chemical engineering research and design 8 7 ( 2 0 0 9 ) 1276–1292

Contents lists available at ScienceDirect

Chemical Engineering Research and Design

journal homepage: www.elsevier.com/locate/cherd

An integrated process to produce vanillin and lignin-based

polyurethanes from Kraft lignin

E.A. Borges da Silva a , M. Zabkova a , J.D. Araújo a , C.A. Cateto b,c , M.F. Barreiro b ,
M.N. Belgacem c , A.E. Rodrigues a,∗
a Laboratory of Separation and Reaction Engineering – LSRE, Associate Laboratory LSRE/LCM, Department of Chemical Engineering,
Faculty of Engineering, University of Porto, Rua Dr. Roberto Frias s/n, 4200-465 Porto, Portugal
b Laboratory of Separation and Reaction Engineering – LSRE, Associate Laboratory LSRE/LCM, Bragança Polytechnic Institute,

Campus de Santa Apolónia, Apartado 1134, 5301-857 Bragança, Portugal

c Matériaux Polymères, École Française de Papeterie et des Industries Graphiques, BP 65, 38402 St. Martin d’Hères, France

a b s t r a c t

The aim of this manuscript is to present an integrated process that includes reaction and separation steps for
producing vanillin and lignin-based polyurethanes from Kraft lignin. It provides details about lignin oxidation and
subsequent vanillin recovery, as well as, the synthesis of lignin-based polyurethanes. The oxidation of Kraft lignin in
alkaline medium has been carried out in a batch reactor and the optimum operational conditions for vanillin produc-
tion obtained. The feasibility of a continuous process for vanillin production has been analyzed using a structured
bubble column reactor. The generated reaction stream (degraded lignin and sodium vanillate) was further subjected
to an ultrafiltration process to recover the vanillate. An ion-exchange process allows recovering the vanillin by pass-
ing the vanillate solution through a column packed with an ion-exchange resin in H+ form. The remaining lignin can
act as a raw material to produce polyurethanes and/or biofuels. In this work the first approach was explored.
© 2009 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

Keywords: Kraft lignin; Vanillin; Polyurethanes; Oxidation; Ultrafiltration; Integrated process

1. Introduction Annevelink, 2007). Hence, there is no doubt about the impact

One of the major priorities of 21st century is the replace-
ment of fossil carbon source by a renewable raw material
(biomass). Nowadays, the international community is strongly
committed with sustainable development and terms such
as bioproducts, bioenergy, bioeconomy are being used
frequently by politicians’ and investors’, R&D projects, com-
panies’ reports, thus corroborating the growing concern, on
sustainable growth, zero-waste industries and environmen-
tally friendly processes (Biomass R&D Technical Advisory
Committee, 2006; European Parliament and Council, 2002;
Ree and Annevelink, 2007). The concept of biorefinery comes
from this state of mind. Biorefineries are structures that inte-
grate processes and technologies for an efficient biomass
conversion into products such as base chemicals, platform
chemicals, fuels and energy (Kamm et al., 2006; Ree and
that biorefineries can bring into our lives in a near future.
Nevertheless the potential associated to industrial biore-
fineries, there are still uncertainties and technical difficulties
(e.g., technology costs, know-how on available processes or
food versus fuel dilemma) that need to be overcome in order
that these novel systems become more competitive against
the current industrial alternatives based on petrochemical
resources. The so-called lignocellulose feedstock (LCF) biore-
finery is particularly favoured by the sustainability and cost of
biomass supply. Moreover, the generated conversion products
are well placed comparatively in the traditional petrochemical
and the future biobased product markets (Kamm and Kamm,
2004).
One of the most abundant aromatic substances in the bio-
sphere is lignin. The main source of lignin is the pulp and
paper industry, where nowadays the Kraft process prevails

