Preparation of synthetic lignins with superior NMR characteristics via isotopically labeled monolignols

Article abstract of DOI:10.1039/a803281e

Synthetic lignins are particularly valuable for studying aspects of lignification, plant cell wall cross-linking, and lignin structure. If they are not too highly polymeric, they are soluble in normal lignin solvents and amenable to solution-state NMR studies. However, in the application of inverse-detected correlation experiments, particularly the popular HMQC and HMBC experiments, the spectra have annoying T₁-noise ridges. These artifacts make it difficult to locate correlation peaks that are near the methoxy signal in the proton dimension. One solution is to use gradient-enhanced NMR but that requires additional hardware that is not yet ubiquitous. An alternative is to produce monolignols in which the atoms of the methoxy group are NMR-invisible. We have accomplished this by preparing coniferyl and sinapyl alcohols using ¹³C-depleted deuterated methyl iodide (¹²C²H₃I). The methods, which incorporate steps simpler than have been used previously for labeled monolignols, are sufficiently low cost and straight-forward that these monomers can be utilized for any synthetic lignins destined for NMR studies. The NMR spectra of lignins derived from these 'methoxy-less' monomers are markedly superior to their normal-monomer counterparts. Several popular NMR experiments are illustrated for synthetic lignins derived from normal vs. isotopically labeled coniferyl alcohol, along with some useful experiments that have not been seen in lignin-related publications to date.

Full text of DOI:10.1039/a803281e

View Article Online / Journal Homepage / Table of Contents for this issue
Preparation of synthetic lignins with superior NMR characteristics
via isotopically labeled monolignols
John Ralph,*,a,b Yingsheng Zhangb,c and Richard M. Eded
a US Dairy Forage Research Center, USDA Agricultural Research Service,
Madison, WI 53706-1108
b Department of Forestry, University of Wisconsin, Madison, WI 53706-1598
c School of Pharmacy, University of Wisconsin, Madison, WI 53706-1515
d Department of Chemistry, University of Waikato, Hamilton, New Zealand
Synthetic lignins are particularly valuable for studying aspects of lignification, plant cell wall cross-linking,
and lignin structure. If they are not too highly polymeric, they are soluble in normal lignin solvents and
amenable to solution-state NMR studies. However, in the application of inverse-detected correlation
experiments, particularly the popular HMQC and HMBC experiments, the spectra have annoying T1-
noise ridges. These artifacts make it difficult to locate correlation peaks that are near the methoxy signal
in the proton dimension. One solution is to use gradient-enhanced NMR but that requires additional
hardware that is not yet ubiquitous. An alternative is to produce monolignols in which the atoms of
the methoxy group are NMR-invisible. We have accomplished this by preparing coniferyl and sinapyl
alcohols using 13C-depleted deuterated methyl iodide (12C2H3I). The methods, which incorporate steps
simpler than have been used previously for labeled monolignols, are sufficiently low cost and straight-
forward that these monomers can be utilized for any synthetic lignins destined for NMR studies. The
NMR spectra of lignins derived from these ‘methoxy-less’ monomers are markedly superior to their
normal-monomer counterparts. Several popular NMR experiments are illustrated for synthetic
lignins derived from normal vs. isotopically labeled coniferyl alcohol, along with some useful
experiments that have not been seen in lignin-related publications to date.
instruments, an attractive solution for synthetic lignins is simple
Introduction
labeling that makes the methoxy NMR-invisible. Here we
report on the synthesis of the two methoxylated monolignols,
coniferyl and sinapyl alcohols, with the methoxy groups deu-
terated and 13C-depleted. Such labeling greatly improves the
appearance and interpretability of spectra as illustrated by
comparison of 2D NMR spectra of synthetic lignins prepared
from normal vs. the labeled coniferyl alcohol.
Despite its rather random and heterogeneous nature, the lignin
polymer discloses many of its intimate structural details to
diagnostic NMR experiments. 13C NMR established itself early
as a method for detailed structural characterization, aided by
the almost exact agreement between chemical shifts of carbons
in good low-molecular-mass model compounds and in the
polymer.1,2 Although this correspondence cannot generally be
expected for proton NMR, good lignin model compounds and
their counterpart units in the polymer also closely match proton
NMR chemical shifts.3–5 Despite the broad featureless 1D
proton spectra, which have nevertheless allowed substantive
interpretation,6 2D spectra, both homo- and hetero-nuclear,
are informationally rich and strikingly useful.
