotyledons included in the Amaranthaceae.1 In monocots,
ferulate acylates the primary hydroxyl at the C-5 position of α-
L-arabinofuranosyl residues and on xylose side-chain residues of
xyloglucans in bamboo, whereas in dicots, such as spinach and
sugar beet, feruloylation occurs on arabinose or galactose side
chains of pectic polysaccharides.2
Although ferulates exist in relatively low amounts in cell
walls, they play very important roles not only for the biology of
the wall but also for their structure. First, peroxidase-mediated
oxidative coupling of ferulates cross-links polysaccharides,
whereas analogous cross-coupling reactions between ferulates
(or the resultant dehydrodiferulates) and monolignols or
growing lignin polymers produce so-called lignin−ferulate−
polysaccharide complexes.3 Second, such cross-links have been
shown to be responsible for the limited digestibility of grass
forages by ruminant bacteria4 and to contribute to the
recalcitrance of grass cell walls to enzymatic hydrolysis by
cellulases or xylanases5 and have also been postulated to be
responsible for controlling cell wall extensibility6 and protection
against pathogen invasion.7 Finally, the ability of ferulates to
cross-couple with monolignols and the finding of evidence for
ferulate−monolignol cross-coupling in grasses8 imply that
ferulates and diferulates may act as nucleation sites for
lignification.9 Meanwhile, as important antioxidants in many
human foods such as vegetables, fruits, and cereals, ferulates are
Received: May 15, 2012
Revised: July 30, 2012
Accepted: July 30, 2012
Published: July 30, 2012
© 2012 American Chemical Society
8272
dx.doi.org/10.1021/jf302140k | J. Agric. Food Chem. 2012, 60, 8272−8277

Journal of Agricultural and Food Chemistry
Article
model substrates for evaluating their potential as antioxidants.14
However, some diferulates are not readily available by
synthesis; for example, our original preparation of 8−O−4-
diferulate involved as many as nine steps3 and is therefore
accessible only to skilled synthetic chemists. Cereal brans such
as corn, wheat, and rice brans are good sources for isolating
ferulic acid dimers. With demanding separation techniques
including Sephadex LH-20 column fractionation and semi-
preparative high-performance liquid chromatography (HPLC),
as many as seven diferulic acids have been isolated with
sufficient purity and in quantities (tens of milligrams) suitable
for being used as standard compounds.15 Electrochemical
oxidative coupling of ferulate was reported16 to produce
isomeric diferulate mixtures containing 8−O−4-, 8−5-, and 8−
8-diferulates (Figure 1). Each diferulate was isolated by flash
catalyst for biomimetic oxidative coupling of ethyl ferulate to
produce 8−O−4-, 8−5-, 8−8-, and 5−5-coupled diferulates
(Figure 1). This biomimetic approach takes advantage of easy-
to-perform coupling reactions (without requiring excessive
levels of synthetic chemistry skill) and compound separation
using simple flash chromatography to produce reasonable
quantities (tens to hundreds of milligrams) of all the major
diferulates found in plant cell walls.
MATERIALS AND METHODS
■
Materials. Ferulic acid was purchased from MP Biomedicals
(Solon, OH, USA). Copper(I) chloride was from Alfa Aesar (Ward
Hill, MA, USA), and tetramethylethylenediamine (TMEDA) was an
Acros product (Fisher Scientific, Pittsburgh, PA, USA). CuCl(OH)−
TMEDA was prepared in situ by adding an equimolar equivalent of
TMEDA into copper(I) chloride suspended in acetonitrile. All
solvents used were of analytical grade and obtained from Fisher.
Silica gel flash chromatography was on a Biotage Isolera system
(Biotage, Charlotte, NC, USA) using prepacked SNAP columns (10,
25, or 100 g of silica gel).
Ferulate Monomer. Ethyl ferulate was trivially prepared from
ferulic acid, as described previously.3 Thus, ferulic acid (25 g) was
dissolved in absolute ethanol (200 mL), and acetyl chloride (15 mL)
was added slowly. The solution was stirred gently for 2 days, after
which the volatiles were removed by rotary evaporation at 40 °C.
Addition of further ethanol and evaporation several times removed the
residual HCl. The oily product was allowed to crystallize from EtOAc/
cyclohexane to produce light yellow crystals (26.4 g, 90%).
Preparation of Dehydrodiferulates. Method A [CuCl(OH)−
TMEDA Oxidation]. An molar equivalent of TMEDA (1.16 g, 10
mmol) was added into anhydrous CuCl (1.0 g, 10 mmol) suspended
in acetonitrile (200 mL) in a 500 mL round-bottom flask. Stirring this
mixture for 1 h produced a solution of the CuCl(OH)−TMEDA
complex (a dark blue solution). To this solution was added ethyl
ferulate (2.22 g, 10 mmol) dissolved in 20 mL of acetonitrile, and this
mixture was stirred for 100 min. The reaction was stopped by adding
150 mL of 1 M HCl, producing a clear green solution. The acetonitrile
was removed by evaporation at 40 °C under reduced pressure, and the
products were recovered by ethyl acetate extraction (150 + 100 mL).
The ethyl acetate solution was washed with saturated NH4Cl (100
mL) solution and dried over anhydrous MgSO4. Filtration through a
sintered glass funnel and evaporation of the ethyl acetate solution
produced the crude product mixture.
