The aerobic oxidative cleavage of lignin to produce hydroxyaromatic benzaldehydes and carboxylic acids via metal/bromide catalysts in acetic acid/water mixtures

Article abstract of DOI:10.1002/adsc.200800614

Roughly 30% of all woody plants is composed of lignin. Five different lignin samples, from wood and bagasse, were oxidized in air with a cobalt/ manganese/zirconium/bromide (Co/Mn/Zr/Br) catalyst in acetic acid as a function of time, temperature, pressure, and lignin and catalyst concentrations. 18 products were identified via gas chromatography-mass spectrometry (GC/MS). The most valuable products from lignin were 4-hydroxybenzaldehyde, 4-hydroxybenzoic acid, 4-hydroxy-3-methoxybenzaldehyde (vanillin), 4-hydroxy-3-methoxybenzoic acid (vanillic acid), 4-hydroxy-3,5-dimethoxybenzaldehyde (syringaldehyde) and 4-hydroxy-3,5-dimethoxybenzoic acid (syringic acid). 10.9 wt% of the lignin was converted to the aromatic products. By the use of model compounds we demonstrate that 1) the presence of the phenolic functionality on an aromatic ring does inhibit the rate of reaction but that the alkyl group on the ring still does oxidize to the carboxylic acid, 2) that the masking of phenol by acetylation occurs at a reasonable rate in acetic acid, 3) that the alkyl group of the masked phenol does very readily oxidize, 4) that an acetic anhydride/acetic acid mixture is a good oxidation solvent and 5) that a two-step acetylation/ oxidation to the carboxylic acid is feasible.

Full text of DOI:10.1002/adsc.200800614

FULL PAPERS
DOI: 10.1002/adsc.200800614
The Aerobic Oxidative Cleavage of Lignin to Produce Hydroxy-
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
aromatic Benzaldehydes and Carboxylic Acids via Metal/Bromide
Catalysts in Acetic Acid/Water Mixtures
a,b,
Walt Partenheimer *
a
Central Research and Development, E. I. DuPont de Nemours & Co., Inc., Experimental Station, Wilmington, Delaware
19880-0262, USA
b
Current address: 7253 SW 26th Ave, Portland, Oregon 97219, USA
E-mail: wpartenheimer@msn.com
Received: October 8, 2008; Revised: January 12, 2009; Published online: February 13, 2009
Abstract: Roughly 30% of all woody plants is com- of model compounds we demonstrate that 1) the
posed of lignin. Five different lignin samples, from presence of the phenolic functionality on an aromat-
wood and bagasse, were oxidized in air with a cobalt/ ic ring does inhibit the rate of reaction but that the
manganese/zirconium/bromide (Co/Mn/Zr/Br) cata- alkyl group on the ring still does oxidize to the car-
lyst in acetic acid as a function of time, temperature, boxylic acid, 2) that the masking of phenol by acety-
pressure, and lignin and catalyst concentrations. 18 lation occurs at a reasonable rate in acetic acid, 3)
products were identified via gas chromatography- that the alkyl group of the masked phenol does very
mass spectrometry (GC/MS). The most valuable readily oxidize, 4) that an acetic anhydride/acetic
products from lignin were 4-hydroxybenzaldehyde, acid mixture is a good oxidation solvent and 5) that
4
-hydroxybenzoic acid, 4-hydroxy-3-methoxybenzal- a two-step acetylation/oxidation to the carboxylic
dehyde (vanillin), 4-hydroxy-3-methoxybenzoic acid acid is feasible.
vanillic acid), 4-hydroxy-3,5-dimethoxybenzalde-
(
hyde (syringaldehyde) and 4-hydroxy-3,5-dimethoxy-
benzoic acid (syringic acid). 10.9 wt% of the lignin Keywords: autoxidation; cobalt; green chemistry; lig-
was converted to the aromatic products. By the use nins; organosolv lignin; oxidation
Introduction
matic portion of lignin is composed of p-hydroxyben-
zene, guaiacyl (4-alkyl-2-methoxyphenol), and syring-
Lignin is a major component of woody plants, the yl (4-alkyl-2,5-dimethoxyphenol) units. The most effi-
second most abundant natural polymer (after cellu- cient and economic method for the conversion of al-
lose), and is produced, largely as a waste-product, kylaromatic compounds to aromatic carboxylic acids
[
1,2]
from the wood-pulp or sugar cane milling industry.
is via the homogeneous air oxidation of these com-
[
4]
The U.S. pulp and paper industry processes approxi- pounds using metal/bromide catalysts. This led us to
[
3]
mately 108 million tons of wood annually. It is most predict that a series of substituted aromatic acids
commonly used for its fuel value, i.e., its unselective might be produced by the aerobic oxidation of the
oxidation to carbon dioxide. Lignin has a complex, the structural units of lignin using metal/bromide cat-
polymeric structure whose exact structure is unknown alysts. For example, the guaiacyl unit could be oxi-
and which varies depending on its source. The aro- dized as shown in Eq. (1).
456
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 456 – 466

The Aerobic Oxidative Cleavage of Lignin to Produce Hydroxyaromatic Benzaldehydes
FULL PAPERS
Figure 1. Types of aromatic functional groups in lignin and types of aromatic acids that can form as a result of oxidation of
these functional groups. This figure is simplified in that biphenyl and biaryl derivatives, known to be present, are not shown.
[1a]
Estimated amounts from ref.