∗
Corresponding author. Tel.: +351 22 5081671; fax: +351 22 5081674.
E-mail addresses: eduardo@fe.up.pt (E.A.B.d. Silva), miriamz@fe.up.pt (M. Zabkova), jaraujo@fe.up.pt (J.D. Araújo), cc@ipb.pt
(C.A. Cateto), barreiro@ipb.pt (M.F. Barreiro), belgacem@efpg.inpg.fr (M.N. Belgacem), arodrig@fe.up.pt (A.E. Rodrigues).
Received 2 October 2008; Received in revised form 29 April 2009; Accepted 18 May 2009
0263-8762/$ – see front matter © 2009 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.cherd.2009.05.008
 chemical engineering research and design 8 7 ( 2 0 0 9 ) 1276–1292
with a quote of approximately 80% of the world chemical pulp
production (Saake and Lehnen, 2004). The black liquor, a by-
product stream of this process, contains, typically, 30–34% of
lignin. In a conventional process, this stream is burned to pro-
vide energy for mill operations, and to recover the cooking
chemicals (Wallberg et al., 2005). Nowadays, this destructive
way to treat the large amount of generated black liquor is
not considered anymore a satisfactory solution. Moreover, it is
worth mentioning that due to the complex energetic integra-
tion of the Kraft process, an expansion in the pulp production
implies a revamp in the burners and also a capacity increase of
the recovery boiler. This introduces a major bottleneck in the
process leading to the appearing of some approaches to debot-
tleneck the recovery boiler. Solutions such as the extraction of
lignin from the black liquor proportionally to the production
increase (sold as a biofuel) or upgrading of the recovery boiler
and the turbine system have been purposed (Axelsson et al.,
2006; Mathias, 1993).
To overcome the disadvantages associated with the con-
ventional black liquor process treatment, alternative methods
are required. One solution is the utilization of black liquor
to generate high-added value products. When black liquor
is burned, many organic compounds are vanished, especially
lignin, from which many valuable substances or materials can
be produced, such as vanillin, vanillic acid, dispersing agents,
synthetic tannins, polymer filter sand binding agents, acti-
vated carbon, ion-exchanger substracts, etc. In this work, the
focus is directed to the production of synthetic vanillin from
lignin obtained from black liquor. The main feedstock is Kraft
lignin and the end products vanillin and polyurethane mate-
rials (plus biofuel).
A flow sheet of the proposed process is shown in Fig. 1. A
portion of the by-product stream, black liquor, is processed
to extract the lignin. This extraction can be done by the
traditional acidification/precipitation method followed by sep-
aration, or by using alternative methods such as the one
developed by the Swedish group and known as LignoBoost
(Öhman et al., 2006). After obtaining the purified lignin, the
Fig. 1 – Valorisation of Kraft lignin by a biorefinery concept: the proposed integrated process for producing vanillin and
biopolymers.
1277
subsequent process is based on three main steps. The first
step is the alkaline lignin oxidation performed in a bubble
column reactor (Araujo, 2008). Thereafter, the obtained mix-
ture passes through a membrane ultrafiltration system where
the lignin higher molecular residues are retained. Sodium
vanillate and other low molecular weight species go to the
permeate stream (Zabkova et al., 2007a). Finally, the perme-
ate containing smaller molecules and excess NaOH flows
through a packed bed with an acid resin, in order to con-
vert the sodium vanillate into vanillin (Zabkova et al., 2007b).
This ion-exchange step is accompanied by a neutralization
reaction thus originating an exit stream with lower pH. The
degraded lignin resulting from the membrane process appears
as a by-product in the flow sheet. Our purpose is to valorise this
material by obtaining lignin-based polyurethane products,
e.g., rigid foams, elastomers, sealants (Cateto et al., 2008a),
and additionally biofuels for the related unit operations.
The aim of this manuscript is to describe an integrated pro-
cess to produce vanillin from Kraft lignin oxidation. Details
about each required unit operation are presented. This work
is supported by a research program dedicated to vanillin pro-
duction from Kraft lignin that has been conducted at the
Laboratory of Separation and Reaction Engineering (LSRE) for
many years. Considering the industrial importance of vanillin,
a brief discussion focusing the market and general aspects of
this compound are presented. In the final section, the utiliza-
tion of lignin as a macromonomer in polyurethane synthesis
is explored.
2. Vanillin: application and world market
Vanillin (C8 H8 O3 ) is the major flavour constituent of vanilla. It
has a wide range of applications in food industry as a flavour
agent and in perfumery as an additive. Other applications
include antioxidant additive, antifoaming agent, vulcaniza-
tion inhibitor and chemical precursor for pharmaceutical and
agrochemical industries (Cerrutti et al., 1997; Fitzgerald et al.,
1278 chemical engineering research and design 8 7 ( 2 0 0 9 ) 1276–1292
2003; Lopez-Malo et al., 1998; Villar et al., 1997). Vanillin market
is mainly constituted by large multinational holders: (i) flavour
and fragrance companies (IFF, Givaudan, Quest, Danisco); (ii)
chocolate and ice cream producers (Unilever, Nestle, Cadbury,
Suchard); and (iii) manufacturer of pharmaceuticals (Merck).
There are two commercial types of vanillin: pure vanillin,
which is obtained by chemical synthesis, i.e., derived from
guaiacol or lignosulphonates of the paper and pulp industry;
and vanilla extract obtained from the pod of tropical Vanilla
orchid (Daugsch and Pastore, 2005; Walton et al., 2003). The
natural vanillin is very expensive comparatively to the syn-
thetic counterpart. The natural vanilla market is characterized
by very volatile prices (Jaeger, 2005), mainly sensitive to sev-
eral events (natural, economic or political) affecting its larger
producer Madagascar (FAOSTAT, 2005). The unstable market of
this natural product has encouraged the development of new
chemicals routes. Hence, synthetic vanillin became widely
used and competition of markets is longstanding and turns
more fierce when prices of natural vanilla rockets.
Vanillin was first produced by Haarmann and Reimer in
the late 1800s, using guaiacol obtained from phenol. Between
1930s and 1980s, the major synthetic vanillin supply market
share was generated by lignin-containing waste produced by
the sulphite pulping process (Triumph Venture Capital Ltd.,
2004a). However, from 1980s, changes introduced in the pulp
and paper industry processes led to a decrease of this raw
material required by vanillin plants. The competing chemical
pulping process, the sulphate or Kraft process, was preferred
since it allowed the recycling of the waste liquors for chemi-
cal recovery. As a result, these by-product streams were not
available for vanillin production and petrochemical guaia-
col has gained relevance. Nowadays the synthesis of vanillin
from petrochemical guaiacol accounts for 85% of the world
supply, with the remaining 15% being produced from lignin
(Triumph Venture Capital Ltd., 2004a). Despite the guaiacol
route produces vanillin with nearly absence of by-products;
this process is dependent on petroleum-derived com-
pounds, in opposition with lignin-based biomass oxidation
process.
Natural vanilla flavouring constitutes less than 5% of the
world market (Triumph Venture Capital Ltd., 2004a). Consid-
ering its application as a flavouring and fragrance ingredient,
the global demand for synthetic vanillin is currently around
16,000 tonnes/year (Triumph Venture Capital Ltd., 2004b).
However, vanillin may be used for the synthesis of several
second-generation fine chemicals and several pharmaceuti-
cal chemicals (Bjorsvik and Liguori, 2002). There are only few
significant manufacturers of this product in the world. Rhodia
SA dominates the vanillin market using the catechol-guaiacol
process. Borregaard (Norway), the second largest vanillin pro-
ducer, is one of the remaining producers of lignin-based
vanillin. Essentially, Borregaard supplies the European market
and its vanillin production is almost exclusively for large-scale
costumers under long-term contracts.
In the middle of 2005, vanillin market prices have reached
$15 kg−1 (Nair, 2005). Since it is mostly produced from guaiacol,
vanillin prices are sensitive to the world oil market (Halliday,
2008). Besides, lignin-based vanillin is in high demand for
certain market sectors, particularly for the perfume indus-
try, European chocolate manufacturers, and Japanese market,
and as such tends to command a price premium. The price
of lignin-based vanillin has consistently maintained at about
$100–200 kg−1 above that of guaiacol-based vanillin (Triumph
Venture Capital Ltd., 2004b).
3. Feedstock: lignin. General aspects
Lignin is a three-dimensional amorphous macromolecule
made of phenylpropane units that arise from the copoly-
merization of three primary precursors: coniferyl alcohol,
sinapyl alcohol and p-coumaryl alcohol. The chemical com-
position of lignin and its content in wood vary greatly with
and within a plant, depending on its species, age, morpholog-
ical location and growth environment. Lignin from softwoods
(gymnosperms) is predominantly based on structural units
derived from the coniferyl alcohol (guaiacyl units), which are
the precursors for vanillin yield (Adachi et al., 1992; Higuchi,
1980). On the other hand, hardwood (angiosperms) lignins
present structures with much broader chemical compositions,
i.e., with various guaiacyl:syringil ratios. Lignin from differ-
ent sources can be oxidized to obtain phenolic compounds:
spent pulp liquor from paper industry, native lignin, sugar-
cane bagasse, Klason lignin, among others (Creighton et al.,
1941; Sales et al., 2002; Villar et al., 1997).
The commercially available lignins comprise two cate-
gories: (1) the sulphur-free lignins, mainly obtained from
biomass conversion technologies focused on biofuel produc-
tion, organosolv pulping processes (Lebo et al., 2001; Sixta,
2006), and soda pulping based on alternative resources like
agricultural residues and non-wood fibres (Lora and Glasser,
2002); and (2) the sulphur containing lignins, which result
essentially from Kraft and sulphite pulping processes. This last
category comprises almost the whole market of commercially
available lignins. The total annual capacity of technical lignins
production is around 785,000 tonnes/year of which approxi-
mately 60% corresponds to the production from Borregaard
LignoTech and the remaining from other companies (as exam-
ple, Tembfibre, Fraser Paper, Tolmozzo, Westvaco). The annual
capacity of the major producers of technical lignins in Europe
and North America can be found elsewhere (Lebo et al., 2001).
It is estimated that 98–99% of the lignin separated from
wood in the sulphite pulping, and especially in the Kraft pulp-
ing, is burned for energy and chemicals recovery or disposed in
waste streams (Lora and Glasser, 2002). The remaining amount
is isolated from the spent pulping liquors and sold for spe-
cialty applications, which is normally around 1 million tonnes
per year worldwide (Lora and Glasser, 2002; Thielemans et
al., 2002). Vanillin and dimethyl sulphoxide are the only
two low molecular chemicals produced in large quantities
from technical lignins, particularly from lignosulphonates.
Industrially, softwood spent sulphite liquors or isolated ligno-
sulphonates are oxidized, in alkaline media, by oxygen or air
to produce vanillin (Bjorsvik, 1999). Our aim fits in this type
of lignin application, but instead of using lignosulphonates,
the vanillin production is based on isolated lignin from the
largely available Kraft pulping waste liquor. Data from the
Food and Agricultural Organization of the United Nations
(FAO)—revealed that, in 2006, 9.8 × 107 tonnes of Kraft pulp
were produced in worldwide developed countries from a total
chemical pulp production of 1.03 × 108 tonnes (FAOUN, 2008).
Kraft lignins can be precipitated and extracted from the
black liquor in a two-step acidification process. First, carbon
dioxide is used to reduce the pH of the liquor till 9–10, and
about 75% of the lignin is precipitated as a sodium salt (Lebo
et al., 2001). For further purification, this lignin is suspended
in water and acidified with H2 SO4 to a pH lower than 3. The
commercial Kraft lignins are generally sold in the sulphonated
form or as lignin amines. The physical properties for softwood
Kraft lignins and softwood lignosulphonates can be found
 chemical engineering research and design 8 7 ( 2 0 0 9 ) 1276–1292
elsewhere (Lebo et al., 2001; Saake and Lehnen, 2004). Lig-
nosulphonates (M.W. 20,000–50,000) have higher molecular
weight and lower content in free phenolic hydroxyl groups
than Kraft lignins (M.W. 2000–3000). The content in free phe-
nolic hydroxyl groups of Kraft lignins is especially rich with
values around 3.1 meq g−1 (Saake and Lehnen, 2004). This
functional group plays an important role in the dissolution
of macromolecules.
The lignin extraction from black liquor represents a very
attractive option, either for energy production or combustion
elsewhere, or to serve as feedstock for chemicals production.
A considerable research effort has taken place in the following
fields (1) improvement of lignin isolation processes from black
liquor, specially using precipitation (Norgren, 2001; Norgren
and Edlund, 2003; Öhman et al., 2007a,b; Sun et al., 1999) and
ultrafiltration methods (Bhattacharjee et al., 2006; Holmqvist
et al., 2005; Liu et al., 2004a,b; Wallberg et al., 2006); and, (2)
as focused earlier, development of new lignin applications. In
Fig. 1, both the squares “Processing” and “Lignin Plant” rep-
resent steps where lignin can be extracted from black liquor
using any suitable technology.
4. The integrated process
The general concept of the integrated process for vanillin
production includes the Kraft lignin oxidation and separa-
tion steps in order to allow the isolation of vanillin from the
oxidized solution. The reaction and purification procedures
related to this integration concept are presented in the fol-
lowing sections. The paper is written within the framework
of the work developed by the Laboratory of Separation and
Reaction Engineering (LSRE) on the area of vanillin chemical
production from Kraft lignin oxidation (Araujo, 2008; Fargues
et al., 1996a,b; Mathias, 1993; Mathias et al., 1995; Mathias
and Rodrigues, 1995; Zabkova et al., 2007a,b, 2006), as well
as, polyurethane synthesis from lignin-based raw materials
(Cateto et al., 2008a,b,c, 2007).
4.1. Lignin oxidation and vanillin production
The only lignin-based process industrially implemented is
based on the alkaline oxidation of lignosulphonates in the
presence of air (Triumph Venture Capital Ltd., 2004a). This
lignin oxidation route has been extensively studied since
the sixties (Bjorsvik, 1999; Diddams and Krum, 1970; Ness,
1983). The use of Kraft black liquor as raw material, instead of
the conventional application of lignosulphonates, is lacking
presently. Since, currently, Kraft process prevails in the pulp
and paper industry, we focus on the Kraft lignin oxidation for
vanillin production in this integrated process.
The chemical oxidation of lignin with oxygen is a reaction
in which typical operating conditions involve high pH values
(close to 14), high temperatures (up to 150 ◦ C) and high total
pressure (e.g., 10 bar (Mathias, 1993)). Under these conditions
it is possible to cleave this large complex molecule thus creat-
ing a favourable medium for the oxidation of their fragments
towards vanillin. In particular, ether bonds in the alfa- and
beta- carbon positions of the phenyl propane units are cleaved,
and there is also the rupture of carbon–carbon bonds in the
propanoid side chains of the phenyl propane units (Lebo et al.,
2001). In general, subsequent to the alkaline oxidative cleav-
age of lignin that provides fragments which can lead to the
production of vanillin, well known stages in the field of reac-
1279
tivity of lignin may be assumed (Tarabanko and Petrukhov,
2003): (1) the formation of phenoxyl radicals (by the detach-
ment of one electron from phenoxyl anion), (2) the formation
of quinone methyde (by desproportionation of phenoxyl rad-
ical), (3) the formation of coniferyl alcohol (by nucleophilic
addition of hydroxide ion), (4) probable formation of gama-
carbonyl (by oxidation of coniferyl alcohol); and, finally, (5) the
formation of vanillin (by the retro-aldol cleavage of the alfa-
and beta-unsaturated aldehydes). There are a few published
works dealing with the oxidation mechanisms of lignin and, in
most cases, the approaches involve lignosulphonates and are
based on different assumptions (Bjorsvik, 1999; Dardelet et al.,
1985; Schultz and Templeton, 1986; Tarabanko et al., 1995a,b,
2004; Tarabanko and Petrukhov, 2003). It is worth mentioning
that due to the molecule complex structure, developing a reac-
tion mechanism is not a straightforward task. Moreover, two
additional reasons make this task harder: the lignin oxida-
tion is associated to complex alkaline hydrolysis steps at high
temperatures; and, under drastic oxidative conditions, peroxyl
radicals (from aromatic aldehydes), which often are inter-
mediate in the oxidation by oxygen, react fast with vanillin
leading to oxidative degradation (Bjorsvik, 1999). A reaction
scheme for vanillin production from lignin and vanillin oxi-
dation in alkaline medium has been discussed previously in
the work of Mathias (1993) and Fargues et al. (1996a). We have
been investigating the vanillin production from chemical oxi-
dation of Kraft lignin in order to develop kinetic models for
this process and, mainly, to analyze the effect of operating
process parameters on the yield of the process. It was veri-
fied that operational parameters such as temperature, lignin
concentration, oxygen partial pressure and sodium hydroxide
concentration have considerable influence on the produced
vanillin amount (Mathias and Rodrigues, 1995). Some results
of lignin oxidation experiments performed at LSRE are sum-
marized in Table 1.
The batch experiments of lignin oxidation have been per-
formed in a jacketed reactor with controlled temperature and
pressure. The reaction mixture (solution of lignin in sodium
hydroxide) was maintained under stirring and oxygen was fed
to the reactor. The total pressure of the system was kept con-
stant by continuous addition of oxygen (Fargues et al., 1996a;
Mathias et al., 1995; Mathias and Rodrigues, 1995).
There is a maximum obtainable vanillin yield from a spe-
cific type of lignin. The formation of phenolic compounds
like vanillin or syringaldehyde is strictly related to the avail-
able percentage of its precursor in the lignin structure. The
less transformations or chemical treatments lignin suffers,
the more useful it becomes for producing phenolic aldehydes
(Tarabanko et al., 1995a,b; Villar et al., 1997).
Mathias and Rodrigues (1995) have reported that a 60 g/L
solution of Kraft lignin from Pinus spp. can provide a maxi-
mum vanillin yield of 13% (w/w); value based on nitrobenzene
oxidation (7 h of reaction, 147 ◦ C, 2N NaOH and 0.84 mL of
nitrobenzene per gram of initial lignin). Villar et al. (1997),
using a lignin precipitated from Kraft black liquor with the
addition of a calcium salt dissolved in a water soluble alco-
hol, have published a maximum yield for the total phenolic
aldehydes (syringaldehyde + vanillin) of 14%, also based on
nitrobenzene oxidation (40 min of reaction, 190 ◦ C, 2N NaOH
and 6 mL of C6 H5 NO2 per gram of initial lignin).
From Table 1, one can observe yields in the range of 3–10%
for the Kraft lignin oxidation with O2 using similar operat-
ing conditions. Under these conditions, Mathias (1993) and
Fargues et al. (1996a) have found a vanillin yield around 10%
1280 chemical engineering research and design 8 7 ( 2 0 0 9 ) 1276–1292

Table 1 – Set of experimental conditions used in batch experiments of Kraft lignin oxidation and obtained vanillin yields.
Lignin type Ta (◦ C) PO2 a (bar) Pa (bar) CLignin (g/L) tmax (min) Vanillin Ref.
yield (%)

121 4.1 9.3 60 43.4 7.3 Mathias (1993)

119 6.5 9.5 60 27.1 7.2
120 3.7 9.7 60 76.5 8.2
141 3.8 10.0 60 64.4 10.5
122 3.8 9.6 30 61.4 10.3
Pinus spp. Westvaco 125 3.0 9.7 60 60.0 8.3 Fargues et al. (1996a)
133 2.8 9.7 60 35.0 10.8
130 1.2 9.7 60 70.0 8.3
135 5.0 9.7 60 20.0 8.3
128 3.0 9.7 30 40.0 6.7
125 4.0 9.7 30 45.0 9.0

123 4.0 9.0 60 75.0 3.7 Araujo (2008)

Indulin AT Westvaco
123 6.5 9.5 60 35.0 3.4

127 3.5 9.1 82.5 92.9 4.1 Mathias (1993)

Pinus pinaster Portucel
140 3.8 9.3 82.5 46.4 2.8

Black liquor Portucel 119 3.5 9.1 50 77.5 8.0

In all experimental runs: CNaOH = 80 g/L (pHinitial ≈ 14).

a
Initial values except for Fargues et al. (1996a) where the values presented result from average conditions.