The advent of inverse-detection, and the enabling of com-
mercial spectrometers to easily carry out these experiments,
provided a welcome sensitivity enhancement for the difficult
lignin polymers, but not without some drawbacks. The major
problems in the inverse experiments, particularly the popular
HMBC7 and HMQC8 experiments, are related to T1-noise arti-
facts which are particularly obstructive in lignins due to their
intense methoxy signals. Methoxy resonances are relatively
sharp and intense in both the proton domain (where the 3–6
methoxy protons per monolignol unit cause the methoxy peak
to dominate proton spectra) and the carbon domain (where the
narrow methoxy chemical shift range causes a similarly domin-
ant methoxy carbon peak). The result in proton-detected 2D
spectra is an almost complete obliteration of information from
protons resonating near the methoxy. The problem can be
somewhat minimized by using excessively long relaxation delays
but only at the expense of valuable signal-to-noise in a given
time. Although this problem can now be obviated by the use
of pulsed field gradients on appropriately configured newer
Results and discussion
An elegant solution to the methoxy region T1-noise NMR arti-
fact problem is possible for synthetic lignins, without requiring
gradient NMR methods. The solution is to make the methoxy
groups NMR-invisible, but normal in every other regard. This
can be accomplished by replacing the normal methoxy carbon
(which has ~1% natural abundance of NMR-active 13C) by
the 12C isotope, which is not NMR-active, and replacing the
protons with deuterons using a 13C-depleted trideuteromethyl
reagent.
Coniferyl alcohol 1a with NMR-invisible methoxy groups
was synthesized as shown in Scheme 1a, using the relatively
cheap 13C-depleted trideuteromethyl iodide as the key reagent.
Other notable steps in the synthesis are the use of the tetra-
hydropyranyl group for the required regioselective phenol
mono-protection of catechol 2, and the direct conversion of
bromoguaiacol 5 to vanillin 6 is achieved via a halogen–metal
interchange and subsequent formylation using dimethyl-
formamide.9 These steps are considerable improvements over
prior methods (directed o-hydroxylation following 2-chloro-5-
nitrobenzophenone protection10 and the arduous route from
bromide 5 via the acid through CO2 addition to the anion,
chlorination, and conversion of the acid chloride to the
required aldehyde 6).11
J. Chem. Soc., Perkin Trans. 1, 1998
2609

View Article Online
EtO
O
OH
γ
(a)
β
H
O
α
Br
1
2
a
b
c
d
e
f
3
OH
OH
O12CD3
O12CD3
O12CD3
O12CD3
O
12CD3
OH
OTHP
OTHP
OH
OH
OH
7a
OH
1a
2
3
4
5
6
EtO
O
(b)
H
O
H
O
H
O
g
c
h
HO
OH
HO
OH
Ph
12CD3O
O
12CD3
12CD3O
O12CD3
Ph
OH
O
O
Ph
O
8
9
10
11
i
OH
EtO
O
γ
β
α
f
1
2
3
12CD3O
O
12CD3
12CD3O
O
12CD3
OH
OH
1b
7b
R
O
O
(c)
HO
Ar
γ
γ
γ
β
β
HO
OCH3
HO
Ar
β
O
α
Ar
α
α
O
Ar
Ar
1a, 1b
A
B
C
R
HO
OH
γ
γ
R
β
OMe
OH
O
Ar′
β
α
Ar′′
O
α
O
O
Ar
Ar
β
α
γ
OMe
Ar
D
E
X
Scheme 1 Synthetic schemes for compounds 1, and polymerization to synthetic lignins containing structures A–E and X. Reagents and conditions: a,
3,4-dihydro-2H-pyran, pyridinium toluene-p-sulfonate (PPTS), CH2Cl2, rt; b, 12CD3I, K2CO3, acetone, 43 ЊC; c, (i) PPTS, EtOH, 55 ЊC, (ii) KBrO3,
HBr, HOAc, rt; d, (i) ethyl vinyl ether (EVE), PPTS, CH2Cl2, rt, (ii) n-BuLi, DMF, THF, Ϫ78 ЊC; (iii) 2  HCl, rt; e, (i) EVE, PPTS, CH2Cl2, rt; (ii)
triethyl phosphonoacetate, NaH, THF, rt; (iii) AcOH, MeOH; f, (i) DIBAL, toluene, 0 ЊC; (ii) HCl, THF; g, PhCH2Cl, acetone, K2CO3, KI, reflux,
overnight; h, (i) triethyl phosphonoacetate, NaH, THF, rt; (ii) AcOH, MeOH; (iii) HCl, THF; i, TMSI, CH3CN; j, peroxidase, H2O2, pH 6.5.