[
4]
In lignin one has a propyl group, as shown in Eq. aromatic carboxylic acids. It is used industrially to
1), which is partially oxygenated and substituted in produce a variety of carboxylic acids including multi-
(
complex ways to form the polymeric unit. The pre- billion pound/year amounts of terephthalic acid, the
dicted acids are given in Figure 1. These 4-carboxy major component of polyethylene terephthalate
and 4-formyl derivatives of phenol have known com- (PET). Thus, most of the engineering problems such
mercial applications for their antibacterial, antifungal as corrosion in bromide/acetic acid mixtures, recycling
[5]
properties and use in food preservation as para- of the solvent and catalyst, and efficient mixing of the
[6]
bens. Thus aerobic oxidation of what is usually con- dioxygen with the acetic acid solvent have been
sidered a waste product, could be a inexpensive green solved. 2) The unit cell of the lignin polymer is based
source of valuable chemicals.
on propylbenzene. It is well-established that the
Our optimism that this could become a viable com- highly active metal/bromide catalysts can convert pro-
mercial source of valuable chemicals arises for the fol- pylbenzene derivatives to benzoic acid in high yield
[
4]
lowing reasons: 1) The metal/bromide catalyzed aero- [Eq. (2)].
bic oxidation of alkylaromatic compounds in acetic Two of the carbon atoms on the propyl side-chain
acid solvent is a well-established method to produce become oxidized to primarily carbon dioxide and
Adv. Synth. Catal. 2009, 351, 456 – 466
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
457

Walt Partenheimer
FULL PAPERS
water. 3) Despite the relative insolubility of polymers
There are a number of reports where oxidation
in most solvents, we have demonstrated that they can methods have been used to prepare valuable alde-
[
10–13]
be aerobically oxidized using metal/bromide homoge- hyde, acids, and phenols from lignin.
There are
[
7]
neous catalysts to valuable products. One particular- two reports of lignin being oxidized by the Co/Mn/Br
ly pertinent example is the oxidation of polystyrene catalyst in acetic acid. Organosolv lignins have been
to benzoic acid where the aliphatic backbone is oxi- oxidized using a Co/Mn/Br catalyst in acetic acid at
[
7]
dized to benzoic acid in 88% yield [Eq. (3)].
2108C to give vanillin and vanillic acid in 1.9 and 3.0
wt% yield, respectively. The same catalyst and sol-
vent was used to measure the rates and activation en-
[
14]
[
15]
ergies of lignins from agricultural wastes.
The re-
search in this paper gives a much more detailed
report on the products found when lignins are oxi-
dized with metal/bromide oxidations in acetic acid
over a wide range of the experimental variables time,
temperature, pressure, water concentration, and con-
centration of lignin.
This demonstrates that the aliphatic backbone of
This is a brief report from a biomass project that
the polymer can be oxidized to produce aromatic car- was pursued at DuPont Central Research. Unfortu-
boxylic acids. 4) One potential method to separate nately, not all of the planned avenues of research
wood pulp from lignin uses an acetic acid/water were completed, especially those in which the lignin
(
80:20 v/v) mixture (the organosolv method) which phenolic groups were masked by the acetate group,
has little pollution and offers the possibility of total but a number of positive scientific and technological
biomass recovery of the wood. In this method the results were obtained and are reported here.
lignin is extracted into the solvent leaving the insolu-
[
8]
ble pulp to be further processed into paper products.
Since acetic acid is the solvent of choice for metal/ Results and Discussion
[
4]
bromide autoxidation, there is a solvent compatibili-
ty in which separation of lignin from the pulp and The Use of Model Compounds to Determine the
then oxidation of the lignin to obtain valuable acids Effectiveness of Metal/Bromide Catalysts to Oxidize
and aldehyde can be performed. 5) Oxidation meth- Lignin
ods are one of the fundamental characterization tools
[
2]
to determine the structure of lignin. Indeed perman- We designed a series of experiments on model com-
ganate oxidation of lignin, after methylation, does pounds to understand the influence of phenols on au-
produce the acids (as their methyated derivatives) toxidation and to determine if acetylation of the phe-
given in Figure 1. 6) We have already shown that 3,4- nolic group would improve the yields to the aromatic
[
9]
dimethoxytoluene, a lignin model compound, can be acids. The oxidation of lignin will generate hydroxyar-
successfully oxidized to its corresponding acid al- omatic acids through the oxidation of the aliphatic
though it did have many unusual characteristics. The propyl group to the benzaldehyde and then to the car-
maximum yield was 51% to the benzaldehyde and boxylic acid [Figure 2, Eq. (1) and Eq. (2)]. The phe-
92% to the carboxylic acid [Eq. (4)].
nolic functionality on the ring (the hydroxy group) is
known to strongly retard autoxidations, i.e., they are
potent anti-oxidants. Also phenols are known to oxi-
dize at a faster rate than alkylbenzenes to give initial-
ly quinones which are subsequently degraded to
[
16]
carbon dioxide.
Thus one can anticipate a signifi-
cant yield loss to the aromatic carboxylic acid via
phenol oxidation. A known method to avoid phenol
oxidation is to first ꢁmaskꢂ the phenolic group by re-
acting it with acetic anhydride to form the acetate,
[
4,17]
[
Figure 2, Eq. (3)].
