Lignin chemistry: Biosynthetic study and structural characterisation of coniferyl alcohol oligomers formed in vitro in a micellar environment

Article abstract of DOI:10.1002/chem.200903302

Model coniferyl alcohol lignin (the so-called dehydrogenative Polymerisate, DHP) was produced in water under homogeneous conditions guaranteed by the presence of a micellised cationic surfactant. A complete study of the activity of the enzymatic system peroxidase/H₂O₂ under our reaction conditions was reported and all the reaction products up to the pentamer were characterised by ¹H NMR spectroscopy and ESI mass spectrometry. Our system, and the molecules that have been generated in it, represent a closer mimicry of the natural microenvironment since an enzyme, under micellar conditions, reproduces the cell system better than in buffer alone. On the basis of the oligomers structures a new biosynthetic perspective was proposed that focused attention on a coni-feryl alcohol dimeric quinone methide as the key intermediate of the reaction. A formal, strictly alternate sequence of a radical and an ionic step underlines the reaction, thus generating ordered oligolignols structures. Alternatively to other model lignins, our olignols present a lower degree of radical coupling between oligomeric units. This offers a closer biosynthetic situation to the observation of a low rate of radical generation in the cell wall.

Full text of DOI:10.1002/chem.200903302

FULL PAPER
DOI: 10.1002/chem.200903302
Lignin Chemistry: Biosynthetic Study and Structural Characterisation of
Coniferyl Alcohol Oligomers Formed In Vitro in a Micellar Environment
Samantha Reale,[a] Francesca Attanasio,[b] Nicoletta Spreti,[a] and
Francesco De Angelis*[a]
Abstract: Model coniferyl alcohol
lignin (the so-called dehydrogenative
polymerisate, DHP) was produced in
water under homogeneous conditions
guaranteed by the presence of a micel-
lised cationic surfactant. A complete
study of the activity of the enzymatic
system peroxidase/H2O2 under our re-
action conditions was reported and all
the reaction products up to the pen-
tamer were characterised by 1H NMR
spectroscopy and ESI mass spectrome-
try. Our system, and the molecules that
have been generated in it, represent a
closer mimicry of the natural microen-
vironment since an enzyme, under mi-
cellar conditions, reproduces the cell
system better than in buffer alone. On
the basis of the oligomers structures a
new biosynthetic perspective was pro-
posed that focused attention on a coni-
feryl alcohol dimeric quinone methide
as the key intermediate of the reaction.
A formal, strictly alternate sequence of
a radical and an ionic step underlines
the reaction, thus generating ordered
oligolignols structures. Alternatively to
other model lignins, our olignols pres-
ent a lower degree of radical coupling
between oligomeric units. This offers a
closer biosynthetic situation to the ob-
servation of a low rate of radical gener-
ation in the cell wall.
Keywords: coniferyl alcohol · lig-
nin · mass spectrometry · polymers ·
surfactants
Introduction
the monomeric building blocks—of which there are only
three (p-hydroxycoumaryl, coniferyl (1) and sinapyl alco-
hols)—of the lignin polymer in a structurally unaltered
Lignin, the most abundant organic substance on earth after
cellulose, is a structural polymer present in all woody plants.
Over the past 170 years of research,[1] lignin has proven to
be quite an intractable macromolecule, and many questions
still exist about its formation, structure, occurrence and
commercial utilisation.
Concerning lignins structural elucidation, the main prob-
lem is related to the difficulty of isolating it from the other
wood components without damaging its structure. In addi-
tion, in contrast to proteins and nucleic acids which, by
simple hydrolysis, give rise to their constituent molecules, no
degradation methods have been developed for generating
form. The high number and variegated forms of linkages
that occur among the three monomeric components and the
resistance to degradation of the ether bonds that frequently
occur limit the extent to which analytical and degrading pro-
cedures can be used to elucidate the lignin structure. It is
also worth observing that the lignin macromolecule exists in
different structural assemblies, depending on its distribution
in a plethora of vegetable species, its age and plant distribu-
tion loci. It is, in fact, more appropriate to refer to the poly-
mer as “lignins”, rather than using the singular form.
[a] Dr. S. Reale, Prof. N. Spreti, Prof. F. De Angelis
Department of Chemistry, Chemical Engeeneering and Material
University of LꢀAquila, Via Vetoio
I-67100 Coppito, LꢀAquila (Italy)
Fax : (+39)0862433753
E-mail: francesco.deangelis@univaq.it
[b] Dr. F. Attanasio
Tumor Cell Invasion Laboratory, Consorzio Mario Negri Sud
Via Nazionale 8 A, I-66030 S. Maria Imbaro, Chieti (Italy)
Owing to the problems associated with the degradation
and analytical studies, investigations on the growth course of
lignin by starting from its precursors—p-hydroxycinnamyl
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200903302.
Chem. Eur. J. 2010, 16, 6077 – 6087
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6077

alcohols—have been used extensively in the attempt to elu-
cidate the lignin structure. However, since the polymeri-
sation reaction of the monomeric lignin precursors cannot
be studied in vivo, many theories on lignin structure and
biosynthesis still rely upon in vitro experiments.[2–5]
A major milestone in lignin chemistry was Freudenbergꢀs
success, almost sixty years ago, in polymerising coniferyl al-
cohol into a lignin-like dehydrogenative polymer (DHP) by
using a fungal laccase and other oxidative enzymes.[6–8] Only
coniferyl alcohol, the most abundant lignin constituent, was
used as the starting material to simplify the structural inves-
tigations on the dehydrogenation products. In fact the
number of products would increase and they would be large-
ly more complex, if the other two monomeric precursors of
lignin (p-hydroxycoumaryl and sinapyl alcohols) were also
present in the reaction mixture.
As to, in particular, the evaluation of the type of intermo-
lecular linkages that are involved in “artificial” lignins
(DHP) devoid of the saccharide moieties, the in vitro studies
are based, in general, on the very first procedure described
by Freudenberg himself in 1958 for the polymerisation of
coniferyl alcohol.[14] According to this procedure, coniferyl
alcohol is reacted in phosphate buffer at pH 6.5, in the pres-
ence of horseradish peroxidase and hydrogen peroxide. The
main problem associated with this procedure is the instanta-
neous separation of the growing oligomers and the conse-
quent formation of a heterogeneous reaction medium for
which the polymerisation process is practically over. A
series of technical solutions to the problem have been put
forward[15–17] for trying to keep in solution the polymeri-
sation products while the reaction proceeds, but they have
only the effect of retarding the unavoidable precipitation of
the growing polymer.