Method B [CuCl(OH)−TMEDA-Catalyzed Oxidation]. A molar
equivalent of TMEDA (116 mg, 1.0 mmol) was added into anhydrous
CuCl (100 mg, 1.0 mmol) in acetonitrile (150 mL) in a 500 mL
round-bottom flask, and this mixture was stirred for 5 min, resulting in
a deep blue solution. Into the solution was added ethyl ferulate (2.22 g,
10.0 mmol), and an oxygen-filled balloon was fitted onto the flask.
After the mixture was stirred for 4.0 h, TLC showed that only small
amounts of ethyl ferulate remained. Then 60 mL of 1 M HCl solution
was added to stop the reaction, and the organic solvent was removed
by evaporation under reduced pressure at 40 °C. The products were
recovered by ethyl acetate (2 × 150 mL) extraction using a separatory
funnel, and the organic phase was washed with acidic NH4Cl solution
(10 mL of 1 M HCl plus 100 mL of saturated NH4Cl). The organic
solution was dried over anhydrous MgSO4, and the inorganic solid was
removed by filtration through a sintered glass funnel. The crude
products were obtained after evaporation of the ethyl acetate solution
at 40 °C under reduced pressure.
When carried out analogously in dichloromethane, the oxidative
coupling reaction can be completed in 3.5 h. However, the total yield
of diferulates was lower because more trimers and higher oligomers
were formed.
Flash Chromatography Purification. Eluting solvents were
solvent A, hexane, and solvent B, ethyl acetate containing 5% (v/v)
ethanol.
Figure 1. Radical coupling of ethyl ferulate 1 by CuCl(OH)−TMEDA
(O2) to produce the ferulate dehydrodimers 2−7. Purification via
simple flash chromatography is described in the text.
chromatography. This approach seems to be a good way to
synthesize diferulates, but the required device for electro-
chemical reactions and the limited amounts of products
obtained prevent it from becoming a convenient and general
protocol. Therefore, there remains a need to develop rapid and
robust methods to produce diferulates in significant quantities
without requiring particularly demanding separation skills and
using readily accessible methods and instruments.
Our group aims to develop efficient and simple methods to
make compounds of interest to aid our own research and that
of others. In this study, we explored the possibility of using a
Cu(II)−amine complex as a biomimetic oxidant or catalyst for
oxidative coupling of ethyl ferulate to produce diferulates. Here
we report on the rapid syntheses of dehydrodiferulates using
the CuCl(OH)−tetramethylethylenediamine [CuCl(OH)−
TMEDA] complex as either a stoichiometric oxidant or a
The crude products from the coupling reactions described above
were loaded with 6−8 mL of dichloromethane onto a Biotage snap
8273
dx.doi.org/10.1021/jf302140k | J. Agric. Food Chem. 2012, 60, 8272−8277

Journal of Agricultural and Food Chemistry
Article
normal-phase prepacked column (100 g of silica gel, 2.5 × 19 cm), and
chromatography was performed on an Isolera purification system
equipped with a UV detector and a sample collector.
7.0 (1H, d, J = 1.8 Hz, B-6), 7.39 (1H, d, J = 1.8 Hz, B2), 7.57 (1H, d,
J = 16.0 Hz, B-7), 7.77 (1H, s, A-7); δC 14.58 (Me), 14.60 (Me), 56.42
(A−OMe), 56.60 (B−OMe), 60.45 (B−CH2), 61.03 (A−CH2),
110.12 (B2), 113.16 (A2), 115.60 (A5), 116.22 (B8), 124.84 (A8),
125.68 (B6), 126.33 (A6), 126.50 (B5), 127.25 (B1), 127.52 (A1),
141.33 (A7), 145.23 (B7), 147.78 (A3), 147.97 (B4), 148.90 (A4),
149.03 (B3), 167.27 (B9), 167.73 (A9) (NMR database,18 compound
2019).
The diferulate products were eluted with solvent mixtures
programmed as follows: flow rate, 40 mL/min; isotropic elution,
1200 mL (70% solvent A + 30% solvent B) followed by gradient
elution increasing the percentage of B to 50% at 2000 mL and held
with that solvent mixture until the end at the 2800 mL elution volume.
The trace profile was monitored by UV detection at 280 nm, and
fractions were autocollected using “mediate slope mode” in 16 × 125
mm test tubes (maximal volume = 18 mL). The initial waste volume
was 400 mL.
Characterization of Diferulate Products. The isolated and
purified diferulate products were fully characterized by the usual series
of NMR experiments (1H, 13C, HSQC, HMBC, COSY). NMR spectra
were recorded on a Bruker Biospin (Billerica, MA, USA) AVANCE
500 (500 MHz) spectrometer fitted with a cryogenically cooled 5 mm
TCI gradient probe with inverse geometry (proton coils closest to the
sample). Bruker’s Topspin 3.1 (Mac) software was used to process
spectra. About 10−30 mg of diferulate in acetone-d6 was used for
NMR characterization, and the central solvent peaks, δH/δC 2.04/29.8,
were used as internal references.
