Stereochemistry of C18 monounsaturated cork suberin acids determined by spectroscopic techniques including 1H-NMR multiplet analysis of olefinic protons

Article abstract of DOI:10.1002/pca.2491

Introduction Suberin is a biopolyester responsible for the protection of secondary plant tissues, and yet its molecular structure remains unknown. The C_18:1 ω-hydroxyacid and the C_18:1 α,ω-diacid are major monomers in the suberin structure, but the configuration of the double bond remains to be elucidated. Objective To unequivocally define the configuration of the C_18:1 suberin acids. Methods Pure C_18:1 ω-hydroxyacid and C_18:1 α,ω-diacid, isolated from cork suberin, and two structurally very close C_18:1 model compounds of known stereochemistry, methyl oleate and methyl elaidate, were analysed by NMR spectroscopy, Fourier transform infrared (FTIR) and Raman spectroscopy, and GC-MS. Results The GC-MS analysis showed that both acids were present in cork suberin as only one geometric isomer. The analysis of dimethyloxazoline (DMOX) and picolinyl derivatives proved the double bond position to be at C-9. The FTIR spectra were concordant with a cis-configuration for both suberin acids, but their unambiguous stereochemical assignment came from the NMR analysis: (i) the chemical shifts of the allylic ¹³C carbons were shielded comparatively to the trans model compound, and (ii) the complex multiplets of the olefinic protons could be simulated only with ³J_HH and long-range ⁴J_HH coupling constants typical of a cis geometry. Conclusion The two C_18:1 suberin acids in cork are (Z)-18-hydroxyoctadec-9-enoic acid and (Z)-octadec-9-enedoic acid. Copyright 2013 John Wiley & Sons, Ltd. The two C_18:1 cork suberin ω-hydroxyacid and α,ω-diacid were proved to have a cis configuration: (Z)-18-hydroxyoctadec-9-enoic acid and (Z)-octadec-9-enedoic acid. Their double bond stereochemistry was elucidated by spectroscopic techniques, namely ¹H-NMR and ¹³C-NMR. The chemical shifts of the olefinic and allylic protons and of the allylic carbons were diagnostic, and the ¹H-NMR multiplets of the olefinic protons could only be simulated using ³J_HH and long-range ⁴J _HH coupling constants typical of the cis configuration. The revealed stereochemistry of this two major suberin monomers brings a new insight into the mostly unknown suberin macromolecular structure. Copyright

Full text of DOI:10.1002/pca.2491

Research Article
Received: 19 September 2012,
Revised: 23 October 2013,
Accepted: 25 October 2013
Published online in Wiley Online Library: 5 December 2013
(wileyonlinelibrary.com) DOI 10.1002/pca.2491
Stereochemistry of C Monounsaturated Cork
1
8
Suberin Acids Determined by Spectroscopic
1
Techniques Including H-NMR Multiplet
Analysis of Olefinic Protons
Sara Santos* and José Graça
ABSTRACT:
Introduction – Suberin is a biopolyester responsible for the protection of secondary plant tissues, and yet its molecular struc-
ture remains unknown. The C18:1 ω-hydroxyacid and the C18:1 α,ω-diacid are major monomers in the suberin structure, but the
configuration of the double bond remains to be elucidated.
Objective – To unequivocally define the configuration of the C18:1 suberin acids.
Methods – Pure C18:1 ω-hydroxyacid and C
α,ω-diacid, isolated from cork suberin, and two structurally very close C
18:1
18:1
model compounds of known stereochemistry, methyl oleate and methyl elaidate, were analysed by NMR spectroscopy,
Fourier transform infrared (FTIR) and Raman spectroscopy, and GC–MS.
Results – The GC–MS analysis showed that both acids were present in cork suberin as only one geometric isomer. The analysis
of dimethyloxazoline (DMOX) and picolinyl derivatives proved the double bond position to be at C–9. The FTIR spectra were
concordant with a cis-configuration for both suberin acids, but their unambiguous stereochemical assignment came from the
1
3
NMR analysis: (i) the chemical shifts of the allylic C carbons were shielded comparatively to the trans model compound, and
3
4
(
ii) the complex multiplets of the olefinic protons could be simulated only with JHH and long-range JHH coupling constants
typical of a cis geometry.
Conclusion – The two C18:1 suberin acids in cork are (Z)-18-hydroxyoctadec-9-enoic acid and (Z)-octadec-9-enedoic acid.
Copyright © 2013 John Wiley & Sons, Ltd.
Supporting information can be found in the online version of this article.
1
Keywords: Cis/trans configuration; H-NMR multiplet analysis; cork suberin; (Z)-18-hydroxyoctadec-9-enoic acid;
(
Z)-octadec-9-enedoic acid
suberins from various sources differs in the relative proportion of
C18 mid-chain substituted acids, the monounsaturated C18:1 are
always present in significant quantities. For instance, in cork
suberin they amount up to 25% of all C18 mid-chain substituted
acids, but there are cases, such as in potato skin suberin, where
they reach close to 100% (Graça and Pereira, 2000). These high
levels of C18:1 α,ω-diacid and C18:1 ω-hydroxyacid monomers
indicate a major role in the macromolecular structure of all
suberins. The stereochemistry of their mid-chain double bonds will
affect the spatial macromolecular arrangements and therefore the
physical behaviour of the suberins.