In the presence of peroxidase and hydrogen peroxide, as
well as in the presence of laccase, coniferyl alcohol (1) un-
dergoes dehydrogenation by losing its phenolic hydrogen
atom to form a phenoxy radical, stabilised by resonance ac-
cording to the mesomeric forms shown in Scheme 1.[8]
For example, Kirk and Brunow proposed a procedure in
which a solution containing both coniferyl alcohol and the
enzyme and a solution containing H2O2 were added simulta-
neously over a long period of time to a stirred solution of
vanillyl alcohol or guaiacyl glyc-
erol.[16] These two substances,
following the suggestion of
Adler,[17] serve as water-soluble
sites of polymerisation and
would retard, but not inhibit,
the precipitation of the starting
material and intermediates or
of the oligomers and polymer
itself.
It is also worth pointing out
Scheme 1. Enzymatic dehydrogenation of coniferyl alcohol (1) and related phenoxy radicals.
that, although DHPs were
found to be good models for
lignins, their structures, which
The reaction continues through the radical coupling of
two of these mesomeric forms to give a dilignol molecule,
and then proceeds by radical and/or ionic steps to form
higher lignols.
are strongly dependent on the polymerisation conditions,
may be quite different from the natural polymer,[2–5,18] prin-
cipally because the wood cell environment, in which the
native lignins are assembled, is dramatically different from
the in vitro environment. In addition, as we have already
pointed out, the oligolignols once formed in vitro subtract
themselves very rapidly from a possibly higher degree of
polymerisation, thus conferring to DHPs very low degrees
of polymerisation.
From the picture outlined above it follows that a major
achievement in the study of lignins and synthetic DHPs
would be the concept of a reaction medium in which the
starting material and the growing polymer would dissolve,
still preserving the catalytic activity of horseradish perox-
idase. In a former preliminary study we have reported on
the in vitro reaction conditions generated by a cationic sur-
factant in phosphate buffer, able to maintain in solution the
growing DHP even for weeks, still permitting horseradish
peroxidase to exert its catalytic activity.[19] Cetyltrimethylam-
moniumsulfate [(CTA)2SO4] was used,[20] at a concentration
higher than the critical micelle concentration (c.m.c.), to
allow the formation of a novel in vitro model of lignin. The
Some problems, however, are connected with Freuden-
bergꢀs oligolignols and DHP synthesis. From the biological
point of view, lignins appear to be synthesised by a non-en-
zymatic process occurring through/outside the living cell
membrane in a polysaccharide gelified environment, finally
resulting in the plant cell wall lignification.[9] This particular
issue has been addressed in the last few years by Monties
and co-workers, who succeeded in the preparation of lignin
polymeric models under conditions that mimic the cell-wall
lignification environment. They studied the b-glucosidase/
peroxidase-triggered polymerisation of coniferin (coniferyl
alcohol glucoside)[10,11] and coniferyl alcohol polymerisation
in the presence of pectin.[12,13] In both cases it has been dem-
onstrated that the structures of the obtained polymers ap-
proximate those of native lignins more closely than the pure
coniferyl lignin model, whereas the latter leads to the isola-
tion of pectin-DHP complex.
6078
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 6077 – 6087

Lignin Chemistry
FULL PAPER
polymerisation process was shown to be an alternate, ordi-
nate sequence of radical and ionic steps. On the other hand,
it is worth nothing that the already cited pectin-associated
DHP[12] shows solubility characteristics that are not dissimi-
lar to our DHP formed in the presence of surfactant.
tration. It is worth emphasising also that at [(CTA)2SO4]=
27ꢂ10À3 m, both coniferyl alcohol and its polymerisation
products are kept in solution (precipitation is never ob-
served), thus creating a homogeneous reaction environment
(see below).
In the present paper, we investigate in detail the polymer-
isation process of coniferyl alcohol catalysed by horseradish
peroxidase, in the presence of cetyltrimethylammoniumsul-
fate: 1) The catalytic activity of the enzyme, under these
conditions, was studied. 2 and 3) The coniferyl alcohol oligo-
mers were analysed by HPLC, isolated and fully character-
To rule out either a spontaneous coniferyl alcohol poly-
merisation under oxidative conditions or any micellar catal-
ysis, two series of control experiments were carried out. In
the first of them, an acetone solution of coniferyl alcohol
was slowly added to a phosphate buffer solution of hydro-
gen peroxide (at three different concentrations: 0.1, 0.8 and
1.6%), in the absence of the enzyme. The solution was then
monitored by TLC for over four days; during this period of
time coniferyl alcohol was absolutely stable. In the second
series of experiments (CTA)2SO4 (same concentrations
listed above) was added, in the absence of the enzyme to ex-
clude any possible micellar catalysis in the polymerisation
process. Also in these cases the coniferyl alcohol polymeri-
sation does not take place at all.
1
ised by H NMR spectroscopy and ESI mass spectrometry.
4) Finally, a possible reaction pathway for coniferyl alcohol
polymerisation under our biosynthetic conditions was pro-
posed.
Results and Discussion
Peroxidase characterisation in the presence of cetyltrimethyl-
ammoniumsulfate and optimisation of the reaction condi-
tions for coniferyl alcohol polymerisation: A surfactant, in
principle, should be able to allow the solubilisation in water
of a lipophylic compound, such as coniferyl alcohol, the
starting material we have chosen for studying in vitro lignin
biosynthesis, as well as its polymerisation products. On the
other hand, its presence should not interfere with the stabili-
ty and activity of horseradish peroxidase, the enzyme
chosen to catalyse the polymerisation reaction.
