Influence of salts during hydrothermal biomass gasification: The role of the catalysed water-gas shift reaction

Article abstract of DOI:10.1524/zpch.219.3.341.59177

Near-critical and supercritical water is an unusual reaction medium with extraordinary properties, varying with temperature and density. This opens the opportunity for new reactions and new technical processes. In the studies of synthesis reactions, total oxidation and biomass gasification interesting effects of salts are observed. These effects are assumed to be caused by complex formation, acidity/basicity or the presence of active hydrogen due to the water-gas shift reaction. The water-gas shift reaction is catalysed by salts and lead to the formation of hydrogen, which might react with other compounds. This article focuses on the salt effects during hydrogen production by biomass gasification in supercritical water. This is a very interesting process to make use of "wet biomass", which is up to now not applied in a technical process. Here salts influenced the main reactions pathways. Possible reasons are discussed. by Oldenbourg Wissenschaftsverlag, Muenchen.

Full text of DOI:10.1524/zpch.219.3.341.59177

Z. Phys. Chem. 219 (2005) 341–366
by Oldenbourg Wissenschaftsverlag, München
Influence of Salts During Hydrothermal
Biomass Gasification: The Role of the
Catalysed Water-Gas Shift Reaction
∗
By Andrea Kruse and Eckhard Dinjus
Institut für Technische Chemie, Bereich Chemisch-Physikalische Verfahren,
Forschungszentrum Karlsruhe, P.O. Box 3640, D-76021 Karlsruhe, Germany
Dedicated to Prof. Dr. Michael Buback on the occasion of his 60th birthday
(Received December 7, 2004; accepted December 15, 2004)
Biomass Gasification / Salt Effects / Supercritical Water /
Water-Gas Shift Reaction
Near-critical and supercritical water is an unusual reaction medium with extraordinary
properties, varying with temperature and density. This opens the opportunity for new
reactions and new technical processes. In the studies of synthesis reactions, total oxidation
and biomass gasification interesting effects of salts are observed. These effects are assumed
to be caused by complex formation, acidity basicity or the presence of active hydrogen
due to the water-gas shift reaction. The water-gas shift reaction is catalysed by salts and
lead to the formation of hydrogen, which might react with other compounds.
This article focuses on the salt effects during hydrogen production by biomass
gasification in supercritical water. This is a very interesting process to make use of “wet
biomass”, which is up to now not applied in a technical process. Here salts influenced
the main reactions pathways. Possible reasons are discussed.
1. Introduction
Near- and supercritical water is a unique and interesting reaction medium. The
unusual properties enable various applications like chemical synthesis, destruc-
tion of hazardous wastes and biomass conversion. During the studies of this
application sometimes unexpected salt effects were observed. Unexpected here
often means different from what is known from reactions in ambient water. The
reason of cause is, that aqueous solutions at high temperatures are different
from aqueous solutions at ambient conditions.
*
Corresponding author. E-mail: andrea.kruse@itc-cpv.fzk.de

342
A. Kruse and E. Dinjus
Here a short overview about the properties of water and aqueous solution
at high pressure and observed salt effects on chemical reactions are given. The
focus of this article is the discussion of salt effects during biomass gasification
in supercritical water.
The properties of water will be discussed from a macroscopic and a mi-
croscopic point of view. It has to be pointed out, that supercritical water
(SCW) has the macroscopic properties like a non-polar solvent but the single
molecules are still polar. This is only apparent a contradiction – and has re-
cent consequences for the structure of aqueous solution and of cause chemical
reaction. Some reactions might take benefit of these special properties of SCW.
A lot of reactions investigated in near-critical and supercritical water are of
technical interest. The synthesis reactions e.g. opens the opportunity to reach
high selectivity with reduced separation or cleaning efforts. For example a pro-
cess with an acid catalyzed reaction, proceeding without the addition of acids
because of the high ionic product of near-critical water, means less cleaning ef-
forts of the product and less waste by acids neutralization. Total oxidation in
supercritical water is an already well-developed process, in which hazardous
substances can be destroyed completely and fast. Here the high solubility of
organic compounds and gases avoids reaction rate limitation by transportation
through phase boundaries. By gasification of wet biomass in supercritical water
hydrogen is produced as a main product by hydrothermal reaction of biomass
with water. Not yet used wet biomass can thus contribute to energy generation
or the production of e.g. synthetic fuels.
1.1 Properties of water and aqueous solutions at high temperatures
and pressures
Water is an ecologically safe substance widespread throughout nature. Be-
low the critical point, the vapor pressure curve separates the liquid and
gaseous phase. The vapor pressure ends at the critical point Tc = 374 ◦C pc =
22 1 MPa, c = 320 kg m3 . Beyond the critical point, the density of the SCW
can be varied continuously from liquid-like to gas-like values without phase
transition [1–4].
Liquid water at standard conditions T = 25 ◦C p = 0 1 MPa is poorly
miscible with hydrocarbons and gases. In contrast, it is a good solvent for
salts because of its high relative dielectric constant of 78.5 at a high dens-
ity of 997 kg m3 [5]. At nearly critical temperature and pressure the relative
dielectric constant is in the range of 10 [5, 6] (Fig. 1), this is nearly the rela-
tive dielectric constant of e.g. methylene chloride at standard conditions, and it
further decreases with rising temperature. Therefore, SCW at low densities be-
comes a poor solvent for ionic species like inorganic salts. On the other hand,
SCW is completely miscible with many organic compounds and gases (see
e.g. [6–8]). As a consequence of the strong dependence of the solvent power on
temperature and density of water (Fig. 1), changes of temperature and pressures

Influence of Salts During Hydrothermal Biomass Gasification:
343
Fig. 1. Selected properties of water at high temperature and high pressure (data taken
from [5]).
near the critical point can be used to precipitate particles of a special struc-
ture or size (see e.g. [9–11]). The complete miscibility of supercritical water
and gases as well as with organic compounds makes SCW to be an excellent
solvent for homogenous reactions of organic compounds with gases, like the
oxidation of organic compounds with oxygen and air. The absence of phase
boundaries leads to a fast and complete reaction there (see e.g. [12–16]).
The good solubility of non-polar substances in SCW is at least a conse-
quence of the high temperature leading to a dominance of the entropy influ-
ence [17]. In water below the critical temperature often no complete miscibility
but an increase solubility of organic compounds is found [18].
Supercritical water is not only a solvent; it is a solvent at high pressure.
Some effects on reactions found in supercritical water are caused by the high
pressure only. Examples are reactions of free radicals: The reaction rates of
small free radicals are increased by the enhanced energy equilibration rate due
to the high collision frequency [19]. This is very important during the oxidation
in supercritical water because all reactions involving OH and HO2 free radicals
are accelerated. OH and HO2 free radicals are central free radicals of the oxi-
dation chain reaction (see e.g. [20]), therefore this enhancement leads to high
reaction rates.
The ionic product increases slightly with temperature up to around 10−11
in the range between 200–300 ◦C [4, 5] (Fig. 1). Above the critical tempera-
ture the ionic product decreases drastically with temperature but increases with
pressure. The ionic product in sub-critical water and in supercritical water at
high pressures can be some orders of magnitude higher than in ambient water
(Fig. 1). In these regions water may play the role of an acid or base catalyst
because of the rather high concentration of H3O+ and OH− ions.
It has to be mentioned here, that the high ionic product is not the only
reason for the appearance of acid or base catalysis in high temperature water.