(defined as the ratio: mass of produced vanillin/mass of ini- With regard to the influence of the operating conditions
tial lignin) for batch oxidations. This value corresponds to a on the Kraft lignin oxidation process, these works reported in
vanillin recovery of almost 80% for Kraft lignin (Pinus spp.). Table 1 contain information to decide the best conditions for
Vanillin recovery is defined as the ratio between the obtained the production of vanillin. In the case of Pinus spp. lignin, an
vanillin yield and the maximum vanillin yield based on optimal production was found for an oxygen partial pressure
nitrobenzene oxidation. However, results from Araujo (2008) of 2.8 bar, under a temperature of 133 ◦ C for a lignin concen-
have shown a lower vanillin yield under equal operating tration of 60 g/L, dissolved in 2N NaOH.
conditions. Here, it is important to notice that two differ-
ent Kraft lignins were used as raw material. Both the Kraft 4.2. The continuous process of lignin oxidation
lignin Indulin AT and the Kraft lignin from the Pinus spp. have
been supplied by Westvaco Co., but the first seems to have From an industrial point of view, the continuous process of
a reduced capacity for vanillin production. Tarabanko et al. lignin oxidation will be a priori more attractive than the batch
(1995a,b) have presented a study where the influence of lignin
origin, conditions of production and type of pre-treatment
on the obtained yields of vanillin and syringaldehyde was
inspected by lignin oxidation. The results have indicated a
competition between lignin fragments condensation (syringil
fragments and guaiacyl fragments) and lignin oxidation into
aldehydes.
The degradation of the produced vanillin is an extremely
important point from a techno-economic approach. To avoid
losses of vanillin by oxidation, two operational parameters
should be taken into account: oxygen partial pressure (pO2 )
and reaction time. There is a conflicting effect on these two
parameters, i.e., a high value of pO2 can reduce reaction
time but leads to vanillin degradation; on contrary, a low
pO2 increases reaction time but it hinders vanillin oxidation.
Moreover, during the Kraft lignin oxidation, formation of acid
compounds can occur, lowering the pH of the medium. At
lower pH values (<11–12) vanillin losses by oxidation become
more significant. Fig. 2 shows the vanillin concentration and
pH as a function of reaction time for two batch experiments
described in Table 1. One can verify that during the lignin
oxidation a maximum amount of vanillin is reached from
where pH value begins to decrease and, as consequence, the
vanillin yield decreases considerably. It is known that at higher
alkali concentration (pH > 12) the rate of vanillin degradation Fig. 2 – Vanillin concentration and pH as a function of
is lower. At lower pH (pH < 12) the vanillin oxidation rate is sec- reaction time for batch lignin oxidation: (a) Pinus spp.
ond order with respect to the vanillin concentration (Fargues lignin; (b) Indulin AT Lignin. Experimental reaction
et al., 1996a,b). conditions are in Table 1.
 chemical engineering research and design 8 7 ( 2 0 0 9 ) 1276–1292 1281

Fig. 3 – (a) Schematic diagram of the pilot set-up for the continuous production of vanillin from lignin (1, safety valve; 2,
on–off valve; 3, electromagnetic valve; 4, needle valve; 5, safety valve; 6, three-way valve; 7, mass flow controller; PT,
pressure transducer; TT, thermocouple. (b) Structured bubble column reactor (both the cylindrical main body and the liquid
stabilization chamber are jacketed). (c) Mellapak 750.Y packing module (made in 316L stainless steel).

alternative, mainly due to the huge Kraft liquor volumes to be
treated in a pulp industry. Also, continuous processes are eas-
ier to control and to attain constant product characteristics as
well as their overall investments and operating costs are usu-
ally lower (Lebo et al., 2001; Rase, 1977). An experimental pilot
set-up was built to promote the concerned gas–liquid reaction
in a continuous operating mode. A schematic diagram of the
operational pilot installation at LSRE is shown in Fig. 3. The
bubble column reactor is made in 316L stainless steel with 8 L
capacity. The main cylindrical body of the reactor is the section
where the gas–liquid reactions take place. It has an internal
diameter of 10 and 70 cm height, and is filled with three mod-
ules of Mellapak 750.Y structured packing (Sulzer Chemtech,
Switzerland). The goal was to enhance the overall mass trans-
fer performance of the system. Each of the modules is rotated
90◦ with respect to the adjacent one to promote radial mixing
of the fluids in the transition between packing elements. In
Fig. 3c a picture of a Mellapak 750.Y packing element is shown.
This type of packing is characterized by a geometric specific
area of 750 m2 /m3 and a void fraction of 0.95 (Siminiceanu et
al., 2007). The corrugated metal sheets that make up these
packings are inclined by an angle of 45◦ with respect to the ver-
tical axis. The separation head at the top of the reactor works
to separate the gas and liquid streams, obtaining a final liquid
stream with no dispersed gas, taking advantage of the impul-
sion forces on the gas bubbles. More details on the reactor
set-up can be found elsewhere (Sridhar et al., 2005).
In a typical run, the liquid mixture (60 g/L lignin + 80 g/L
sodium hydroxide) was pumped to the reactor by a pis-
ton pump (work range 1.3–13 L/h). The temperature inside
the reactor was regulated to 403 K. At the beginning of the
reaction, nitrogen was admitted to the reactor to pressurize
the system. Once the operating temperature, pressure and
flow rates stabilized, lignin oxidation could be initiated. The
desired N2 and O2 flow rates were controlled with two mass
flow controllers.
With the purpose of determining the reactor performance
for lignin oxidation in terms of vanillin formation, the experi-
ments of Kraft lignin oxidation were performed in two different
reactor configurations: structured packed bubble column reac-
tor (SPBCR) and bubble column reactor (i.e., with no internals;
BCR). The operating conditions used in these experiments are
summarized in Table 2.
These continuous experiments were performed with Kraft
lignin Indulin AT supplied by Westvaco Co. The steady state
was reached at approximately 6 h of operation for both types
of reactor configuration. Vanillin concentration in the BCR exit
stream reached a steady state value of around 0.56 and 0.67 g/L
for the run BCR1 and BCR2, respectively. The vanillin yield had
a slight improvement in experiment BCR2 due to an increase
in gas flow rate (higher mass transfer coefficients) and also a
higher pO2 for the oxygen solubility in the liquid phase. Con-
sidering the batch experiment, these results are about 25–30%
of the maximum level of vanillin concentration obtained in
the batch reactor.
In the second configuration, the hydrodynamics environ-
ment (i.e., dispersion coefficient, phase hold-up or even heat
transfer coefficients) for the structured packed bubble col-
1282 chemical engineering research and design 8 7 ( 2 0 0 9 ) 1276–1292
Table 2 – Operating conditions of the lignin oxidation experiments performed in the structured packed bubble column
reactor (SPBCR) and bubble column (BCR).
Run Cin
L (g/L) pHina QL (L/h)
1 60 14 2.12
BCR
2 60 14 2.50
3 60 14 2.12
SPBCR
4 60 14 0.99
NTP: Normal temperature and pressure: 273 K and 1 atm.
a
Liquid feed containing 2 mol/L of sodium hydroxide (pH of the reaction media controlled by a carbonate buffer system in the liquid feed).
b
The set point of the thermostatic bath (TFset ) was changed at a certain time of reaction (as given in table) since there was an increase in the
reactor temperature (about 7–8 K) due to the heat liberated by the reaction.
umn reactor is quite different from the bubble column reactor,
because of the structured packing (Mellapak 750.Y) above the
gas distributor. At steady state, the concentration of vanillin
obtained from the SPBCR configuration was 0.73 and 0.89 g/L
for runs 3 and 4, respectively (i.e., corresponding to 33–40%
vanillin yield from batch oxidation of Indulin AT Kraft lignin).
One can expect that the improvement on the vanillin content
at the exit of the SPBCR in relation to the BCR is due to an
increase in the mass transfer of oxygen. Araujo (2008) has
shown that the mass transfer coefficient was 8.1 × 10−4 s−1
using packing against 6.0 × 10−4 s−1 without packing (that
means a 1.35-fold increase). However, in both cases, the lignin
mass conversion was substantially lower than that obtained
for the batch operation. The main reason of this difference is
the poor mass transfer of oxygen from the gas phase to the
liquid phase.
To improve the performance of the continuous reactor and
reach the production levels obtained in batch mode, the influ-
ence of some operating conditions was studied using a model
developed by Araujo (2008). Since the oxygen mass transfer
was the limiting step to vanillin formation, the limiting situa-
tion of pure oxygen in the gas feed was considered. Of course,
the liquid residence time should be decreased as the oxy-
gen mass transfer rate increases in order to avoid excessive
vanillin oxidation. From the structured reactor operating at
steady state, it was possible to achieve a value of the vanillin
concentration in the exit stream of approximately 1.8 g/L,
which is 85% of the maximum levels of vanillin concentration
obtained in the batch reactor.
It should be noticed that vanillin yield obtained in batch
experiments with Insulin AT Kraft lignin was quite low, lim-
iting strongly the operation of the continuous reactor. These
levels of vanillin yield seem to be limited by the used raw mate-
rial and by the low rates of oxygen transfer to the liquid in the
reaction unit. These two points are crucial for improving the
vanillin formation from Kraft lignin.
Despite the technical details to be solved or improved in
envisaging the industrial vanillin production from Kraft lignin,
an optimistic sight might be considered in view of the avail-
able Kraft lignin resources, reflecting on its low cost and its
supply sustainability; crucial aspects that could make the eco-
nomic application of this technology commercially feasible.
One should also keep in mind that the profitability of the lignin
extraction from the black liquor must be taken into account
(Olsson et al., 2006).
The commercial practicability of this process might be
attained if special attention is directed to the proper biomass
QO2 (mLNTP /min) QN2 (mLNTP /min) P (bar) TFset (K)b
1000 1000 10 443 (time < 40.2 min),
440 (time > 40.2 min)
2000 1000 10 436 (time < 66.5 min),
432 (time > 66.5 min)
1000 1000 10 443
1000 1000 10 433 (time < 66.5 min),
428 (time > 66.5 min)
(Kraft lignin) to be used. Improving the pre-treatment meth-
ods in order to preserve the essential fragments needed for
vanillin and improving the knowledge on biomass, and how
to produce more homogeneous lignin raw materials (for exam-
ple, using membrane processes before the reaction step), could
provide important technological advances for the application
of this alternative procedure. Anyway, it corresponds to a con-
siderable progress towards solving the mentioned bottleneck
in pulp and paper industries.
In this process, the energy requirements for the conversion
of lignin into vanillin can be provided by the exceeding Kraft
liquor from the pulp industry or even from the by-products of
lignin oxidation process (degraded lignin), according the cur-
rent biorefinery concept. We must remark that surplus energy
obtained from burning lignin is, nowadays, sold as electricity
in current Kraft pulp mills; thus, the cost of producing energy
for vanillin production instead of electricity should be added
to the final product.
In conclusion, considering the mentioned factors and the
achievable yields of Kraft lignin oxidation process, a favourable
balance could be obtained. The success of this process will
largely depend on the nature and composition of this complex
raw material.
4.3. Membrane separation
In the process of Kraft lignin oxidation to produce vanillin,
the reaction media contains degraded molecules of lignin and
sodium salt of vanillin as well as other species. Some other
oxidation products can be acetovanillone, dehydrodivanillin,
guaiacol, p-hydroxybenzaldehyde, and aromatic acids (Pearl,
1942). Mathias and Rodrigues (1995) have shown, using HPLC
and GC analyses, that after chemical oxidation of pine Kraft
lignin with molecular oxygen the lignin molecule is degraded
producing vanillin and, in some cases, acetovanillone (only
under strong oxidation conditions). The isolation of vanillate
from the oxidized solution is an important stage in the vanillin
production. Industrial companies as Monsanto (U.S.A.), Salvo
Chemical Co. (U.S.A.), and Ontario Paper Co. (Canada) produce
vanillin from the air oxidation of sulphite lignin and carry
out the extraction of vanillin from the reactant medium by
using liquid–liquid extraction with suitable solvents (benzene
or toluene) after acidification of the medium. The acidification
step is to precipitate the compounds of high molecular weight
(MW) and a large amount of acidic solution is required. Fur-
thermore, it is possible that vanillin is lost during this process,
since vanillin may be co-precipitated during the acidification
 chemical engineering research and design 8 7 ( 2 0 0 9 ) 1276–1292
step of the lignin (Craig and Logan, 1962; Fargues et al., 1996a;
Zabkova et al., 2006). Accordingly, the traditional precipitation
procedure, either by using toxic chemicals or by the waste
generation, is an option to be changed by more satisfactory
solutions.
In our proposal, the first step towards recovering the
vanillin from Kraft lignin oxidation is the ultrafiltration (UF)
process. In a simple approach, vanillin goes to the perme-
ate stream because of its low molecular weight, whilst the
lignin as a macromolecule is retained by the membrane during
the ultrafiltration. Membrane processes have all the require-
ments needed to act as key separation units in biorefineries
due their excellent fractionation capability, low chemical con-
sumption and low energy requisites (Jonsson et al., 2008). The
energy requirement is directly related to the volumetric flux
of the process. The cost of filtration processes can be reduced
if optimal operating conditions are defined for this operation.
Since 1960s the pulp and paper industry has paid attention to
membrane processes focused on the fractionation and con-
centration of sulphite liquor (to reduce the discharge of bleach
plant effluent or to use as biofuel) and also on the re-use of
water from purification of mill process effluents. The treata-
bility of spent sulphate liquor has also been considered in a
commercial-sized membrane plant (Jonsson et al., 2008) and
many researches have been focused on the filtration of Kraft
black liquor (Keyoumu et al., 2004; Liu et al., 2004a,b; Wallberg
et al., 2006).
The application of membrane technology on black liquor
has been largely limited due to its high temperature, high pH
value and high suspension solids concentration. Nowadays,
there is a high effort towards lignin extraction by ultrafiltration
of the Kraft black liquor (Barnier et al., 1987; Kirkman et al.,
1986; Tanistra and Bodzek, 1998; Wallberg et al., 2006; Woerner
and McCarthy, 1984). The cost of this operation is related to
the flux, cleaning frequency, membrane life-time and energy
requirement.
A number of works dealt with the use of UF and other fil-
tration techniques in view of the fractionation of chemical
compounds obtained from spent liquor (Afonso and Pinho,
1991; Bhattacharya et al., 2005; Drouin and Desroches, 1988;
Tsapiuk et al., 1989). The UF technique has been evaluated at
LSRE to recover vanillin from Kraft lignin oxidation (Zabkova
et al., 2007a). As seen, the temperatures of the solution at
the exit stream of the oxidation reactor are above 100 ◦ C and
the pH values are around 9–12. Since ceramic membranes can
withstand hard conditions of temperature, pH and alkalinity,
they are appropriate for this application. Therefore, tubular
ceramic membranes with different molecular weight cut-offs
were tested for the continuous fractionation of well-defined
high and low molecular weight fractions from lignin and vanil-
late mixtures.
The experiments were carried out in the batch mode,
in which the retentante was recirculated to the feed tank
and the permeate was withdrawn to a separated reservoir.
The membranes were Kerasep type with molecular weight
cut-off of 15 kDa supplied by Orelis (France) and Filtanium
type with molecular weight cut-off of 1 and 5 kDa supplied
by Tami Industries (France). Kerasep membranes are made
of Al2 O3 –TiO2 active layer on the ceramic support (working
transmembrane pressure (P): 0–4 bars; total membrane area:
0.008 m2 ; 400 mm length; 6 and 10 mm inside (Øi ) and out-
side (Øe ) diameter, respectively). Filtanium membranes are
made of TiO2 active layer on the ceramic support (working
P: 0–10 bars; membrane surface: 0.011 m2 ; 400 mm length;
1283
Øi = 6 mm and Øe = 10 mm). Although the mixtures of vanillin,
lignin and sodium hydroxide used on the experiments have
been prepared (i.e., they did not come from the exit stream of
the oxidation step), these experiments allow understanding
the involved phenomena, performances and expected prob-
lems related to this operation. A detailed description of the
lab-scale membrane set-up is shown in Zabkova et al. (2007a).
Considering the composition of 60 g/L lignin and 6 g/L
vanillin (a mixture artificially prepared assuming a vanillin
yield of around 10% from oxidation process), Tables 3a and 3b
show the experimental conditions together with the mea-
sured values (permeate flow, rejection, volume reduction
(permeate volume withdrawn/initial volume of the feed tank))
in UF of the lignin/vanillate mixture using tubular ceramic
membranes. Vanillin and lignin concentration were measured
by UV–vis spectrophotometer after pre-treatment of the sam-
ples (Zabkova et al., 2007a,b). The value of the permeate flux
refers to the first measurement of the flux in which, at this
case, concentration polarization layer was already formed.
This is due to the fact that the feed solution was circulated
through the system without opening the permeate valve in
order to fill the membrane system and to adjust the feed flow
rate. Afterwards, the permeate valve was opened and the first
measurement of the flux was performed. During the batch UF
experiments, the permeate flow rate, the retentate flow rate
and transmembrane pressure were measured. For the case
of pH 12.5, the percentage of reduction of the permeate flux
during the UF experimental run was 24% in 1100 min for the
case of 1 kDa cut-off membrane (Run 4), 20% in 780 min for
5 kDa cut-off membrane and 22% in 570 min for 15 kDa cut-off
membrane.
Lignin is mostly rejected and predominantly responsi-
ble for the surface fouling, when a progressive growth of a
deposited layer at the surface membrane takes place dur-
ing the ultrafiltration operation. The smaller molecules of
vanillin (152 Da MW) can diffuse entirely through the pore.
Once the vanillin transport across the membrane is facili-
tated, no vanillin rejection has been observed. It seems that
there is no occurrence of pore internal fouling (completely or
partially pore clogging), since there was no changing in mem-
brane rejection characteristics. It should be mentioned that
if a real stream was employed as feed, some other oxidation
products can be present. We can expect that compounds hav-
ing similar molecular weight to the vanillin will be recovered
in the permeate stream. In the crystallization step, the vanillin
can be purified. UF experiments using mixtures resulting from
the oxidation reaction must be considered in future work. In
current case, the cleaning step of the membrane using alkaline
solution has restored the original membrane permeability.
As one can see from Table 3b, the optimal membrane cut-
off used to recover vanillin from lignin/vanillin mixture is not
obvious since there are two important parameters involved
that should be taken into account: the purity of vanillin in
the permeate and the optimal permeate flux. The highest
lignin rejection value was obtained with a membrane of 1 kDa
L
cut-off (Robs = 0.972), but, due to the small cut-off, a signifi-
cant decrease of permeate flux was observed comparatively
to other membranes with bigger cut-off.
Since membranes with bigger cut-off might combine
higher flux with acceptable rejection values (as a comparison
between the results obtained from the 1 and 5 kDa cut-off),
a scheme of staging ultrafiltration membranes starting from
larger cut-off is proposed in this step of the integrated process.
In fact, the permeate of a first membrane might be taken as the
1284 chemical engineering research and design 8 7 ( 2 0 0 9 ) 1276–1292