To investigate this latter issue, we have examined a series
of commercially available and synthetic surfactants with dif-
ferent head-group charge (i.e. anionic, cationic and zwitter-
ionic) at various concentrations,[21] looking for the stability/
activity of horseradish peroxidase against o-phenylendi-
In our initial experiments, carried out on 5 mg of coniferyl
alcohol each, we decided to follow the substrate polymeri-
sation, catalysed by horseradish peroxidase/hydrogen perox-
ide in the presence of the surfactant, by TLC (eluent:
CHCl2/MeOH 20:1). The reaction medium is constituted by
5 mm phosphate buffer at pH 6.5, hydrogen peroxide
(0.1%), and 1.75 U of horseradish peroxidase (Sigma type II
200 UmgÀ1), in the presence of (CTA)2SO4 at 3ꢂ10À3 m, the
minimum surfactant concentration able to maintain both
coniferyl alcohol and all the incoming oligomerisation prod-
ucts in solution. A control reaction in buffer solution alone,
without the surfactant, was also carried out. It is worth
noting that, soon after the enzyme addition, the pure buffer
reaction medium becomes immediately heterogeneous with
the formation of a precipitate, whereas the solution created
in the presence of the surfactant remains clear due to the
surfactants ability to aid the solubilisation of the starting
material and of all the incoming polymerisation products.
To evaluate the effect of the surfactant concentration on
the kinetics of the coniferyl alcohol polymerisation, the re-
action course was followed by UV spectroscopy. According-
ly, three experiments at increasing surfactant concentration
(see above) were carried out. The reaction progress was
monitored by following the increase of relative maxima in
the range from 300 to 500 nm (in particular at 330 and
390 nm), which are clearly related to the reaction products.
Incidentally, the absorption maximum at 330 nm is due to
pinoresinol (a coniferyl alcohol dimer) or the pinoresinyl
moiety present in any of the produced oligomers eventually
(our own measurements on reference samples). The conifer-
yl alcohol disappearance could not be followed, since its ab-
sorption maximum (lmax =260 nm) is in the same absorption
range of the enzyme and hydrogen peroxide.
ACHTUNGTRENNUNGamine, a water-soluble model substrate that is oxidised by
hydrogen peroxide in the presence of the enzyme. The reac-
tion course was followed by UV spectroscopy (see the Ex-
perimental Section).
Finally, (CTA)2SO4 was selected as the surfactant of
choice. It is actually a surfactant synthesised from commer-
cial CTABr.[20] CTABr, in fact, is not suitable when used in
the presence of an oxidative reagent as it can also act as a
substrate under such reaction conditions.
The influence of the surfactant on the enzyme activity
was evaluated, in a phosphate buffer, at different surfactant
concentrations (3ꢂ10À3, 9ꢂ10À3 and 27ꢂ10À3 m—c.m.c.
being 2.7ꢂ10À4 m). It is worth noting that the minimum sur-
factant concentration that is able to guarantee a clear solu-
tion of coniferyl alcohol is 1.0ꢂ10À4 m. This concentration,
however, is not sufficient to avoid the precipitation of the
polymerisation products. A buffered solution of o-phenylen-
diamine was used as a reference for the evaluation of the
enzyme activity, following the substrate oxidation by UV
spectroscopy. It appears that the enzyme works for over
24 h at any surfactant concentration. Its activity is at a maxi-
mum at the minimum surfactant concentration, whereas the
activity is reduced to 50% at the highest surfactant concen-
In contrast to the test reactions carried out by using o-
phenylendiamine as the substrate, with coniferyl alcohol the
maximum velocity is registered at a 9ꢂ10À3 m surfactant con-
centration.
Chem. Eur. J. 2010, 16, 6077 – 6087
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6079

F. De Angelis et al.
A clear picture came out when the reaction was followed
by RP-HPLC. At the lowest surfactant concentration (i.e.,
3ꢂ10À3 m), after only 1 min, eight well-separated peaks are
present in the chromatographic profile (see the Experimen-
tal Section for the chromatographic details). At the same re-
action time, with the surfactant concentration at 27ꢂ10À3 m,
only three peaks are present, the one related to coniferyl al-
cohol being the most abundant. On the other hand, under
these last conditions (which we considered safer as they
allow a perfect solubility of the reaction products) the sub-
strate is practically totally consumed after 5 min of reaction
time. A nice series of products, well-separated by HPLC, is
produced (Figure 1). This chromatographic profile, however,
fully reproduces that obtained after 1 min at a surfactant
concentration of 3ꢂ10À3 m.
This allows the hypothesis of a substantial product instabili-
ty and interconversion phenomenon (de-polymerisation and
re-polymerisation).
The reaction products (the analytical HPLC profile is
shown in Figure 1) were then separated by using semi-prep-
arative RP-HPLC. Five well-separated abundant peaks are
in fact present in the chromatographic profile, together with
some other minor peaks. However, also in this case an inter-
conversion reaction occurs when concentrating the solutions
of the products, always leading again to complex mixtures in
which, finally, two main products could be identified. They,
even if already present, were not largely abundant in the
original reaction mixture but appear to be generated as the
separation and workup procedures are carried out. These
two “end-products”, which were revealed to be stable, have
been finally isolated by preparative TLC, purified and char-
acterised by 1H NMR spectroscopy as pinoresinol (2) and
dehydrodiconiferyl alcohol (3) (see the Supporting Informa-
1
tion for their H NMR spectra).
They are both well-known coniferyl alcohol dimers.[22] Pi-
noresinol (2) is characterised by a b–b structure, since it is
formed by a b-radical/b-radical coupling reaction, after coni-
feryl alcohol oxidation (see Schemes 1 and 2). The dehydro-
diconiferyl alcohol (3) has the so-called b-5 structure, origi-
nating from the coupling of a b-radical from one coniferyl
alcohol unit with a 5-radical formed from a second coniferyl
alcohol molecule.
To isolate and identify also the other coniferyl alcohol
polymerisation products, it was imperative to stabilise them.
The reaction mixture was thus treated with acetic anhydride
and pyridine so that all of the hydroxy groups present on
the coniferyl alcohol polymerisation products were acetylat-
ed. Their high reactivity, in fact, would depend on the pres-
ence of the free OH groups, the easily oxidisable phenolic
ones in particular.
The HPLC analysis of the reaction mixture after acetyla-
tion (Figure 2) gives rise to a chromatographic profile that
parallels that obtained for the non-acetylated reaction mix-
ture, the main difference being a positive shift in the reten-
tion times. The co-eluting free alcohol oligomers were sepa-
rated after acetylation as a result of the different degrees/
nature of acetylation (free phenolic vs. hydroxyl groups—
vide infra).
Figure 1. RP-HPLC profile of the reaction mixture after 5 min (numbers
refer to structures 1–8).