Standard lignin numbering is used. Lignin structures are: A, β-aryl ether (β–O–4); B, phenylcoumaran (β–5); C, resinol (β–β); D, dibenzodioxocin
(5–5 and β–O–4/α–O–4); E, α,β-diether (β–O–4/α–O–4); X, cinnamyl alcohols.
Specifically labeled sinapyl alcohol 1b was prepared from
commercial 3,4,5-trihydroxybenzaldehyde 8 by regioselective
mono-benzylation (directed by the aldehyde group) followed
by similar methylation and subsequent conversions, Scheme
1b. Removal of the benzyl group used TMSI;12 normal hydro-
genolysis is incompatible with the double bond.
As shown in Fig. 1, a variety of 2D NMR spectra of a
synthetic lignin produced from the specialized coniferyl alcohol
1a are markedly superior to an equivalent lignin made from
normal coniferyl alcohol. The spectra also illustrate the value
of spectral editing in 2D NMR experiments that has not
yet attracted much attention from lignin chemists. The DEPT-
HMQC13 experiments can aid assignment by selectively invert-
ing (for example) -CH2 resonances (Fig. 1C). The HMQC-
TOCSY14 experiment is still under utilized in plant cell wall
research. It provides beautifully redundant correlations that
greatly facilitate rapid assignment of lignin spectra, particularly
in the sidechain region.
The freedom from rapid-pulsing artifacts in these spectra
allows detailed assignments to be made; all of the major
subunits are assigned in selected spectra of Fig. 1. Such
assignments aid the characterization of real plant lignins.
2610
J. Chem. Soc., Perkin Trans. 1, 1998

View Article Online
Fig. 1 2D NMR spectra of synthetic lignins made using (left) normal coniferyl alcohol and (right) ‘methoxy-less’ coniferyl alcohol. A, HMQC; B,
DEPT-HMQC with a π/3 editing pulse (CH’s up, CH3’s and CH2’s down); C, DEPT-HMQC with a π editing pulse (CH’s and CH3’s up, CH2’s
down); D, HMQC-TOCSY; E, HMBC. Dashed lines are negative contours. Assignments: on HMQC, structure (A–E, X) followed (smaller size) by
sidechain C–H pair (e.g. Bβ is the structure B C-β/H-β correlation); on HMQC-TOCSY, structure followed by carbon excited in HMQC part followed
by proton from TOCSY part (e.g. Bβα is the structure B C-β/H-α correlation); on HMBC, structure followed by carbon followed by coupled proton
within 3-bonds (e.g. Bβα is the structure B C-β/H-α long-range correlation).
Rendering the normally intense methoxy groups NMR-
invisible also dramatically improves homonuclear 2D experi-
ments. Such isotopically labeled lignins have further value for
studies using mass spectrometry, where the Mϩ3 or Mϩ6 ions
(from the three deuterons in 1a or the 6 deuterons in 1b) allow
use of these synthetic lignins as internal standards or markers
for natural lignin studies.
J. Chem. Soc., Perkin Trans. 1, 1998
2611

View Article Online
and HBr (2.58 mL of 30% HBr in HOAc, 12.97 mmol) were
Experimental
added. The mixture was stirred at room temperature for
4 h, diluted with 100 ml H2O, and then extracted twice with
EtOAc. The organic phase was washed with sat. aq. NaHCO3
to neutralize the solution, then washed with aq. NaCl, dried
over MgSO4, and the solvent evaporated to give a syrup (1.21 g,
5.87 mmol, 77%).
General
Reagents were all purchased from Aldrich Chemical Co.