. Below we will demonstrate
458
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 456 – 466

The Aerobic Oxidative Cleavage of Lignin to Produce Hydroxyaromatic Benzaldehydes
FULL PAPERS
Figure 2. Oxidation/acetylation of the aromatic portion of lignin.
that 1) the presence of the phenolic functionality does first hour. At 1.8 h a weak rate of oxygen uptake was
inhibit the rate of reaction but that the reaction still observed along with a sizable rate of carbon dioxide
À1
does proceed to the carboxylic acid, 2) that acetyla- formation (0.31 mL/min; 0.0124 mmolmin ). The re-
tion of a phenol occurs at a reasonable rate in acetic action was continued for 4.2 h at which the conver-
acid, 3) that the resultant phenolic acetate does readi- sion was 22% with an acid yield of of 4-hydroxyben-
ly oxidize, 4) that an acetic anhydride/acetic acid mix- zoic acid of 12% Assuming all of the carbon dioxide
ture is a good oxidation solvent and 5) that a two-step came from complete oxidation of 4-hydroxybenzalde-
acetylation/oxidation is feasible.
hyde, then 4% of the 4-hydroxybenzaldehyde was lost
4
-Hydroxybenzaldehyde was oxidized because it as CO2.
contains the highly reactive benzaldehyde functional
Vanillin acetate was oxidized, see Eq. (6), to con-
group simultaneously with the 4-hydroxy (phenolic) trast its behavior to that of 4-hydroxybenzaldehyde
group [Eq. (5)]. The initial conditions and results are discussed above. The compounds are structurally simi-
given in Table 1. The reaction exhibited autocatalytic
behavior in which no oxygen uptake occurred in the
[
a]
Table 1. Summary of reactions of lignin model compounds.
Reaction Reagent, initial conc.
o
[c]
À1
5 [b]
T [ C] Catalyst
Type of reaction Rate [s ꢃ10 ]
t1/2 [h]
1
2
3
4
5
4-hydroxybenzaldehyde, 0.41M
3-methoxy-4-acetoxybenzaldehyde, 0.52M 80
4-methoxytoluene, 0.82M
3-methoxy-4-hydroxytoluene, 0.37M
3-methoxy-4-acetoxytoluene, 0.37M
80
Co/Mn/Zr/Br Oxidn
–
–
Co/Mn/Br
Co/Mn/Br
None
Oxidn
Oxidn
Acetylaction
81 (16)
27 (1.3)
4.54 (0.45)
7.2 (0.62)
0.23
0.71
4.2
2.8
80
95
95
Co/Mn/Br/Zr Oxidn
[
[
[
a]
Reactions 1 and 2 are in acetic acid; reactions 3–5 in a acetic anhydride/acetic acid mixture.
Rate of disappearance of the reagent. Standard deviation in parenthesis.
Concentration of catalyst components in reaction 1 is 0.02/0.02/0.0005/0.04M, in reactions 2, 3 it is 0.01/0.01/0.02M, and
in entry 5 0.01/0.01/0.005/0.02M.
b]
c]
Adv. Synth. Catal. 2009, 351, 456 – 466
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
459

Walt Partenheimer
FULL PAPERS
lar except that in vanillin acetate the phenolic group
has been ꢁmaskedꢂ by the acetate group. The 3-me-
thoxy group only slightly deactivates the oxidation of
benzaldehyde so the comparison of vanillin acetate
[
9]
with 4-hydroxybenzoic acid is valid. The oxidation
of vanillin acetate is not autocatalytic as the reaction
immediately initiated. Also the rate of reaction of va-
nillin acetate is much faster than that of 4-hydroxy-
benzaldehyde. The rate of disappearance of the re-
agent has a half-life of only 0.23 h in contrast to a
22% conversion of 4-hydroxytoluene in 4.2 h, see residual 4-hydroxy-3-methoxytoluene in the mixture.
Table 1. The reaction of 4-acetoxy-3-methoxybenzal- The half-life of 4-hydroxy-3-methoxytoluene was 2.8
dehyde was complete in two hours. A 90% isolated h after the reaction initiated. This rate at 958C is
yield of the 4-acetoxy-3-methoxybenzoic acid was ob- slower than that of 3,4-dimethoxytoluene, a com-
tained by evaporation of the reactor effluent and pound of similar reactivity, at 808C which has a half-
water-washing the product.
life of 1.3 h. Thus the reaction is being inhibited by
4
-Methoxytoluene is a model compound of similar the presence of the residual phenol as expected.
reactivity to the guaiacyl and syringyl groups in lignin.
It was oxidized in a 30% Ac O/HOAc mixture to de-
2
termine if this was a viable oxidation solvent, see ex- Selection and Characterization of the of Lignins Used
ample 3 of Table 1. A rapid oxidation of this substrate in this Study
occurred with a half-life of 0.71 h confirming that
indeed it is a good oxidation solvent. The rate con- The lignins were used as received from Aldrich
stant for the disappearance of 3,4-dimethoxytoluene Chemical.
À4 À1
in acetic acid, as reported by us, is 1.6ꢃ10 s (t =
1) Lignin, organosolv was isolated from a commer-
1
/2
[9]
1
.2 h). The reactivity of 4-methoxytoluene is about cial pulp mill using mixed hardwood (50% maple,
6% higher than that of 3,4-dimethoxytoluene.
[9]
1
35% birch, 15% poplar). Analysis of the sample gave
Thus, in an acetic anhydride/acetic acid mixture it is 18.9 wt% methoxy groups, <0.5 wt% sugars, <1 wt%
1
at least as active, if not more so, than in acetic acid ash. H NMR analysis indicates almost equal amounts
[
18]
alone.
of phenolic and aliphatic hydroxy groups.
This
The following reactions demonstrate that one can lignin was selected because one of the organosolv
first acetylate a phenol and then oxidize it to its car- pulping methods uses 80% acetic acid/20% water
boxylic acid. The rate of acetylation of 4-hydroxy-3- which is the same solvent system used in this work.
methoxytoluene, see Eq. (7), in 20% Ac O/HOAc
2) Lignin, organosolv, acetate – this is a derivative
of organosolv lignin, (see 1 above) prepared by dis-
solving the lignin in acetic acid and reacting it with
acetic anhydride using sodium acetate as the catalyst.
This procedure acetylates the phenol groups. Prior to
isolation, it was partially bleached with hydrogen per-
2
1
oxide. H NMR analysis indicates almost equal
[18]
amounts of aromatic and aliphatic acetoxy groups.