344
A. Kruse and E. Dinjus
Recent studies [21] show, that the chemical kinetics of some acid or base cat-
alyzed reactions are less sensitive to changes of the ionic product by pressure
and temperature than expected. Here the water itself, may play the role of a pro-
ton donator or acceptor.
In most cases ([22, 23]) the acid or base catalyzed reactions in water at
high pressures and high temperatures show a characteristic non-Arrhenius ki-
netic behavior near the critical point of water: Below the critical temperature of
water the reaction rates usually increase with temperature until the critical tem-
perature is reached. At the critical point the reaction rate decreases drastically.
Diffusion rates are high and viscosity is low in a supercritical aqueous
mixture. Transport properties and miscibility are important parameters, which
influence the rate of chemical reactions. High diffusion rates and low vis-
cosity, together with the complete miscibility with many substances, make
supercritical water to an excellent medium for homogeneous, fast, and efficient
reactions. In addition SCW is an excellent reaction medium with heteroge-
neous catalysts, because the high diffusion rate avoids mass transfer limitation
and the good solubility prevents coke formation on or poising of the catalyst.
The low relative dielectric constant of SCW (Fig. 1) corresponds to a low
solubility of salts, which is three or more order of magnitudes lower than in
ambient water (see e.g. [24]).
Water is also at high temperatures and pressures a polar molecule1, which is
able to interact especially with polar species. This means that water molecules
e.g. can take part in an activated complex, which is assumed for some important
reaction steps in total oxidation in SCW. By forming a complex the activation
energy is lowered, which means that water is a catalyst for these reactions. And
water is still able to react. E.g. in hydrolysis reactions or in the water gas shift
reaction water reacts with the organic compound and carbon monoxide, re-
spectively. The high compressibility of SCW from a macroscopic view means,
that water is structured to an increased extent by a solute compared with am-
bient water. This means for example that the influence on activated complexes
might be greater than predicted by classic thermodynamic approaches using the
dielectric constant.
Especially for very fast reactions, like free radical reactions, dynamic as-
pects are important like solute–solvent collisions, diffusion limitations, and
cage effects.
What is the structure of water at high temperatures and pressures?
Both measurements and computer simulation shows that the pair cor-
relation functions changes significantly with temperature and pressure (see
e.g. [25–30]). In simple words the change with increasing temperature cor-
responds to the loss of tetrahedral coordination found in ambient water. With
decreasing density water as medium becomes more and more like a normal
1
Indeed, while the dipole moment of water molecule is 1 85 D in the dilute gas
phase, it is considered to be about 2 5 D in the ambient liquid phase due to intramolecular
polarization ([17] and citation there).

Influence of Salts During Hydrothermal Biomass Gasification:
345
gas with isolated molecules due to the increased repulsive interaction [25].
On average, water molecules are bonded to each other almost half as strongly
under supercritical conditions compared with ambient water [25]. On the other
hand water at high temperatures and pressures shows a large tendency to form
clusters, therefore these on average effects give no complete view on the struc-
ture of water.
The amount of hydrogen bonds as function of temperature and density was
the subject of a lot of discussions over the years. Often it is mentioned in art-
icles that the number of hydrogen bonds is much smaller than in ambient water
and vanishes with increasing temperature and decreasing pressure. A more pre-
cise formulation might be that in water at high temperatures and pressures
the average number of water molecules connected by hydrogen bonds to form
clusters and the lifetime of a single hydrogen bond decreases with increas-
ing temperature and decreasing density [31]. Also at very low density because
of thermodynamic stability constraints bonded and unbonded water molecules
must exist together [25]. The existence of hydrogen bonds in supercritical wa-
ter at low densities is confirmed by measurements e.g. NMR [32, 33] and rapid
X-ray diffraction [34] and newer neutron scattering results [35]. Molecular
simulations have shown that although at high temperatures like 500 ◦C or more
the average number of molecules connected by hydrogen bonds are five or
less; a small number of larger clusters of more than twenty molecules also
exist. These “difference of local density” or density heterogeneities are very ex-
treme of water near the critical point. This “glass-like” behavior with relative
slow fluctuation of these inhomogeneities may have an influence on very fast
dynamic processes by the effect of “critical slowing down” [36, 37] because
a high density environment is “frozen” relative to very fast dynamic processes
of solutes like relaxing in spectroscopic measurements. Anyway the effect is
small and therefore this would be difficult to measure.
This changes in respect of hydrogen bonds leads to an increased “mobility”
of single water molecules. The barrier for translation and rotational motions is
therefore reduced, which leads to an increased self-diffusivity [38] and influ-
ences the energy transfer between very fast reacting solutes and solvent.
Molecular simulations and measurements [26, 39–44] show that the sol-
vent density about an ion remains high in the vicinity of the ions. The primary
solvation shell remains largely intact, but for mono-valent ions the density
of solvent molecules beyond the primary solvation shell is weaklier corre-
lated with the presence of the ion. The second shell of di-valent cations
persists to high temperatures, although it is not as prominent as at ambi-
ent conditions (e.g. [45]). This means that around ions the “local density”
is higher than the bulk density. For mono-valent cations (here Na+, K+ and
Rb+, at water densities between 0 29–0 087 g cm−3 [45]) the local density is
calculated to be a factor 9–20 higher than the bulk density. The calculation
leads for bi-valent cations (here Ca2+ and Sr2+, at water densities between
0 29–0 087 g cm−3 [45]) to density, which is increase by a factor of 30–100.

346
A. Kruse and E. Dinjus
And this means that the “local relative dielectric constant” is also higher than
the bulk relative dielectric constant, which has to be considered for calculation
of solvation energies. Up to now, a direct characterization of a solvations shell
was only achieved by XAFS for Rb+ (at 400 ◦C [46]), Ag+ (at 350 ◦C [47]) and
Sr2+ (bis 300 ◦C [48]). The results of the simulations are mainly confirmed by
the experimental data, also some difference exists (see [45]). Also conductivity
measurements (e.g. [49]) hints to a strong solvation shell. In contrast to ambient
water, salts like NaCl partially are solved as ion pairs, or even higher clusters.
This is calculated by simulations [44, 50–52] and confirmed by conductance
measurements (e.g. [49]). The degree of dissociation rises with decreasing con-
centration and increasing density.
Also polar molecules, in particular if the formation of hydrogen bonds is
possible leads to a local density enhancement, as e.g. measured by UV absorp-
tion solvatochromic shift [53, 54].
The local density around an un-polar compound like a noble gas or me-
thane is rather low or “gas-like” [17, 27].
1.2 Acidity and basicity in water at high temperatures
The acidity and basicity behavior of salts in near- and supercritical water is
very complex. In general the dissociation of at ambient conditions weak acids
like acetic acid or water is endothermic and the dissociation of strong acids like
HCl or HNO3 is exothermic. These differences between “weak” and “strong”
acids vanish with increasing temperature. In the order of acidity two acids may
indeed invert positions [55]. As an example, the dissociation constant of HCl
decreases from ambient to the critical temperature of water by thirteen orders
of magnitude [55]. Using data from simulations and thermodynamics, the rea-
son for this behavior is [56]: In case of anions only the first solvations shell
weakly persevered (confirmed by neutron scattering results [57]), therefore the
differences in solvation energy of ions with different ion diameters, here OH−
and Cl−, vanishes and approximates the gas phase value. As a consequence
the acidities of water and HCl become alike and HCl becomes a much weaker
acid in hot compressed water, respectively. Also in the case of sulfuric acid the
acidity decreases with temperature. This decrease is in particular strong in the
supercritical region. Sulfuric acid (in regard of the first dissociation constant) is
a weaker acid than HCl in the sub-critical region, but in the supercritical region
their acidity is very similar and weak [58]. The second dissociation constant
is very low at high temperatures [59]. The behavior of bases is different. The
dissociation of KOH and other hydroxides shows a very low dependence on
temperature in the sub-critical region and decreases very much at the critical
point of water [55]. Likely the reason is, that the interaction of cations with the
oxygen atoms of the surrounding water molecules is very strong, stronger than
the interaction between anions with the hydrogen atoms of water, and there-
fore the solvation energies are nearly independent in the temperature region of