(Milwaukee, WI). Liquid reagents were distilled prior to use;
solids were used as supplied. Light petroleum was the 40–60 ЊC
boiling fraction. Reactions were monitored by thin layer
chromatography (TLC) using Alugram Sil-G/UV254 plates
(Macherey-Nagel); visualization was by UV light. Merck silica
gel 60 (70–230 mesh) was used for flash chromatography.
Melting points were on an Electrothermal digital mp apparatus
and are uncorrected. High-resolution MS was on a Kratos
MS-80RFA spectrometer.
2
Vanillin, 4-hydroxy-3-[12C, H3]methoxybenzaldehyde 6. The
phenol in 5 was first protected as its ethoxyethylidene deriv-
ative.11 Compound 5 (1.21 g, 5.87 mmol) and a catalytic
amount of pyridinium toluene-p-sulfonate were dissolved in
dichloromethane, and then ethyl vinyl ether (1.8 mL, 18.8
mmol) was added. The mixture was stirred at room temperature
for 4 h, after which time TLC indicated that the reaction was
complete. Work-up11 gave the protected bromoguaiacol as a
syrup (1.567 g, 5.63 mmol, 96%), which was then formylated.9
Protected 5 was dissolved in dry THF under nitrogen and
cooled to Ϫ78 ЊC, and n-BuLi (3.38 mL, 2.5  in hexane, 16.9
mmol) added. The solution was stirred for 45 min at Ϫ78 ЊC,
following which dry DMF (1.31 mL, 16.9 mmol) was added.
Vigorous stirring was maintained for 2 h, then the solution was
allowed to warm to room temperature, and the reaction
quenched with 2  HCl (5 mL). After an additional 1 h, the
solvent was evaporated and the syrup was extracted with
EtOAc, washed with sat. aq. NaHCO3 and NaCl sequentially,
dried over MgSO4, and the solvent evaporated under reduced
pressure. Crystallization from acetone–H2O gave pure 6 as
white needle crystals (706 mg, 4.56 mmol, 78%).
NMR Spectra were taken on a Bruker AMX 360 MHz
instrument (Bruker instruments, Billerica, MA) fitted with a
5 mm 4-nucleus (QNP) probe with normal geometry (proton
coils furthest from the sample) using standard Bruker pulse
programs. Samples were in 0.4 mL of acetone-d6, with the
central solvent peak as internal reference (δH 2.04, δC 29.8).
J Values are given in Hz. The carbon/proton designations are
based on the standard lignin numbering system (Scheme 1).
NMR assignments were authenticated by the usual comple-
ment of 1D and 2D NMR experiments. Benzyl group aromatic
resonances are not reported. Phase-sensitive short-range 2D
correlation spectra (HMQC,8 DEPT-HMQC13 and HMQC-
TOCSY14) were acquired from ~10–140 ppm using 1k FIDs
with 256 increments and 96–128 scans per increment. These
acquisitions were for high S/N spectra and were run excessively
long–remarkably good HMQC spectra with all correlations
fully represented can be obtained in under 20 minutes from
these samples (under 4 minutes using gradient selected experi-
ments). The TOCSY sequence used an 80 ms mixing period.
Apodization was by Gaussian multiplication, and the final
matrix data size was 1k by 1k real points (corresponding to
resolutions of 3.6 and 11.4 Hz per pt in the proton and carbon
domains). Magnitude-mode long-range correlation (HMBC7)
spectra were acquired from ~10–200 ppm using 2k FIDs
with 256 increments and 480 scans per increment, and a long-
range coupling delay of 100 ms (J = 5 Hz). Apodization was by
Gaussian multiplication, and the final matrix data size was
1k by 1k real points (corresponding to resolutions of 3.6 and
17.4 Hz per pt in the proton and carbon domains).
Ethyl ferulate, ethyl 4-hydroxy-3-[12C, 2H3]methoxycinnamate
7a. The ethyl ferulate was made via a Horner–Emmons reaction
as described previously, again following protection as the
ethoxyethylidene derivative.11 Thus, triethyl phosphonoacetate
(0.90 mL, 4.54 mmol) was ionized with NaH (163 mg, 6.8
mmol) and the ethoxyethylidene derivative of 6 (1.03 g, 4.54
mmol) added to give 7a (915 mg, 4.07 mmol, 90%) which was
directly used in the next step without further purification.