8−O−4-Coupled Diferulate 4 (Z)-Ethyl 2-[4-((E)-3-Ethoxy-3-
oxoprop-1-enyl)-2-methoxyphenoxy]-3-(4-hydroxy-3-
methoxyphenyl)acrylate, 8−O−4-Diferulate 4. Flash chromatogra-
phy of CuCl(OH)−TMEDA oxidation products produced a fraction
(380 mg) containing diferulates 2 and 4, which was allowed to
crystallize from ethyl acetate/cyclohexane to produce 220 mg of
diferulate 2. The mother liquor solution was evaporated, and the
resulting residual oil was dissolved in 10 mL of acetonitrile and treated
with 100 mg of TBAF for 30 min. After the addition of 5 mL of 1 M
HCl solution, the mixture was evaporated at 40 °C under reduced
pressure to remove the organic solvent. The residue was partitioned
between ethyl acetate (15 mL) and 10 mL of 1 M HCl and the
extracted ethyl acetate phase was washed with saturated NH4Cl and
dried over MgSO4. Filtration through sintered glass removed the
solids, and evaporation gave the product mixture as a yellow oil, which
was purified by flash chromatography using a Biotage snap silica gel
(10 g) prepacked column eluted with hexane/ethyl acetate (3:1, v/v)
to produce 100 mg of pure diferulate 4 (eluting solvent volume from
120 to 200 mL) and 20 mg of diferulate 3 (eluting solvent volume
from 360 to 560 mL). Diferulate 4: NMR (acetone-d6) δH 1.21 (3H, t,
J = 7.0 Hz, Me), 1.26 (3H, t, J = 7.0 Hz, Me), 3.73 (3H, s, A−OMe),
3.99 (3H, s, B−OMe), 4.15−4.23 (4H, m, A/B−CH2−), 6.45 (1H, d,
J = 16.0 Hz, B-8), 6.78 (1H, d, J = 8.5 Hz, B-5), 6.81 (1H, d, J = 8.5
Hz, A-5), 7.12 (1H, dd, J = 8.5, 1.8 Hz, B-6), 7.23 (1H, dd, J = 8.5, 1.8
Hz, A-6), 7.38 (1H, s, A-7), 7.46 (1H, d, J = 1.8 Hz, B-2), 7.23 (1H, d,
J = 1.8 Hz, A-2), 7.58 (1H, d, J = 16.0 Hz, B-7); δC 14.46 (Me), 14.58
(Me), 55.84 (A−OMe), 56.40 (B−OMe), 60.57 (B−CH2), 61.70 (A−
CH2), 112.17 (B2), 113.65 (A2), 114.39 (B5), 115.94 (A5), 117.46
(B8), 122.91 (B6), 125.14 (A1), 126.08 (A6), 128.08 (A7), 130.08
(B1), 138.29 (A8), 144.79 (B7), 148.22 (A3), 148.76 (B4), 149.42
(A4), 150.16 (B3), 163.70 (A9), 167.63 (B9). These data were
consistent with those published (NMR database,18 compound 2041,
ethyl/methyl ester).
8−8-Coupled Diferulates 5 and 6 Diethyl 7-Hydroxy-1-(4-
hydroxy-3-methoxyphenyl)-6-methoxy-1,2-dihydronaphthalene-
2,3-dicarboxylate, 8−8-Diferulate (Cyclic) 5. Pure diferulate 5 (250
mg), eluted at solvent volume from 1360 to 1600 mL, was obtained by
a flash chromatography fractionation of crude products from
CuCl(OH)−TMEDA catalyzed aerobic oxidation of ethyl ferulate.
Diferulate 5: NMR (acetone-d6) δH 1.10 (3H, t, J = 7.0 Hz, A−Me),
1.24 (3H, t, J = 7.0 Hz, B−Me), 3.75 (3H, s, A−OMe), 3.88 (3H, s,
B−OMe), 3.93 (1H, d, J = 3.5 Hz, A-8), 3.97−4.05 (2H, m, A−
CH2−), 4.13−4.16 (2H, m, B−CH2−), 4.51 (1H, d, J = 3.5 Hz, A-7),
6.42 (1H, dd, J = 8.0, 1.8 Hz, A-6), 6.64 (1H, s, B-5), 6.66 (1H, d, J =
8.0 Hz, A-5), 6.78 (1H, d, J = 1.8 Hz, A-2), 7.10 (1H, s, B-2), 7.64
(1H, s, B-7); δC 14.36 (Me), 14.56 (Me), 46.42 (A7), 48.24 (A8),
56.13 (A−OMe), 56.30 (B−OMe), 60.86 (B−CH2), 61.18 (A−CH2),
111.98 (A2), 113.08 (B2), 115.52 (A5), 116.70 (B5), 120.91 (A6),
123.59 (B8), 124.39 (B1), 132.11 (B6), 135.28 (A1), 137.97 (B7),
146.22 (A4), 147.50 (B3), 148.12 (A3), 149.43 (B4), 167.03 (B9),
172.72 (A9). These data were consistent with those published (NMR
database,18 compound 2035, methyl analogue).
8−5-Coupled Diferulates 2 and 3 (E)-Ethyl 5-(3-ethoxy-3-
oxoprop-1-enyl)-2-(4-hydroxy-3-methoxyphenyl)-7-methoxy-2,3-di-
hydrobenzofuran-3-carboxylate, 8−5-Diferulate 2. Diferulate 2
(Figure 1) was one of the major dimeric products from the
CuCl(OH)−TMEDA catalyzed aerobic oxidative coupling reaction
of ethyl ferulate as described above. Flash chromatography of the
crude oxidation products gave a fraction (eluting solvent volume from
740 to 850 mL) containing a mixture of diferulate 2 and diferulate 4.