Introduction
Cork is a tissue found in the outer bark of plants that exhibit
secondary growth (as trees typically are), ensuring protection
against physical and biotic aggressions, preventing water loss
and affording insulation to internal tissues. Cork cell walls have
a specific polymeric component, suberin, which accounts for
up to 50% of its dry sweight, and is thus believed to be the main
component responsible for the vital barrier properties. Commer-
cial cork is extracted from the cork-oak (Quercus suber L.), the
only known tree able to produce thick layers of continuous cork
tissue in a few years growth. Cork, as a material, has a wide array
of properties of technical interest and, because of that, is used in
an enormous number of applications, some of which are without
viable substitutes. Despite its fundamental role in trees and
plants, and its relevance for cork properties, suberin is still very
poorly understood.
Stereospecific syn-hydroxylation provided indirect evidence
that unsaturated cork suberin acids, that is, the ω-hydroxyacid,
18-hydroxyoctadecenoic acid (Hyd18:1) and the α,ω-diacid,
octadecene-1,18-dioic acid (Di18:1) were present both as cis
and trans isomers (Ribas and Seoane, 1954a; Seoane, 1966).
Suberin is a polyester that, after ester-breaking depolymerisation,
releases a mixture of monomers, including long-chain (C16 to C24
)
* Correspondence to: Sara Santos, Universidade Técnica de Lisboa, Instituto
Superior de Agronomia, Centro de Estudos Florestais, 1349–017 Lisboa,
Portugal.
α,ω-diacids and ω-hydroxyacids and glycerol. The composition of
suberin from different plant sources is variable, but is typically
dominated by C18 α,ω-diacids and ω-hydroxyacids with mid-chain
substituents, which can be an unsaturated, an epoxide or a vicinal-
diol group (Kolattukudy, 2002). Although the composition of
E-mail: sarasantos@isa.utl.pt
Universidade Técnica de Lisboa, Instituto Superior de Agronomia, Centro
de Estudos Florestais, 1349-017 Lisboa, Portugal
Phytochem. Anal. 2014, 25, 192–200
Copyright © 2013 John Wiley & Sons, Ltd.

Stereochemistry of C18:1 Cork Suberin Acids
However, these authors pointed out that the results would be
reliable only if the naturally present suberin acids with mid-
chain diol groups were completely removed prior to the syn-
hydroxylation. In a later work, these same C18:1 acids, but from
potato skin suberin, were tentatively assigned as cis, based
solely in the fact that they were liquid at room temperature
Methyl 18-hydroxyoctadecenoate (Hyd18:1_Me) and dimethyl
,18-octadecenodioate (Di18:1_Me). Di18:1_Me (1.6 mg, 0.54 mmol)
and Hyd18:1_Me (1 mg, 0.31 mmol) were dissolved in pyridine (202 μL
and 130 μL respectively) and derivatised with BSTFA (202 μL and
1
1
30 μL respectively), for 30 min, in an oven at 60 °C.
GC–MS analysis. The analytical samples were injected on a 7890A gas
chromatograph coupled to a 5975C mass spectrometer detector (Agilent
Technologies, Santa Clara, CA, USA) using the following conditions: col-
umn DB5-MS (60 m, internal diameter 0.25 mm, film thickness 0.25 μm);
oven temperature programme was from 200 °C to 300 °C with a heating
rate of 0.3 °C/min, and a helium flow rate of 1 mL/min. Injections were
made in splitless mode, with an injector temperature of 300 °C. Mass
spectrometer conditions: electron ionisation 70 eV; source temperature
(
Rodríguez and Ribas, 1972).
The cis or trans configuration also can be determined using
physical methods (Eliel and Wilen, 1994). The objective of the
present work was the determination of the double bond config-
uration of Hyd18:1 and Di18:1 using two NMR approaches: the
1
13
analysis of H and C chemical shifts of the olefinic protons
and carbons; and the analysis of the splitting pattern of the
complex multiplets that arise from the olefinic protons in the
2
30 °C and quadrupole temperature 150 °C; transfer line temperature
was kept at 310 °C.
1
H-NMR spectra, extracting the coupling constants by computer
simulation. Two C18:1 model compounds, structurally similar to
the C18:1 suberin acid methyl esters, with known double bond
cis and trans stereochemistry were co-analysed, namely methyl
oleate and methyl elaidate. In addition to the one-dimensional
NMR analyses, both C18:1 suberin acids were fully characterised
by two-dimensional NMR, Fourier transform infrared (FTIR) and
Raman spectroscopy, electron-induced mass spectrometry
Double bond position
Picolinyl esters. Picolinyl esters were prepared using a modified
method of Gunstone (1999). Thus, each of the suberin acids (2 mg) was
reacted with oxalyl chloride (1 mL) overnight at room temperature. The
excess of oxalyl chloride was then removed in a warm water bath under
a nitrogen flow.
To the cooled (0 °C, ice bath) acid chlorides was added a cooled (0 °C,
ice bath) solution (1 mL) of 3-hydroxymethylpyridine (HMP, 34.8 mg) in
dichloromethane (1.7 mL). After 30 min at 0 °C, the reaction mixtures
were allowed to warm to room temperature over 2.5 h. The
dichloromethane and excess of HMP were removed in a warm water
bath under a nitrogen flow and the resultant picolinyl esters dried in a
vacuum oven, at 45 °C, over phosphorus pentoxide, overnight. Each
picolinyl ester was then derivatised with pyridine and BSTFA (120 μL of
each per mg of ester), prior to GC–MS analysis.
(EIMS), and the double bond position confirmed using picolinyl
and dimethyloxazoline (DMOX) derivatives.
Experimental
Chemicals and reagents
Methyl oleate (cis-octadec-9-enoic acid methyl ester) and methyl
elaidate (trans-octadec-9-enoic acid methyl ester) with purity above
9
9% were obtained from Sigma-Aldrich (Sintra, Portugal) and used as
DMOX derivatives. The procedure to synthesise the DMOX deriva-
tives was adapted from Harvey (1992). Thus, each suberin acid (3 mg)
was reacted with 2-amino-2-methylpropanol (15.5 mg) for 6 h in an oil
bath at 160 °C. Both reaction mixtures were allowed to cool to room
temperature and subsequently derivatised with BSTFA and pyridine as
described for the picolinyl esters prior to GC–MS analysis.
received. All solvents used were of HPLC grade (Merck, Germany).