This lignin was chosen because the acetylation proce-
dure ꢁmasksꢂ the phenolic groups, see above.
3) Lignosulfonic acid, sodium salt, acetate – this is
prepared by reacting the sodium salt of lignosulfonic
acid with acetic anhydride in acetic acid. The lignin is
and that of 4-hydroxytoluene was determined, see ex- believed to be derived from Norway spruce. It con-
ample 4 in Table 1. The acetic anhydride/reagent ratio tains 11.6 wt% sodium, 2.6 wt% sulfur, 6.5 wt% me-
is 5.4 mol/mol. The half-life was 4.2 h at 958C. The re- thoxy groups, and 15.8 wt% carboxyl groups. The mo-
[18]
action was continued at 958C for 7 h and then cooled lecular weight is 174,400 (MW) and 16,000 (MN).
to room temperature overnight. 4) Lignin, hydrolytic – this is a polymeric autohy-
The GC analysis of the material in the Ac O/HOAc drolysis lignin material isolated from a commercial
2
indicated that the reaction was 94.3% complete that hydrolysis pilot plant using predominatly sugar cane
is, there was 5.7% unreacted 4-hydroxy-3-methoxyto- bagasse as raw material. 9–-11 wt% of the sample is
1
luene. This mixture was oxidized with a Co/Mn/Zr/Br methoxy groups. H NMR analysis indicates a pre-
catalyst at 958C, see Eq. (8) and example 5 of dominance of phenolic over aliphatic hydroxy groups
[
18]
Table 1. The reaction was autocatalytic because of the and shows evidence of methyl groups.
460
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 456 – 466

The Aerobic Oxidative Cleavage of Lignin to Produce Hydroxyaromatic Benzaldehydes
FULL PAPERS
5
) Lignin, hydrolytic, hydroxymethyl – prepared Table 3. Results from GC/MS and silation GC/MS analysis
from the products of the autoxidation of Aldrich ꢁlignin or-
ganosolvꢂ.
from lignin, hydrolytic (see 4 above) by treating lignin
in a homogeneous aqueous alkali solution with form-
1
aldehyde. H NMR analysis indicates that a significant
Product(s)
Probable origin
increase in aliphatic hydroxy groups occurred during
[
18]
hydroxymethylation.
Methyl bromide, glycolic
acid, acetoxyacetic acid,
maleic acid, fumaric acid
From acetic acid solvent. By-
products normally found in
metal/bromide oxidation in
acetic acid
The elemental analyses of the lignin samples used
are given in Table 2. This was done to ascertain the
amount of nitrogen and sulfur in the samples since
Methyl formate, methyl ace- Products expected from oxi-
tate, formic acid
dation of acetic acid and
lignin
Table 2. Elemental analysis of lignin compounds evaluated
in this work.
Propionic acid
Impurity in the acetic acid
solvent
Description
C
H
N
S
Acetone, ethyl acetate, 2-fur- From lignin oxidation
aldehyde, 2-furoic acid
%
%
%
%
p-Hydroxybenzoic acid
Expected product from au-
toxidation of p-hydroxylphen-
yl units in lignin
Expected product from au-
toxidation of guaiacyl units in
lignin
lignin hydrolytic
lignin hydrolytic – hydroxymethyl 57.64 5.66 0.41 <0.20
derivative
lignosulfonic acid, sodium salt ace- 48.70 4.86 3.54 4.87
tate
lignin organosolv
57.73 5.54 0.54 <0.20
Vanillin, vanillic acid
66.05 5.67 0.17 <0.20
62.84 5.49 0.11 <0.20
Syringaldehyde, syringic acid Expected product from au-
lignin organosolv acetate
toxidation of syringyl units in
lignin
often compounds containing these elements are
strong antioxidants. Except for the expected high
sulfur in the lignosulfonic acid salt, the samples did furaldehyde probably come from oxidation of the pi-
not contain significant amounts of sulfur or nitrogen. noresinol linkages which constitute 8–14% of hard-
[4]
[19]
Sulfonic acids themselves are not antioxidants.
wood lignin, see Eq. (9).
Formic acid has been shown to come from the oxi-
dation of the aliphatic alcohol on the propyl side
[
1,2]
Product Characterization by GC/MS and Silation GC/ chain.
Products observed during the oxidation of
MS Analysis hydrocarbons in acetic acid form from both the oxida-
[
7,9,20]
tion of the hydrocarbon and the solvent.
The
Characterization of products from the aerobic oxida- same by-products associated with the oxidation of the
tion of Aldrich ꢁlignin organosolvꢂ for 2 h at 1508C acetic acid solvent (methyl acetate, methyl bromide,
and 70 bar catalyzed by a Co/Mn/Zr/Br catalyst in glycolic acid, maleic acid) have been detected during
acetic acid is given in Table 3. Four compounds of lignin autoxidation, see Table 3. The amount of ethyl
high intensity corresponded to vanillin, vanillic acid, acetate varied widely from the same sample which
syringaldehyde, and syringic acid. 4-Hydroxybenzoic may suggest the the samples were non-uniform in
acid was also detected but was much less intense. character, i.e., truly representative samples were not
These are indicated in the table along with a number always obtained.
of other minor components associated with lignin and
acetic acid oxidation. It is significant that no ring bro-
minated products were detected considering the high Expected Products based on Lignin Structure
electron density of the phenolic-methoxylated rings.