Diferulate 2 was crystallized from ethyl acetate−hexane as white
needles (0.48 g, 21.6%). It has the same NMR data as those previously
published:17(NMR database,18 compound 2019) NMR (acetone-d6)
δH 1.27 (3H, t, J = 7.0 Hz, B-Me), 1.29 (3H, t, J = 7.0 Hz, A-Me), 3.83
(3H, s, ArOMe), 3.91 (3H, s, ArOMe), 4.18 (2H, q, J = 7.0 Hz, B−
CH2−), 4.26 (2H, m, A−CH2−), 4.43 (1H, d, J = 8.0 Hz, A-8), 6.03
(1H, d, J = 8.0 Hz, A-7), 6.42 (1H, d, J = 16.0 Hz, B-8), 6.84 (1H, d, J
= 8.0 Hz, A-5), 6.91 (1H, br d, J = 8.0 Hz, A-6), 7.09 (1H, br s, A-2),
7.28 (1H, br s, B-6), 7.34 (1H, br s, A-2), 7.62 (1H, d, J = 16.0 Hz, B-
7); δC 14.46 (B-Me), 14.61 (A-Me), 56.00 (A7), 56.22 (OMe), 56.40
(OMe), 60.54 (A−CH2), 62.18 (B−CH2), 88.32 (A8), 110.65 (A2),
113.10 (B2), 115.76 (A5), 116.60 (B8), 118.90 (B6), 120.17 (A6),
127.34 (B5), 129.40 (B1), 132.00 (A1), 145.22 (B7), 145.78 (B3),
147.91 (A4), 148.50 (A3), 150.92 (B4), 167.26 (B9), 171.07 (A9).
The mother liquor after crystallization contained mainly 8−O−4-
coupled diferulate 4 that could be isolated from the residual diferulate
2 following a tetrabutylammonium fluoride (TBAF) treatment as
described below for the preparation of diferulate 4. Although
CuCl(OH)−TMEDA oxidative coupling of ethyl ferulate produced
8−5-diferulate 2 in a lower (10%) yield, it still could also be
crystallized from a mixture fractionated by flash chromatography. If
diferulate 2 is the primary required compound, peroxidase-catalyzed
coupling remains perhaps the most straightforward and highest
yielding method.17
(E)-Ethyl 2-[5-((E)-3-ethoxy-3-oxoprop-1-enyl)-2-hydroxy-3-me-
thoxyphenyl]-3-(4-hydroxy-3-methoxyphenyl)acrylate, 8−5-Diferu-
late (Noncyclic) 3. Diferulate 3 is not a coupling product from
oxidation of ethyl ferulate, nor is its feruloyl polysaccharide analogue
present in plant cell walls.19 However, the free acid of 3 is often
detected in alkaline hydrolysates of grass cell walls because alkaline
hydrolysis of diferulate 2, the only 8−5-coupled diferulate from radical
coupling reactions, produces significant amounts of the acid of 3 in
addition to the free acid of 2. Diferulate 3 is therefore a valuable
diferulate for cleanly preparing its acid analogue. In this study,
diferulate 3 was obtained from TBAF treatment of diferulate 2 as
described below for the isolation of diferulate 4. Diferulate 3: NMR
(acetone-d6) δH 1.20 (3H, t, J = 7.0 Hz, Me), 1.24 (3H, t, J = 7.0 Hz,
Me), 3.44 (3H, s, B−OMe), 3.96 (3H, s, A−OMe), 4.13−4.19 (4H,
m, A/B−CH2−), 6.48 (1H, d, J = 16.0 Hz, B-8), 6.70 (1H, d, J = 1.8
Hz), 6.71 (1H, d, J = 7.8 Hz, A-5), 6.83 (1H, dd, J = 1.8, 7.8 Hz, A-6),
(2E,3E)-Diethyl 2,3-Bis(4-hydroxy-3-methoxybenzylidene)-
succinate, 8−8-Diferulate (Noncyclic) 6. Diferulate 6 was isolated
in a fraction (eluting solvent volume from 1000 to 1180 mL) from
flash chromatography of oxidation products of ethyl ferulate with
CuCl(OH)−TMEDA as oxidant. This crude fraction (215 mg) was
further purified by secondary chromatography using a 25 g SNAP
column, resulting in pure diferulate 6 (130 mg, 5.8% yield) as a pale
yellow oil. Diferulate 6: NMR (acetone-d6) δH 1.10 (3H, t, J = 7.0 Hz,
Me), 3.73 (3H, s, Ar−OMe), 4.11 (2H, q, J = 7.0 Hz, −CH2−), 6.78
(1H, d, J = 8.2 Hz, A-5), 7.09 (1H, br d, J = 8.2 Hz, A-6), 7.25 (1H, br
8274
dx.doi.org/10.1021/jf302140k | J. Agric. Food Chem. 2012, 60, 8272−8277

Journal of Agricultural and Food Chemistry
Article
s, A-2), 7.81 (1H, s, A-7); δC 14.48 (Me), 55.97 (OMe), 61.19
(−CH2−), 113.26 (2), 115.92 (5), 125.56 (6), 125.60 (8), 127.80 (1),
142.42 (7), 148.14 (3), 149.30 (4), 167.53 (9). These data were
consistent with those published.16
5−5-Coupled Diferulate 7 (2E,2′E)-Diethyl 3,3′-(6,6′-Dihydroxy-
5,5′-dimethoxybiphenyl-3,3′-diyl)diacrylate, 5−5-Diferulate 7. Di-
ferulate 7, one of the major dimeric products resulting from
CuCl(OH)−TMEDA oxidation of ethyl ferulate, was isolated (220
mg, 10%) (eluting solvent volume from 1720 to 1800 mL) by flash
chromatography. More diferulate 7 could be obtained by TLC
separation of the other mixture fractions containing 7. Diferulate 7:
NMR (acetone-d6) δH 1.26 (3H, t, J = 7.0 Hz, Me), 3.96 (3H, s, Ar−
OMe), 4.17 (2H, q, J = 7.0 Hz, −CH2−), 6.44 (1H, d, J = 16.0 Hz, H-
8), 7.21 (1H, d, J = 1.5 Hz, H-6), 7.53 (1H, d, J = 1.5 Hz, H-2), 7.64
(1H, d, J = 16.0 Hz, H-7); δC 14.60 (Me), 56.46 (OMe), 60.45 (CH2),
109.76 (2), 116.20 (8), 125.52 (5), 126.20 (6), 126.50 (1), 1145.58
(7), 147.46 (4), 148.84 (3), 167.39 (9). These data were consistent
with those in our NMR database,18 compound 2057.
Figure 2. GC−total ion chromatography of TMS-derivatized dimeric
products from (A) CuCl(OH)−TMEDA oxidation and (B) oxidative
coupling of ethyl ferulate by CuCl(OH)−TMEDA catalyzed aerobic
oxidation.
RESULTS AND DISCUSSION
■
It is well-known that Cu(II)−amine complexes can react with
phenols generating radical species and that coupling of the
resulting radicals produces dimers, trimers, oligomers, and even
polymers with C−C or C−O bonds depending upon the
Cu(II)−amine complex and the coupling conditions used.20,21
Because Cu(II)−amine complexes are readily available and a
variety of Cu(II)−amine complexes can be prepared by various
combinations of copper salts and tertiary amines, they have
been widely used to make diphenols, phenolic oligomers, and
polyphenols.22 The commercially available CuCl(OH)−
TMEDA is a stable, free-flowing solid, soluble in chlorinated
solvents, ethanol, diethyl ether, methanol, and THF. It was first
introduced by Noji et al. in 1994 as a convenient catalyst for
aerobic oxidative coupling of 2-naphthols to make binaph-
thols.23 Most oxidative coupling reactions of phenols catalyzed
by CuCl(OH)−TMEDA produced biaryl compounds; that is,
such coupling reactions favor forming carbon−carbon bonds at
the para- or ortho-positions to the phenol.24
To date, CuCl(OH)−TMEDA has not been used for the
coupling of phenols with conjugated double bonds at the para-
position, such as isoeugenol, hydroxycinnamates, or hydrox-
ycinnamyl alcohols. Although the CuCl(OH)−TMEDA
complex is commercially available and can be readily prepared,
CuCl(OH)−TMEDA generated in situ in this study was
equally effective and even more convenient to use. Thus,
CuCl(OH)−TMEDA generated in situ was used as either an
oxidant or a catalyst to effect oxidative coupling of ethyl
ferulate, aimed at developing simple and convenient protocols
to prepare diferulate isomers.
In method A, in which an molar equivalent CuCl(OH)−
TMEDA was used as the sole oxidant, the coupling reaction
carried out in acetonitrile went to completion in about 1.5 h,
producing various dimers resulting from 8−5-, 8−8-, 5−5-, and
8−O−4-couplings, as indicated by GC-MS analysis of the crude
products (Figure 2). However, in method B, in which such
oxidative coupling was performed catalytically (10% CuCl-
(OH)−TMEDA), using an oxygen atmosphere to regenerate
the catalyst in situ, the total yields of all isolated diferulates were
higher than those obtained in method A (Table 1). Thus, a
total of five diferulates can be formed in a CuCl(OH)−
TMEDA-catalyzed aerobic oxidative coupling of ferulate, all
isolable in pure form, in yields from 4 to 21%, following flash
chromatography. At this moment, it is not clear why these two
methods produced different results, especially in the yields of
8−8-coupled products. It may be speculated that, at relatively
Table 1. Isolated Yields of Dimeric Products from Oxidation
of Ethyl Ferulate 1
product description
method A
(noncatalytic)
method B
(catalytic)
diferulate
no.
8−5-c
8−5-nc
8−O−4-
8−8-c
8−8-nc
5−5-
2
3
4
5
6
7
10.0
1.5
21.6
0.5
a
4.0
4.3
b
−
12.0
4.5
5.6
6.7
2.5
total
27.8
43.4
a
Compound 3 was not produced directly from the coupling reaction
b
of ethyl ferulate. No attempt was made to isolate compound 5 due to
its low yield via method A.
low concentrations of CuCl(OH)−TMEDA (method B),
coupling of monomeric ethyl ferulate is favored, whereas at
relative high concentrations of CuCl(OH)−TMEDA (method
A) such preference (selectivity) is low so that trimers or
oligomers are also produced significantly. It was observed that
higher amounts of trimers or oligomers were indeed formed by
method A than by method B.