Table 3a – Experimental conditions in ultrafiltration of the lignin/vanillin mixture using tubular ceramic membranes
(vc-flow , cross-flow velocity; C0f , feed concentration; V0 , initial feed volume).
Run Cut-off membrane (kDa) vc-flow (m s−1 ) C0f (g L−1 ) pH V0 (L)

Lignin Vanillin

1 0.992 60 6 8.5
15 5
2 0.992 60 6 12.5

3 5 1.179 60 5 12.5 4
4 1.277 60 6 8.5 5
5 1 1.179 60 6 12.5 4
6 1.179 2 6 12.5 4

Transmembrane pressure: P (1 and 5 kDa cut-off) = 1.55 bar; P (15 kDa cut-off) = 1.30 bar. Viscosity of solution:  = 1.190 × 10−3 Pa s (except to
the mixture of 2 g/L lignin and 6 g/L vanillin). The ultrafiltration experiments have run at ambient temperature.

feed of the second membrane of smaller cut-off, which means
a lower concentration of degraded lignin on the feed mixture
and, consequently, a higher permeate flux. From Table 3b, the
permeate flux with 1 kDa cut-off in staging mode was almost
5-fold larger in comparison to that when high concentration
was filtrated (Run 6). Considering the low concentration of
lignin in the solution, the flux decline does not represent sig-
nificant limitation in the process and this could be a way to
optimize the recovery of vanillin from Kraft lignin oxidation
media using ultrafiltration. In order to increase the volume
reduction, higher transmembrane pressures and, also, higher
temperatures of feed mixture could be applied. It is worth
mentioning that if an acidification step is used to precipitate
the compounds of high molecular weight (MW) from the reten-
tate in order to recover its vanillin content, a reduced amount
of acidic solution will be required.
Despite the limitation of permeate flux caused by the sur-
face fouling attributed to concentration polarization and gel
formation, a very high cleaning efficiency can be obtained
using a simple alkali cleaning medium (0.1–0.2 M NaOH). This
fact, in addition to membrane cleaning frequency and the
membrane lifetime, will be determinant parameters to decide
on the feasibility of this process.
4.4. Ion-exchange process purification
The mixture leaving the oxidation reactor and the membrane
separation step will contain vanillin and other vanillin related
low molecular weight phenolates. The vanillin is in the salt
form due to the used excess of sodium hydroxide required in
the oxidation reaction.
According to a patent of Hibbert and Tomlinson (1937),
the mixture can be acidified and vanillin extracted with suit-
able solvents such as benzene or toluene. A drawback of
this technique is related to the large amount of acid needed
to neutralize the mixture with high alkalinity leading to
lignin precipitation that further complicates the extraction
and causes vanillin losses. The precipitation can be avoided
by direct extraction of sodium vanillate with n-butyl alcohol
or isopropyl alcohol (Bryan, 1956; Sandborn et al., 1936). The
disadvantage of this process is also related to the high alkalin-
ity since sodium vanillate presents limited solubility in such
organic solvents.
An alternative method to recover vanillin without needing
acidification of the oxidized liquor consists on using cationic
ion-exchange resin as suggested by Forss et al. (1986) and
Fargues et al. (1996a). It was found out that eluting the strongly
alkaline effluent with water through a cationic ion-exchange
resin in the sodium form provides good separation of vanillin,
and other oxidation products, from lignin, which elutes first.
Moreover, more than 60% of the sodium contained in the efflu-
ent is eluted together with lignin and could be reused without
any neutralization, a step always performed prior to chemical
recovery in pulp mills.
According to Forss et al. (1986), in vanillin production, the
treatment involving ion exchange in sodium form should be
carried out between the oxidation and extraction steps. This
method has some advantages such as separation of 80% of
dry matter, lignin and sodium from the vanillin reactor efflu-
ent and the quantity of acid needed to neutralize the vanillin
fraction is small. Moreover, the ion-exchange resin does not
require a regeneration step and the lignin can be returned to
the chemical recovery of the pulp mill.
In 1971, Craig and Logan (1971) have reported the use of
a weak ion-exchange resin in acid form for the vanillin iso-
lation. The same procedure is mentioned by Logan (1965) as
one of the steps in a cyclic process for the recovery of vanillin.
The exchange capacity strongly depends on the effluent pH.