Cork suberin C18 monounsaturated acids
Methyl 18-hydroxyoctadecenoate (Hyd18:1_Me) and dimethyl 1,18-
octadecenodioate (Di18:1_Me) were obtained from cork suberin after
methanolysis depolymerisation, followed by a multi-step isolation and
purification process, which is now part of a patent submission (data
not shown). The purity of these suberin acids was checked by GC–MS
GC–MS analysis. The TMS-derivatised solutions of the picolinyl esters
and DMOX derivatives were injected into the GC–MS system previously
described. The oven temperature programme was: 5 min at 100 °C,
followed by a temperature increase from 100 °C to 250 °C at a rate of
(> 99.9%) and their structure confirmed by EIMS, one-dimensional
8
°C/min, and then from 250 °C to 300 °C at 3 °C/min; the final tempera-
1 13
NMR ( H,
C) and two-dimensional correlation NMR (correlation
ture was kept for 20 min. Injection and mass spectrometer conditions
were as described above.
spectroscopy (COSY), heteronuclear single-quantum coherence (HSQC),
heteronuclear multiple bond correlation (HMBC)) (online Supporting in-
formation, Fig. SM1).
For the preparation of picolinyl esters and DMOX derivatives (see fol-
lowing sections), the two C18:1 suberin acid methyl esters (Hyd18:1_Me
and Di18:1_Me) were hydrolysed to free carboxylic acids in a 0.5 M KOH
ethanol:water 9:1 solution and recovered after acidification to an organic
phase. After drying and trimethylsilyl (TMS) derivatisation, the complete-
ness of the methyl esters hydrolysis to free acids was checked by GC–MS.
FTIR analysis
The FTIR absorption spectra (32 scans per spectrum) were acquired on
an Alpha-P spectrometer (Bruker Optik, Karlsruhe, Germany) with a spec-
À1
À1
tral resolution of 4 cm and a wavenumber range from 4000 cm to
À1
4
00 cm . The spectra were obtained by attenuated total reflectance
(
ATR), with a diamond cell and the pressure clamp applied directly over
the samples, which were liquid at room temperature.
Cis/trans separation by GC–MS analysis
Analytical samples were prepared in comparable pyridine and N,
O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) solutions, although only
Hyd18:1_Me needed to be derivatised, to keep the chromatographic
conditions identical between model compounds and suberin acids.
Raman analysis
Vibrational Raman spectra were obtained in an apparatus consisting of a
double monochromator Spex 1403 (Horiba, Kyoto, Japan), with an argon
ion laser line at 514.5 nm, model 2016 (Spectra-Physics, Santa Clara, CA,
USA) and a R928 photomultiplier detector (Hamamatsu Photonics,
Shizuoka, Japan). The spectra were acquired at room temperature, with
90° geometry, a resolution of 4 cm , a time of integration of 1 s and
an exit power of 1 W. The liquid suberin acid methyl ester samples were
placed in glass tubes and analysed at room temperature.
Methyl oleate and methyl elaidate mixtures. Methyl oleate (3 μL, ca.
2.6 mg, 0.77 mmol) and methyl elaidate (3 μL, ca. 2.6 mg, 0.77 mmol)
À1
were each separately diluted in pyridine (340 μL) and BSTFA (340 μL).
An aliquot (100 μL, ca. 0.4 mg) of each of these solutions was taken
and mixed in a vial to which pyridine (800 μL) was added.
Phytochem. Anal. 2014, 25, 192–200
Copyright © 2013 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/pca

S. Santos and J. Graça
NMR analysis
if both, in what relative proportions. Furthermore, assignment
of cis or trans configuration by NMR is facilitated by the presence
of different substituents close to the double bond. However,
both Hyd18:1 and Di18:1 are highly symmetrical molecules in
relation to the double bond, with identical alkyl substituents
up to seven carbons away.
The NMR spectra were recorded on an Avance II + 600 spectrometer (Bruker
Biospin, Rheinstetten, Germany), operating at 600.13 MHz for protons and
150.96 MHz for carbons, equipped with a cryoprobe and pulse gradient
units, capable of producing magnetic field pulsed gradients in the z
direction of 56.0 G/cm. All NMR spectra were acquired in deuterated chloro-
form with 0.03% of TMS at a temperature of 300 K. The H spectra chemical
1
13
shifts were referenced to TMS (0.00 ppm) and the C spectra to chloroform
77.00 ppm). The sample concentration used was 5 mg/500μL placed in
mm or 5 mm diameter NMR tubes suitable for 600 MHz. The NMR spectra
Determination of cis/trans character by GC–MS
(
3
The C18:1 cis and trans model compounds were separated in this
present study with methyl oleate eluting 30 s earlier than methyl
elaidate (Fig. 1A). Using the same chromatographic conditions,
each of the two C18:1 suberin acid methyl esters showed a single
chromatographic peak (Fig. 1B and C). Due to the structural
similarity between the model compounds and the C18:1 suberin
acids, these results indicate that the C18:1 cork suberin acids,
Hyd18:1 and Di18:1 exist only as one of the isomers, either cis
or trans.
were further processed, and spin simulations made using MestReNova,
Version 7.1.2 (Mestrelab Research, Santiago de Compostela, Spain).