As one of the major dimeric products in the CuCl(OH)−
TMEDA oxidation system, the 5−5-coupled diferulate 7 was
readily isolated in a modest yield by simple flash chromatog-
raphy. More diferulate 7 could be obtained if the rest of the
mixture containing 7 was fractionated again by chromatog-
raphy. Diferulate 7 (or 5−5-diferulic acid) was the first
dehydrodimer detected and isolated from plant cell walls.25
The free acid of diferulate 7, often referred to simply as
“diferulic acid”, had been the sole diferulic acid found from
plants until 1994, when other dehydrodimers including 8−5-,
8−O−4-, and 8−8-coupled dimers were identified thanks to
syntheses and authentication of these dimers;3 a further minor
4−O−5-dimer was later synthesized and authenticated from
plants.26 Although diferulate 7 was synthesized in a relatively
easy way from vanillin,3,27 several steps were required, and
modest synthetic chemistry skills are still required to
accomplish such a task. The protocol described here provides
8275
dx.doi.org/10.1021/jf302140k | J. Agric. Food Chem. 2012, 60, 8272−8277

Journal of Agricultural and Food Chemistry
Article
a simple alternative way to synthesize diferulate 7. To the best
of our knowledge, although it has been demonstrated that 5−5-
diferulate 7 can be formed via coupling reactions initiated by
various oxidation methods, this is the first time that diferulate 7
has been isolated in quantity, that is, on a preparative scale,
directly from radical coupling of ferulate.
diferulate 3 is routinely found in alkaline hydrolysis products of
grass plant cell walls and sometimes was mistakenly considered
as a direct 8−5-coupling product from radical coupling of
ferulate, a notion we have repeatedly tried to correct.19 As its
acid is a product resulting from cell wall saponification,
however, it is still very useful to synthesize 3 for use as a
standard compound to quantitate 8−5-coupled diferulate in
plant cell walls. The free acid of 3 was one of the products
produced by alkaline hydrolysis of diferulate 2, whereas 1,8-
diazabicyclo[5.4.0]undec-7-ene (DBU) treatment of diferulate
2 produced diferulate 3.3 In the current study, it was found that
TBAF treatment of diferulate 2 in acetonitrile also resulted in
diferulate 3 quantitatively, whereas diferulate 4 remained intact.
Thus, TBAF can be used not only to aid in the isolation of
diferulate 4 from a coeluting mixture containing 2 and 4 but
also to obtain 3 from the readily available diferulate 2,17 from
the peroxidase−H2O2 system, if quantities of diferulate 3 are
needed.
In conclusion, it has been demonstrated that CuCl(OH)−
TMEDA in acetonitrile is a robust biomimetic oxidation system
for oxidative coupling of ferulate. Five diferulates, including 8−
5-coupled 2, 8−O−4-coupled 4, 8−8-coupled 5-6, and 5−5-
coupled 7 diferulates, can all be formed and isolated in a pure
form from CuCl(OH)−TMEDA-catalyzed aerobic oxidative
coupling of ferulate in acetonitrile. To the best of our
knowledge this is the first time that up to five pure diferulates
have been isolated from one oxidative coupling reaction by
flash chromatography; a simple TBAF treatment allows
chromatographic separation of the 8−O−4-diferulate 4 from
8−5-diferulate 2, with which it normally cochromatographs,
and conveniently converts 2 to another valuable product, the
open form of 8−5-diferulate 2, that is, 3, which is useful for
cleanly producing its required carboxylic acid analogue. This
synthetic approach allows diferulates needed for plant research
to be prepared and isolated on a useful scale, without the need
for skilled synthetic organic chemistry expertise.
The 8−8-coupled diferulates (5 and 6) and their
corresponding acids have been synthesized via multiple-step
routes starting with ferulic acid;3 the syntheses involved making
the intermediate dilactone via oxidation of ferulic acid following
a laborious workup. It has been reported that diferulate 5 can
be obtained in 30% yield from radical coupling of ferulate by
ferric chloride oxidation in aqueous acetone.28 However, we
were not able to obtain the claimed yield for diferulate 5 by
following the exact procedure described in that paper. Instead,
only 5% of diferulate 5 was isolated because huge amounts of
unreacted starting materials remained. Diferulate 5 was
detected in the coupling reaction products resulting from
CuCl(OH)−TMEDA oxidation by method A, but isolation of
pure 5 by flash chromatography was unsuccessful because of
the extremely low yield produced. However, the CuCl(OH)−
TMEDA catalyzed oxidative coupling of ethyl ferulate (method
B) produced a much higher yield of diferulate 5, and the pure 5
was isolated here in 12% yield by flash chromatography.