Coniferyl alcohol, 4-hydroxy-3-[12C,2H3]methoxycinnamyl
alcohol 1a. Compound 7a was reduced as has been described for
the preparation of coniferyl alcohol from ethyl ferulate.16 Thus,
7a (915 mg, 4.07 mmol) was reduced with DIBAL (11.4 mL of
1.5  solution in toluene, 17.1 mmol) and worked up to give
crude 1a as an oil (730 mg, 3.99 mmol, 98%). Crystallization
from CH2Cl2–light petroleum gave 1a as colorless plates (540
mg, 2.95 mmol, 73%); δH 3.69 (1H, t, J 5.7, γ-OH), 4.18 (2H, td,
J ~5.6, 1.6, H-γ), 6.21 (1H, dt, J 15.9, 5.6, H-β), 6.49 (1H, dt, J
15.9, 1.6, H-α), 6.75 (1H, d, J 8.1, H-5), 6.84 (1H, dd, J 8.1, 2.0,
H-6), 7.04 (1H, d, J 2.0, H-2), 7.53 (1H, s, Ar-OH) ; δC 63.4 (γ),
109.9 (2), 115.5 (5), 120.6 (6), 128.0 (β), 130.2 (1), 130.5 (α),
Synthesis of labeled coniferyl alcohol 1a
2-Tetrahydropyranyloxyphenol 3. Catechol 2 (10 g, 91 mmol),
3,4-dihydro-2H-pyran (8.56 mL, 91 mmol) and pyridinium
toluene-p-sulfonate (297 mg, 1.2 mmol) were dissolved in
dichloromethane. The mixture was stirred at room temperature
for 3 h, after which time TLC (CHCl3–EtOAc, 9:1) indicated
that the reaction was complete. The solvent was removed in
vacuo and the syrup dissolved in EtOAc, washed twice with sat.
aq. NaHCO3, once with sat. aq. NaCl, dried over MgSO4 and
the solvent evaporated under vacuum to give compound 2 as a
light yellow liquid (16.65 g, 86.2 mmol, 95%).
2
147.2 (4) 148.5 (3); m/z 183.0967 (69%) (C10H9 H3O3 requires
183.0975).
Synthesis of labeled sinapyl alcohol 1b
4-Benzyloxy-3,5-dihydroxybenzaldehyde 9. 3,4,5-Trihydroxy-
benzaldehyde 8 (3.0 g, 17.4 mmol) was dissolved in anhydrous
acetone. Benzyl chloride (2.0 mL, 17.4 mmol) was added, fol-
lowed by anhydrous K2CO3 (4.0 g, 29 mmol) and a cat. amount
of KI. The mixture was refluxed overnight and completeness
was checked by TLC (EtOAc–light petroleum, 2:1). The
inorganic salts were filtered off and the filtrate was diluted with
water and extracted with EtOAc (3 × 100 mL), washed with sat.
aq. NaCl, dried over Na2SO4 and the solvent evaporated under
reduced pressure to give a yellow solid. Flash chromatography
(EtOAc–light petroleum 2:1) afforded 9 as light yellow crystals
(3.78 g, 15.5 mmol, 89%): δH 5.22 (2H, s, CH2OPh), 6.97 (2H, s,
H-2/6), 8.53 (2H, s, 3,4-OHs), 9.77 (1H s, α); δC 74.6 (ArOCH2),
109.7 (2/6), 133.6 (1), 140.1 (4), 152.1 (3/5), 191.7 (α).
1-Tetrahydropyranyloxy-2-[12C, 2H3]methoxybenzene 4. Com-
pound 3 (2.00 g, 10.35 mmol) and powdered anhydrous K2CO3
(2.29 g, 16.6 mmol) were dissolved in anhydrous acetone and
12CD3I (0.71 mL, 11.4 mmol) was added. The mixture was
stirred at 41 ЊC overnight. The reaction was checked for com-
pleteness by TLC (CHCl3–EtOAc, 9:1). The inorganic salts
were filtered off, and the filtrate evaporated to give 4 as a syrup
(1.77 g, 8.56 mmol, 83%).