Results and discussion
In this analysis of the unsaturated suberin acids from cork,
without commercial standards available for reference, it was
not known if one or both geometric isomers were present and,
(
A)
1
00
2
8.50
Methyl elaidate
2
8.03
Methyl oleate
5
0
0
8
.00 10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00 34.00
Time (min)
(
B)
1
00
62.740
50
0
Time(min)
80.00
1
5.00
20.00
25.00
30.00
35.00
40.00
45.00
50.00
55.00
60.00
65.00
70.00
75.00
(
C)
1
00
66.152
50
0
1
5.00
20.00
25.00
30.00
35.00
40.00
45.00
50.00
55.00
60.00
65.00
70.00
75.00 80.00 Time(min)
Figure 1. GC–MS ion chromatograms of: (A) 1:1 mixture of methyl oleate (t
R
= 28.03min) and methyl elaidate (t
R
= 28.50 min); (B) purity control of dimethyl
(Z)-octadec-9-enedioate (Di18:1); and (C) methyl (Z)-18-hydroxyoctadec-9-enoate (Hyd18:1) (analysed as TMS derivative), isolated from cork suberin.
wileyonlinelibrary.com/journal/pca
Copyright © 2013 John Wiley & Sons, Ltd.
Phytochem. Anal. 2014, 25, 192–200

Stereochemistry of C18:1 Cork Suberin Acids
+
•
Double bond position
at m/z 320 (M À 103) until m/z 222. The 14 amu sequence re-
starts at m/z 196 until the end of the alkyl chain, at m/z 98.
The ions of m/z 222 and m/z 196 had a difference of 26 amu,
thus indicating the double bond position at C À 9.
Chemical degradation studies indicated that the double bond
was at C–9, for both Hyd18:1 (Ribas and Seoane, 1954b) and
Di18:1 (Ribas and Seoane, 1954a). To confirm this assignment,
the double bond position was determined here by mass
spectrometry, after derivatisation of the C18:1 cork suberin acids
to the corresponding picolinyl esters and 4,4-dimethyloxazoline
derivatives (Harvey, 1992; Graça and Pereira, 2000). In these
derivatives, the presence of the nitrogen atom close to the
carboxyl group of the monounsaturated long-chain acids, under
Similarly, the spectrum of Di18:1 picolinyl ester, TMS
derivative (Fig. 2B) showed ions at m/z 92, m/z 108, m/z 151
and m/z 164 indicative of the pyridine ring being present. A
+
•
series of 14 amu losses was found from m/z 358 (M À CO TMS)
2
until m/z 260 and restarted at m/z 234, until the end of the alkyl
chain. As with the case of the DMOX derivatives, this gap of
6 amu between m/z 260 and m/z 234, indicated the double
bond position at C–9. Previous work with Di18:1 bis-DMOX
derivative, but extracted from the suberin of potato periderm,
also led to the same assignment of the double bond position
at C–9 (Graça and Pereira, 2000).
2
typical GC–MS electron ionisation conditions, gives
a
characteristic pattern of molecule fragmentation that allows
the recognition of the double bond position, particularly when
it is located mid-chain (Harvey, 1992).
The GC–EIMS spectra of Hyd18:1 DMOX–TMS derivative and
Di18:1 picolinyl–TMS derivative are presented in Fig. 2A and B
respectively. The mass spectra of Hyd18:1 and Di18:1 TMS deriv-
atives, together with the mass spectra of Hyd18:1 picolinyl–TMS
derivative, Di18:1 DMOX–TMS and bis-DMOX derivatives, are
shown as online Supporting information (Figs SM2 to SM6,
respectively). The Hyd18:1 DMOX–TMS spectrum (Fig. 2A)
showed the ions at m/z 113 and m/z 126 typical of the DMOX
FTIR and Raman analysis
The FTIR spectra of the two C18:1 suberin acids, Hyd18:1_Me and
Di18:1_Me and of the two C18:1 model compounds are com-
pared in Fig. 3, with the most diagnostic bands for cis and trans
assignment annotated. The Raman spectra of the Hyd18:1_Me
and Di18:1_Me suberin acids, together with the complete list
2
moiety, and the successive losses of 14 amu (CH group) starting
(
A)
Hyd18:1 DMOX/TMS derivative
4
23
100
1
13
98
126
154
182
236
264
292
320
1
26
O
OTMS
1
12
140
168
196
222
250
278
306
N
4
08
7
3
5
50
7
+
.
M
55
1
82
4
23
2
50
336
2
36
3
66
380
1
8
03
140
1
54
68
222
196208
9
1
2
292 320
64278
3
06
352
394
0
40
60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420
m/z
(
B)
100
Di18:1 Picolinyl/TMS derivative
9
2
4
75
9
2
O
150
178
206
234
260
288
316
344
O
O
OTMS
1
08
164
192
220
274
302
330
358
N
50
7
3
1
08
164
4
60
2
74
88
+
.
3
44
M
2
4
75
5
5
1
29
1
51
2
06
20
1
78
1
2
302 330
316
92
2
34
260
3
58 386 418
0
m/z
50
100
150
200
250
300
350 400
450
Figure 2. Assignment of the double bond position. Electron ionisation mass spectrum and fragmentation pattern of: (A) (Z)-18-hydroxyoctadec-9-
enoic acid (Hyd18:1) DMOX–TMS derivative; (B) (Z)-octadec-9-enedioic acid (Di18:1) picolinyl–TMS derivative.