Diferulate 6 is another 8−8-coupled product, and the acid
analogue has been frequently found in alkaline hydrolysates of
grass cell walls and cereal bran since it was found and identified
using the synthetic compound as a reference.3 It was reported
that diferulate 6 could be produced in 9% yield from oxidation
of ethyl ferulate with alkaline potassium ferricyanide in a
benzene−water two-phase system.29 In this work, we were able
to obtain pure diferulate 6 in 5.6% yield by a simple flash
chromatographic separation of products from CuCl(OH)−
TMEDA oxidation of ethyl ferulate (method A), in addition to
8−5-coupled diferulate 2, 8−O−4-coupled diferulate 4, and 5−
5-coupled diferulate 7 produced from the same reaction.
Similar yields of diferulate 6 can be obtained by executing the
coupling reaction catalytically (method B), although a second
purification by chromatography was needed to obtain pure 6.
The synthesis of 8−O−4-coupled diferulate 4 was formerly
the most challenging task, where up to nine steps were required
to make diferulate 4 from vanillin and coniferaldehyde.3 So far,
this synthetic strategy remains the only way to produce 4 on a
preparative scale. Although silver(I) oxide has been used
frequently to make diferulate, there was only one report that
diferulate 4 was obtained, in 17% yield from silver(I) oxide
oxidation of methyl ferulate.30 Unfortunately, we were not able
to repeat the claimed results by following the published
protocol with methyl or ethyl ferulate as substrate. In this study
we demonstrated that diferulate 4, produced by CuCl(OH)−
TMEDA oxidation of ferulate 1, can be isolated in two steps.
Specifically, flash chromatography was first used to fractionate
the oxidation products, producing a fraction containing
diferulates 2 and 4 (Figure 1). This fraction mixture was
allowed to crystallize in hexane−ethyl acetate, resulting in white
needles of diferulate 2. Then the residues after crystallization
were treated with TBAF in acetonitrile to convert diferulate 2
into 3 while diferulate 4 remained intact. Diferulate 4 is readily
separated from 3; flash chromatographic separation led to
isolation of the pure diferulate 4 in a yield of 4%, along with a
small amount of diferulate 3.
AUTHOR INFORMATION
■
Corresponding Author
*Telephone: +1 (608) 890-2552. Fax: +1 (608) 265-2904. E-
mail: fachuanglu@wisc.edu.
Funding
This work was funded by the DOE Great Lakes Bioenergy
Research Center (DOE BER Office of Science DE-FC02-
07ER64494).
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Dr. Ifat Parveen (Aberystwyth University, UK) for
using and testing this method and Dr. Mirko Bunzel
(University of Minnesota) for his constructive comments.
REFERENCES
■
(1) Hartley, R. D.; Ford, C. W. Phenolic constituents in plant cell
walls and wall biodegradability. In Plant Cell Wall Polymers, Biogenesis
and Biodegradation; Lewis, N. G., Paice, M. G., Eds.; ACS Symposium
Series 399; American Chemical Society: Washington, DC, 1989; pp
137−145.
(2) Hatfield, R. D.; Ralph, J.; Grabber, J. H. Cell wall cross-linking by
ferulates and diferulates in grasses. J. Sci. Food Agric. 1999, 79, 403−
407.
Although diferulate 3 is not a product resulting directly from
coupling reactions of ferulate, the free acid derived from
8276
dx.doi.org/10.1021/jf302140k | J. Agric. Food Chem. 2012, 60, 8272−8277

Journal of Agricultural and Food Chemistry
Article
(3) Ralph, J.; Quideau, S.; Grabber, J. H.; Hatfield, R. D.
Identification and synthesis of new ferulic acid dehydrodimers present
in grass cell walls. J. Chem. Soc., Perkin Trans. 1 1994, 3485−3498.
(4) Grabber, J. H. How do lignin composition, structure, and cross-
linking affect degradability? A review of cell wall model studies. Crop
Sci. 2005, 45, 820−831.
(23) Noji, M.; Nakajima, M.; Koga, K. A new catalytic system for
aerobic oxidative coupling of 2-naphthol derivatives by the use of
CuCl-amine complex − a practical synthesis of binaphthol derivatives.
Tetrahedron Lett. 1994, 35, 7983−7984.
(24) Lipshutz, B. H.; James, B.; Vance, S.; Carrico, I. A potentially
general intramolecular biaryl coupling approach to optically pure 2,2′-
BINOL analogs. Tetrahedron Lett. 1997, 38, 753−756.
(5) Buanafina, M. M. D. Feruloylation in grasses: current and future
perspectives. Mol. Plant 2009, 2, 861−872.
(25) Markwalder, H. U.; Neukom, H. Diferulic acid as a possible
crosslink in hemicelluloses from wheat germ. Phytochemistry 1976, 15,
836−837.
(6) Fry, S. C. Phenolic components of the primary cell wall and their
possible role in the hormonal regulation of growth. Planta 1979, 146,
343−351.
(7) Ikegawa, T.; Mayama, S.; Nakayashiki, H.; Kato, H. Accumulation
of diferulic acid during the hypersensitive response of oat leaves to
Puccinia coronata f. sp. avenae and its role in the resistance of oat
tissues to cell wall degrading enzymes. Physiol. Mol. Plant Pathol. 1996,
48, 245−255.
(26) Bunzel, M.; Ralph, J.; Marita, J. M.; Steinhart, H. Identification
of 4-O-5′-coupled diferulic acid from insoluble cereal fiber. J. Agric.
Food Chem. 2000, 48, 3166−3169.