Ring bromination would convert the active ionic bro- Figure 1 has been adapted from the structure of lignin
[
1a]
mide in an inactive catalytic form and hence reduce given in ref. If the metal/bromide catalysts can suc-
the activity of the catalyst. The 2-furoic acid and 2- cessfully oxidize the ꢁuncondensed phenolic ligninꢂ,
Adv. Synth. Catal. 2009, 351, 456 – 466
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
461

Walt Partenheimer
FULL PAPERS
then approximately 15 wt% of the lignin will be con- which were extracted from the lignin into the solvent
verted to aromatic benzaldehydes and acids, see which made the reactions autocatalytic (see above).
Figure 1. The products differ only in the number of To go to higher temperatures and to maintain the
methoxy groups on the aromatic ring that is, n=0, 1, system in the liquid phase, small vials were used at
2
in Figure 1. The ꢁuncondensed, non-phenolic ligninꢂ elevated pressures to keep the acetic acid in the
can be converted to ꢁuncondensed, phenolic ligninꢂ by liquid phase. The experiments were performed on
hydrolysis of the PhÀOÀX ether bond which would batch in which the contents of the vials were analyzed
then give the same products as the former. Solvolysis after cooling to room temperature. This same tech-
[
2,21,22]
of the OÀX bond has been extensively studied.
nique has been successfully used previously to oxidize
The aromatic acids from uncondensed lignin will be a number of selected reagents including p-xylene and
[
27]
different from those from condensed lignin since the 4-hydroxylmethylfurfural to the expected products
condensed lignin has an additional carbon atom that in high yield. The experiments are designed so there
can be oxidized to a carboxylic acid to produce iso- is sufficient O in the vials to oxidize the reagents.
2
penic and ortho-phthalic acids. The GC/MS analysis
The experimental variables studied were 1) temper-
of the organosolv lignin gives products characteristic ature (100–2208C), 2) time (1–10 h), lignin concentra-
of only uncondensed phenolic compounds, that is, hy- tion (3.8–38 wt% in solvent), catalyst concentration
droxybenzoic acid, vanillin, vanillic acid, syringalde- (6.9–30 mM Co), concentration of water in acetic acid
hyde, and syringic acid.
(0, 8, 20 wt%), type of catalyst (Co/Mn/Zr/Br, HBr)
and pressure (69, 136 bar). Only a selected number of
these experiments are reported in Table 4 and
Table 5. The Co/Mn/Zr/Br catalyst was chosen be-
cause it is one of the most active and selective metal/
Results from the Aerobic Oxidation of Various
Lignins
[
4,9,20]
bromide catalysts known.
HBr was chosen be-
Initial attempts to oxidize lignin were made in a glass cause it is an active and selective catalyst [23], but
autoclave at ambient pressure and 958C as previous also provides acidity which may aid hydrolysis of the
[9,16,20]
described.
The reaction did not initiate probably lignin. Increasing the catalyst concentration increases
because of the presence of phenolic compounds activity but also selectivity in alkylaromatic aerobic
Table 4. Initial conditions for the aerobic oxidation of selected lignins.
[a]
o
Entry Name of lignin
Catalyst type, [Co] mM H O [wt%] T [ C] Time [h] P [bar] AHCTUNGRETNNUNG[ lignin] wt%
2
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
organosolv
organosolv
organosolv
organosolv
organosolv
organosolv
organosolv
organosolv
organosolv
organosolv
organosolv
organosolv
organosolv
organosolv
organosolv
organosolv
organosolv
hydrolytic
hydrolytic
hydrolytic
organosolv acetate
organosolv acetate
organosolv acetate
organosolv acetate
hydrolyic hydroxymethyl
1, 6.9
1, 6.9
1, 6.9
1, 6.9
1, 6.9
1, 6.9
1, 6.9
1, 6.9
1, 6.9
1, 6.9
1, 6.9
1, 30
1, 6.9
1, 6.9
1, 6.9
1, 6.9
2, 55
1, 30
1, 6.9
1, 6.9
1, 30
1, 6.9
1, 6.9
1, 6.9
1,30
8.00
8.00
8.00
8.00
8.00
8.00
0.00
0.00
20.00
20.00
8.00
0.00
8.00
8.00
8.00
8.00
0.00
0.00
8.00
8.00
0.00
8.00
8.00
8.00
0.00
0.00
180
140
140
140
140
140
140
140
140
140
140
140
220
140
140
140
140
140
140
100
140
140
180
140
140
140
2.00
8.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
10.00
2.00
2.00
2.00
2.00
8.00
2.00
2.00
2.00
2.00
138
138
69
69
69
69
69
69
69
69
138
69
138
69
69
69
69
69
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
4.00
16.00
24.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
138
138
69
138
138
138
69
sulfonic acid, sodium salt, acetate 1, 30
69
[
a]
Catalyst type 1 is a Co/Mn/Zr/Br mixture at a ratio of 1.00/1.00/0.10/2.00 mol/mol. Catalyst type 2 is HBr.
462
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 456 – 466

The Aerobic Oxidative Cleavage of Lignin to Produce Hydroxyaromatic Benzaldehydes
FULL PAPERS
Table 5. Aromatic benzaldehydes and benzoic acids from the autoxidation of lignin.
Entry Hydroxybenz- Hydroxybenzoic Vanillin, Vanillic
Syringaldehyde Syringic
Total Syringic/vanillic
acid [mol/mol]
aldehyde
wt%]
acid [wt%]
[wt%]
acid [wt%] [wt%] acid [wt%]
[
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
0
0
0
0
0
0
0
0
0
0
0
0
0.84
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1.22
0
0
0
0
0
0
0
0
0.99
1.12
0.98
0.88
0.83
0.39
0.73
0.70
1.36
1.17
0.92
0.63
0.0
1.02
0.79
0.79
0.51
1.12
1.45
0.62
0.37
1.27
1.02
1.01
0
2.91
2.34
2.35
1.05
1.46
0.96
1.58
1.20
1.13
0.96
1.82
1.03
0.0
1.04
0.95
0.26
1.09
1.35 [
N.M.