Influence of Salts During Hydrothermal Biomass Gasification:
347
interest here [45]. In the supercritical region, there are significant differences
in regard of the relation of dissociation constants of different alkali hydroxides
and halides found [55]. These differences of the dependencies of the strength of
bases and acids have an interesting consequence on the properties of salts. For
example NaCl is a base in the near critical region below the critical point [60],
and in the supercritical region at higher concentrations and lower densities [55].
Because of the very low second dissociation constant (and the relative low fist
dissociation constant) e.g. Na2SO4 is a strong base. Bi-valent cations behave
different, because they become strong cation acids, at conditions of interest
here. This is a consequence of the very strong interaction between bivalent
cations and hydration shell. For example at 400 ◦C Mg2+ (data from [61]) has
a similar acidity like sulfuric acid (data from [58]). Therefore salts of bi-valent
cations are much more acidic, or less basic than the corresponding salts of
mono-valent cations.
In the studies of biomass gasification in supercritical water (see Sect. 1.6)
with model compounds, KHCO3 or K2CO3 are added as a substitute for natural
salts in biomass. Here it is very difficult to predict the excess of H+ or OH−,
because of two main reasons:
•
Organic acids and CO2, which are strong acids at near- and supercritical
water conditions, are formed during the reaction.
•
There are only very few data available. There are some data measured in
regard of geothermal processes, but these data are obtained at much higher
pressures (e.g. [62, 63]).
In regard of the acidic or basic influences on the gasification of biomass or
model compounds, two facts are of special importance:
•
CO2 becomes a much stronger acid, therefore its salts are less basic.
K2CO3 has a relative low dissociation constant [62], which means it is less
basic.
•
2−
−
In summary it can be assumed, that the CO3 HCO3 CO2-system is a buffer
in the highly acid region, also if high amounts of KHCO3 or K2CO3 are added.
1.3 Salt effects in synthesis reactions
Investigations of reaction of lactic acid in the temperature range 320–400 ◦C
[64, 65] show a significance of the nature of salts.
Lactic acid can react via three reaction pathways (see Fig. 2):
1. Decarbonylation: Formation of CO, H2O and acetaldehyde. This reaction
pathway was assumed to be catalysed by acids, because the addition of
sulfuric acid increases the yields of these products [65].
2. Decarboxylation: Formation of CO2, H2 and acetaldehyde. This was as-
sumed to be a homolytic reaction [65].
3. Water elimination: Formation of acrylic acid. This reaction was assumed to
process via an cyclic intermediate, -propiolactone [65].

348
A. Kruse and E. Dinjus
Fig. 2. Reaction paths of lactic acid at near- and supercritical water ([65]).
Fig. 3. Assumed cyclic intermediate complex during decarboxylation of malonic acid [66].
The addition of NaCl (1 mol L) leads to an increase of the reaction rate of all
three reactions [65]. In the paper [65], no reason for the effect of NaCl is dis-
cussed. This is likely not a consequence of the basic action of NaCl mentioned
above at these conditions because bases are found to decrease the rates of reac-
tion pathways I and II at low concentrations [64] and a decrease of the rate of
reaction pathway III at higher base concentrations [65].
Studying the decarboxylation of malonic acid not only the influence of
pH value but also the effect of the cation (here Li+, Na+, K+, Rb+, Cs+,
Mg2+, Ca2+ and Sr2+ [66]) is measured by the CO2 formation rate, determined
by IR-spectroscopy. In general the rate of decarboxylation decreases with the
charge density but Mg2+ is an exception. The dependence on charge density
was explained by an increased substitution share of the cation in the cyclic in-
termediate complex (Fig. 3). Here it takes the position of the carboxyl group
hydrogen. The decarboxylation with Mg2+ as cation is much faster than ex-
pected likely due to the dominance of a new reaction pathway because of com-
plex formation [66]. In addition the decarboxylation of other organic acids and
its dependence on alkali cations are investigated in the same equipment [67].
In the case of trifluoracetic acid the abstraction of fluoride was dependent on
the nature of the cation and decrease in the direction Li+ Na+ K+ [67].

Influence of Salts During Hydrothermal Biomass Gasification:
349
The decarboxylation of cyanoacetic acid was found to be independent of the
nature of the alkali ion. Lithium propiolate (HCCOOLi) shows a lower de-
carboxylation rate than the other propiolates. Here complexation by Li+ was
assumed [67].
The elimination of water from alcohols (here different propanols and bu-
tanols [68]) shows an interesting effects of the salts (Na2SO4, MgSO4, or
ZnSO4) added. In general and simplified it can be concluded, that Na2SO4
decreases and MgSO4, and ZnSO4, respectively increases the rate of water
elimination [68]. The water elimination is assumed to be an acid catalyzed re-
action (see e.g. [69, 70]), therefore Na2SO4, as a strong base (see above), would
reduce the reaction rate. Bi-valent cations are different (see above) and become
cation acids, therefore MgSO4 and ZnSO4 may be more or less neutral. On the
other hand, Mg2+ and Zn2+ may perhaps form a complex, which may lead to
an activated complex of lower energy, which e.g. is part of an other reaction
pathway, corresponding to a higher reaction rate. This assumption is alike to
the assumption made for the decarboxylation of malonic acid (see above).
An interesting effect of NaCl addition is found for the hydrolysis of
diphenyl ether: At 430 ◦C and 0 46 g L water density (around 48 MPa [71]).
Here the hydrolysis rate at low salt concentration decreases and at higher in-
creases up to the double, compared with measurements without salts. In the
paper [71] is assumed, that at low NaCl concentrations HCl is formed, corres-
ponding to the basic character of NaCl at these conditions. At higher concentra-
tions the properties as cation acid and anion base, respectively are supposed to
be dominant, leading to an increased concentration of H+aq and OHa−q. This hy-
drolysis is expected to be acidic catalyst [72], therefore the reaction rate would
increase.
In the studies of the hydrolysis of dibenzylether to benzylalcohol (with
minor yields of other, here pyrolysis products) and the hydrolysis of ben-
zylphenylamine mainly to benzylalcohol and aniline at lower temperature
(380 ◦C, 21 6 MPa, 10–60 min [73]) the opposite effect was found: The add-
ition of salts to the reaction mixture increased the hydrolysis rate at low salt
concentration, while having no effect on a back ground pyrolysis reaction rate.
At higher salt concentrations, the hydrolysis rate reached a maximum then
decreased, approaching the rate observed in the absence of salts [73]. The
authors indicate from this, that the reaction proceeds via a SN2-mechanism
with a nucleophilic attack of water. Here the activated complex is more po-
lar than the educts, therefore the additions of salts, corresponding with an
increase of ionic strength and relative dielectric constant, would decrease
the energy of the activated complex more than the energy of the educts.
That means, that the reaction rate increases. The decrease in reaction rate
at high salt concentration was assumed to be a concentration effect; on the
macroscopic or microscopic scale. The first explanation is, that at higher salt
concentration phase separation occurs. Hydrolysis would occur in the “or-
ganic” phase, in which the water concentration is low and therefore also the

350
A. Kruse and E. Dinjus
Fig. 4. Mechanisms of ethyl acetate hydrolysis [22].
hydrolysis rate [73]. The second, microscopic explanation given is, that at
high salt concentration to much water is bonded in hydration shells, which
leads to a decreased concentration of “free” water and decreased hydrolysis
rate [73].
The hydrolysis of ethylacetate is studied in detail [22, 74] and it was con-
cluded, that the mechanism changes, when the temperature is increased from
sub- to supercritical. In the sub-critical range an AAc2-mechanism is assumed
(Fig. 4). Here the ester is first protonated and then water is added. The add-
ition of water is assumed to be the rate-limiting step. In the supercritical region
a SN2-mechanism with a nucleophilic attack and substitution of alcohol by
water is supposed (Fig. 4). This idea is in accordance with the unusual shape
of the Arrhenius-plot, with a maximum in the sub-critical region and a mini-
mum near the critical temperature of water [68]. In experiments with addition
of Na2SO4 and ZnSO4 the maximum and the minimum vanishes and the con-