Phytochem. Anal. 2014, 25, 192–200
Copyright © 2013 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/pca

S. Santos and J. Graça
(A)
(B)
(C)
(D)
Methyl elaidate
Methyl oleate
Hyd18:1_Me
Di18:1_Me
Figure 3. The FTIR spectra of: (A) methyl (E)-octadec-9-enoate (methyl elaidate); (B) methyl (Z)-octadec-9-enoate (methyl oleate); and cork suberin
acids (C) methyl (Z)-18-hydroxyoctadecenoate (Hyd18:1_Me) and (D) dimethyl (Z)-1,18-octadecenodioate (Di18:1_Me). The main absorption bands
assignable to cis or trans configurations are annotated.
of tentative band assignments for all FTIR and Raman spectra are
presented as online Supporting information (Fig. SM7 and
Tables SM1 and SM2, respectively). Up to five modes of vibration
can be used to differentiate cis and trans isomers, that is,
the = C–H stretching vibration, the C = C–H in-plane and out-
of-plane deformation vibrations, and the C = C stretching
and skeletal vibrations (Günzler and Gremlich, 2002).
trans isomers absorbing at a higher wavenumber than their
equivalent cis isomers (Socrates, 2001). This was observed in
À1
the C18:1 model compounds with absorptions at 3005 cm
À1
and 3016 cm for methyl oleate and methyl elaidate respec-
tively (Fig. 3). In the case of Hyd18:1_Me and Di18:1_Me this
À1
À1
same band was at 3002 cm
and 3003 cm
respectively,
values close to the one observed in the cis model compound.
Absorptions for C = C stretching in cis isomers typically are
A
weak to medium intensity absorption band at
À1
À1
3
090–3010 cm is typical of the = C–H stretching vibration, with
below 1665 cm , while the trans isomers absorb above this
wileyonlinelibrary.com/journal/pca
Copyright © 2013 John Wiley & Sons, Ltd.
Phytochem. Anal. 2014, 25, 192–200

Stereochemistry of C18:1 Cork Suberin Acids
À1
frequency (Socrates, 2001). A weak band at ca. 1655 cm was
observed for methyl oleate, Hyd18:1_Me and Di18:1_Me
supporting the cis configuration. No such absorption band was
observed for methyl elaidate (Fig. 3).
deformation vibrations and the skeletal vibrations respectively
of the C–H bond in cis–CH = CH– groups (Socrates, 2001)
(Fig. 3).
The analysis of the FTIR and Raman therefore indicate that
the C18:1 acids from cork suberin have, at least, a dominant
cis configuration.
The Raman analysis of methyl oleate and methyl elaidate
showed a C = C stretching band at ca. 1656 cm for the former
cis isomer and ca. 1670 cm for the latter trans (Bailey and
À1
À1
Horvat, 1972). The Raman spectra of Hyd18:1_Me and Di18:1_Me
had strong bands at 1662 and 1660 cm respectively (Fig. SM7),
À1
NMR analysis
supporting the cis configuration.
The relative symmetry of these unsaturated acids results in an
overlapping complexity for the olefinic proton resonances that
precludes a straightforward analysis of coupling constants to as-
sign configuration. Therefore, an examination of the chemical
shifts of the allylic C carbons (Fig. 4) and the estimation of
the JHH olefinic coupling constants through spectra simulation
were used to determine configuration about the double bonds.
Another possible diagnostic vibration band for cis and trans
discrimination comes from the C = C–H in-plane deformation,
which is not intense in the FTIR spectra, but is relatively strong
in Raman spectra of cis isomers. This band is present at ca.
13
À1
1
267 cm for methyl oleate but not observed in methyl elaidate
(Sadeghi-Jorabchi et al., 1991). In the Hyd18:1_Me and
Di18:1_Me Raman spectra, a medium to strong band assignable
to this vibration is found at 1266 and 1270 cm respectively
À1
Chemical shifts and cis and trans configuration. The assign-
1
13
(Fig. SM7), therefore supporting a cis configuration for these
ment of H and C chemical shifts of the suberin acid methyl es-
ters, Hyd18:1_Me, and Di18:1_Me, and of the model compounds
methyl oleate and methyl elaidate, are presented as online
Supporting information (Tables SM3 and SM4), with the two
latter compared with published results. The molecular schemes
of the cis and trans double bonds are presented in Fig. 4, with
the olefinic (A and B) and allylic (X and Y) positions identified.
suberin acids.
In the FTIR spectra, the relatively strong absorption band at
À1
9
80–955 cm , typical of the C = C–H out-of-plane deformation,
is known to be unique to trans isomers (Chapman, 1965).
This band is prominent in the FTIR spectrum of methyl elaidate
À1
at 967 cm (Fig. 3) but is not observed for methyl oleate or
1
13
the C18:1 suberin acids spectra, thereby indicating a cis configu-
ration. In Raman spectra, however, the most intense band for
the C = C–H out-of-plane deformation is found for cis-isomers.
This band was observed at 882 cm for methyl oleate, but
was not well defined in methyl elaidate (Beattie et al., 2004).
The observation of a band at 882 cm in the Raman spectra
of Hyd18:1_Me and Di18:1_Me supports a cis configuration.
Further support for the cis configuration was the observation
The H and C chemical shifts of the olefinic and allylic posi-
tions are shown in Table 1. In this table, the differences
between the chemical shifts of the cis and trans model com-
À1
cis
trans
pounds, Δ(ppm) = δ À δ
, are also given.
cis
trans
In the model compounds, a small difference of δ À δ
=
À1
À0.03 ppm was observed in the olefinic protons, and of
+0.05 ppm in the allylic protons (Table 1) showing that the cis
olefinic protons are slightly more shielded than the trans olefinic
protons, and that the cis allylic protons are more deshielded
than the trans allylic protons. Similar differences were observed
À1
of a weak band (shoulder) around 695 cm and another weak
À1
band at ca. 965 cm assigned to the wagging (out-of-plane)
Figure 4. Molecular schemes of double bonds of cis and trans configuration with reference to the coupling constants.