N.M.
0.84
N.M.
N.M.
N.M.
0
2.52
2.86
3.24
1.83
1.41
0.75
2.36
2.34
3.48
3.19
2.36
0.13
0.0
3.03
2.66
2.63
0.95
2.33
1.73
0.61
0
4.51
4.4
10.9 1.80
10.7 2.10
10.2 2.06
6.61 2.42
6.53 1.86
4.41 2.27
7.65 2.32
7.27 2.83
9.35 2.76
8.43 2.95
8.93 2.26
4.52 1.72
3.63
2.84
2.84
2.31
2.99
3.03
3.39
3.10
3.83
2.73
0.0
2.85
2.39
0.65
2.11
1.51
N.M.
N.M.
1.01
N.M.
N.M.
N.M.
0.28
0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
0.84
–
7.94 2.85
6.79 2.90
4.33 3.12
4.67 1.91
7.53 1.55
a]
–
–
1.19
0.98
2.22 0.83
3.24
2.79
2.51
0.64
0
–
–
–
0.92
0.2
2.55
2.74
2.49
–
0.15
0.05
0.00
[
a]
N.M. =not measured.
[
16,20a]
oxidations.
The water concentration was varied acid is a catalyst-modified free radical chain mecha-
[16,25]
to enhance possible hydrolysis reactions, see Figure 1. nism, commonly called autoxidation.
It is known that increasing water concentration de- autoxidation is often idiosyncratic
creases the activity of the catalyst.
The rate of
[
26]
because it is
[
4,16]
highly sensitive to the presence of trace amounts of
One does not measure the oxygen uptake or specific metals and organic compounds such as phe-
[
4]
carbon dioxide formation from the oxidation experi- nols and amines. The presence of these substances
ments given in Table 4 and Table 5 which represents can make the reactions autocatalytic with varying in-
the usual evidence that aerobic oxidation is occurring. duction times which can make reproducibility in yield
However, direct evidence that oxidation is occurring difficult. This was observed occasionally in this work.
is from the expected products of oxidation, see
Table 3, and from the changes in methyl acetate for-
mation. Methyl acetate is a characteristic by-product Results from Oxidation of Organosolv Lignin
of aerobic oxidation which, just like carbon dioxide,
originates from the catalytic decarboxylation of acetic The yields to hydroxybenzaldehyde, hydroxybenzoic
[24]
acid.
Figure 3 illustrates the expected increase in acid, vanillin, vanillic acid, syringaldehyde and syrin-
rate of formation of methyl acetate as a function of gic acid are given on Table 5. The yields are given in
temperature. It also increased with reaction time weight percent because the exact composition of
during lignin oxidation. For example, when organo- these aromatic moieties in the lignin are unknown.
solv acetate was oxidized at 1508C for 1, 2 and 5 h Organosolv lignin gave the highest yields of aromatic
the yields of methyl acetate were 0.98, 1.8, 2.2 wt%, compounds with a total of 10.9 wt% at 2 h at 1808C
respectively. One cannot preclude the possibility that with 8% water, see example 1 of Table 4 and Table 5.
methyl acetate is also forming from the lignin by oxi- This yield could be closely duplicated by going to
dation of its aliphatic methyl groups.
lower temperatures (1408C) and longer reaction time
The mechanism of air oxidation of hydrocarbons (8 h), see example 2. The reproducibility in examples
using homogeneous metal/bromide catalysts in acetic 7 and 8 and 9 and 10 was good but is quite poor in ex-
Adv. Synth. Catal. 2009, 351, 456 – 466
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
463

Walt Partenheimer
FULL PAPERS
Figure 3. Methyl acetate formation as a function of temperature using different lignins. All data using 8.0 wt% lignin in 8%
water/acetic acid for 8 h.
amples 3–6 where the total yield to aromatic benzal- lyst gave an inferior yield of 4.7% yield, see example
dehydes and acids varied from 4.4 to 10.2 wt% with 17 in Table 4.
an average value of 6.9%. Increasing the water con-
centration at 1408C for 2 h from 0 to 8% increased
the yield from 7.4% (examples 7 and 8) to 10.2% (ex- Results from Oxidation of Hydrolytic Lignin
ample 3) but slightly decreased the yield to 8.8 at
20% water (examples 9 and 10). Slower reaction rates This and the other lignins discussed below were not
are normally observed with increasing water concen- as extensively investigated as organosolv lignin. Ex-
[
16]
tration in metal/bromide oxidations hence the de- ample 18 of Table 4 gives a total aromatic product
crease in yield in example 10 may be because the re- yield of 7.6% when oxidized at 1508C for 2 h in anhy-
action was incomplete. Unusually low acid/aldehyde drous acetic acid. The comparable benzaldehyde
ratios probably means that the reactions were not yields in examples 19 indicate an equally high yield at
driven to completion since benzaldehydes are very re- 1408C for 2 h. The measured methoxy content of hy-
[
20a]
active intermediates in autoxidation reactions.
unusually low acid/aldehyde ratio in examples 9 and solv lignin (10% vs. 15%, respectively) hence a lower
0 could therefore give significantly higher yields if weight percent yield for hydrolytic vs. organosolv
The drolytic lignin is considerable less than that of organo-
1
the reaction time was increased. Doubling the pres- lignin is expected.