4-Bromoguaiacol, 4-bromo-2-[12C,2H3]methoxyphenol 5.
Direct bromination with molecular bromine in HOAc is satis-
factory, but we preferred to generate the bromine in situ.15
Compound 4 (1.77 g, 8.56 mmol) was dissolved in EtOH and a
catalytic amount of pyridinium toluene-p-sulfonate was added.
The mixture was stirred at 55 ЊC for 3 h and evaporated at 35 ЊC
in vacuo to afford a syrup (0.97 g, 7.63 mmol), which was dis-
solved in 20 mL glacial acetic acid. KBrO3 (427 mg, 2.54 mmol)
4-Benzyloxy-3,5-bis([12C,2H3]methoxy)benzaldehyde
10.
Compound 9 (0.98 g, 8.0 mmol) and anhydrous K2CO3 (2.43 g,
17.6 mmol) were dissolved in anhydrous acetone and 12CD3I
(1.0 mL, 15.95 mmol) was added. The mixture was refluxed
2612
J. Chem. Soc., Perkin Trans. 1, 1998

View Article Online
overnight and the inorganic salts filtered off. The filtrate was
concentrated in vacuo to give 10 as a syrup (2.1 g, 7.6 mmol),
which was directly used in next step without further purifi-
cation. TLC-purification was used before characterizing the
compound. Compound 10: δH 5.09 (2H, s, CH2OPh), 7.21 (2H,
s, H-2/6), 9.88 (1H, s, α); δC 75.1 (ArOCH2), 107.3 (2/6), 133.1
(1), 143.1 (4), 154.9 (3/5), 191.6 (α).
Preparation of a synthetic lignin from labeled coniferyl alcohol
1a
The synthetic lignins were prepared from coniferyl alcohol
using horseradish peroxidase at pH 6.5, as previously
described.17,18 NMR Spectra here are of acetylated samples,
95 mg, in 400 µl acetone-d6.
Ethyl 4-benzyloxy-3,5-bis([12C, 2H3]methoxy)cinnamate 11.
The ethyl sinapate was made via a Horner-Emmons reaction as
described previously.11 Thus, triethyl phosphonoacetate (1.7
mL, 8.4 mmol) was ionized with NaH (273.6 mg,11.4 mmol)
in dry THF (15 mL) and compound 10 (2.1 g, 7.6 mmol)
added. Work-up gave 11 as a light yellow solid (2.58 g, 7.1
mmol), which was directly used in the next step without further
purification. Compound 11: δH 1.28 (3H, t, J 7.1, CH3CH2O),
4.20 (2H, q, J 7.1, CH3CH2O), 5.01 (2H, s, CH2OPh), 6.49
(d, 1H, J 15.9, H-β), 7.02 (2H, s, H-2/6), 7.59 (1H, d, J 15.9,
H-α); δC 14.6 (CH3CH2O), 60.6 (CH3CH2O), 75.1 (ArOCH2),
106.8 (2/6), 118.3 (β), 131.0 (1), 140.0 (4), 145.2 (α), 154.8 (3/5),
167.1 (γ).
Acknowledgements
The authors are grateful to prior work by Stéphane Quideau
(now at Texas Tech. University, Lubbock, TX) which produced
the normal synthetic lignin used for comparisons, to Paul
Schatz (U. Wisconsin-Madison) for alerting us to the bromin-
ation method using KBrO3, to USDA-NRI competitive grants
(# 96-35304 in the plant growth and development section) for
partial funding, and to the USDA-ARS for funding of the
NMR instrumentation essential to this work.