R = [CH CH
À CO
Suberin acids
2
]
6
2
3
R
R
1
= [CH
= [CH
2
]
]
7
À OH: methyl (Z)-18-hydroxyoctadecenoate (Hyd18:1_Me)
1
2
6
À CO CH : dimethyl (Z)-1,18-octadecenodioate (Di18:1_Me)
2
3
Model compounds
R
R
1
= [CH
= [CH
2
]
6
À CH
3
: methyl (Z)-octadec-9-enoate (methyl oleate)
1
2
] À CH : methyl (E)-octadec-9-enoate (methyl elaidate)
3
Phytochem. Anal. 2014, 25, 192–200
Copyright © 2013 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/pca

S. Santos and J. Graça
1
13
Table 1. Chemical shifts (ppm) of the olefinic and allylic H and C in cork suberin acids methyl (Z)-18-hydroxyoctadecenoate
Hyd18:1_Me) and dimethyl (Z)-1,18-octadecenodioate (Di18:1_Me), and model compounds methyl (Z)-octadec-9-enoate (methyl
oleate) and methyl (E)-octadec-9-enoate (methyl elaidate)
(
cis
trans
NMR
Compound type
Proton
Hyd18:1_Me
Di18:1_Me
Methyl oleate
Methyl elaidate
Δ = δ À δ
1
H
Olefinic
Allylic
Olefinic
H
H
C
C
C
C
9
and H10
and H11
5.34
2.01
129.74
129.85
27.11
27.09
5.34
2.01
129.79
129.79
27.11
27.11
5.35
2.01
129.74
129.99
27.20
27.14
5.38
1.96
130.20
130.47
32.59
32.54
À0.03
+0.05
À0.46
À0.48
À5.39
À5.40
8
13
C
9
10
8
Allylic
11
in other alkyl disubstituted olefins, as in cis- and trans-2-pentene
and cis- and trans-3-hexene, which were simulated through
theoretical calculations taking into account the double bond
magnetic anisotropy and steric effects (Abraham et al., 2001).
In these respects, for the suberin acids, Hyd18:1_Me and
Di18:1_Me, the chemical shifts of their olefinic and allylic pro-
tons were almost coincident with those observed in methyl ole-
ate (Table 1) and, therefore, compatible with a cis configuration.
Also diagnostic for the cis or trans configuration assignment in
conformation of the allylic protons can change by rotation
4
around the C–C bond. More relevant can be the J long-range
HH
allylic coupling, due to the transmission of spin information
through the π-electrons, which adds a J(π) contribution to the J
(σ) component (Pavia et al., 2009). These long-range couplings
are dependent on the configuration and conformation, namely
on the torsional angle (Φ) of the allylic C–H bond (as defined
4
in Fig. 4). The sign of the J(π) component of the
JHH allylic
coupling is negative, whereas the sign of the J(σ) component
13
these olefins are the C chemical shifts. Comparing methyl ole-
ate and methyl elaidate, the difference between the olefinic car-
bons was relatively minor, ca. À0.47 ppm, but a much larger
difference was observed in the allylic carbons, ca. À5.40 ppm,
showing a significant upfield shift for the cis isomer (Table 1).
This comparative shielding of the cis-allylic carbons is probably
due to the so-called γ-effect, attributed to the van der Waals in-
teractions between spatially close groups located at three-bond
distance (Günther, 2001). In the cis double bond configuration,
this effect is much more important due to the forced spatial
proximity of the opposite methylene allylic groups (Fig. 4). This
same shielding of À5.40 ppm in the cis allylic carbon was also
observed in cis- and trans-2-butene (Allman, 2010).
is positive (Günther, 2001), and therefore the sign and magni-
4
tude of the
JHH allylic coupling constant reflects the stereo-
4
chemistry. Typically, in alkyl disubstituted olefins, the
J
HH
allylic coupling constant is higher in magnitude in trans isomers
(Rummens and Haan, 1970).
To determine the coupling constants discussed above, the
olefinic proton multiplets from the suberin acids and C18:1 model
compounds were simulated (Fig. 5, Table 2). In Fig. 5, the
acquired signals are compared with the simulated multiplets,
with their corresponding splitting diagrams. The spin systems,
the chemical shifts and the coupling constants used in the
simulation are shown in Table 2.
2 2
Two spin systems were used in the simulation. One, ABX Y ,
13
The absolute value of the C chemical shifts of allylic carbons
can be by itself diagnostic of the cis or trans configuration. In
several C18:1 fatty acids, when the double bond is close to mid-
chain, the chemical shifts of the allylic carbons were found to
be very close to 27 ppm for the cis isomer, and 32 ppm for the
trans isomer (Gunstone, 1993). This reasoning can be applied
to both 9-unsaturated C18:1 suberin acids, Hyd18:1_Me and
Di18:1_Me analysed here, which showed allylic carbon chemical
shifts of ca. 27 ppm (Table 1), and therefore is concordant with a
cis configuration.
for the two model C18:1 fatty acids and Hyd18:1_Me, which have
non-identical alkyl substituents in relation to the double bond.
The other, an AA’X X ’ system was used for Di18:1_Me, although
2
2
one could expect it to have an A X spin system, due to the com-
2
4
plete symmetry of the molecule in relation to the double bond.
The simulation showed that relatively minor variations in the
coupling constants had a major impact in the multiplet patterns
(Table 2). A simulation of the olefinic protons signals of the
model compounds has already been published (Schaumburg
and Bernstein, 1968), but the acquired spectra did not
match ours, due to differences in the NMR frequencies
3
Coupling constants and cis/trans configuration. The JHH
used (60 MHz).