Thus, to guarantee the homogeneity of the reaction mix-
ture and at the same time to preserve an adequate enzyme
activity for the subsequent experiments, the following reac-
tion conditions were chosen: Phosphate buffer (5 mm) at
pH 6.5,
[(CTA)2SO4]=27ꢂ10À3 m,
coniferyl
alcohol
(5 mgmLÀ1), hydrogen peroxide (0.1%), horseradish perox-
idase (1.75 UE).
Separation and identification of coniferyl alcohol polymeri-
sation products: The coniferyl alcohol (1) polymerisation re-
action (see above for conditions) was followed by RP-
HPLC. After 5 min, the coniferyl alcohol is no longer pres-
ent and the reaction was stopped by the addition of a few
drops of 5% Na2S2O7. A first separation attempt was carried
out by using TLC (a number of chromatographic runs, com-
bining the bands with the same Rf). Six well-separated
bands are in fact present on the chromatographic plate. Un-
fortunately, once these products are re-analysed by TLC, or
even by HPLC, they are revealed to be an even more com-
plicated mixture than the starting reaction mixture itself.
6080
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 6077 – 6087

Lignin Chemistry
FULL PAPER
that, according to a well-estab-
lished biosynthetic process for
lignin generation,[23] the poly-
merisation process would pro-
ceed through an oxidative
phenol coupling reaction and
an addition reaction to an inter-
mediate
quinone
methide
moiety (vide infra).
First of all, for each product,
the polymerisation degree was
established by counting the me-
thoxy group content, in the
1H NMR spectra. Monomeric
coniferyl alcohol only possesses
one methoxy group, which is
obviously repeated n times in
the n-mer. In addition, the
count of the acetyl groups gives
the number of the free hydroxy
functions present in the non-
acetylated molecule. Three ace-
tylated oxygen atoms in
a
Scheme 2. Dimer formation by one radical step. a) Radical coupling between a coniferyl alcohol radical at the
5-position (I5) and one at the b-position, which gives rise to a dehydrodiconiferyl alcohol dimer. b) Radical
coupling between two coniferyl alcohol radicals at the b-position to give a pinoresinol dimer.
dimer, for instance, means that
one of the two hydroxy groups
of one coniferyl alcohol unit is
engaged in an ether linkage.
Moreover, by examining the chemical shifts of the acetoxy
groups, it is also possible to attribute their aliphatic or aro-
matic nature. Finally, by considering that three aromatic
protons are present in coniferyl alcohol, any aromatic
proton missing in an oligomer implies that one extra aro-
matic position is involved in an interunit linkage.
The examination of the olefinic proton region gives infor-
mation on how many propenoic side-chain moieties survive
after oligomerisation. Protons that are lacking in the overall
olefinic balance are now on sp3 carbon atoms, involved in
b–b, b–O-4, b-5 or a-O-4 interunit linkages, and should be
looked for in the corresponding spectral regions. According-
ly, also the methylene groups adjacent to the oxyacetyl moi-
eties produce chemical shifts relevant to their molecular en-
vironment, depending whether they are at an allylic posi-
tion.
Figure 2. RP-HPLC profile of the solvent extracted/acetylated reaction
mixture after 5 min of reaction (numbers refer to structures 3–8).
The separation of the acetylated products was carried out
by semi-preparative HPLC, giving rise to six pure products:
each of them appears as a single peak once re-analysed by
analytical HPLC.
All the above considerations have been used to identify
the coniferyl alcohol oligomers, which have been isolated by
semi-preparative RP-HPLC as fully acetylated derivatives.
Besides the dehydrodiconiferyl alcohol (3), a-O-4/b-O-4
coniferyl alcohol trimer 4, two tetramers (5 and 6; one in-
cluding a pinoresinyl and the other a dehydrodiconiferyl al-
cohol unit) and a pentamer with a-O-4/b-O-4 interunit
bonds (7) have been fully characterised, again, as peracety-
The characterisation of the acetylated products of conifer-
1
yl alcohol polymerisation was carried out by H NMR spec-
troscopy. Acetylated dehydrodiconiferyl alcohol was easily
1
identified (see the Supporting Information for the H NMR
spectrum). Its spectrum together with those of the non-ace-
tylated dimers (2 and 3) were used as references for the
study of the spectra of the more complex molecules (the
spectrum of acetylated coniferyl alcohol is also reported in
the Supporting Information as a reference datum). At first
examination it is apparent that in the high-order oligomeri-
sation products the pinoresinyl and/or dehydrodiconiferyl al-
cohol moieties are also present. It is important to consider
1
lated compounds (the H NMR spectra of their acetylated
congeners, together with signal attributions, are reported in
the Supporting Information). It could be appreciated that
the structural complexity of the coniferyl alcohol oligomers
increases with the degree of polymerisation, as does the ap-
pearance of the proton spectra of their acetyl esters.
Chem. Eur. J. 2010, 16, 6077 – 6087
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6081

F. De Angelis et al.
We are now able to give
more information on the oligo-
mers structures, looking also at
their ESIMS spectra (both in
the positive- and in the nega-
tive-ion modes) obtained after
HPLC separation, as well as of
their peracetylated derivatives.
Table 1 reports the molecular
weights and the mass spectro-
metric data of the oligomers 1–
9
(besides coniferyl alcohol
itself), together with their fully
acetylated derivatives.
The mass spectra (in positive-
and negative-ion modes) ob-
tained by direct infusion of the
whole reaction mixture are re-
ported in Figure 3a and b (see
also Table 1 for peaks attribu-
tions). Dimers appear mainly as
protonated or deprotonated
ions in the positive and nega-
tive modes, respectively. Con-
versely the higher oligomers
Mass spectrometry: To collect more information on the con-
iferyl alcohol oligomers, a series of ESI mass spectrometric
experiments have been carried out.
form ammonium clusters in the positive-ion spectrum,
whereas they appear as deprotonated ions in the negative-
ion one. It is worth observing that signals up to the octamer
(not listed in Table 1) are present in both spectra. The satel-
We already reported on the ESI mass spectrometric anal-
ysis of the polymerisation products, as obtained by direct in-
fusion of the reaction mixture, after doping with ammonium
acetate to get a better response in the positive-ion mode.[19]
All the oligomers up to the octamer were revealed, their
masses differing one from the other, alternatively, by 178
and 180 Da. This was interpreted in the light of a regular re-
action mechanism that operates for the coniferyl alcohol
polymerisation, under our reaction conditions (see the fol-
lowing for more details).