sure did not improve the yields (examples 3 and 11)
nor did increasing the catalyst concentration (com-
pare examples 7 and 8 to 12). Increasing temperatures Results from Oxidation of Organosolv Acetate
at 2 h to 140, 180 and 2208C gave yields of 8.9, 10.9
and 0.84% (examples 11, 1, 13) respectively. The low The total benzaldehyde+aromatic acid yield of orga-
yield at 2208C is not surprising considering the high nosolv acetate at 1408C for 2 h at 0% water gives a
reactivity of these compounds which will result in ex- total yield of only 2.2% (example 21). Since the phe-
tensive decarboxylation and subsequent degredation nols were acetylated in this sample and the phenolic
of the products. Increasing lignin concentrations of 4, acetates of the benzoic acids were not determined,
8
, 16, 24 wt% gave yields of 7.9, 10.2, 6.8 and 4.3 wt% the yields could be considerably higher. Examples 22–
when oxidized for 2 h at 1408C (compare examples 24 in Table 4 give benzaldehyde yields of 3.5–4.5%
4, 3, 15, 16). This suggests that an optimum lignin which suggests that substantially higher yields are
1
concentration may exist. This optimum may exist be- possible. This Aldrich sample was partially bleached
cause the concentration of phenols in solution in- with hydrogen peroxide. There is the possibility that
crease from 0.016 to 0.055M which will decrease the the bleaching may have significantly changed the
rate of reaction. The acid/aldehyde ratios are unusual- structure of the lignin and destroyed some of the aro-
ly low for examples 15 and 16 suggesting that longer matic groups via oxidation also accounting for lower
reaction times may give a higher yield. The HBr cata- yields.
464
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 456 – 466

The Aerobic Oxidative Cleavage of Lignin to Produce Hydroxyaromatic Benzaldehydes
FULL PAPERS
Results from the oxidation of hydrolytic hydroxy- Experimental Section
methyl lignin gave a total yield of only 0.9%, see ex-
ample 25 in Table 4. A number of other samples were General Remarks
evaluated as their benzaldehydes only, and a maxi-
Aldrich cobalt(II) and Fluka manganese(II) acetate tetrahy-
drates, Baker hydrobromic acid, Aldrich zirconium(IV) ace-
tate (16% solution in acetic acid/water solution) were used
as received. All lignin samples were obtained from Aldrich
mum yield of only 1.1% was obtained. Only one ex-
periment was performed using the lignosulfonic acid
derivative which gave a yield of only 0.2 wt%, see ex-
ample 26. The lignin sulfonic acid derivative has only and used as received. The catalyst used was a mixture of co-
6
.5 wt% methoxy groups hence a low yield is antici- balt(II) acetate tetrahydrate, manganese(II) acetate tetrahy-
pated.
drate, zirconium(IV) acetate and hydrogen bromide at a 1.0/
.0/0.1/2.0 (mol/mol) ratio, respectively. The catalyst was
1
prepared by dissolving the compounds in acetic acid. The
lignins were soluble at room temperature to at least 7.6
wt%. Reported yields are in weight percent based on the
amount of lignin used that is, weight productꢃ100/weight
lignin.
Absolute Yields of the Hydroxyaromatic
Benzaldehydes and Acids
As discussed above the GC/MS analysis of the orga-
nosolv lignin gives products characteristic of only ꢁun-
condensed phenolic ligninꢂ Since roughly 15% of lig-
nins are composed of uncondensed phenolic species
and the maximum observed yield for organosolv
lignin was 10.9 wt%, then one has an approximate
yield of about 70% of these groups that have been
successfully oxidized to the aromatic phenols. This as-
sumes that no hydrolysis of the PhÀOÀX ether bond
Apparatus
[27]
A glass cylindrical reactor, as previously described,
was
used for the reactions at ambient atmospheric pressure. Au-
toxidation at 70 bar in air were performed in 20-mL cylin-
drical glass reactors. The samples were analyzed after re-
moval from the reactor.
Caution: The use of high pressures and the use of dioxy-
gen/nitrogen mixtures is potentially explosive and dangerous.
has occurred. The organosolv lignin is reported to They should be performed only with adequate barriers for
have 18.9 wt% of methoxy groups. Since we know the protection. The source of oxygen was air which contains
moles of methoxy groups in the oxidation products, 20.9% O
2
one finds that 15.9% of the methoxy groups in entry 1
of Table 5 can be accounted for in the isolated prod- Analysis of Products using GC
ucts. This also suggests that most of the uncondensed,
phenolic part of the lignin is being oxidized.
After and sometimes before autoxidation, lignin samples
were analyzed via GC by direct injection onto the column
using an HP6890 gas chromatograph equipped with a 25 m
HP-FFAP column and FID detector using terphenyl as the
internal standard. Response factors were obtained from au-
thentic compounds using terphenyl as the internal standard.
The following compounds were analyzed by GC: methyl for-
mate, methyl acetate, ethyl acetate, 3,3-dimethoxy-2-buta-
none, vanillin acetate, vanillin, syringaldehyde acetate, syrin-
galdehyde.
Conclusions
We have successfully isolated a number of expected
products from the autoxidation of various lignins
using the well-established industrial method with
metal/bromide catalysts in acetic acid. We suspect
that some of the ꢁcondensed, phenolic ligninꢂ has
been oxidized to isopenic and phthalic acids but their
identification requires further analytical development.
The acids were analyzed using silation gas chromatogra-
phy (SGC). After evaporation of the acetic acid/water at
5
08C, the samples were treated with BSTFA and analyzed
on an HP5890 with an MDN-5S fused silica capilary
Further development of this approach to optimize column. Compounds analyzed were 4-hydroxybenzaldehyde,
yields would involve a study of the initial acetylation 4-hydroxybenzoic acid, vanillin, vanillin acetate, vanillic
of the phenols followed by oxidation. Also the rate of acid, syringaldehyde, and syringic acid using terphenyl as an
internal standard.
hydrolysis of the uncondensed and condensed ether
linkages should be studied apart from oxidation.