Influence of Salts During Hydrothermal Biomass Gasification:
351
version increases with the reaction temperature [68]. Below 340 ◦C Na2SO4
addition decreases and above it increases the conversion [68]. ZnSO4 add-
ition leads to an increase of the reaction rate in the hole temperature range
investigated [68]. The AAc2 is an acidic catalyzed reaction and the presence
of a base leads to a decrease in the reaction rate. One explanation would be,
that the addition of Na2SO4 leads to a discrimination of the AAc2-mechanism,
because of its basic properties and the increase of ionic strength and relative
dielectric constant. In the SN2-mechanism the activated complex is more po-
lar than educts, therefore this reaction is accelerated by a higher dielectric
constant in the presence of a salt. This would explain, that in the presence
of Na2SO4, the hydrolysis could only occur via the SN2-mechanism; with
lower reaction rate in the high density, low temperature region and with an
increased reaction rate in the low density region. In the case of the ZnSO4 add-
ition, perhaps complex formation is the reason for the increased reaction rate.
A coordination of the oxygen atoms around the Zn-cation leads to a decrease
electron density at the carbon atom of the ester functional group. This would
make the nucleophilic attack easier and the energy of the activated complex
lower.
To conclude, the influence of salts can be strong and may be based on dif-
ferent effects. Basicity of salts, which are neutral at ambient conditions and
complex formation has to be taken in account. Concentration effects are more
or less not understood.
1.4 Salt effects in the water-gas shift reaction and supercritical water
oxidation
The so-called “un-catalyzed” water-gas shift reaction is assumed to proceed
via formic acid as intermediate. (A detailed description about studies about the
“un-catalyzed” water-gas shift reaction and the difficulties to measure it are
presented in [20]).
Up to 400 ◦C investigations about the catalytic effect of alkali, earth-alkali
and transition metal carbonates exist [75]. Alkali carbonates were found to be
most and earth-alkali carbonates less effective. The catalytic effect of transi-
tion metal carbonates ranges from similar to alkali to similar to earth-alkali.
A mechanism was assumed, in which formaldehyde, formate and OH− are
intermediates (Fig. 5, [76]). This mechanism is not sufficient to explain dif-
ferences in the properties of Na-salts like benzoate, maleate, oxalate, formate,
carbonate, hydrogencarbonate, hydroxide and others [75], because there order
of catalytic activity corresponds not with the order of basicity.
An other mechanism [77] assumes the formation of alkali formates as inter-
mediates, like formic acids were assumed to be the intermediate of the uncatal-
ysed reaction before (reaction (1)–(3)). Here the fact is confusing, that formate
was found to be more stable at similar conditions than formic acid [122]. Usu-
ally catalysis cycles have to include a very reactive intermediate.

352
A. Kruse and E. Dinjus
Fig. 5. Postulated mechanism of the alkali catalyzed water-gas shift reaction [75].
Obviously the catalytic effect of salts on the water-gas shift reaction is not
understood (see also [20]).
Second mechanism of the alkali catalyzed water-gas shift reaction [77]:
K2CO3 +H2O → KHCO3 +KOH
KOH+CO → HCOOK
(1)
(2)
(3)
HCOOK+H2O → KHCO3 +H2
New studies [78] of the oxidation of 2-chlorophenol indicates an cata-
lytic effect of cations on supercritical water oxidation. The conversion of
2-chlorophenol to CO2 was accelerated by different cations (Li+, Na+, K+,
Ca2+, Fe2+, Fe3+). This effect increases with increasing charge density. Here
complex formation with an interaction between the cation and the Cl-atom of
chlorophenol was assumed.
1.5 Salt effects in “dry” biomass conversion processes
Also during classical “dry” biomass conversion, e.g. gasification processes ef-
fects of salts are known for a long time, but usually not understood. Generally
even no assumption about the mechanism or reason of these effects is published
together with the experimental results. For example salts increase the rate of
degradation of sugars at low temperature and increase the yield of liquid prod-
ucts during pyrolysis [79]. During the studies to produce selected products by
pyrolysis, it was found that e.g. NaCl addition increases the yield of coke and
glyceraldehyde and decreases the yield of anhydroglucose [79]. A similar ef-
fect was found for ZnCl2 addition; here also the furfurals yield was higher. This
high furfural yields were also achieved in the presence of CoCl2 (see [79] and
literature cited there). During gasification alkali or earth alkali salts were used
to improve the yield and quality of gaseous products (see [79, 80] and litera-
ture cited therein). In some cases the influence of alkali salts was assumed to
be a change in polymerization and depolymerization processes [81, 82], which
may also lead to a higher phenols yield [83].

Influence of Salts During Hydrothermal Biomass Gasification:
353
In the LBL-process (Lawrence Berkeley Laboratory, Berkely, USA),
Na2CO3 and CO is added. The carbonate in aqueous solution acts as a catalyst
for the water-gas shift reaction. The “active hydrogen” (formate? See Fig. 5)
formed, reacts with free radicals and leads to a preference of scission reac-
tions compared with polymerization reactions. Also during steam reforming
of biomass an improvement of gas products yield and quality was found when
alkali salts were added [83].
1.6 Supercritical water gasification
The gasification of biomass in supercritical water (“supercritical water gasifi-
cation – SCWG” or “hydrothermal gasification”) is a new process to produce
hydrogen from wet biomass.
This kind of mainly unused biomass is wet biomass with a water content of
more than 50%, in most cases more than 80%. In the case of residual biomass
from e.g. agriculture and food industry this biomass may cause high negative
costs (disposal costs), which are a benefit for the economics of such a gasifica-
tion process.
These wet biomass can not be used directly in “dry gasification processes”,
based on conventional coal gasification techniques (see e.g. [84]). The reasons
are the high water content and the relative high ash content. The drying of the
biomass and the handling of the high ash content, e.g. causing severe corrosion
problems, leads to additional financial efforts.
The water in the wet biomass plays an important role in the SCWG process:
It is reaction medium and reactant at the same time. With regard to biomass
gasification in supercritical water the complete miscibility of gases and most
of the organic compounds is of special importance. This high solubility avoids
the formation of unwanted by-products like tar and coke by dissolution of the
precursors. This effect is supported by the high reactivity of water at this condi-
tions leading to a very fast decomposition of the polymeric structure of biomass
by fast hydrolysis of cellulose to sugar units. The high reactivity of water leads
also to a high hydrogen yield. In fact, nearly half of the hydrogen formed orig-
inates from water. At best, for one mol glucose 12 mol of hydrogen can be
obtained:
C6H12O6 +6H2O → 6CO2 +12H2
(4)
This equation gives the maximum hydrogen yield stoichiometrically possible,
in the experiments usually, depending on reaction conditions and feedstock
minor amounts of methane are formed, which leads to lower H2-yields.
The main advantages of the wet biomass gasification in supercritical water
are:
•
Less tar and coke is formed as a consequence of the special properties of
supercritical water.

354
A. Kruse and E. Dinjus
•
High yields of hydrogen and very low CO yields are obtained. Usually the
CO content in the product gas is below 1 vol %. In “dry processes” synthe-
sis gas with usually CO as the main product is formed, which has to pass
an additional water-gas-shift process to obtain such high H2 and low CO
yields.
•
•
H2 is produced at high pressure: therefore no energy for compression e.g.
to transport and store H2 is necessary.
The solubility of CO2 at ambient temperatures and high pressures is much
higher than the solubility of H2. Therefore H2 and CO2 can easily be sepa-
rated in the process.
•
•
•
The high density due to the high pressure and the high reactivity of biomass
in supercritical water leads to a high space-time-yield.
No drying of wet biomass prior to the process is necessary; on the contrary
the water in wet biomass is essential for the chemical reaction.
Inorganic ingredients, like salts remain in the aqueous solution. Corrosion
problems during gas treating due to salts in the gas flow are avoided.
In the last two decades the biomass gasification in supercritical water was in-
tensively investigated. After first studies in the USA [85–90], now the main
activities are located in Europe [91–94] and Japan [95–99]. Most of these stud-
ies were conducted with biomass model compounds like glucose (e.g. [95, 97]),
but also investigations with real biomass exist (e.g. [86, 87, 92, 100]). The in-
fluence of reaction parameters and different ingredients of biomass as well as
that of different catalyst (e.g. [93, 97, 101, 102]) has been investigated. Not all
of these studies focus on the production of H2. At lower temperatures CH4 or
bio-oil are the main products (e.g. [101–104]).
Based on the promising lab-scale results, a process development unit was
constructed at the Research Center Karlsruhe. In the Institute for Technical
Chemistry CPV a plant with a throughput of 100 kg h, which can be oper-
ated up to 700 ◦C and 35 MPa, was built up and is in operation [105]. This
plant consists of components, which can be scaled-up. The first results of the
plant are very promising. Thus, a technical and economical feasible industrial
application can be expected.
Here a review of studies concerning salt effects is given. The experiments
were conducted in three different reactors and the results are composed in the
view of salt effects and the role of the water-gas shift reaction.
2. Experimental set-up
The experiments were carried out using three different reactors: A batch oper-
ated tumbling reactor (1 l inner volume, up to 500 ◦C, 50 MPa), a spiral tubular
reactor (6 m long, about 20 ml inner volume, up to 600 ◦C, 30 MPa), and a con-
tinuous stirred tank reactor (CSTR, about 190 ml inner volume, up to 600 ◦C,