+
lite peaks at higher masses with respect to the MNH4 ions
(5 Da increment) are due to sodiated ions, as demonstrated
also by the CID (collision induced dissociation) spectra ob-
tained by increasing the orefice potential. The sodiated ions,
in fact, being more stable than the ammoniated ones, frag-
ment to a lower extent and survive the dissociation process
more easily.
It is worth observing that in the CID negative-ion spec-
trum the small satellite peaks at 2 mass units lower than
those relevant to the main olig-
omers are selected by the in-
duced fragmentation process
Table 1. MS data for the coniferyl alcohol oligomers (bold numbers refer to structures 1–9, roman numerals
refer to the polymerisation degree—pedix “P” indicates the presence of a pinoresynil moiety in the structure
and pedix “D” indicates the presence of a dehydrodiconiferyl moiety—and numbers in italic indicate that the
relevant peaks are not present in the mass spectra).
(see inset in Figure 3b, in which
the case of tetramers is shown).
The corresponding molecules,
which are shown to be more
robust than the relevant high-
mass congeners, could be possi-
bly formed by a further dehy-
drogenation—cyclisation pro-
cess to give trimer 8 and tetra-
mer 9 (tentative structures are
shown). It is well-known that
cyclic ions are more resistant to
fragmentation than the corre-
sponding acyclic ones.
+[a]
+[a,c]
MW
MH+[a]
MNH4
[MÀH]À[b]
[M+AcO]À[b]
MAcNH4
[MAc+AcO]À[b,c]
ACHTUNGTRENNUNG
1
180
358
358
538
536
716
716
714
894
896
896
1074
1074
181
359
359
539
537
717
717
715
895
897
897
1075
1075
198
376
376
556
554
734
734
732
912
914
914
1092
1092
179
357
357
537
535
715
715
713
893
895
895
1073
1073
239
417
417
597
595
775
775
773
953
955
955
1133
1133
240
460
502
724
680
902
944
942
1122
1122
1167
1344
1386
281
501
543
765
721
943
985
983
1163
1163
1208
1385
1427
2 IIP
3 IID
4 III
8 IIIP
5 IVP
6 IVD
9 IV[d]
V[d]
[d]
VP
7 VD
VIP
VID
All the results described
[a] Positive-ion mode. [b] Negative-ion mode. [c] Peracetylated products. [d] Further dehydrogenated product. above were confirmed by the
6082
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 6077 – 6087

Lignin Chemistry
FULL PAPER
prise, in their molecular struc-
tures, a pinoresinyl or, alterna-
tively,
a
dehydrodiconiferyl
moiety. They are, in fact, isobar-
ic molecules that, however, pos-
sess a different number of hy-
droxy functions (see structures
1–3). It follows that their fully
acetylated derivatives differ by
one acetyl unit and, according-
ly, their molecular weights are
separated by 42 Da (see
Table 1), the one which includes
the pinoresinyl unit with
lower molecular weight.
a
The ESI mass spectra of the
separated acetylated oligomers
of coniferyl alcohol are report-
ed in the Supporting Informa-
tion. The molecules always
appear cationised by the ammo-
nium ion (i.e., 18 mass units
higher than the molecular
weight) in the positive-ion spec-
tra, whereas in the negative-ion
spectra they get the negative
charge by virtue of being com-
plexed by the acetate anion
(i.e., 59 mass units higher than
Figure 3. ESIMS spectra in the positive- (a) and negative-ion (b) mode of the reaction mixture. In the inset of
the negative-ion spectrum a comparison between the mass spectrum of the tetramers and the relevant CID-
MS is shown.
the molecular weight). The oligomers up to the tetramer
also form dimeric clusters, cationised by the ammonium ion
in the positive or by the acetate anion in the negative-ion
mode, respectively.
Acetylated dehydrodiconiferyl alcohol is the first com-
pound to be eluted after HPLC separation (structure 3, for
the free alcohol). Its cationised molecular ion appears at
m/z: 502, corresponding to the free alcohol molecular
weight (358 Da) plus three acetyl moieties (42 Daꢂ3) plus
the ammonium cation (18 Da). Also the cationised dimeric
cluster appears at m/z: 986. In the negative-ion mode, the
relevant peaks (clusters with the acetate anion) are at m/z:
543 and 1028, respectively. It is worth noting that the acety-
lated derivative of pinoresinol (its cationised molecular ion
should be at m/z: 460) does not appear to be separated
under these chromatographic conditions.
The acetylated trimer (compound 4 is the free alcohol)
presents the cationised molecular ion at m/z: 724 (the free
alcohol plus four acetyl moieties and the ammonium ion—
m/z: 765 in the negative-ion mode). The relevant chromato-
graphic peak is preceded by a small signal corresponding to
a new trimer that, according to the cationised molecular-ion
signal that occurs at m/z: 680, should comprise the pinore-
sinyl moiety (only three acetyl moieties are in fact present)
and must also be formed by way of a further dehydrogena-
tion process (two mass units less than expected). These
types of compounds have also been revealed by the CID-
MS experiments, performed on the direct infusion of the
RP-HPLC–ESIMS analysis (the relevant HPLC profile, to-
gether with the mass spectra (insets) are reported in
Figure 4). Eluent doping with ammonium acetate was em-
ployed for a more sensitive mass spectrometric response,
thus giving rise to the formation of pronounced clusters with
the ammonium cations in the positive-ion mode and with
the acetate anions in the negative-ion mode.
The analysis of the mass spectra of the fully acetylated
oligomers, obtained after RP-HPLC separation (see
Figure 2 for the HPLC profile—refer to Table 1 for the
mass peaks attributions) give further structural information,
in particular with respect to the differentiation between the
oligomers with the same polymerisation degree which com-
Chem. Eur. J. 2010, 16, 6077 – 6087
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6083

F. De Angelis et al.
sinyl substructure present
a
lower chromatographic reten-
tion time than the correspond-
ing congeners with the dehy-
drodiconiferyl substructure.
Biosynthesis: On the grounds
of the structural information
1
obtained by the H NMR spec-
troscopic analysis and the ESI
mass spectrometric investiga-
tions of the coniferyl alcohol
oligomers, formed by horse-
radish peroxidase/H2O2 oxida-
tive polymerisation in a surfac-
tant doped reaction medium, it
is possible to draw a reasona-
ble hypotheses on the biosyn-
thetic pathways leading to
their formation.