Table 3b – Measured values in ultrafiltration of the lignin/vanillin mixture using tubular ceramic membranes (VR, volume
reduction; J, permeate flux; Robs , observed rejection; Rm , membrane hydraulic resistance).
Run Cut-off membrane (kDa) VR Robs (lignin)a J × 10−3 (L m−2 h−1 ) Rm × 1013 (m−1 )

1 0.313 0.943 28.12

15 0.2806
2 0.340 0.865 16.87

3 5 0.275 0.968 13.50 1.222

4 0.102 0.951 4.75

5 1 0.103 0.972 3.75 2.806
6 0.410b 0.981b 18.00b

a
Robs (Vanillin) ≈ 0 for all ultrafiltration run.
b
Considering this filtration on the 1 kDa cut-off membrane as a second stage in a staging membrane operation mode.
 chemical engineering research and design 8 7 ( 2 0 0 9 ) 1276–1292
Fig. 4 – History of the vanillate (VNa), vanillin (VH) and pH
in an ion-exchange process using a column packed with
Amberlite IR120H (150 × 20 mm i.d.). Feed solution:
0.05 Eq L−1 NaOH plus vanillin (pH 11.9). Conditions:
19 cm3 min−1 ; pHinitial = 6.1; 293 K.
Furthermore, the process rate is accelerated with temperature
increase. This fact stands for an important task in continu-
ous overall process for vanillin production from Kraft liquor.
Since the vanillin effluents leave the reactor at high tempera-
ture there is no need for cooling the media before it enters the
ion-exchange column.
This particular method, where a strong cationic resin in H+
form is applied, was studied considering the influence of both
alkalinity and vanillin concentration of the treated solution by
using static and dynamic approaches (Zabkova et al., 2007b).
In the overall process of vanillin recovery from Kraft liquor,
the ion-exchange step should be designed after the mem-
brane module where the high molecular weight compounds
and dry matter are safely removed. Therefore, the applica-
tion of an appropriate membrane cut-off, temperature and
high alkalinity resistance materials are essential. The per-
meate stream containing low molecular weight phenolates
and sodium vanillate can be directly introduced to the col-
umn filled with a strong cation ion-exchange resin in H+ form.
Since the eluent which is collected in the outlet of the col-
umn contain all compounds in phenolic form there is no need
for further acidification or neutralization. Moreover, if the
ion-exchange process is placed after the membrane, the pre-
cipitation of lignin, which could be motivated by a decrease in
pH in the opposite case, can be avoided. The main advantages
of the integrated process include: (1) no need of acid to neu-
tralize the vanillin fraction (is only needed to regenerate the
resin bed); (2) no lignin precipitation during the ion-exchange
process; (3) no need of high cost equipment neither equipment
working under high pressure is needed; and (4) ion-exchange
resin regeneration easy to perform.
When the solution contains only a base (NaOH), the H+ ion
from the resin is exchanged with the Na+ ion from the solution
and converted into water due the irreversible reaction with the
OH− co-ion in solution. When the solution contains vanillate
and vanillin in solution (or only vanillate if the pH value is
high enough), the H+ ion from the resin exchanges for Na+ ion
from the solution and vanillin is produced. Fig. 4 shows the
typical ion-exchange profiles of the Na+ ions, vanillate (VNa),
vanillin (VH) and pH for a feed solution of vanillate and vanillin
with pH 11.9 using a fixed bed (resin: Amberlite IR120H; 293 K;
19 mL min−1 ). The resin will be converted to sodium form and
the vanillin appears in the outlet of the column. The behaviour
of pH profile depends on the buffer vanillate/vanillin formed
during the ion-exchange process.
1285
4.5. Vanillin crystallization—process description
The vanillin obtained by the Kraft liquor oxidation and later
recovered at a certain yield using a combination of membrane
and ion-exchange processes can still contain small amounts of
impurities. The final purification step of vanillin represents a
difficult task and multistage crystallization is needed to obtain
a high purity product (Makin, 1984).
The main impurities of vanillin consist of low molecular
weight substances, e.g., o-vanillin, 5-formyl vanillin, vanillin
acid, acetovanillone, etc. The number of crystallization steps
should be always based on further vanillin applications but
it also depends on impurity contents and type. It is well
established that the needed vanillin purity for technical appli-
cations is over 97%, while the vanillin grade suitable for
foodstuffs applications should be higher than 99.8%.
Schoeffel (1962) reported that vanillin can be purified using
multiple crystallization steps from water–methanol solutions.
The vanillin obtained from lignin oxidation is dissolved in a
solvent mixture of water and methanol with a composition
preferably of 40% of the methanol (w/w). In the beginning of
the crystallization process the solution is stirred and rapidly
cooled to reach a temperature between 23 and 25 ◦ C. The first
indicator that crystallization begins is the slight increase in
temperature which must be measured inside the reactor.
As already mentioned, the main target in batch crystalliza-
tion is to control the supersaturation and nucleation especially
during the initial stage. This can be performed by seeding
using purified crystalline vanillin to obtain a better crys-
tallization performance. The temperature is maintained, for
approximately one hour, within a range of ±2 ◦ C of the ini-
tial crystallization temperature. Afterwards, the reactor with
the vanillin solution is continuously stirred and the content
slowly cooled during about three hours to reach a tempera-
ture around 5 ◦ C. The crystallization is completed following a
stirring period of about 2 h of the reactor mixture maintaining
a temperature of 5 ◦ C.
The following step in vanillin purification includes filtra-
tion of the crystallized mixture. The solids filter cake must be
washed with successive amounts of cold water (under 5 ◦ C)
and thereafter carefully dried. The vanillin crystals obtained
in first crystallization step should be analyzed for the men-
tioned impurities and methanol content in order to plan the
needed multiple crystallization steps.
In the case that the vanillin was not obtained with the
desired purity, the recrystallization can be performed using
a similar procedure but with a solvent mixture containing
methanol at 50% (w/w). The vanillin in the crystalline form is
first diluted at higher temperature (40–50 ◦ C) and later rapidly
cooled to 35 ◦ C under continuous stirring. The content of
the reactor is again seeded using finished pure vanillin and
the crystallization is carried on at 30–34 ◦ C under continuous
stirring during a time period of 1 h. Afterwards, the reactor
is cooled slowly during 2 h to reach a temperature of 13 ◦ C.
Finally, the crystals are filtered and dried using the previ-
ously described procedure. The filtrates containing methanol
obtained from multiple crystallization steps, and from the per-
formed washings, can be lately reused. The methanol content
could be easily distilled and recycled to the crystallization.
Apart from multiple water–methanol crystallization pro-
cess, the purified vanillin can be also obtained by one or
more crystallizations from water, using charcoal to adsorb
the remaining impurities traces (Major and Nicolle, 1977;
Schoeffel, 1962). First, the vanillin solution is prepared by dis-
1286 chemical engineering research and design 8 7 ( 2 0 0 9 ) 1276–1292

solving vanillin in hot water at around 90 ◦ C. The solution

is thereafter cooled till the crystallization begins (at around
50 ◦ C). The crystallization continues for a time period of 2 h
at 45 ◦ C. Afterwards, the temperature is slowly decreased to
reach 13 ◦ C during 3 h. Finally, the crystallization is completed
after an additional period of 1 h maintaining the same con-
ditions. Finally, the obtained product is carefully washed and
dried.

4.6. Lignin-based polyurethane synthesis

This section discusses lignin potential as a raw material for

polyurethane synthesis having in view the valorisation of a
co-product generated during the biorefinery process. In this
context, two lignin sources are available: (1) polymeric lignin
obtained after processing the black liquor generated dur-
ing the papermaking process (lignin plant product) and (2)
degraded lignin obtained as a by-product of the vanillin syn-
thesis process (vanillin plant by-product).
Polyurethanes are considered as one of the most versatile
polymeric materials offering a wide range of products with
various applications in diverse sectors. Rigid polyurethane
(RPU) foams and elastomers belong to this class of prod-
ucts. The interest to explore lignin as a raw material for
polyurethane synthesis has lead, in the past few decades,
to some research works where various types of materials
(rigid foams, elastomers, sealants) have been produced using
a wide range of chemical systems. The exhibited proper-
ties were found promising and, in some cases, similar to
those of conventional polyurethanes. The importance of these
achievements leads to the appearing of some patented results
(Glasser et al., 1991; Kurple, 2000; Yoshida et al., 1986). It is
worth to mention that the interest for developing lignin-based
applications, including lignin-based materials, became more
notorious during the last few years. Two major factors are
related to this situation, namely (1) the availability of new
lignin sources, such as, sulphur-free lignins and (2) the grow-
ing interest on biorefinery processes where lignin valorisation
offers impact (it represents up to 30% of biomass weight).
The utilization of lignin in polyurethane synthesis often
follows two global approaches: (1) the direct utilization of
lignin without any preliminary chemical modification, alone
or in combination with other polyols (Cateto et al., 2008c;
Evtuguin et al., 1998; Vanderlaan and Thring, 1998; Yoshida
et al., 1990) or, (2) by making hydroxyl functions more read-
ily available by chemical modification, such as esterification
and etherification reactions (Cateto et al., 2008b; Gandini et
al., 2002; Glasser, 1989; Hatakeyama, 2002; Nadji et al., 2005). A
summary of possible synthetic routes is shown schematically
in Fig. 5. Path (1) shows lignin direct use as macromonomer,
path (2) followed by path (3) illustrates lignin liquefaction
by chemical modification and its subsequent use in rigid
polyurethane foams preparation, path (2) followed by path
(4) illustrates the preparation of an intermediate lignin-based
polyol and its subsequent incorporation into polyurethane
elastomers, and path (5) corresponds to the direct use of lignin
as a reactive filler in polyurethane foam chemical systems.
Table 4 shows a summary of some representative research
works putting in evidence synthesis details and the used
approach according to the synthetic routes defined in Fig. 5.
Within this work we report the research conducted by LSRE
in cooperation with LGP2 in this field during the past five years.
Briefly, in the first approach lignin was used directly as a co-
monomer in combination with a linear polyester-polyol. In the
Fig. 5 – Synthetic routes for lignin incorporation in
polyurethane materials: (1) direct use, (2) synthesis of
liquid polyol, (3) synthesis of rigid polyurethane foams
using lignin-based polyols, (4) synthesis of polyurethane
elastomers using lignin-based polyols and (5) synthesis of
rigid polyurethane foams using lignin as reactive filler.
second approach lignin was oxypropylated and the resulting
liquid polyols incorporated into RPU foams, alone or in com-
bination with other commercial polyether-polyols. In order
to develop a competitive green process, our objective was
mainly centred on developing systems where lignin is used
as received and where bulk processes are favoured.
4.6.1. Lignin characterization
For polyurethane synthesis, the knowledge of lignin hydroxyl
content is needed to establish a formulation. Typical analytical
procedures used for that purpose are based on spectroscopic
techniques (NMR, UV, FTIR), or on wet chemical methods
(acetylation, permanganate oxidation, aminolysis, etc.), or in
a combination of both. A review on lignin hydroxyl groups
determination was reported in previous work using some
commercially available technical lignins (Indulin AT (Mead-
westvaco), Alcell (Repap), Curan 27-11P (Borregaard LignoTech)
and Sarkanda (Granit SA)) (Cateto et al., 2008a). The hydroxyl
contents were determined using NMR techniques, namely: H,
C and P NMR and a titration procedure based on a standard
method (ISO 14900:2001 (E) method). The combination of the
used analytical techniques allowed for a complete hydroxyl
characterization: type and content. Table 5 shows total and
phenolic hydroxyl contents determined by C NMR using acety-
lated samples according to the procedure ISO 14900:2001(E).
Acid content was determined by P NMR, and ash content gravi-
metrically after subject lignin samples to a muffle furnace at
525 ◦ C during 5 h.
4.6.2. Lignin use without chemical modification
Lignin can be directly incorporated into polyurethane for-
mulations. It contains, within its structure, both aliphatic
and aromatic hydroxyl groups which can potentially act as
reactive sites for isocyanate groups (formation of urethane
linkages). This strategy has a main drawback; the produced
materials are, most often, rigid and brittle. Lignin molecules
are relatively stiff polyhydroxy macromolecules thus gener-
 chemical engineering research and design 8 7 ( 2 0 0 9 ) 1276–1292 1287

Table 4 – Summary of some representative works in the field of lignin-based polyurethane synthesis.
Work Synthesis details