Influence of Salts During Hydrothermal Biomass Gasification:
355
100 MPa). These three reactors correspond to the three classical types of reac-
tors used in reaction technology and do not only differ by the reaction times
reached, but also by the back-mixing and heating rates. (For details concern-
ing the experimental set-up and procedure, see [106] for the batch, [107] for the
tubular and [108] for the CSTR.)
In the experiments with the addition of gases described here, an aque-
ous solution of glucose or glucose with K2CO3 is used. In experiments with
Raney-Nickel, it was suspended in the aqueous glucose solution. The solu-
tion or suspension was fed in the autoclave and nitrogen was used to refill the
space above the liquid. The concentration of glucose was 5 and 7 5 wt %, of
K2CO3 and Raney-Nickel, respectively 0 5 wt %. In some experiments addi-
tionally 2 7 MPa CO or H2 were pressed into the autoclave. With a heating
rate 1 K min and 3 K min, respectively, the temperature of the autoclave was
increased up to 500 ◦C. The reaction time at 500 ◦C was one hour [109].
3. Results and discussion
As biomass is a complex mixture of variable composition, it is very difficult
to identify individual reaction paths. For this reason, experiments were carried
out with model substances (e.g. glycerin [110], glucose [106–108] or pyro-
catechol [111]) first. Key substances, usually typical intermediate products,
were allocated to each reaction path identified as “indicators” [91, 100, 106–
108]. With the help of these key substances, these reaction paths were identified
again in the conversion of biomass. Variations of the concentrations of key sub-
stances indicate a changed selectivity of the respective reaction path, e.g. due
to salts. These reactions are combined to a mechanism for the degradation of
cellulose, the main component of biomass (see Fig. 6; [91, 100, 106–108]).
The first reaction step of cellulose degradation (see Fig. 6) is the hydro-
lysis of cellulose to sugar units. This reaction step is very fast [112, 113].
From sugars, e.g. glucose or fructose furfurals may be formed by multiple
water elimination [114, 115]. The key substances for this reaction path are
5-hydroxymethylfurfural, methylfurfural, and furfural. Their concentrations
are particularly high in the sub-critical range and in the presence of acids (see
e.g. [91]). But also at higher temperatures and short residence times, these sub-
stances can be detected. It is demonstrated by experiments using glucose in the
tubular reactor at 400 ◦C (see e.g. [107]) that the addition of KHCO3 leads to
a significant reduction of the furfural concentration measured (Fig. 7). If pos-
sible, furfural formation should be reduced, as furfurals polymerize easily and
the thus formed solids react further very slowly only. In the experiments with
the batch reactor, always small amounts of coke were formed in contrast to
the studies in the other reactors. The coke formation in the batch reactor de-
creases, if potassium salts are added [106]. The reduced coke formation in the
presence of alkali salts may possibly be attributed to a reduced formation of

356
A. Kruse and E. Dinjus
Fig. 6. Simplified reaction mechanism of cellulose degradation during biomass conversion
in near- and supercritical water.
furfurals and other polymerizable compounds. Both sugars and furfurals may
form smaller molecules, above all short-chained aldehydes and acids, which
are partly unsaturated. These reaction steps may be an aldol cleavage of the
sugars [95, 115, 116] or an acid cleavage of furfurals [117]. The key substances
for these reaction paths are e.g. formic acid, acetic acid, and laevulinic acid
as well as formaldehyde and acetaldehyde. From these low-molecular inter-
mediate products, gases and molecules of higher molar mass may form. The
experimental finding of phenol formation having a reaction order close to 2
among others allows the assumption to be made that the phenols detected are at
least partly formed from smaller fragments by a ring closure reaction [92]. Ac-
cording to studies performed in a batch-type reactor with glucose at 500 ◦C, the
addition of alkali salts causes the phenol concentration to be increased slightly
(Fig. 8) [106].
The addition of alkali salts strongly affects the gas composition, as was
demonstrated by the use of various model substances. By the addition of alkali
salts, the CO yield was reduced drastically, while the yields of CO2 and H2 in-
creased accordingly (Figs. 9, 10) [107, 111, 118]. This certainly is a result of
the catalysis of the water-gas shift reaction by alkali salts (see above).
It is obvious from the results presented that various reaction steps of
biomass decomposition are influenced by potassium salts. The reasons are not
yet understood.

Influence of Salts During Hydrothermal Biomass Gasification:
357
Fig. 7. Concentration of 5-hydroxymethylfurfural (HMF), methylfurfural (MF), and fur-
fural (FU) after reaction of glucose with and without KHCO3 (tubular reactor; 400 ◦C,
25 MPa, 16 s reaction time, 1 5% (g/g) Glucose, 0 2% (g/g) KHCO3 [107]).
Fig. 8. Concentration of all phenols (phenols index), and of selected single phenols after
reaction of glucose with and without K2CO3 (batch reactor heated up to 500 ◦C (32 MPa)
with 1 K min, 7 5% (g/g) Glucose [106]).
There is experimental evidence, however, that these effects are closely
linked with the catalysis of the water-gas shift reaction by alkali salts [91].
Apparently, not only the gas composition, but other reaction paths also may
be affected by the occurrence of reactive intermediate stages of the water-gas
shift reaction or by hydrogen in statu nascendi as “active hydrogen”. In the ex-
periments with glucose and other model substances for biomass, the potassium
salts were added, because “wet biomass” as the initial product desired for gasi-
fication possesses a fairly high salt content and no information was available
on how these salts act on biomass decomposition. The question now arising
is whether the salts in real biomass actually have an effect similar to that of
KHCO3 or K2CO3 and whether the cellulose of the biomass shows a behav-
ior similar to that of glucose under the experimental conditions. This question
may be settled by comparative experiments with biomass only. For the work

358
A. Kruse and E. Dinjus
Fig. 9. Composition of the gas phase after reaction of glucose with and without KHCO3
(tubular reactor; 400 ◦C, 25 MPa, 16 s reaction time, 1 5% (g/g) Glucose, 0 2% (g/g)
KHCO3 [107]).
Fig. 10. Composition of the gas phase after reaction of pyrocatechol with various amount
of KOH (batch reactor heated up to 500 ◦C (25 MPa), 5% (g/g) pyrocatechol [111]).
presented here, a pap made of carrots and potatoes was used (phytomass, for
details see [91]).
The differing concentrations of the key substances reflect varying selec-
tivities of the reaction paths. Fig. 11 shows the concentrations of hydroxy-
methylfurfural, total phenol, phenol, formic acid, and acetic acid after reaction
of phytomass, glucose, and glucose K2CO3 at the same experimental con-
ditions. As already mentioned, the addition of alkali salts reduces furfural
concentration. Similar to glucose K2CO3, the phytomass conversion leads to
very small furfural concentrations. The total phenol concentration of the phy-
tomass experiments is somewhat higher than after the conversion of glucose,
but not as high as for the glucose K2CO3 system. From this, it may be con-