Figure 4. RP-HPLC–ESIMS profile of the coniferyl alcohol polymerisation reaction mixture.
In general, it should be con-
sidered that the reaction con-
whole reaction mixture (vide supra). We have assigned to
this compound the tentative structure of 8, as a free alcohol.
It should be emphasised that oligomers with a higher oxida-
tion degree, with respect to those we are studying herewith,
have already been reported by us in a study by MALDI-MS
of the coniferyl alcohol polymerisation products, carried out
in buffer alone.[24]
The case of the two tetramers, the one comprising the pi-
noresinyl unit and the other the dehydrodiconiferyl one, is
paradigmatic (see structures 5 and 6 for the corresponding
free alcohols). The former, in fact, takes four acetyl groups
(at m/z: 902 once cationised by the ammonium ion—m/z:
943 in the negative-ion mode), whereas the latter is acetylat-
ed five times (m/z: 944 and 985, respectively). The mass dif-
ference between the two acetylated isomers is of 42 mass
units. The acetylated trimer to which we assigned the tenta-
tive structure 9 (m/z: 942 in the positive-ion spectrum) com-
prises the dehydrodiconiferylic subunit, and it is further de-
hydrogenated with respect to 6.
ditions we have developed allow the contemporary pres-
ence, in solution, of all the reactive species and reaction in-
termediates. No precipitation occurs. Also the already just-
formed oligomeric molecules are present in solution, behav-
ing then as building blocks for the growing polymer.
As to the two dimeric lignols, the mechanism for their for-
mation is already well established,[23] and reported in
Scheme 2. Pinoresinol (2), in fact, derives from a b–b radical
coupling (see Scheme 1 for radical locations). Dehydrodico-
niferol (3) is formed by way of a radical coupling reaction,
involving position 5 of one coniferyl alcohol unit and the b
radical of a second monomer, with successive O4-a ring clo-
sure by way of an intramolecular electrophilic addition. In
both cases, one radical step is involved.
The bond formation between a b-radical and an O4-radi-
cal, on two different coniferyl alcohol molecules, produces
the reactive key intermediate quinone methide 10, which
then undergoes conjugate electrophillic addition by the
other alcohols that are present in solution (Scheme 3).[25]
Only one acetylated penta-
ACHTUNGTRENNUNGmer was then revealed (7 is
the corresponding alcohol), its
cationised molecular ion occur-
ring at m/z: 1167 (m/z: 1208 in
the negative-ion mode).
Finally, the signals of two
acetylated
hexamers
also
occur, as a non-resolved chro-
matographic peak. The one in-
cluding the pinoresinyl moiety
is at m/z: 1345 and the other
with the dehydrodiconiferyl
unit at m/z: 1386. In general,
however, the acetylated oligo-
mers that possess the pinore-
Scheme 3. Formation of a quinone methide intermediate by radical coupling of a b- radical and an O4-radical
to give the so-called b-O4 substructure. The addition to the quinone methide intermediate of a water molecule
(with R=H) gives rise to the b-O4 dimer, whereas the addition of a coniferyl alcohol molecule or oligomer
will give the trimer 4 or higher oligomers (see structures 4–7 and the text).
6084
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 6077 – 6087

Lignin Chemistry
FULL PAPER
The addition of a new coniferyl alcohol unit gives rise to the
trimer 4, with an ionic step that follows the radical one. 2
and 3 additions to quinone methide generate the two iso-
meric tetramers 5 and 6. In total, two radical steps and an
ionic one are involved in this latter case. In all the cases, a
new a-O4 bond is formed.
been fully characterised, can form a radical at the phenolic
oxygen atom of the former quinone methide moiety (see
Scheme 3), which then stabilises itself by releasing the group
located at the a-carbon atom. The quinone methide 10 is
formed again, whereas the unpaired electron is carried away
by a phenolic fragment that in most cases will produce the
two stable dimers, eventually. This decomposition reaction,
in principle, could be also promoted by acidic or basic catal-
ysis.
According to this hypothesis, the addition of the trimer to
10 forms the only pentamer 7 (in total, two radical and two
ionic steps). On the contrary, two isomers are formed by the
addition of the two tetramers giving rise to the hexamers
(three radical and two ionic steps are overall involved).
The pathway that has been outlined above firmly relies
upon the mass spectral data. The oligomeric units differ
from each other by 178 and 180 mass units, alternatively.
With the molecular weight of coniferyl alcohol at 180 Da,
the coupling reaction of two radicals, that is, one radical
step, would give a mass increment of 178 mass units with re-
spect to the starting material. This is the case, for the forma-
tion of the two dimeric alcohols and of the quinone methide
10. Conversely, the addition of one coniferyl unit (i.e., an
ionic step) to 10 would result in a 180 mass unit increment.
As the polymerisation goes on, one radical step and then an
ionic one follow, formally alternating each other and follow-
ing a regular fashion. This will produce, finally, a mixture of
oligomers that, as demonstrated by their molecular weights,
differ alternatively by 178 and 180 mass units, according to
their increasing degree of polimerisation.
The same considerations hold for the fully acetylated de-
rivatives. They differ from each other by 220 and 222 mass
units corresponding, respectively, to 178 plus 42 and 180
plus 42 mass units; 42 Da, again, is the mass increment due
to the presence of one extra acetyl moiety. As already ob-
served, however, the presence in the molecule of a pinore-
sinyl unit would introduce a defect in the mass increment of
42 Da.
On this basis it can be proposed that in our reaction
medium radical coupling mainly occurs for the formation of
the dimeric structures, that is, pinoresinol (2), dehydrodico-
niferyl alcohol (3) and the quinone methide (10). The poly-
merisation reaction then proceeds through the addition of
coniferyl alcohol itself or of the oligomers already present in
the reaction medium to the quinone methide 10. Since no
water addition to the quinone methide is observed, it is con-
ceivable that the addition of the coniferyl alcohol oligomers
is favoured over the water molecule by the micellar lipophil-
ic environment from which water is excluded. Moreover, as
already pointed out by Brunow and co-workers,[4] our work-
ing (almost neutral) pH favours the formation of benzyl aryl
ethers over the benzylic alcohols, as products of the phenolic
addition to the quinone methide.
Conclusion
Model coniferyl alcohol lignins can be created in water
under homogeneous conditions guaranteed by the presence
of a suitable cationic surfactant at a concentration higher
than c.m.c. The reaction takes place in minutes. No precipi-
tation occurs even by prolonging to hours the reaction time.