(27) Yamamoto, H.; Hoshino, T.; Uchiyama, Y. Convenient
preparation and quantification of 5,5′-diferulic acid. Biosci., Biotechnol.,
Biochem. 1999, 63, 390−394.
(28) Kuo, Y. H.; Wu, C. H. Synthesis of 5-(3-hydroxypropyl)-7-
methoxy-2-(3′-methoxy-4′-hydroxyphenyl)-3-benzo[b]-
furancarbaldehyde, a novel adenosine A(1) receptor ligand from the
root of Salvia miltiorrhiza. J. Nat. Prod. 1996, 59, 625−628.
(29) Wang, Q.; Yang, Y.; Li, Y.; Yu, W.; Hou, Z. H. An efficient
method for the synthesis of lignans. Tetrahedron 2006, 62, 6107−6112.
(30) Rakotondramanana, D. L. A.; Delomenede, M.; Baltas, M.;
Duran, H.; Bedos-Belval, F.; Rasoanaivo, P.; Negre-Salvayre, A.;
Gornitzka, H. Synthesis of ferulic ester dimers, functionalisation and
biological evaluation as potential antiatherogenic and antiplasmodial
agents. Bioorg. Med. Chem. 2007, 15, 6018−6026.
(8) Jacquet, G.; Pollet, B.; Lapierre, C.; Mhamdi, F.; Rolando, C.
New ether-linked ferulic acid-coniferyl alcohol dimers identified in
grass straws. J. Agric. Food Chem. 1995, 43, 2746−2751.
(9) Ralph, J.; Grabber, J. H.; Hatfield, R. D. Lignin-ferulate crosslinks
in grasses: active incorporation of ferulate polysaccharide esters into
ryegrass lignins. Carbohydr. Res. 1995, 275, 167−178.
(10) Ou, S. Y.; Kwok, K. C. Ferulic acid: pharmaceutical functions,
preparation and applications in foods. J. Sci. Food Agric. 2004, 84,
1261−1269.
(11) Somerville, C.; Youngs, H.; Taylor, C.; Davis, S. C.; Long, S. P.
Feedstocks for lignocellulosic biofuels. Science 2010, 329, 790−792.
(12) Grabber, J. H.; Mertens, D. R.; Kim, H.; Funk, C.; Lu, F.; Ralph,
J. Cell wall fermentation kinetics are impacted more by lignin content
and ferulate cross-linking than by lignin composition. J. Sci. Food Agric.
2009, 89, 122−129.
(13) Barberousse, H.; Roiseux, O.; Robert, C.; Paquot, M.;
Deroanne, C.; Blecker, C. Analytical methodologies for quantification
of ferulic acid and its oligomers. J. Sci. Food Agric. 2008, 88, 1494−
1511.
(14) Neudorffer, A.; Desvergne, J. P.; Bonnefont-Rousselot, D.;
Legrand, A.; Fleury, M. B.; Largeron, M. Protective effects of 4-
hydroxycinnamic ethyl ester derivatives and related dehydrodimers
against oxidation of LDL: radical scavengers or metal chelators? J.
Agric. Food Chem. 2006, 54, 1898−1905.
(15) Bunzel, M.; Funk, C.; Steinhart, H. Semipreparative isolation of
dehydrodiferulic and dehydrotriferulic acids as standard substances
from maize bran. J. Sep. Sci. 2004, 27, 1080−1086.
(16) Neudorffer, A.; Deguin, B.; Hamel, C.; Fleury, M. B.; Largeron,
M. Electrochemical oxidative coupling of 4-hydroxycinnamic ester
derivatives: a convenient methodology for the biomimetic synthesis of
lignin precursors. Collect. Czech. Chem. Commun. 2003, 68, 1515−
1530.
(17) Ralph, J.; Garcia-Conesa, M. T.; Williamson, G. Simple
preparation of 8−5-coupled diferulate. J. Agric. Food Chem. 1998, 46,
2531−2532.
(18) Ralph, S. A.; Landucci, L. L.; Ralph, J. NMR Database of Lignin
and Cell Wall Model Compounds, available at http://ars.usda.gov/
Services/docs.htm?docid=10429 (previously http://www.dfrc.ars.usda.
gov/software.html), updated at least annually since 1993 (accessed
May, 2012).
(19) Ralph, J.; Bunzel, M.; Marita, J. M.; Hatfield, R. D.; Lu, F.; Kim,
H.; Schatz, P. F.; Grabber, J. H.; Steinhart, H. Peroxidase-dependent
cross-linking reactions of p-hydroxycinnamates in plant cell walls.
Phytochem. Rev. 2004, 3, 79−96.
(20) Feringa, B.; Wynberg, H. Oxidative phenol coupling with cupric-
amine complexes. Tetrahedron Lett. 1977, 4447−4450.
(21) Shibasaki, Y.; Suzuki, Y.; Ueda, M. Copper-catalyzed regio-
controlled oxidative coupling polymerization of 2,5-dimethylphenol.
Macromolecules 2007, 40, 5322−5325.
(22) Kushioka, K. Autoxidation of phenols catalyzed by copper(II)-
ethylenediamine complexes − the reaction mechanism. J. Org. Chem.
1984, 49, 4456−4459.
8277
dx.doi.org/10.1021/jf302140k | J. Agric. Food Chem. 2012, 60, 8272−8277