When the hydrolysis reaction has been optimized,
then one should hydrolyze the lignins first, then acety- Acknowledgements
late them, and finally oxidize the mixture. The
chemistry of lignins is very complex and one must rec-
ognize that condensation reactions are also known to
occur and these must obviously be minimized.
Dr. Michael Mawn is acknowledged for the GC-MS results,
David E. Rothfuss and Susan Solek for their excellent experi-
mental assistance.
Adv. Synth. Catal. 2009, 351, 456 – 466
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
465

Walt Partenheimer
FULL PAPERS
References
[15] A. R. Goncalves, G. A. A. Labat, Appl. Biochem. Bio-
technol. 2007, 148, 151–161.
[
1] a) D. W. Goheen, C. H. Hoyt, ꢁLigninꢂ in: Kirk-
[16] a) W. Partenheimer, Adv. Synth. Catal. 2004, 346, 297–
306; b) W. Partenheimer, Adv. Synth. Catal. 2005, 347,
580–590.
rd
Othmer, Chemical Encyclopedia,
3
edn., Wiley,
London, 1997, Vol. 14, p 294; b) S. Y. Lin, S. E. Lebo,
th
Jr., in: Kirk-Othmer, Chemical Encyclopedia, 4 edn.,
[17] K. A. Fjare, (to Amoco), US Patent 4,873,361, 1989.
Wiley, London, 2001, Vol. 15, pp 268–289.
[
[
[
[
[
[
18] From the data sheets provided by Aldrich Chemical
[
[
2] E. Adler, Wood Sci. Technol. 1977, 11, 169–218.
3] A. J. Ragauskas, M. Nagy, D. H. Kim, C .A. Eckert,
J. P. Hallett, C. L. Liotta, Industrial Biotechnology,
company.
19] D. V. Evtuguin, C. Pascoal Neto, J. Rocha, J. D. Pedro-
sa de Jesus, Applied Catal. A 1998, 167, 123–139.
20] a) W. Partenheimer, J. Mol. Catal. 2003, 206, 105–119;
b) W. Partenheimer, J. Mol. Catal. 2003, 206, 131–144.
21] K. Lundquist, R. Lundgren, Acta Chem. Scand. 1972,
2
006, 2, 55.
[
[
4] W. Partenheimer, Catalysis Today 1995, 23, 69–158.
5] Disinfectants and antiseptics, in: Kirk-Othmer, Chemi-
cal Encyclopedia, Wiley, London, 1993, Vol. 8, p 252.
2
6, 2005–2023.
22] T. P. Schultz, T. H. Fisher, Holzforschung 2002, 56,
92–593.
23] a) F. F. Rust, W. E. Vaughan, Ind. Eng. Chem. 1949, 41,
595; b) E. R. Bell, F. J. H. Raley, F. F. Rust, W. E.
[
6] Salicylic acid and related compoundsꢂ, in: Kirk-
rd
Othmer, Chemical Encyclopedia,
London, 1997, Vol. 21, p 602.
3
edn., Wiley,
5
[
[
7] W. Partenheimer, Catalysis Today 2003, 81, 117–135.
8] C. P. Neto, D. Evtuguin, A. Robert, J. Wood Chem.
Tech. 1994, 14, 383–402, and references quoted therein.
9] W. Partenheimer, Adv. Synth. Catal. 2004, 346, 1495–
2
Vaughan, Ind. Eng. Chem. 1949, 41, 2597; c) P. J. Na-
wrocki, J. H. Raley, F. F. Rust, W. E. Vaughan, Ind.
Eng. Chem. 1949, 41, 2598; d) E. R. Bell, G. E. Irish,
J. H. Raley, F. F. Rust, W. E. Vaughan, Ind. Eng. Chem.
[
1
500.
[
10] F. G. Sales, L. C. A. Maranhao, N. M. Lima Filho,
1
949, 41, 2609; e) B. Barnett, E. R. Bell, F. H. Dickey,
F. F. Rust, W. E. Vaughan, Ind. Engin. Chem. 1949, 41,
609.
C. A. M. Abreu, Ind. Eng. Chem. Res. 2006, 45, 6627–
6
631.
11] M. Hepditch, R. Thring, Can. J. Chem. Eng. 2007, 75,
108–1111.
2
[
[
[
[
[
24] a) P. Roffia, P. Callini, L. Motta, Ind. Eng. Chem. Prod.
Res. Dev. 1984, 23, 629; b) R. Roffia, P. Calini, S. Tonti,
Ind. Eng. Chem. Res. 1988, 27, 765.
25] R. A. Sheldon, J. K. Kochi, Metal Catalyzed Oxidations
of Organic Compounds, Academic Press, New York,
N. Y. , 1981.
1
12] I. P. Deineko, E. V. Evtyugin, G. M. Dmitriev, Mezh-
vuz. Sb. Nauch. Tr., Khim. Pererab. Drev. Otkhodov
1
987, 28–32 (abstract only).
[
[
13] H.-M. Chang, J. F. Kadla, Y. S. Kim, J. Wood Chem.
Technol. 2007, 27, 225–241.
26] S. W. Benson, P. S. Nangia, Acc. Chem. Res. 1979, 12,
14] a) A. R. Goncalves, U. Schuchardt, Proc. Braz. Symp.
Chem. Lignins Other Wood Compounds 1993, 253;
b) U. Schuchardt, A. R. Goncalves, Appl. Biochem.
Biotechnol. 1999, 77–79.
2
24.
[27] W. Partenheimer, V. V. Grushin, Adv. Synth. Catal.
2001, 343, 102–111.
466
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 456 – 466