In the past, many in vitro lignins have been produced, but in
all cases the immediate precipitation of the growing polymer
was the major drawback to be overcome. A complete study
of the activity of the enzymatic catalyst, under our reaction
conditions, has been carried out, and all the reaction prod-
ucts up to the pentamer have been characterised by
1H NMR spectroscopy and ESI mass spectrometry. This soft
ionisation technique, which usefully complements the well-
established pyrolysis mass spectrometry, only recently has
been used in lignin studies opening new intriguing perspec-
tives for the structural elucidation and biosynthesis of this
molecule.[26]
Based on the oligomers structures that we have identified
in the reaction mixture, a new biosynthetic point of view has
been proposed now, which focuses attention on the quinone
methide 10, as the key intermediate of the reaction progress.
Addition reaction of the other phenolic oligomers that are
created in the reaction mixture on this conjugated olefin
allows the production of new oligomers with a higher
degree of polymerisation, according to a regularly defined
chain-reaction module. A formal, strictly alternate sequence
of a radical and an ionic step underlines the reaction prog-
ress, thus generating ordered oligolignol structures formed
by coniferyl alcohol subunits (according to the generally ac-
cepted practice of using coniferyl alcohol as the only starting
material for the in vitro experiments[4]). This consideration
marks an important difference with what is generally report-
ed in the current literature, for which lignin formation is
known to proceed in a random fashion by radical and ionic
steps.[27,28] Our experiments, in fact, are more in favour of
some alternative points of view[29,30] that assume an ordered
structure for the natural lignins.
The decomposition reaction that all the oligomers, with
the exception of the dimers, easily undergo also merits con-
sideration. Preparative TLC and solvent evaporation to iso-
late the products after HPLC separation produces a thor-
ough decomposition that finally leads to the principal for-
mation of pinoresinol and dehydrodiconiferyl alcohol. All
the oligomers from 4 to 7, just to cite only those that have
We have also demonstrated that oligomeric lignins, as
produced under our reaction conditions, are not stable mol-
ecules, at least when attempts for their isolation are carried
out. This observation fully parallels all those reported in the
literature,[5] in which natural lignins are described as chemi-
cally and enzymatically highly unstable molecules, a situa-
tion that generally makes their isolation as unaltered mole-
Chem. Eur. J. 2010, 16, 6077 – 6087
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6085

F. De Angelis et al.
about 2 mm, was eluted (by an HPLC HP1100, Agilent) by a gradient of
ammonium acetate (0.1%) (eluent A) and acetonitrile (eluent B) with B
ramping from 0 up to 40% within 15 min, flow rate=0.2 mLminÀ1 in a
RP-18 (5 mm) packaged 30–2 mm chromatographic column (Luna, Phe-
nomenex). The column eluent was then split and a 5 mLminÀ1 flow was
analysed by the ESI mass spectrometer (Quattro LC-Z-SPRAY ESI
source-triple quadrupole, Micromass) with the following source parame-
ters: capillary: 2.87 V, cone voltage: 20 and 35 V for in source fragmenta-
tion experiments, extractor: 3 V; source temperature: 1208C, desolvation
temperature: 1208C. Also the DAD spectra were acquired.
cules difficult and prevents their complete structural charac-
terisation.
In our in vitro model the lignin growth to higher polymers
appears to be limited by the coniferyl alcohol supply. This,
of course, is not the case of lignin formation in the natural,
wooden cells. A common observation, in both cases, is that
our oligomers, and the natural polymer, are “alive” mole-
cules, ready to restart a new polymerisation process accord-
ing to the environmental, natural or even laboratorial condi-
tions.
1H NMR analysis: The DHP fractions collected after HPLC separation
were dissolved in CDCl3 and analysed in a Bruker 200 MHz NMR spec-
trometer.
Finally, three points are to be underlined.[19] It has been
considered that an enzyme, under micelle conditions, repro-
duces better the cell system than in buffer alone.[31,32] As a
consequence, we think that our system represents a closer
mimicry of the natural microenvironment with respect to
the in vitro experiments that have been proposed in the
past. Secondly, the lack of water addition to quinone me-
thide-like intermediates, as is the case for our model, marks
a difference with the other in vitro studies. Last, but not
least, alternatively to other model lignins, our olignols pres-
ent a lower degree of radical coupling between oligomeric
units. This offers a biosynthetic circumstance closer to the
observation of a low rate of radical generation in the cell
wall,[33–36] thus reproducing in an in vitro experiment what is
generally observed in the in vivo lignifications process.
We would also comment, however, that in such a compli-
cated “lignin world” our biosynthetic approach and the re-
sults we have obtained usefully complement all the other re-
lated in vitro studies. It might be interesting, in fact, to look
at the effect of carbohydrates and cellulose, for instance, in
the biosynthetic outcome in the presence of a micellised cat-
ionic surfactant.
Acknowledgements
Encouragement and helpful discussions with Prof. R. Nicoletti (Univer-
sity Sapienza of Rome) are gratefully acknowledged.
[1] A. Payen, Compt. Rend. 1838, 7, 1052–1056.
[2] N. G. Lewis, L. B. Davin, S. Sarkanen in Lignin and Lignan Biosyn-
thesis, (Eds.: N. G.Lewis, S. Sarkanen) ACS, Washington, 1998,
pp. 1–27.
[3] B. Cathala, B. Saake, O. Faix, B. Monties, Pol. Deg. Stab. 1998, 59,
65–69.
[4] G. Brunow, I. Kilpelainen, J. Sipila, K. Syrjanen, P. Karhunen, H.
Setala, P. Rummakko in Lignin and Lignan Biosynthesis (Eds.: N. G.
Lewis, S. Sarkanen), ACS, Washington, 1998, pp. 131–147.
[5] J. Ralph, K. Lundquist, G. Brunow, F. Lu, H. Kim, P. F. Schatz, J. M.
Marita, R. D. Hatfield, S. A. Ralph, J. Holst Christensen, W. Boer-
jan, Phytochem. Rev. 2004, 3, 29–60.
[6] K. Freudenberg, W. Heimberger, Chem. Ber. 1950, 83, 519–530.
[7] K. Freudenberg, Nature 1959, 183, 1152–1155.