Influence of Salts During Hydrothermal Biomass Gasification:
359
Fig. 11. “Fingerprints” of various initial substances: Concentrations of selected key sub-
stances following the conversion of phytomass, glucose, and glucose K2CO3 (batch-type
reactor, heating to 500 ◦C, and 1 h reaction at 500 ◦C, 5% dry matter content, if applicable,
0 5 wt % K2CO3).
cluded that the behavior of alkali salts in the biomass is similar to that of added
potassium salt and that glucose K2CO3 is a good model system for salt-rich
biomass.
In spite of all differences, it is surprising why the “fingerprints” of the dif-
ferent model substances are so similar. The phytomass is a solid and in both
other cases, solutions are fed into the reactor. In addition, the composition of
the biomass is completely different. It mainly consists of cellulose, and not
of glucose. Apart from alkali salts, also other salts are included in biomass,
and not in the form of carbonates. The similarity of the chemical processes in
spite of the varying compositions demonstrates that the first step of biomass
decomposition, namely, the hydrolysis of the cellulose, must be a very fast re-
action. Consequently, cellulose exhibits a behavior that is very similar to that
of the model substance glucose. This very rapid decomposition of polymer
biomass structures is the major difference from “dry biomass gasification pro-
cesses”.
Interesting findings were also obtained from the comparison of the results
achieved with various reactors. For instance, small amounts of coke were found
for each gasification experiment in the batch-type reactor. In the continuous
stirred vessel, coke was never found, irrespective of the initial product used. In
the tubular reactor, experiments were carried out with model substances only
and no coke formation was observed. As the biomass or model substance are
heated relatively rapidly in the tubular reactor and in particular in the continu-
ous stirred vessel – here, heating takes place nearly promptly by the mixing of

360
A. Kruse and E. Dinjus
the cold educt with the hot reactor contents. In contrast the heating rate in the
batch-type reactor is relatively slow (1 or 3 K min), it may be assumed that the
heating rate is of decisive importance to coke formation. A possible explana-
tion is that at lower temperatures and sufficiently high residence times, many
compounds with double bonds are formed under sub-critical conditions. These
also include furfurals that polymerize with increasing temperature and starting
radical formation. This effect could be enhanced by a phase separation in the
sub-critical range, as it was predicted for these conditions [119]. In this case,
the polymerization rate in the organics-rich phase would be much faster than
in a homogeneous mixture due to the concentration effect. Most probably, the
relatively good gasification results obtained in a continuous stirred vessel with
respect to residual carbon and gas yield may not only be attributed to rapid
heating. Due to back mixing, for instance, the “active hydrogen” formed by
the water-gas shift reaction could be effective in each reaction step and not
only at the end of the reaction chain as it is the case in reactions without back
mixing. Hydrogen is a late consecutive product of biomass gasification (see
Fig. 6)!
To prove the assumption, that “active hydrogen” plays an important role in
SCWG, especially in view of the effect of salts, additional experiments were
conducted [109]. Glucose was gasified in the batch reactor with two different
heating rates. In the first set of experiments, only glucose in water was filled
into the reactor. In other experiments Raney-Nickel or K2CO3 was added to the
aqueous glucose solution. Finally Raney-Nickel and H2 or alternatively K2CO3
and CO were used to influence the glucose degradation in SCW. The idea of the
experiments with added gases was to create “active hydrogen” which is present
in every reaction step, from furfural up to gas formation (see Fig. 6). Here two
different types of “active hydrogen” are created. The first is a species formed
by H2 and a hydrogenation catalyst, here Raney-Nickel. Here likely a com-
plex of hydrogen with Nickel on the surface or in solution is the active species.
(Nickel was found to show corrosion effects at these conditions; therefore it
is not clear how a catalytic effect as a consequence of Raney-Nickel addition
is possible.) The addition of K2CO3 and CO leads to the formation of H2 via
the water-gas shift reaction. Here an intermediate, may be formate could be the
active reducing compound. In fact, the CO added was nearly completely con-
verted to H2 at reaction conditions. The results were surprising (Figs. 12, 13).
The effects found with Raney-Nickel or K2CO3 were increased by the addition
of the gases mentioned. The addition of K2CO3 decreases furfural (Fig. 12)
and increases phenols formation (Fig. 12). CO enhanced both effects. The add-
ition of Raney-Nickel decreases the yield of furfurals; if also H2 is present
the concentration of furfurals was lower, below the detection limit (Fig. 12).
The combination of Raney-Nickel and H2 also decreases the yield of phenols
(Fig. 13).
These experiments show, that the formation of “active hydrogen” can in-
fluence different reaction paths of biomass degradation in SCW. Especially the

Influence of Salts During Hydrothermal Biomass Gasification:
361
Fig. 12. Concentration of hydroxymethylfurfural at two different heating rates (batch reac-
tor 5 wt % glucose, 500 ◦C).
Fig. 13. Concentration of phenols at two different heating rates (batch reactor, 5 wt % glu-
cose, 500 ◦C).
reduction of furfurals concentration may be a consequence of hydrogenation of
furfurals precursors or furfurals.
The formation of phenols is reduced by Raney-Nickel H2 and increased
by K2CO3 CO. Likely phenols precursors are hydrogenated during the Raney-
Nickel H2 experiments, and therefore the phenols formation is lower. It is not
clear, why the phenols formation is enhanced in the K2CO3 CO experiments,
in which also the medium is highly reductive.

362
A. Kruse and E. Dinjus
4. Conclusions
A lot of different effects of salts on reactions in near-critical and supercritical
water are observed. Usually two different main explanations are given:
•
The salts show acidic or basic properties. Therefore a salt, reacting as
a base, inhibits acid catalyzed reactions and enhances base catalyzed reac-
tions.
•
Complex formation decreases or increases the enthalpy of the activated
complex. Sometimes the reaction mechanism may change because of this.
One may assume that the effects observed during biomass gasification in SCW,
may be simply a consequence of the basicity of alkali salts, leading to an accel-
eration of reactions catalyzed by bases. The aldol condensation e.g. is known
to be a base catalyzed reaction proceeding also in near- and supercritical water.
This is in fact not likely because,
•
•
salts in biomass and e.g. KHCO3 are no bases at reaction conditions (see
above),
and during SCWG acid (organic acid and CO2) were formed.
On the other hand there exist hints, that the catalysis of the water-gas shift re-
action plays the important role during SCWG and that the different effects of
salts are connected with the water-gas shift reaction. The most important ex-
perimental hint is that the addition of CO enlarges the effects found by alkali
salts.
Additional effects by complex formation, found in studies on synthesis re-
action probably influence the SCWG also. Here the knowledge is poor. E.g.
concerning phenols formation a lot of questions are not answered up to now.
5. Outlook
To get further information and to verify the assumptions mentioned above, ad-
ditional, but very expensive and time consuming investigations are necessary.
Measurements of the pH value at SCWG conditions would be useful to get data
about the influence of salts (and compounds of biomass, e.g. proteins) on the
acid or base catalysed reactions. These kind of measurements are experimen-
tially very difficult and carried out only up to 400 ◦C until now (e.g. [120, 121]).
In order to get information about complex formation, spectroscopic meas-
urement are necessary. The fundamental question is, where are organic
molecules during reactions in near- and supercritical water? Are they near the
salt ions and do they form complexes? This question is very difficult to answer
at SCWG conditions. Up to now only some measurements about solvation of
non-polar compounds or salts in near and supercritical water exist. New experi-
ments with EXAFS on systems, which do not only include salt and water, but
additionally an organic compound are in preparation.