Saraf et al. (1985) Lignin type: Kraft

Chemical system: Polyol: Oxypropylated lignin + PBD. Isocyanate: TDI. Catalyst: DBTDL
Polymerization details: Polymerization in solution (THF)
Synthetic route: (2) followed by (4)

Tan (1996) Lignin type: Kraft

Chemical system: Polyol: Lignin-cardanol based polyol. Isocvanate: TDI. Catalyst: SnOct
Polymerization details: Polymerization in solution (THF)
Synthetic route: (2) followed by (4)

Hirose et al. (1998) Lignin type: Kraft

Chemical system: Polyol: lignin + PEG400. Isocyanate: MDI. Catalyst: DBTDL
Polymerization details: Polymerization in solution (THF)
Synthetic route: (1)

Evtuguin et al. (1998) Lignin type: Oxygen-organosolv lignin

Chemical system: Polyol: Oxygen-organosolv lignin. Isocyanate: ODI. Catalyst: DBTDL
Polymerization details: Polymerization in bulk
Synthetic route: (1)

Sarkar and Adhikari (2001) Lignin type: Kraft

Chemical system: Polyol: lignin + HTPB. Isocyanate: TDI. Catalyst: TEA
Polymerization details: Polymerization in solution (THF and THF:DMF (80:20))
Synthetic route: (1)

Hatakeyama (2002) Lignin type: Organosolv and Kraft

Chemical system: Polyol: Lignin-caprolactone based polyol. Isocyanate: MDI. Catalyst: DBTDL
Polymerization details: Polymerization in solution (THF)
Synthetic route: (2) followed by (4)

Ni and Thring (2003) Lignin type: Organosolv

Chemical system: Polyol: Lignin + PEG400. Isocyanate: MDI. Catalyst: DBTDL
Polymerization details: Polymerization in solution (THF)
Synthetic route: (1)

Chahar et al. (2004) Lignin type: Kraft

Chemical system: Polyol: Lignin + PEG (MW: 200, 600, 1000, 1500 and 4000). Isocyanate: TDI.
Catalyst: SnOct
Polymerization details: Polymerization in solution (THF)
Synthetic route: (1)

Thring et al. (2004) Lignin type: Organosolv

Chemical system: Polyol: Lignin + PEG (MW: 200, 600, 1000, 1500 and 2000). Isocyanate: MDI,
Catalyst: SnOct
Polymerization details: Polymerization in solution (THF)
Synthetic route: (1)

Nadji et al. (2005) Lignin type: Kraft and organosolv

Chemical system: Polyol: Oxypropylated lignin Isocyanate: polymeric MDI. Catalyst:
NIAX + DMCHA. Others: Foam additives (blowing agents and surfactants)
Polymerization details: Polymerization in bulk (foam preparation)
Synthetic route: (2) followed by (3)

Hatakeyama and Hatakeyama (2005) Lignin type: Kraft and sulphite

Chemical system: Polyol: lignin + PEG200. Isocyanate: polymeric MDI. Catalyst: DBTDL, Others:
foam additives (water and surfactants)
Polymerization details: Polymerization in bulk (foam preparation)
Synthetic route: (5)

Cateto et al. (2008a,b,c) Lignin type: Kraft

Chemical system: Polyol: lignin + PCL (MW: 400, 750 and 1000). Isocyanate: MDI. Catalyst: n.a.
Polymerization details: Polymerization in bulk
Synthetic route: (1)

Polyols: polybutadiene glycol (PBD), poly(ethylene glycol) (PEG), hydroxyl-terminated polybutadiene (HTPB), polycaprolactone (PCL). Isocyanates:
Toluene diisocyanate (TDI), methylene diphenyl diisocyanate (MDI), octadecyl isocyanate (ODI). Catalysts: Stannous octoate (SnOct), dibutyltin
dilaurate (DBTDL), Triethylamine (TEA), NIAX and DMCHA are amine based catalysts. Solvents: Tetrahydrofuran (THF), dimethylformamide
(DMF).

ating highly branched three-dimensional polyurethanes, in
which the lignin residues act as crosslinking sites. To over-
come this restriction and generate more attractive materials,
like elastomers, it is a current practice to incorporate a linear
polyol, such as PEG, in the original formulation. However, the
efficiency of this approach strongly depends on lignin type,
molecular weight and content, and polyol type and molecular
weight.
Within this context, we have conducted some studies
using a three-component system, namely lignin (Indulin AT),
1288 chemical engineering research and design 8 7 ( 2 0 0 9 ) 1276–1292
Table 5 – Properties of the characterized technical lignins. Values are represented in mmol/g on a free ashes basis.
Lignin sample Ash content (%, w/w) Phenolic OH content (mmol/g)
Alcell 0.05 3.82
Indulin AT 3.06 3.90
Sarkanda 3.26 2.46
Curan27-11P 17.0 3.53
4,4-methylene-diphenylene isocyanate (MDI) and a linear
polycaprolactone (PCL) of three different average-molecular
weights (1000, 750 and 400). PCL, a biodegradable polyester,
was introduced in the formulation to impart flexibility to final
products and enable polymerization in bulk. Lignin was used
as received at contents of 10%, 15%, 20%, and 25% (w/w) in
the polyol mixture (PCL + lignin). Preparation details can be
found elsewhere (Cateto et al., 2008b,c). Samples were char-
acterized by Dynamic Mechanical Analysis (DMA) using an
RSA3-TA instrument working in tensile modulus.
The obtained results, using FTIR-ATR monitoring and
swelling studies, pointed out for effective lignin incorporation
into final three-dimensional polyurethane networks (Cateto
et al., 2008c). For Indulin AT, our results support the con-
clusion that the use of PCL400 favours lignin incorporation
but, in terms of mechanical properties, it showed to be too
short to impart flexibility into final products. The produced
materials were brittle and consequently, not useful, even for
structural applications. For longer PCL chains, like PCL1000,
a decrease in lignin incorporation efficiency was observed,
mainly justified as a result of the increasing dilution factor
of hydroxyl groups with increasing PCL molecular weight. It
is also expected that for too short PCL segments, and depend-
ing on lignin type and molecular weight, the PCL segments
attached to a lignin become difficult to react with isocyanates
appended to other lignin molecules, thus hindered the for-
mation of a three-dimensional network (Thring et al., 2004).
For the used PCL molecular weight range, 400, 750 and 1000,
the effect was not noticed for Indulin AT based systems but
clearly appeared for systems based on Alcell lignin. Compar-
atively to Indulin AT, Alcell lignin, an organosolv lignin, has
lower molecular weight and lower hydroxyl content. These
results pointed out that, lower molecular weight lignins like
degraded lignins obtained as by-products of the vanillin pro-
cess, will need higher molecular weight PCL to produce useful
materials. In this situation we may assume that lignin has a
main role of chain extender.
Fig. 6 shows the effect of Indulin AT content on the storage
modulus (E ) and loss tangent (tan ı) with temperature. As can
be observed, the maximum of tan ı shifts towards higher tem-
peratures and the peak become broader as the lignin content
increased. Thus, glass transition temperature (Tg ) increases
with the increase of lignin content reflecting an increasing
of lignin incorporation. For the rubbery region the E value
increases with lignin content until a maximum of 20%. E value
for the sample prepared with 25% has shown no variation
comparatively to the sample with 20% lignin content.
4.6.3. Lignin use after chemical modification
Within this approach we have studied lignin oxypropylation
and its subsequent incorporation into RPU foams (Cateto et
al., 2007). Oxypropylation has been recognized as a viable and
promising approach to overcome the technical limitations and
constrains imposed by the polymeric nature of lignin when
directly used as a macromonomer for synthesis purposes.
By means of oxypropylation, the hydroxyl groups, in partic-
Total OH content (mmol/g) COOH (mmol/g)
5.24 0.23
6.89 0.62
5.39 0.39
5.99 0.47
ular the phenolic ones entrapped inside the molecule and of
difficult access, are liberated from steric and/or electronic con-
strains and, at the same time, the solid lignin becomes a liquid
polyol, thanks to the introduction of multiple ether moieties.
The optimal oxypropylation conditions were fixed on the
basis of the requirements which a given polyol should fill
when used in rigid polyurethane foam formulations, i.e.,
a hydroxyl index between 300 and 800 and a viscosity
below 300 Pa s. Table 6 shows the technical properties of the
produced Indulin AT based polyols using lignin/propylene-
oxide/catalyst (L/PO/C, w/v/(%, w/w)) formulations of 20/80/5
and 30/70/2, where the catalyst content is based on lignin
weight. The oxypropylation reaction was carried out in bulk
using a 450 mL Parr reactor. The desired amount of lignin,
propylene oxide and the catalyst (KOH) were placed in the
reactor that was thereafter closed and heated under stirring
till 160 ◦ C. After reaching a maximum, the pressure decreased
almost instantaneously (1–3 min) reflecting the propylene
oxide consumption. When the relative pressure reached zero,
the reactor was cooled under stirring and the ensuing polyols
recovered.
Oxypropylation is always accompanied by homopolymer-
ization of propylene oxide arising from transfer reactions
during the anionic grafting mechanism. The produced poly-
ols are in fact a mixture of oxypropylated lignin and some low
molecular weight products, polypropylene oxide oligomers.
These oligomers are normally left in the final mixture
since they constitute a very useful bifunctional co-monomer,
decreasing viscosity and the glass transition temperature. The
prepared lignin-based polyols were introduced into RPU foam
formulations without performing any type of purification in
which regards homopolymer and catalyst (KOH) presence.
A typical RPU foam formulation included a polyol prepared
with the lignin-based polyol (alone or mixed with 25–75%
Fig. 6 – Storage modulus (E ) and loss tangent (tan ı) as a
function of temperature for elastomers containing Indulin
AT contents of 10%, 15%, 20% and 25% (w/w in the polyol
mixture). Samples have been prepared with PCL750.
 chemical engineering research and design 8 7 ( 2 0 0 9 ) 1276–1292 1289

Table 6 – Technical properties of the produced Indulin AT based polyols.

Lignin-based polyol Viscosity (Pa s) IOH (mg KOH/g) Homopolymer M̄n (g/mol) M̄w (g/mol) PI
(L/PO/C, w/v/(%, w/w)) content (%, w/w)

20/80/5 4.34 326.2 36.7 1224 3829 3.1

30/70/2 66.5 348.8 24 2 1453 5331 37

Molecular weight data (M̄n , M̄w and polydispersity index (PI)) based on PS standards.