References
Ethyl sinapate, ethyl 4-hydroxy-3,5-bis([12C, 2H3]methoxy)-
cinnamate 7b. Compound 11 (2.58 g, 7.1 mmol) was dissolved
in anhydrous CH3CN (50 mL) and iodotrimethylsilane (5.1 mL,
35 mmol) was added under nitrogen. The mixture was stirred
at 50 ЊC for 3 h, after which time TLC (EtOAc–hexane, 1:1)
indicated that the reaction was complete. The reaction was
quenched with methanol (4.5 mL, 3 equiv.), taken up in H2O
(100 mL) and extracted with EtOAc (3 × 100 mL) and the com-
bined organic layer was dried over MgSO4 and concentrated
in vacuo to afford a brown yellow solid, which was purified by
flash chromatography (EtOAc–hexane, 1:1) to give 7b as color-
less plate crystals (1.31 g, 4.9 mmol, 70%); δH 1.27 (3H, t, J 7.1,
CH3CH2O), 4.18 (2H, q, J 7.1, CH3CH2O), 6.40 (1H, d, J 15.9,
H-β), 7.00 (2H, s, H-2/6), 7.56 (1H, d, J 15.9, H-α), 7.69 (1H,
s, 4-OH); δC 14.6 (CH3CH2O), 60.5 (CH3CH2O), 106.8 (2/6),
116.2 (β), 126.2 (1), 139.4 (4), 145.8 (α), 148.9 (3/5), 167.4 (γ).
1 H.-D. Lüdemann and H. Nimz, Makromol. Chem., 1974, 175,
2393.
2 H.-D. Lüdemann and H. Nimz, Makromol. Chem., 1974, 175,
2409.
3 I. Kilpeläinen, J. Sipilä, G. Brunow, K. Lundquist and R.M. Ede,
J. Agric. Food Chem., 1994, 42, 2790.
4 R. M. Ede and J. Ralph, Magn. Reson. Chem., 1996, 34, 261.
5 J. Ralph, Magn. Reson. Chem., 1993, 31, 357.
6 K. Lundquist, A. Paterson and L. Ramsey, Acta Chem. Scand.,
Ser. B, 1983, 37, 734.
7 A. Bax and M. F. Summers, J. Am. Chem. Soc., 1986, 108, 2093.
8 A. Bax and S. Subramanian, J. Magn. Reson., 1986, 67, 565.
9 E. L. Stogryn, J. Med. Chem., 1973, 16, 1399.
10 K. Kratzl and F. W. Vierhapper, Monatsh. Chem., 1971, 90, 771.
11 J. Newman, R. N. Rej, G. Just and N. G. Lewis, Holzforschung,
1986, 40, 369.
12 F. Nicotra, R. Perego, F. Ronchetti, G. Russo and L. Toma,
Carbohydr. Res., 1984, 131, 180.
13 H. Kessler, P. Schmieder and M. Kurz, J. Magn. Reson., 1989, 85,
400.
2
Sinapyl alcohol, 4-hydroxy-3,5-bis([12C, H3]methoxy)cinna-
myl alcohol 1b. Compound 7b was reduced as has been
described for the preparation of normal sinapyl alcohol from
ethyl sinapate.16 Thus, compound 7b (1.31 g, 4.9 mmol) was
reduced with DIBAL (13 mL of 1.5  solution in toluene, 19.6
mmol) and worked up to give 1b as an oil (1.05 g, 4.8 mmol).
Crystallization from CH2Cl2–hexane gave pure 1b as pale yel-
low needles (864 mg, 4.0 mmol, 82%); δH 3.78 (1H, t, J 5.5, γ-
OH), 4.20 (2H, td, J 5.5, 1.5, H-γs), 6.24 (1H, dt, J 15.8, 5.5, β),
6.49 (1H, dt, J 15.8, 1.5, α), 6.71 (2H, s, H-2/6), 7.20 (1H, s,
ArOH); δC 63.4 (γ), 104.8 (2/6), 128.4 (β), 129.1 (1), 130.6 (α),
14 L. Lerner and A. Bax, J. Magn. Reson., 1986, 69, 375.
15 P. F. Schatz, J. Chem. Ed., 1996, 73, 267.
16 S. Quideau and J. Ralph, J. Agric. Food Chem., 1992, 40, 1108.
17 J. Ralph, R. F. Helm, S. Quideau and R. D. Hatfield, J. Chem. Soc.,
Perkin Trans. 1, 1992, 2961.
18 J. Ralph and Y. Zhang, Tetrahedron, 1998, 54, 1349.
Paper 8/03218E
Received 30th April 1998
Accepted 25th June 1998
2
136.7 (4), 148.8 (3/5); m/z 216.1270 (89%) (C11H8 H6O4 requires
216.1269), 173.1099 (100%).
J. Chem. Soc., Perkin Trans. 1, 1998
2613

View Article Online