Influence of Salts During Hydrothermal Biomass Gasification:
363
The most important aspect is, that the knowledge about single or elemen-
tary reactions steps in near- and supercritical water is very poor. In regard of the
special properties of near-critical and supercritical water, here huge differences
compared to reactions in ambient water or in the gas phase can be expected.
Informations about how a single reaction proceeds at near- and supercritical
conditions are necessary to understand the influence of salts on the different
reactions also relevant for biomass gasification in the future.
References
1. E. U. Franck, Endeavour 27 (1968) 55.
2. E. U. Franck, Fluid Phase Equilibria 10 (1983) 211.
3. K. Heger, M. Uematsu, and E. U. Franck, Ber. Bunsen-Ges. Phys. Chem. 84 (1980)
758.
4. W. L. Marshall and E. U. Franck, J. Phys. Chem. Ref. Data 10 (1981) 295.
5. C. A. Meyer, R. B. McClintock, G. J. Silvestri, and R. C. Spencer, Jr., Steam Tables,
ASME, New York (1992).
6. E. U. Franck, S. Rosenzweig, and M. Christoforakos, Ber. Bunsen-Ges. Phys.
Chem. 94 (1990) 199.
7. M. Heilig and E. U. Franck, Ber. Bunsen-Ges. Phys. Chem. 94 (1990) 27.
8. M. L. Japas and E. U. Franck, Ber. Bunsen-Ges. Phys. Chem. 89 (1985) 793.
9. Y. Hakuta, T. Adschiri, H. Hirakoso, and K. Arai, Fluid Phase Equilibria 158–160
(1999) 733.
10. Y. Hakuta, T. Haganuma, K. Sue, T. Adschiri, and K. Arai, Mater. Res. Bull. 38
(2003) 1257.
11. K. Sue, K. Kimura, and K. Arai, Mater. Lett. 58 (2004) 3229.
12. H. E. Barner, C. Y. Huang, T. Johnson, G. Jacobs, M. A. Martch, and W. R. Killilea,
J. Hazard. Mater. 31 (1992) 1.
13. E. F. Gloyna, L. Li, and R. N. McBrayer, Water Sci. Technol. 30 (1994) 1.
14. E. F. Gloyna and L. Li, Waste Manag. 13 (1993) 379.
15. M. Hodes, P. A. Marrone, G. T. Hong, K. A. Smith, and J. W. Tester, J. Supercrit.
Fluids 29 (2004) 265.
16. P. Kritzer and E. Dinjus, Chem. Eng. J. 83 (2001) 207.
17. N. Matubayasi and M. Nakahara, J. Chem. Phys. 112 (2000) 8089.
18. J. Lu, J. S, Brown, C. L. Liotta, and C. A. Eckert, Chem. Commun. (2001) 665.
19. H. Hippler, Proceedings of the 4th Italian Conference on Supercritical Fluids and
their Applications, E. Reverchon (1997) 279.
20. M. Watanabe, T. Sato, H. Inomata, R. L. Smith, Jr., K. Arai, A. Kruse, and E. Din-
jus, Chem. Rev. 104 (2004) 5803.
21. S. E. Hunter and P. E. Savage, Chem. Eng. Sci. 59 (2004) 4903.
22. P. Krammer and H. Vogel, J. Supercrit. Fluids 16 (2000) 189.
23. G. L. Huppert, B. C. Wu, S. H. Townsend, M. T. Klein, and S. C. Paspek, Ind. Eng.
Chem. Res. 28 (1989) 161.
24. J. W. Tester, A. Marrone, M. M. DiPippo, K. Sako, M. T. Reagan, T. Arias, and
W. A. Peters, J. Supercrit. Fluids 13 (1998) 225.
25. A. G. Kalinichev, Molecular Modeling Theory: Applications in the Geoscience,
R. T. Cygan, J. D. Dubicki, Mineral. Soc. Am., Rev. Mineral. Geochem. 42 (2001)
83.
26. P. T. Cummings, A. A. Chialvo, and H. D. Cochran, Chem. Eng. Sci. 49 (1994)
2735.

364
A. Kruse and E. Dinjus
27. P. T. Cummings, H. D. Cochran, J. M. Simonson, R. E. Mesmer, and S. Karaborni,
J. Chem. Phys. 94 (1991) 5606.
28. N. Matubayasi, C. Wakai, and M. Nakahara, J. Chem. Phys. 110 (1999) 8000.
29. H. Sato and F. Hirata, J. Chem. Phys. 111 (1999) 8545.
30. P. Jedlovszky and J. Richardi, J. Chem. Phys. 110 (1999) 8019.
31. M. Boero, T. Ikeshoji, C. C. Liew, K. Terakura, and M. Parrinello, J. Am. Chem.
Soc. 126 (2004) 6280.
32. N. Matubayasi, C. Wakai, and M. Nakahara, Phys. Rev. Lett. 78 (1997) 2573.
33. D. Sebastiani and M. Parrinello, Chemphyschem: Eur. J. Chem. Phys. Phys. Chem.
3 (2002) 675.
34. T. Yamaguchi, J. Molec. Liq. 78 (1998) 43.
35. M.-C. Bellissent-Funel, J. Molec. Liq 90 (2001) 313.
36. G. Goodyear and S. C. Tucker, J. Chem. Phys. 111 (1999) 9673.
37. S. C. Tucker and M. W. Maddox, J. Phys. Chem. B 102 (1998) 2437.
38. I. A. Beta, J.-C. Li, and M.-C. Bellissent-Funel, Chem. Phys. 292 (2003) 229.
39. A. A. Chialvo, P. T. Cummings, J. M. Simonson, and R. E. Mesmer, J. Molec. Liq.
73–74 (1997) 361.
40. H. D. Cochran, P. T. Cummings, and S. Karaborni, Fluid Phase Equilibria 71
(1992) 1.
41. J. J. de Pablo, J. M. Prausnitz, H. J. Strauch, and P. T. Cummings, J. Chem. Phys.
93 (1990) 7355.
42. S. H. Lee, P. T. Cummings, J. M. Simonson, and R. E. Mesmer, Chem. Phys. Lett.
293 (1998) 289.
43. S. T. Cui and J. G. Harris, Chem. Eng. Sci. 49 (1994) 2749.
44. T. Driesner, T. M. Seward, and I. G. Tironi, Geochim. Cosmochim. Acta 62 (1998)
3095.
45. B. Balbuena Perla, K. P. Johnston, and P. J. Rossky, J. Phys. Chem. 100 (1996)
2706.
46. J. L. Fulton, D. M. Pfund, S. L. Wallen, M. Newville, E. A. Stern, and Y. Ma,
J. Chem. Phys. 105 (1996) 2161.
47. T. M. Seward, C. M. B. Henderson, J. M. Charnock, and B. R. Dobson, Geochim.
Cosmochim. Acta 60 (1996) 2273.
48. T. M. Seward, C. M. B. Henderson, J. M. Charnock, and T. Driesner, Geochim.
Cosmochim. Acta 63 (1999) 2409.
49. M. G. E. Goemans, T. J. Funk, M. A. Sedillo, S. J. Buelow, and G. K. Anderson,
J. Supercrit. Fluids 11 (1997) 61.
50. A. A. Chialvo, P. T. Cummings, H. D. Cochran, J. M. Simonson, and R. E. Mesmer,
J. Chem. Phys. 103 (1995) 9379.
51. Z. Zhang and Z. Duan, Chem. Phys. 297 (2004) 221.
52. J. P. Brodholt, Chem. Geol. 151 (1998) 11.
53. H. Oka and O. Kajimoto, Phys. Chem. Chem. Phys. 5 (2003) 2535.
54. N. Kometani, K. Takemiya, Y. Yonezawa, F. Amita, and O. Kajimoto, Chem. Phys.
Lett. 394 (2004) 85.
55. K. Sue and K. Arai, J. Supercrit. Fluids 28 (2004) 57.
56. B. Balbuena Perla, K. P. Johnston, and P. J. Rossky, J. Phys. Chem. 100 (1996)
2716.
57. T. Yamaguchi, M. Yamagami, H. Ohzono, H. Wakita, and K. Yamanaka, Chem.
Phys. Lett. 252 (1996) 317.
58. T. Xiang, K. P. Johnston, W. T. Wofford, and E. F. Gloyna, Ind. Eng. Chem. Res. 35
(1996) 4788.
59. J. L. Oscarson, R. M. Izatt, P. R. Brown, S. Gillepie, and J. J. Christensen, J. Solu-
tion Chem. 17 (1988) 841.
60. K. P. Johnston and J. B. Chlistunoff, J. Supercrit. Fluids 12 (1998) 155.