Table 7 – Density, thermal conductivity and compressive modulus of Indulin AT based RPU foams.
Lignin-based polyol Lignin-based polyol Density Conductivity Compressive
(L/PO/C, w/v/(%, w/w)) content (%, w/w) (kg/m3 ) (mW/mK) modulus (MPa)

20/80/5 100 19.2 26.8 2.6

50 22.4 32.9 2.4

30/70/2 100 23.1 27.4 4.0

50 23.7 29.1 3.6

Reference foam (100% commercial polyol) 31.1 30.3 4.6

(w/w) of Lupranol 3323, a commercial polyether-polyol) in
combination with 10% (w/w) of glycerol (a co-crosslinking
agent). Additionally, a physical blowing agent (n-pentane at
20% (w/w)), a catalyst combination (2% (w/w) of a mixture with
equal amounts of DMCHA and NIAX-A1), a surfactant (SR-321
NIAX at 2% (w/w)) and water at 2% (w/w) were introduced.
Weight contents of the blowing agent, catalyst, surfactant and
water were determined with respect to the total weight of
polyol. Polymeric isocyanate (p-MDI with an average function-
ality of 2.7) was added at a NCO/OH ratio of 1.1. RPU foams
were prepared according to the following steps: (1) the polyol
was stirred together with the surfactant and the catalyst mix-
ture, for 60 s, (2) the physical blowing agent was added and the
mixture stirred for more 30 s and (3) the isocyanate was added
in one portion and the reactive mixture stirred for more 15 s.
After foam rise samples were left to cure during 24 h at room
temperature. Density, conductivity and compressive modulus
were determined according to ASTM D1622, ASTM D1621 and
ASTM C177 standards, respectively.
Table 7 shows the characterization results obtained for den-
sity, conductivity and compressive modulus. Properties of the
reference RPU foam, produced with commercial polyols, are
also presented. The produced lignin-based RPU foams (partic-
ularly those based on 20/80/5 polyols) exhibit slightly lower
mechanical properties and density comparatively to the ref-
erence foam. Nevertheless, we must highlight some added
value properties imparted by lignin, such as, improvement of
moisture and flame resistance.
5. Conclusions
Sustainable growth involves safe raw materials resources for
the industrial production of goods. In this context, Kraft lignin,
as a large available biomass resource (by-product stream of
pulp mills), is a very attractive and promising raw material
to produce vanillin, biopolymers and biofuels to be used in
process unit operation. This idea concept was explored by
designing an integrated process for producing vanillin and
lignin-based polyurethanes.
Vanillin is produced using Kraft lignin oxidation under alka-
line medium generating a stream containing degraded lignin
and salt of vanillin. For the separation of these two fractions,
and due to the presence of other contaminants, the use of
intensive purification procedures is required. We have dis-
cussed one alternative route to allow vanillin recovering using
consecutive membrane, ion-exchange and crystallization pro-
cesses. These processes have a key role on the economical
competitiveness of lignin-based process, and breakthroughs
in this field will lead to major positive impacts towards its
expansion in the industrial scenery.
A complementary approach considering the production of
lignin-based polyurethanes elastomers and foams has been
also explored. The incorporation of lignin into materials is
in fact an attractive approach since it can take advantage of
macromolecular properties. Both approaches of using lignin
(as such or after chemical modification) yielded materials with
quite promising properties. A reliable economic evaluation
could be noteworthy in this context, but is difficult to achieve
at this stage of development. In a first approach it needs to
be based on existing technology (that sometimes is not suit-
able) and on the knowledge of market product acceptance.
Furthermore, assessment of the degree of difficulty for devel-
oping new technology must also be needed. In any case, some
lignin-based polyurethanes with quite interesting properties,
in some cases quite similar to those prepared with conven-
tional counterparts, have been produced, which corroborate
the interest and viability of this approach both as a primary or
secondary product.
Acknowledgments
Financial support for projects POCTI/EQU/33198/99, POCI/EQU/
61738/2004, CYTED IV.17/2002-2006 and ACÇÃO LUSO-
FRANCESA F13/06 is acknowledged.
References
Adachi, S., Tanimoto, M., Tanaka, M. and Matsuno, R., 1992,
Kinetics of the alkaline nitrobenzene oxidation of lignin in
rice straw. Chem Eng J, 49(2): B17–B21.
Afonso, M.D. and Pinho, M.N.D., 1991, Membrane separation
processes in the pulp and paper industry. Desalination, 85(1):
53.
Araujo, J.D.P., 2008, Processo de Produção de Vanilina a Partir do
Licor Negro das Indústrias da Pasta de Papel, Ph.D. thesis,
University of Porto, Porto, Portugal.
Axelsson, E., Olsson, M.R. and Berntsson, T., 2006, Increased
capacity in kraft pulp mills: Lignin separation and reduced
steam demand compared with recovery boiler upgrade. Nord
Pulp Pap Res J, 21(4): 485–492.
1290 chemical engineering research and design 8 7 ( 2 0 0 9 ) 1276–1292
Barnier, H., Maurel, A. and Pichon, M., 1987, Separation and
concentration of lignosulfonates by ultrafiltration on mineral
membranes. Pap Puu-Pap Tim, 69(7): 581–583.
Bhattacharjee, S., Datta, S. and Bhattacharjee, C., 2006,
Performance study during ultrafiltration of kraft black liquor
using rotating disk membrane module. J Clean Prod, 14(5):
497–504.
Bhattacharya, P.K., Todi, R.K., Tiwari, A., Bhattacharjee, C.,
Bhattacharjee, S. and Datta, S., 2005, Studies on ultrafiltration
of spent sulfite liquor using various membranes for the
recovery of lignosulphonates. Desalination, 174(3):
287–297.
Biomass R&D Technical Advisory Committee, 2006, Vision for
bioenergy and biobased products in the United
States—bioeconomy for a sustainable future, Available from:
http://www1.eere.energy.gov/biomass/pdfs/final 2006 vision.
pdf.
Bjorsvik, H.R., 1999, Fine chemicals from lignosulfonates. 1.
Synthesis of vanillin by oxidation of lignosulfonates. Org
Process Res Dev, 3(5): 330–340.
Bjorsvik, H.R. and Liguori, L., 2002, Organic processes to
pharmaceutical chemicals based on fine chemicals from
lignosulfonates. Org Process Res Dev, 6(3): 279–290.
Bryan, C.C., 1956, Canadian Patent 5,280,837.
Cateto, C.A., Barreiro, M.F. and Rodrigues, A.E., 2008, Monitoring
of lignin-based polyurethane synthesis by FTIR-ATR. Ind
Crops Prod, 27(2): 168–174.
Cateto, C.A., Barreiro, M.F., Rodrigues, A.E. and Belgacem, M.N.,
2007, Oxypropylation of lignins and characterization of the
ensuing polyols, In Proceedings of the 8th ILI Forum ,
pp. 115–119.
Cateto, C.A., Barreiro, M.F., Rodrigues, A.E. and Belgacem, M.N.,
2008, Lignin-based polyurethane materials, In Proceedings of
the 10th International Chemical and Biological Engineering
Conference—CHEMPOR , pp. 231–236.
Cateto, C.A., Barreiro, M.F., Rodrigues, A.E., Brochier-Solan, M.C.,
Thielemans, W. and Belgacem, M.N., 2008, Lignins as
macromonomers for polyurethane synthesis: a comparative
study on hydroxyl group determination. J Appl Polym Sci,
109(5): 3008–3017.
Cerrutti, P., Alzamora, S.M. and Vidales, S.L., 1997, Vanillin as an
antimicrobial for producing shelf-stable strawberry puree. J
Food Sci, 62(3): 608–610.
Chahar, S., Dastidar, M.G., Choudhary, V. and Sharma, D.K., 2004,
Synthesis and characterisation of polyurethanes derived from
waste black liquor lignin. J Adhes Sci Technol, 18(2):
169–179.
Craig, D. and Logan, C.D., 1971, Pulp Pap, 45: 137.
Craig, D. and Logan, C.D., 1962, Method of producing vanillin and
other useful products from lignosulfonic acid compounds,
U.S. Patent 3,054,825; Br. Patent 935,305.
Creighton, R.H.J., McCarthy, J.L. and Hibbert, H., 1941,
Syringaldehyde from Spruce and Maple Woods. J Am Chem
Soc, 63: 312.
Dardelet, S., Froment, P., Lacoste, N. and Robert, A., 1985,
Aldihyde syringique, possibilitis de production a partir de bois
de feuillus. Rev ATIP, 39(5): 267–274.
Daugsch, A. and Pastore, G., 2005, Production of vanillin: a
biotechnological opportunity. Quim Nova, 28(4): 642–645.
Diddams, D.G. and Krum, J.K., 1970, in Mark, H.F., Mark, H.F., et al.
(eds) (p. 180).
Drouin, M.P. and Desroches, M.J., 1988, Isolation of lignin from
spent kraft liquor by hyper and ultrafiltration. AIChE Forest
Prod Div, 2: 58.
European Parliament and Council, 2002, Towards a European
Strategy for the Security of Energy Supply, European
Parliament, Strasbourg.
Evtuguin, D.V., Andreolety, J.P. and Gandini, A., 1998,
Polyurethanes based on oxygen-organosolv lignin. Eur Polym
J, 34(8): 1163–1169.
Food and Agriculture Organization of the United Nations, 2008,
Forestry FAO home, Food and Agriculture Organization—FAO
site, Available from: www.fao.org/forestry.
Food, Food and Agriculture Organization of The United Nations,
FAOSTAT, Available from http://www.fao.org/forestry
(accessed in: July, 2005).
Fargues, C., Mathias, A. and Rodrigues, A., 1996, Kinetics of
vanillin production from kraft lignin oxidation. Ind Eng Chem
Res, 35(1): 28–36.
Fargues, C., Mathias, A., Silva, J. and Rodrigues, A., 1996, Kinetics
of vanillin oxidation. Chem Eng Technol, 19(2): 127–136.
Fitzgerald, D.J., Stratford, M. and Narbad, A., 2003, Analysis of the
inhibition of food spoilage yeasts by vanillin. Int J Food
Microbiol, 86(1–2): 113–122.
Forss, K.G., Talka, E.T. and Fremer, K.E., 1986, Isolation of vanillin
from alkaline oxidized spent sulfite liquor. Ind Eng Chem Prod
Res Dev, 25(1): 103–108.
Gandini, A., Belgacem, M.N., Guo, Z. and Montanari, S., 2002, in
Hu, T.Q. (ed) (Kluwer Academic Plenum Publishers, New York),
pp. 57–80.
Glasser, W.G., 1989, in Kennedy, J.F., Phillips, G.O., & Williams, P.A.
(eds) (Ellis Hordwood Ltd, Chichester), pp. 163–173.
Glasser, W.G., Oliveira, W., Kelley, S.S. and Niehl, S., 1991, U.S.
Patent 5,066,790.
Halliday, J., 2008, Borregaard raises vanillin prices, will work on
efficiencies, FoodNavigator—Financial & Industry, Available
from: http://www.foodnavigator.com/Financial-Industry/
Borregaard-raises-vanillin-prices-will-work-on-efficiencies.
Hatakeyama, H., 2002, in Hu, T.Q. (ed) (Kluwer Academic Plenum
Publishers, New York), pp. 41–56.
Hatakeyama, H. and Hatakeyama, T., 2005, Environmentally
compatible hydrid-type polyurethane foams containing
saccharide and lignin components. Macromol Symp, 224:
219–226.
Hibbert, H. and Tomlinson, G.H.J., 1937, Manufacture of vanillin
from waste sulphite liquor, US Patent 2,069,185, United States.
Higuchi, T., 1980, Lignin structure and morphological distribution
in plant cell wall, in Lignin Biodegradation: Microbiology,
Chemistry, and Potential Applications, Kirk, T.K., Higuchi, T., &
Chang, H.M. (eds) (CRC Press, Boca Raton), pp. 1–20.
Hirose, S., Kobashigawa, K., Izuta, Y. and Hatakeyama, H., 1998,
Thermal degradation of polyurethanes containing lignin
studied by TG-FTIR. Polym Int, 47(3): 247–256.
Holmqvist, A., Wallberg, O. and Jonsson, A.S., 2005, Ultrafiltration