[8] K. Freudenberg, A. C. Neish, Constitution and Biosynthesis of
Lignin, Springer, Berlin, 1968, and references therein.
[9] R. H. Atalla in Lignin and Lignan Biosynthesis (Eds.: N. G. Lewis,
S. Sarkanen), ACS, Washington, 1998, pp. 172–179.
[10] N. Terashima, R. H. Atalla, S. A. Ralph, L. L. Landucci, C. Lapierre,
B. Monties, Holzforschung 1995, 49, 521–527.
Experimental Section
[11] N. Terashima, R. H. Atalla, S. A. Ralph, L. L. Landucci, C. Lapierre,
B. Monties, Holzforschung 1996, 50, 9–14.
[12] B. Cathala, B. Monties, Int. J. Biol. Macromol. 2001, 29, 45–51.
[13] B. Cathala, B. Chabbert, C. Joly, P. Dole, B. Monties, Phytochemistry
2001, 56, 195–202.
Materials: Horseradish peroxidase type II, 200 UmgÀ1 (Sigma); coniferyl
alcohol HPLC purity grade (Aldrich); HPLC grade solvents (Riedel-
deHaꢃn); (CTA)2SO4 synthesised according to reference [10].
Methods
[14] K. Freudenberg, J. Harkin, M. Reichert, T. Fukuzumi, Chem. Ber.
1958, 91, 581–590.
[15] T. K. Kirk, W. J. Connors, R. D. Bleam, W. F. Hackett, J. G. Zeikus,
Proc. Natl. Acad. Sci. USA 1975, 72, 2515–2519.
[16] K. T. Kirk, G. Brunow in Methods in Enzimology, Biomass, Vol. 161,
Part B (Eds.: W. A. Wood, S. T. Kellog), Academic Press, San Diego,
1988, pp. 65–72.
Test for horseradish peroxidase activity in the presence of (CTA)2SO4:
H2O2 (0.03%) and horseradish peroxidase (3.5 mUE) were sequentially
added to phospate buffer (1 mL, 9 mm at pH 6.5) containing o-phenilen-
diammine (8.8ꢂ10À5 m) and (CTA)2SO4 (3 mm or 9 mm or 27 mm). The
reaction mixture was maintained at 308C and the UV absorbance at
492 nm (Shimadzu UV 160 A double ray spectrophotomer) was mea-
sured after 2,5 min.
[17] E. Adler, S. Yllner, Acta Chem. Scand. 1953, 7, 570–581.
[18] Y. Z. Lay, K. V. Sarkanen, Cellul. Chem. Technol. 1975, 9, 239.
[19] F. De Angelis, R. Nicoletti, N. Spreti, F. Verꢄ, Angew. Chem. 1999,
111, 1364–1367; Angew. Chem. Int. Ed. 1999, 38, 1283–1285.
[20] L. A. Bunton, L. H. Gan, J. R. Hoffatt, L. S. Romsted, G. Savelli, J.
Phys. Chem. 1981, 85, 4118–4125.
[21] N. Spreti, A. Bartoletti, P. Di Profio, R. Germani, G. Savelli, Bio-
technol. Prog. 1995, 11, 107–111.
[22] Y. Katayama, T. Fukuzumi, Mokuzai Gakkaishi 1978, 24, 664–667.
[23] D. Fengel, G. Wegener, Wood—Chemistry, Ultrastructure, Reactions.
Walter de Gruyter. Berlin, 1989, pp. 137–181.
TLC: Eluent mixture: CH2Cl2/CH3OH 20:1
Coniferyl alcohol polymerisation in the presence of (CTA)2SO4: Coniferyl
alcohol (5 mg) was dissolved in sodium phosphate buffer (1.0 mL, 5 mm,
pH 6.5) containing (CTA)2SO4 (2.7·10À2 m). H2O2 (10 mL, 10%) and
buffer (5 mL) containing two purpurogallin units of horseradish perox-
idase (Type II, 200 UmgÀ1, Sigma, St. Louis MO) were sequentially
added to this solution. The clear solution was vigorously stirred over a
period of 5 min, the reaction was then stopped by the addition of one
drop of Na2S2O7 (5%) and the mixture extracted with ethyl acetate. The
organic layer was washed with brine, dried and the solution evaporated
under vacuum.
[24] F. De Angelis, P. Fregonese, F. Verꢄ, Rapid Commun. Mass Spec-
trom. 1996, 10, 1304–1308.
HPLC and HPLC–ESIMS analysis: The sample, dissolved in a solution
of ammonium acetate (0.1%)/acetonitrile 50:50 at a concentration of
[25] G. J. Leary, Wood Sci. Technol. 1980, 14, 21–34.
6086
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 6077 – 6087

Lignin Chemistry
FULL PAPER
[26] S. Reale, A. Di Tullio, N. Spreti, F. De Angelis, Mass Spectrom. Rev.
2004, 23, 87- 126.
[32] G. Savelli, N. Spreti, P. Di Profio, Curr. Opin. Colloid Interface Sci.
2000, 5-5, 111–117.
[27] R. R. Sederoff, J. J. MacKay, J. Ralph, R. D. Hatfield, Curr. Opin.
Plant Biol. 1999, 2, 145–152.
[28] W. Boerjan, J. Ralph, M. Baucher, Annu. Rev. Plant Biol. 2003, 54,
519–546.
[33] E. Adler, Wood Sci. Technol. 1977, 11, 169–218.
[34] K. V. Sarkanen in Lignin: Occurrence, Formation, Structure and Re-
action (Eds.: K. V. Sarkanen, C. H. Ludwig), Wiley, New York, 1971,
pp. 95–163.
[29] K. Forss, K. E. Fremer, B. Stenlund, Pap. Puu 1966, 48, 565–568.
[30] D. A. I. Goring in Lignin Properties and Materials (Eds.: W. G.
Glasser, S. Sarkanen), ACS, Washington, 1985, pp. 11–28.
[31] C. Tanford, The Hydrofobic Effect: Formation of Micelles and Bio-
logical Membranes, Wiley, New York, 1980.
[35] M. Tanahashi, T. Higuchi, Wood Res. 1981, 67, 29–42.
[36] K. Syrjꢅnen, G. Brunow, J. Chem. Soc. Perkin Trans. 1 2000, 183–
187.
Received: December 2, 2009
Published online: April 15, 2010
Chem. Eur. J. 2010, 16, 6077 – 6087
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6087