Influence of Salts During Hydrothermal Biomass Gasification:
365
61. X. Chen, R. M. Izatt, and J. L. Oscarson, Chem. Rev. 94 (1994) 467.
62. J. D. Frantz, Chem. Geol. 152 (1998) 211.
63. I. Martinez, C. Sanchez-Valle, I. Daniel, and B. Reynard, Chem. Geol. 207 (2004)
47.
64. C. T. Lira and P. J. McCrakin, Ind. Eng. Chem. Res. 32 (1993) 2608.
65. W. S.-L. Mok, M. J. Antal, and M. Jones, J. Org. Chem. 54 (1989) 4596.
66. N. R. Gunawardena and T. B. Brill, J. Phys. Chem. A 105 (2001) 1876.
67. D. Miksa, J. Li, and T. B. Brill, J. Phys. Chem. A 106 (2002) 11107.
68. W. Henk, Doctoral thesis, Technical University Darmstadt (2004).
69. M. J. J. Antal, M. Carlsson, X. Xu, and D. G. M. Anderson, Ind. Eng. Chem. Res.
37 (1998) 3820.
70. S. Ramayya, A. Brittain, C. DeAlmeida, W. Mok, and J. Antal, Fuel 66 (1987)
1364.
71. J. M. L. Penninger, R. J. A. Kersten, and H. C. L. Baur, J. Supercrit. Fluids 17
(2000) 215.
72. J. M. L. Penninger, R. J. A. Kersten, and H. C. L. Baur, J. Supercrit. Fluids 16
(1999) 119.
73. L. A. Torry, R. Kaminsky, M. T. Klein, and M. R. Klotz, J. Supercrit. Fluids 5
(1992) 163.
74. P. Krammer, Doctoral thesis, Technical University Darmstadt (1998).
75. D. C. Elliott and L. J. Sealock, Ind. Eng. Chem. Prod. Res. Devel. 22 (1983) 426.
76. D. C. Elliott, R. T. Hallen, and L. J. Sealock, Ind. Eng. Chem. Prod. Res. Devel. 22
(1983) 431.
77. O. T. Onsager, M. S. Brownrigg, and R. Lodeng, Int. J. Hydrogen Energy 21 (1996)
883.
78. K. S. Lin and H. P. Wang, Environ. Sci. Technol. 33 (2004) 3278.
79. A. V. Bridgwater, Appl. Catal. A: General 116 (1994) 5.
80. D. Sutton, B. Kelleher, and J. R. H. Ross, Fuel Process. Technol. 73 (2001) 155.
81. E. G. Baker and L. K. Mudge, J. Anal. Appl. Pyrolysis 6 (1984) 285.
82. L. K. Mudge, L. J. J. Sealock, R. J. Robertus, D. H. Mitchell, E. G. Baker, and P. C.
Walkup, Report Conf-7809125-1 (1978).
83. D. C. Elliott, R. T. Hallen, and J. Sealock, J. Anal. Appl. Pyrolysis 6 (1984) 299.
84. P. McKendry, Bioresource Technol. 83 (2002) 55.
85. M. Modell, Fundamentals of Thermochemical Biomass Conversion, R. P. Overend,
T. A. Milne, L. K. Mudge, Elsevier Applied Science Publisher, London, New York
(1985) 95.
86. X. Xu, Y. Matsumura, J. Stenberg, and M. J. Antal, Jr., Ind. Eng. Chem. Res. 35
(1996) 2522.
87. X. Xu and M. J. Antal, Jr., Environ. Prog. 17 (1998) 215.
88. D. Yu, M. Aihara, and M. J. Antal, Jr., Energy Fuels 7 (1993) 574.
89. D. C. Elliott, L. J. Sealock, Jr., and E. G. Baker, Ind. Eng. Chem. Res. 32 (1993)
1542.
90. D. C. Elliott, M. R. Phelps, L. J. Sealock, Jr., and E. G. Baker, Ind. Eng. Chem. Res.
33 (1994) 574.
91. A. Kruse and A. Gawlik, Ind. Eng. Chem. Res. 42 (2003) 267.
92. A. Kruse, T. Henningsen, J. Pfeiffer, and A. Sinag, Ind. Eng. Chem. Res. 42 (2003)
3711.
93. A. Kruse, D. Meier, P. Rimbrecht, and M. Schacht, Ind. Eng. Chem. Res. 39 (2000)
4842.
94. H. Schmieder, J. Abeln, N. Boukis, E. Dinjus, A. Kruse, M. Kluth, G. Petrich,
E. Sadri, and M. Schacht, J. Supercrit. Fluids 17 (2000) 145.
95. B. M. Kabyemela, T. Aschiri, R. M. Malaluan, and K. Arai, Ind. Eng. Chem. Res.
36 (1997) 1552.

366
A. Kruse and E. Dinjus
96. C. C. Liew, H. Inomata, and K. Arai, Fluid Phase Equilibria 144 (1998) 287.
97. M. Watanabe, H. Inomata, and K. Arai, Biomass Bioenergy 22 (2002) 405.
98. M. Saisu, T. Sato, M. Watanabe, T. Adschiri, and K. Arai, Energy Fuels 17 (2003)
922.
99. M. Osada, M. Watanabe, K. Sue, T. Adschiri, and K. Arai, J. Supercrit. Fluids 28
(2004) 219.
100. A. Kruse, J. Abeln, N. Boukis, E. Dinjus, M. Kluth, G. Petrich, E. Sadri,
M. Schacht, and H. Schmieder, Proceedings of the International Meeting of the
GVC-Fachausschuss “Hochdruckverfahrenstechnik”, published as Berichte des
Forschungszentrums Karlsruhe FZKA 6271, Forschungszentrum Karlsruhe, Ger-
many (1999) 111.
101. Z. Fang, T. Minowa, R. L. Smith, Jr., T. Ogi, and J. A. Koziski, Ind. Eng. Chem.
Res. 43 (2004) 2454.
102. T. Minowa and T. Ogi, Fuel Energy Abstr. 39 (1998) 200.
103. M. Goto, R. Obuchi, T. Hirose, T. Sakaki, and M. Shibata, Bioresource Technol. 93
(2004) 279.
104. F. Goudriaan and D. G. R. Peferoen, Chem. Eng. Sci. 45 (1990) 2729.
105. N. Boukis, U. Galla, V. Diem, P. D’Jesus, and E. Dinjus, Chem. Eng. Transact.
(2004) 131.
106. A. Sinag, A. Kruse, and J. Rathert, Ind. Eng. Chem. Res. 43 (2004) 502.
107. A. Sinag, A. Kruse, and V. Schwarzkopf, Eng. Life Sci. 3 (2003) 469.
108. A. Sinag, A. Kruse, and V. Schwarzkopf, Ind. Eng. Chem. Res. 42 (2004) 3519.
109. J. Pfeiffer, A. Kruse and N. N., In preparation.
110. W. Bühler, E. Dinjus, H. J. Ederer, A. Kruse, and C. Mas, J. Supercrit. Fluids 22
(2002) 37.
111. A. Kruse, D. Meier, P. Rimbrecht, and M. Schacht, Ind. Eng. Chem. Res. 39 (2000)
4842.
112. M. Sasaki, B. Kabyemela, R. Malaluan, S. Hirose, N. Takeda, T. Adschiri, and
K. Arai, J. Supercrit. Fluids 13 (1998) 261.
113. M. Sasaki, T. Adschiri, and K. Arai, AIChE J. 50 (2004) 192.
114. M. J. Antal, Jr., W. S. Mok, and G. N. Richards, Carbohydrate Res. 199 (1990) 91.
115. Z. Srokol, A. G. Bouche, A. van Estrik, R. C. J. Strik, T. Maschmeyer, and J. A. Pe-
ters, Carbohydrate Res. 339 (2004) 1717.
116. B. M. Kabyemela, T. Adschiri, R. M. Malaluan, and K. Arai, Ind. Eng. Chem. Res.
38 (1999) 2888.
117. G. C. A. Luijkx, W. van der Horst, S. O. A. Koskinen, F. van Rantwijk, and H. van
Bekkum, J. Anal. Appl. Pyrolysis 28 (1994) 245.
118. A. Kruse, D. Meier, P. Rimbrecht, and M. Schacht, Ind. Eng. Chem. Res. 39 (2004)
4842.
119. W. Feng, H. J. van der Kooi, and J. Swaan Arons, Chem. Eng. Process. 43 (2004)
1459.
120. D. J. Wesolowski, D. A. Palmer, and R. E. Mesmer, Water–rock interaction. Proc.
Symposium, Vladivostok, 1995, published by Balkema. Y. K. Kharaka, O. V. Chu-
daev (Eds.) (1995) 51.
121. K. Sue, K. Murata, Y. Matsuura, M. Tsukagoshi, T. Adschiri, and K. Arai, Fluid
Phase Equilibria 194–197 (2002) 1097.
122. P. G. Maiella and T. B. Brill, J. Phys. Chem. A 102 (1998) 5886.