of kraft black liquor from two Swedish pulp mills. Chem Eng
Res Des, 83(A8): 994–999.
Jaeger, P.S., 2005, Market News Service, International Trade Centre
UNCTAD/WTO, Available from http://www.intracen.org/mas.
Jonsson, A.S., Nordin, A.K. and Wallberg, O., 2008, Concentration
and purification of lignin in hardwood kraft pulping liquor by
ultrafiltration and nanofiltration. Chem Eng Res Des, 86(11A):
1271–1280.
Kamm, B. and Kamm, M., 2004, Principles of biorefineries. Appl
Microbiol Biotechnol, 64(2): 137–145.
Kamm, B., Kamm, M., Gruber, P.R. and Kromus, S., 2006, in
Kamm, B., Gruber, P.R., & Kamm, M. (eds) (Wiley-VCH Verlag
GmbH & Co. KGaA, Weinheim), pp. 3–40.
Keyoumu, A., Sjodahl, R., Henriksson, G., Ek, M., Gellerstedt, G.
and Lindstrom, M.E., 2004, Continuous nano- and
ultra-filtration of kraft pulping black liquor with ceramic
filters - a method for lowering the load on the recovery boiler
while generating valuable side-products. Ind Crops Prod,
20(2): 143–150.
Kirkman, A.G., Gratzl, J.S. and Edwards, L.L., 1986, Kraft lignin
recovery by ultrafiltration: economic feasibility and impact on
the kraft recovery system. Tappi J, 69(5): 110–114.
Kurple, K.R., 2000, U.S. Patent PCT/US96/20140, WO97/24362.
Lebo, S.E., Gargulak, J.D. and Timothy, J.M., (2001). Lignin in
Kirk-Othmer Encyclopedia of Chemical Technology (4th ed.). (John
Wiley & Sons, New York).
Liu, G.L., Liu, Y.S., Ni, J.R., Shi, H.C. and Qian, Y., 2004, Treatability
of kraft spent liquor by microfiltration and ultrafiltration.
Desalination, 160(2): 131–141.
Liu, G.L., Liu, Y.S., Shi, H.C. and Qian, Y., 2004, Application of
inorganic membranes in the alkali recovery process.
Desalination, 169(2): 193–205.
 chemical engineering research and design 8 7 ( 2 0 0 9 ) 1276–1292
Logan, C.D., 1965, Cyclic process for recovering vanillin and
sodium values from lignosulphonic waste liquors by ion
exchange, US Patent 3,197,359, United States.
Lopez-Malo, A., Alzamora, S.M. and Argaiz, A., 1998, Vanillin and
ph synergistic effects on mold growth. J Food Sci, 63(1):
143–146.
Lora, J.H. and Glasser, W.G., 2002, Recent industrial applications
of lign sustainable alternative to nonrenewable materials.
Polym J Environ, 10(1–2): 39–48.
Major, F.W. and Nicolle, F.M.A., 1977, Vanillin recovery process, US
Patent 4,021,493, United States.
Makin, E.C., 1984, Purification of vanillin, US Patent 4,474,994,
United States.
Mathias, A.L., 1993, Produção de vanilina a partir da lenhina:
estudo cinético e do processo, Ph.D. thesis, University of Porto,
Porto, Portugal.
Mathias, A.L., Lopretti, M.I. and Rodrigues, A.E., 1995, Chemical
and biological oxidation of pinus-pinaster lignin for the
production of vanillin. J Chem Tech Biotechnol, 64(3): 225–234.
Mathias, A.L. and Rodrigues, A.E., 1995, Production of vanillin by
oxidation of pine kraft lignins with oxygen. Holzforschung,
49(3): 273–278.
Nadji, H., Bruzzese, C., Belgacem, M.N., Benaboura, A. and
Gandini, A., 2005, Oxypropylation of lignins and preparation
of rigid polyurethane foams from the ensuing polyols.
Macromol Mater Eng, 290(10): 1009–1016.
Nair, G.K., 2005, Vanilla prices fall on demand, rising output,
Business Line—Financial Daily from THE HINDU group of
publications; Internet Edition, Available from: http://www.
blonnet.com/2005/06/14/stories/2005061401521000.htm.
Ness, J.H.V., 1983, in Standen, A. (ed) (p. 704).
Ni, P. and Thring, R.W., 2003, Synthesis of polyurethanes from
solvolysis lignin using a polymerization catalyst: mechanical
and thermal properties. Int J Polym Mater, 52: 685–707.
Norgren, M., 2001, On the physical chemistry of Kraft lignin:
Fundamental and applications, Ph.D. thesis, Lund University,
Sweden.
Norgren, M. and Edlund, H., 2003, Ion specific differences in salt
induced precipitation of kraft lignin. Nord Pulp Pap Res J,
18(4): 400–403.
Öhman, F., Wallmo, H. and Theliander, H., 2006, An improved
method for washing lignin precipitated from Kraft black
liquor – The key to a new bio-fuel
(http://www.lignoboost.com). In: 5th Conference on Chemical
Industry and Environment, Vienna, Austria.
Öhman, F., Wallmo, H. and Theliander, H., 2007, An improved
method for washing lignin precipitated from kraft black
liquor—the key to a new biofuel. Filtration, 7(4): 309–315.
Öhman, F., Wallmo, H. and Theliander, H., 2007, Precipitation and
filtration of lignin from black liquors of different origin. Nord
Pulp Pap Res J, 22(2): 188–193.
Olsson, M.R., Axelsson, E. and Berntsson, T., 2006, Exporting
lignin or power from heat-integrated kraft pulp mills:
techno-economic comparison using model mills. Nord Pulp
Pap Res J, 21(4): 476–484.
Pearl, I.A., 1942, Vanillin from lignin materials. J Am Chem Soc,
64: 1429–1431.
Rase, H.F., (1977). Chemical Reactor Design for Process Plants. Volume
One: Principles and Techniques. (John Wiley & Sons, New York).
Ree, R.V. and Annevelink, E., 2007, Status Report Biorefinery 2007
(Agrotechnology and Food Sciences Group), Available from:
http://www.biorefinery.nl/fileadmin/biorefinery/docs/
publications/StatusDocumentBiorefinery2007final211107.pdf.
Saake, B. and Lehnen, R., 2004, Lignin, in: Ullmans’s Encyclopedia
of Industrial Chemistry, 7th Edition (Internet version 2003,
CD-ROM 2004), Karin Sora (ed.) (Wiley-VCH Verlag GmbH &
Co. KGaA, Weinheim, Germany).
Sales, F.G., Abreu, C.A.M. and Pereira, J., 2002, Catalytic wet-air
oxidation of lignin in a three-phase reactor with aromatic
aldehyde production, In 14th Brazilian Congress of Chemical
Engineering Natal, BRAZIL,
Sandborn, L.R., Salvesen, J.R. and Howard, G.C., 1936, Process of
making vanillin, US Patent 2,057,117, United States.
1291
Saraf, V.P., Glasser, W.G. and Wilkes, G.L., 1985, Engineering
plastics from lignin. 7. Structure property relationships of
poly(butadiene glycol)-containing polyurethane networks. J
Appl Polym Sci, 30(9): 3809–3823.
Sarkar, S. and Adhikari, B., 2001, Synthesis and characterization
of lignin-htpb copolyurethane. Eur Polym J, 37(7): 1391–1401.
Schoeffel, E.W., 1962, Vanillin purification, US Patent 3,049,566,
United States.
Schultz, T.P. and Templeton, M.C., 1986, Proposed mechanism for
the nitrobenzene oxidation of lignin. Holzforschung, 40(2):
93–97.
Siminiceanu, I., Friedl, A. and Drăgan, M., 2007, A simple equation
for the effective mass transfer area of the Mellapak 750Y
structured packing, Available from: http://www.upg-ploiesti.
ro/sescom/pdf/s09/s09-l21-is.pdf.
Sixta, H., 2006, in Sixta, H. (ed). (WILEY-VCH Verlag GmbH & Co.
KGaA, Weinheim)
Sridhar, P., Araujo, J.D. and Rodrigues, A.E., 2005, Modeling of
vanillin production in a structured bubble column reactor, In
2nd International Conference on Structured Catalysts and Reactors
(ICOSCAR-2) Delft, Netherlands,
Sun, R.C., Tomkinson, J. and Bolton, J., 1999, Effects of
precipitation ph on the physico-chemical properties of the
lignins isolated from the black liquor of oil palm empty fruit
bunch fibre pulping. Polym Degrad Stab, 63(2): 195–200.
Tan, T.T.M., 1996, Cardanol-lignin-based polyurethanes. Polym
Int, 41(1): 13–16.
Tanistra, I. and Bodzek, M., 1998, Preparation of high-purity
sulfate lignin from spent black liquor using ultrafiltration and
diafiltration processes. Desalination, 115(1): 111–120.
Tarabanko, V.E., Fomova, N.A., Kuznetsov, B.N., Ivanchenko, N.M.
and Kudryashev, A.V., 1995, On the mechanism of vanillin
formation in the catalytic-oxidation of lignin with oxygen.
React Kinet Catal Lett, 55(1): 161–170.
Tarabanko, V.E., Koropatchinskaya, N.V., Kudryashev, A.V. and
Kuznetsov, B.N., 1995, Influence of lignin origin on the
efficiency of the catalytic-oxidation of lignin into vanillin and
syringaldehyde. Russ Chem Bull, 44(2): 367–371.
Tarabanko, V.E. and Petrukhov, D., 2003, Study of mechanism and
improvement of the process of oxidative cleavage of lignins
into the aromatic aldehydes. Chem Sust Dev, 11:
655–667.
Tarabanko, V.E., Petukhov, D.V. and Selyutin, G.E., 2004, New
mechanism for the catalytic oxidation of lignin to vanillin.
Kinet Catal, 45(4): 569–577.
Thielemans, W., Can, E., Morye, S.S. and Wool, R.P., 2002, Novel
applications of lignin in composite materials. J Appl Polym
Sci, 83(2): 323–331.
Thring, R.W., Ni, P. and Ahoroni, S.M., 2004, Molecular weight
effects of the soft segment on the ultimate properties of
lignin-derived polyurethanes. Int J Polym Mater, 53:
507–524.
Triumph Venture Capital Ltd., 2004b, Part 3—Aroma chemicals
from petrochemical feedstocks, in: Study into the
establishment of an aroma and fragrance fine chemicals value
chain in South Africa (Final Report), FRIDGE, NEDLAC, South
Africa, Available from: http://www.nedlac.org.za.
Triumph Venture Capital Ltd., 2004a, Part 2—Aroma chemicals
derived from effluent from the paper and pulp industry, in:
Study into the establishment of an aroma and fragrance fine
chemicals value chain in South Africa (Final Report), FRIDGE,
NEDLAC, South Africa, Available from:
http://www.nedlac.org.za.
Tsapiuk, E.A., Bryk, M.T., Medvedev, M.I. and Kochkodan, V.M.,
1989, Fractionation and concentration of lignosulfonates by
ultrafiltration. J Membr Sci, 47(1–2): 107–130.
Vanderlaan, M.N. and Thring, R.W., 1998, Polyurethanes from
alcell (r) lignin fractions obtained by sequential solvent
extraction. Biomass Bioenergy, 14(5–6): 525–531.
Villar, J.C., Caperos, A. and GarciaOchoa, F., 1997, Oxidation of
hardwood kraft-lignin to phenolic derivatives. Nitrobenzene
and copper oxide as oxidants. J Wood Chem Technol, 17(3):
259–285.
1292 chemical engineering research and design 8 7 ( 2 0 0 9 ) 1276–1292
Wallberg, O., Holmqvist, A. and Jonsson, A.S., 2005, Ultrafiltration
of kraft cooking liquors from a continuous cooking process.
Desalination, 180(1–3): 109–118.
Wallberg, O., Linde, M. and Jonsson, A.S., 2006, Extraction of
lignin and hemicelluloses from kraft black liquor.
Desalination, 199(1–3): 413–414.
Walton, N.J., Mayer, M.J. and Narbad, A., 2003, Molecules of
interest—vanillin. Phytochemistry, 63(5): 505–515.
Woerner, D.L. and McCarthy, J.L., 1984, Ultrafiltration of kraft
black liquor. AIChE Symp Series, 80(232): 25–33.
Yoshida, H., Mörck, R. and Kringstad, K.P., 1986, Patent
PCT/SE86/00237, WO 86/07070.
Yoshida, H., Morck, R., Kringstad, K.P. and Hatakeyama, H., 1990,
Kraft lignin in polyurethanes. 2. Effects of the
molecular-weight of kraft lignin on the properties of
polyurethanes from a kraft lignin polyether triol polymeric
MDI system. J Appl Polym Sci, 40(11–12): 1819–1832.
Zabkova, M., da Silva, E.A.B. and Rodrigues, A.E., 2007, Recovery
of vanillin from kraft lignin oxidation by ion-exchange with
neutralization. Sep Purif Technol, 55(1): 56–68.
Zabkova, M., da Silva, E.A.B. and Rodrigues, A.E., 2007, Recovery
of vanillin from lignin/vanillin mixture by using tubular
ceramic ultrafiltration membranes. J Membr Sci, 301(1–2):
221–237.
Zabkova, M., Otero, M., Minceva, M., Zabka, M. and Rodrigues,
A.E., 2006, Separation of synthetic vanillin at different ph onto
polymeric adsorbent sephabeads sp206. Chem Eng Prog, 45(7):
598–607.