3
between the two olefinic protons is typically the main diagnostic
coupling constant used for cis or trans configuration assignment.
The magnitude of this coupling constant depends on the
overlapping extent of the orbitals of the two = C–H bonds, which
is highest when they are parallel (Pavia et al., 2009). The orbital
parallelism depends on the dihedral angle of = C–H bonds and
on the H–C = C bond (valence) angles (α and α´, Fig. 4). In the
trans configuration, the dihedral angle is 180°, due to the anti-
parallel arrangement of the = C–H orbitals (Pavia et al., 2009),
whereas in the cis configuration, the parallelism is smaller due
to the bond angles (Fig. 4), thus explaining the comparatively
higher coupling constants observed in trans isomers.
In our simulation, the J olefinic coupling constants used
HH
were unequivocally within the cis range for methyl oleate
(11.0 Hz), Di18:1_Me (10.5 Hz) and Hyd18:1_Me (10.5 Hz), and
4
within the trans range for methyl elaidate (16.0 Hz). The JHH
allylic coupling constants, including their magnitude and signal,
had a significant impact on the simulated spectra. In fact, only
negative coupling constants could approach the acquired split-
ting patterns of the olefinic protons. As discussed above, trans
4
isomers have higher JHH allylic coupling constants in absolute
value, which was confirmed by the values used to simulate the
methyl elaidate olefinic multiplet (À2.6 and À2.3 Hz), compared
with the ones used in methyl oleate (À1.5 Hz) and in both
suberin acids Di18:1_Me (À1.5 Hz) and Hyd18:1_Me (À2.0 Hz)
(Table 2).
3
The JHH between the olefinic and allylic vicinal protons (Fig. 4)
is not affected by the double-bond configuration, because the
wileyonlinelibrary.com/journal/pca
Copyright © 2013 John Wiley & Sons, Ltd.
Phytochem. Anal. 2014, 25, 192–200

Stereochemistry of C18:1 Cork Suberin Acids
Hyd18:1_Me
Di18:1_Me
Methyl oleate
Methyl elaidate
(
A) Simulated
spectra
(
B) Acquired
spectra
Figure 5. NMR olefinic proton multiplets of: (A) simulated and (B) acquired from methyl (Z)-18-hydroxyoctadecenoate (Hyd18:1_Me) and dimethyl
(
(
Z )-1,18-octadecenodioate (Di18:1_Me) from cork suberin; and the model compounds methyl (Z)-octadec-9-enoate (methyl oleate) and methyl
E)-octadec-9-enoate (methyl elaidate).
Table 2. Simulation of olefinic proton NMR multiplets: spin systems, chemical shifts (ppm) and coupling constants (Hz) used in
the simulation. Cork suberin acids methyl (Z)-18-hydroxyoctadecenoate (Hyd18:1_Me) and dimethyl (Z)-1,18-octadecenodioate
(Di18:1_Me), and model compounds methyl (Z)-octadec-9-enoate (methyl oleate) and methyl (E)-octadec-9-enoate (methyl
elaidate)
Variable
Hyd18:1_Me
ABX2Y2
Di18:1_Me
AA´X X ´
Methyl oleate
ABX2Y2
Methyl elaidate
ABX2Y2
Spin system
2
2
Chemical shifts
A
B
X
Y
5.3525
5.3460
2.0150
2.0050
10.5
5.35
5.35
2.01
2.01
5.356
5.344
2.015
2.005
11.0
5.388
5.379
2.015
2.005
16.0
3
3
Coupling constants
J
J
HH olefinic
HH olefinic/allylic
10.5
= 7.3
H H = 7.0
H
A
H
X
= 6.5
H
A
H
X
7.5
H
A
H
X
= 8.0
H H = 6.5
H H = 7.9
B
Y
B
Y
B Y
4
JHH allylic
À2.0
À1.5
À1.5
H H = À2.3
A
Y
H H = À2.6
B
X
The configuration of C18:1 suberin acids and the molecular
structure of suberin
arrangement in suberin would be in line with cork properties,
such as low rigidity or a glass transition temperature estimated
at 18 °C (Dionisio et al., 1995). However, a glass transition phe-
nomenon in cork might apply to only a part of the material, be-
cause it also includes in its composition, besides suberin, other
biopolymers such as lignin and polysaccharides (Mano, 2002).
Also, an ordered molecular arrangement in suberin could justify
the regular lamellar structure known to exist in suberised cell
walls (Sitte, 1962).
Taking into account that the C18:1 acids also exist in suberin as
glycerol esters (Graça and Santos, 2007) a fairly comparable situ-
ation of cis-C18:1 fatty acids esterified to glycerol can be found in
plant fats and oils, namely in the triacylglycerol of oleic acid. The
molecular packing of triolein in the solid phase has been thor-
oughly studied and is known to have up to three different
The above results demonstrated that both C18:1 suberin acids are
found, at least in cork, only in the cis configuration. These two
C
18:1 suberin acids play a major role in the suberin structure: they
have been found in significant quantities in all suberins studied
so far, sometimes in overwhelming proportions, such as in
potato periderm, where they represent more than 70% of all
long-chain monomers. The cis stereochemistry of these mono-
mers can be determinant for the macromolecular structure of
suberin. An important point of discussion is whether the suberin
acids are organised in a totally amorphous manner, or if they
have some order at molecular or macromolecular level, as in
semi-crystalline polymers. A disordered amorphous molecular
Phytochem. Anal. 2014, 25, 192–200
Copyright © 2013 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/pca

S. Santos and J. Graça
crystalline forms, each associated with different conformations
around the cis double bond of the oleic acid units (Larsson
et al., 2006). Altogether, this shows that the C18:1 acids in suberin
can also be part of an ordered molecular arrangement.
Gunstone FD. 1993. The composition of hydrogenated fats by high-
resolution C-13 nuclear-magnetic-resonance spectroscopy. J Am Oil
Chem Soc 70: 965–970.
Gunstone FD. 1999. The preparation of derivatives for lipid analysis. In
Lipid Synthesis and Manufacture, Gunstone FD (ed.). Sheffield
Academic Press: Sheffield; 321–346.
Günther H. 2001. NMR Spectroscopy. John Wiley & Sons, Inc: New York;
1
08–128.
Günzler H, Gremlich H-U. 2002. IR Spectroscopy. Wiley-VCH: Weinheim;
85–190.
Harvey D. 1992. Mass spectrometry of picolinyl and other nitrogen-
containing derivatives of lipids. In Advances in Lipid Methodology –
One, Christie WW (ed.). The Oily Press: Dundee; 19–80.
Kolattukudy P. 2002. Suberin from plants. In Biopolymers – Polyesters I.
Doi Y, Steinbüchel A (ed.). Wiley-VCH: Weinheim; 41–73.
Larsson K, Quinn P, Sato K, Tiberg F. 2006. Lipids: Structure, Physical
Properties and Functionality. The Oily Press: Brigwater; 18–25.
Mano JF. 2002. The viscoelastic properties of cork. J Mater Sci 37: 257–263.
Pavia D, Lampman G, Kriz G, Vyvyan J. 2009. Introduction to Spectroscopy.
Brooks/Cole: Belmont; 233–247.
Acknowledgements
1
This work was supported by the Portuguese Technology and
Science Foundation (FCT) Project PTDC/QUI/70589/2006 and
Grant SFRH/BD/31033/2006 and is part of the activities of Forest
Research Centre (Centro de Estudos Florestais). We also acknowl-
edge Eurico Cabrita and Ângelo Figueiredo (Universidade Nova
de Lisboa) and the Portuguese National NMR network (RNRMN),
funded by FCT. We are grateful for the help of: José Rodrigues
(
Flor, IICT) in the FTIR analysis, and Luís Santos (Materials Engineer
Department, IST/TULisbon) in the Raman analysis.
Ribas I, Seoane E. 1954a. Química del corcho. XI. Ácidos no saturados.
Anal Real Soc Fís Quím 50B: 963–970.
Ribas I, Seoane E. 1954b. Química del corcho. XII. Ácido 18-hidroxi-
References
9
-octadecenoico y derivados. Anal Real Soc Fís Quím 50B: 971–976.
Abraham RJ, Canton M, Griffiths L. 2001. Proton chemical shifts in NMR:
Part 17. Chemical shifts in alkenes and anisotropic and steric effects
of the double bond. Magn Reson Chem 39: 421–431.
Allman R. 2010. Nuclear magnetic resonance. CMR: Peak splitting. http://
www.wavesignal.com/forensics/NMR.html (accessed October 2012).
Bailey GF, Horvat RJ. 1972. Raman spectroscopic analysis of the cis/trans
isomer composition of edible vegetable oils. J Am Oil Chem Soc 49:
Rodríguez B, Ribas I. 1972. Investigaciones químicas sobre el corcho de
solanium tuberosum L. (patata). Anal Quím 68: 303–308.
Rummens FHA, Haan JW. 1970. Spectroscopic studies on olefins – III:
NMR of cis- and trans- disubstituted olefins. Org Magn Resonance 2:
351–355.
Sadeghi-Jorabchi H, Wilson RH, Belton PS, Edwardswebb JD, Coxon DT.
1991. Quantitative analysis of oils and fats by Fourier transform
Raman spectroscopy. Spectrochim Acta A 47: 1449–1458.
Schaumburg K, Bernstein HJ. 1968. Calculation of the NMR spectrum of
double-bond protons in aliphatic systems. Lipids 3: 193–198.
Seoane E. 1966. Ácido 18-hidroxi-trans-9-octadecenoico, aislado del
corcho en forma de ácido treo-9,10,18-trihidroxioctadecanoico. Anal
Real Soc Fís Quím 62B: 563–568.
494–498.
Beattie JR, Bell SEJ, Moss BW. 2004. A critical evaluation of Raman
spectroscopy for the analysis of lipids: Fatty acid methyl esters. Lipids
3
9: 407–419.
Chapman D. 1965. The Structure of Lipids. Methuen: London; 52–93.
Dionisio MSC, Correia NT, Mano JF, Ramos JJM, Fernandes AC, Saramago
B. 1995. Absorbed water in the cork structure – a study by thermally
stimulated currents, dielectric-relaxation spectroscopy, isothermal
depolarization experiments and differential scanning calorimetry.
J Mater Sci 30: 4394–4400.
Sitte P. 1962. Zum feinbau der suberinschichten im flaschenkork.
Protoplasma 54: 555–559.
Socrates G. 2001. Infrared and Raman Characteristic Group Frequencies –
Tables and Charts, 3rd edn. John Wiley & Sons, Ltd: Chichester; 68–81.
Eliel EL, Wilen SH. 1994. Stereochemistry of Organic Compounds. John
Wiley & Sons, Inc: New York; 555–573.
Graça J, Pereira H. 2000. Suberin structure in potato periderm: Glycerol,
long-chain monomers, and glyceryl and feruloyl dimers. J Agric Food
Chem 48: 5476–5483.
Graça J, Santos S. 2007. Suberin: a biopolyester of plants’ skin. Macromol
Biosci 7: 128–135.
Supporting information
Supporting information can be found in the online version of
this article.
wileyonlinelibrary.com/journal/pca
Copyright © 2013 John Wiley & Sons, Ltd.
Phytochem. Anal. 2014, 25, 192–200