Reaction Pathways During Hydrothermal
Upgrading of Biomass
Zbigniew Wojciech Srokol
Reaction Pathways During Hydrothermal
Upgrading of Biomass
Proefschrift
ter verkrijging van de graad van doctor
aan de Technische Universiteit Delft,
op gezag van de Rector Magnificus prof. dr. ir. J.T. Fokkema,
voorzitter van het College voor Promoties,
in het openbaar te verdedigen
op 26 februari 2009 om 15:00 uur
door
Zbigniew Wojciech SROKOL
magister inżynier van de Technische Universiteit Krakau
geboren te Bielsko-Biała, Polen
Dit proefschrift is goedgekeurd door de promotoren:
Prof. dr. R. A. Sheldon
Copromotor: dr. ir. J. A. Peters
Samenstelling promotiecommissie:
Rector Magnificus,
voorzitter
Prof. dr. R. A. Sheldon
Technische Universiteit Delft, promotor
Dr. ir. J. A. Peters
Technische Universiteit Delft, copromotor
Prof. dr. ir. H. van Bekkum
Technische Universiteit Delft
Prof. dr. ir. W.P.M. van Swaaij
Universiteit Twente
Prof. dr. ir. W. Prins
Universiteit Gent, België
Dr. J.-P. Lange
Shell Global Solutions
Dr. ir. F. Goudriaan
Biofuel B.V.
Prof. dr. H. Th. Wolterbeek
Technische Universiteit Delft, reserve lid
The research described in this thesis was supported with a grant from the Netherlands
Organization for Scientific Research (NWO) with financial contributions of Shell Global
Solutions and The Dutch Ministry of Economic Affairs (Senter – Novem).
Copyright  2009 by Z.W. Srokol
All rights reserved. No part of the material protected by this copyright notice may be reproduced or utilized in any
form or by any other means, electronic or mechanical, including photocopying, recording or by any information
storage and retrieval system, without written permission from the author.
Printed in The Netherlands
Voor mijn ouders
For my parents
Dla moich rodziców
Table of contents
CHAPTER 1
GENERAL INTRODUCTION ................................................................................................................1
1.1
ENERGY IN THE FUTURE ................................................................................................2
TRADITIONAL SOURCES OF ENERGY ................................................................................4
1.2.1
Coal.......................................................................................................................4
1.2.2
Oil .........................................................................................................................5
1.2.3
Natural gas.............................................................................................................6
1.3
ALTERNATIVE SOURCES OF ENERGY ...............................................................................6
1.3.1
Hydroelectricity .....................................................................................................6
1.3.2
Wind energy ..........................................................................................................7
1.3.3
Solar energy...........................................................................................................9
1.3.4
Nuclear energy.....................................................................................................11
1.3.5
Biomass ...............................................................................................................12
1.4
BIOMASS AS A SOLID FUEL ...........................................................................................17
1.5
CONVERSION OF BIOMASS INTO FUELS ..........................................................................18
1.5.1
Gasification..........................................................................................................19
1.5.2
Pyrolysis processes...............................................................................................21
1.5.3
Fermentation and ethanol production ....................................................................25
1.5.4
Biogas..................................................................................................................29
1.5.5
Hydrothermal upgrading (HTU)............................................................................29
1.6
WATER PROPERTIES UNDER HTU CONDITIONS ..............................................................29
1.7
MODEL COMPOUNDS FOR HTU REACTION ....................................................................31
1.8
AIM OF THE THESIS .....................................................................................................32
REFERENCES .............................................................................................................................33
1.2
CHAPTER 2
EXPERIMENTAL SET-UPS FOR STUDYING REACTION PATHWAYS DURING HYDROTHERMAL
REACTIONS ...............................................................................................................................37
2.1
INTRODUCTION ...........................................................................................................38
2.2
REPORTED SYSTEMS IN THE LITERATURE .......................................................................40
2.3
THE EXPERIMENTAL SET-UPS APPLIED IN THE RESEARCH DESCRIBED IN THIS THESIS ..........48
2.3.1
Small autoclave....................................................................................................48
2.3.2
Batch micro reactors.............................................................................................49
2.3.3
Continuous reactor set-up .....................................................................................49
2.3.4
Batch reactor set-up..............................................................................................51
REFERENCES .............................................................................................................................54
CHAPTER 3
HYDROTHERMAL UPGRADING OF BIOMASS TO BIOFUEL; STUDIES ON SOME MONOSACCHARIDE
MODEL COMPOUNDS .................................................................................................................55
3.1
INTRODUCTION ...........................................................................................................56
3.2
RESULTS AND DISCUSSION ...........................................................................................57
3.2.1
D-Glucose (1), D-mannose (2), D-fructose (3), D-galactose (4), and D-sorbitol .........57
3.2.2
D-Arabinose (8)....................................................................................................63
3.2.3
Reaction pathways via 5-hydroxymethylfurfural (7) ..............................................63
3.2.4
Reaction pathways via glycolaldehyde (10) and D-erythrose (11) ...........................66
3.2.5
Reaction pathways via glyceraldehyde (13)...........................................................67
3.2.6
Effects of added acid and base ..............................................................................69
Contents
3.3
3.4
CONCLUSIONS ............................................................................................................71
EXPERIMENTAL ..........................................................................................................72
3.4.1
Chemicals ............................................................................................................72
3.4.2
Apparatus and method ..........................................................................................73
3.4.3
Analyses ..............................................................................................................73
ACKNOWLEDGEMENTS ...............................................................................................................74
REFERENCES .............................................................................................................................75
CHAPTER 4
HYDROTHERMAL REACTIONS OF GLUCOSE; THE INFLUENCE OF CONCENTRATION ON REACTION
PATHWAYS ...............................................................................................................................77
4.1
INTRODUCTION ...........................................................................................................78
4.2
RESULTS AND DISCUSSION ...........................................................................................79
4.2.1
Water phase .........................................................................................................81
4.2.2
Tar phase .............................................................................................................85
4.2.3
Gas phase.............................................................................................................90
4.3
CONCLUSIONS ............................................................................................................91
4.4
EXPERIMENTAL ..........................................................................................................94
4.4.1
Chemicals……….……………………………………………………………….…94
4.4.2
Apparatus and methods...........…...............................................................................94
4.4.3
Analyses....................……………………………………………………………….95
ACKNOWLEDGEMENTS…………………………………………………………………………95
REFERENCES .............................................................................................................................96
CHAPTER 5
HYDROTHERMAL REACTIONS OF GLYCOLALDEHYDE .................................................................99
5.1
INTRODUCTION .........................................................................................................100
5.2
RESULTS AND DISCUSSION .........................................................................................101
5.2.1
Effect of added acid and base..............................................................................105
5.2.2
Autoclave experiments .......................................................................................106
5.3
CONCLUSIONS ..........................................................................................................107
5.4
EXPERIMENTAL ........................................................................................................107
5.4.1
Chemicals ..........................................................................................................107
5.4.2
Apparatus and methods.......................................................................................108
5.4.3
Analysis.............................................................................................................108
ACKNOWLEDGEMENTS .............................................................................................................108
REFERENCES ...........................................................................................................................109
CHAPTER 6
HYDROGEN TRANSFERS DURING HYDROTHERMAL REACTIONS OF GLYCOLALDEHYDE AND
GLUCOSE ...............................................................................................................................111
6.1
INTRODUCTION .........................................................................................................112
6.2
RESULTS AND DISCUSSION .........................................................................................113
6.2.1
Reactions of glycolaldehyde ...............................................................................113
6.2.2
Reactions of glucose...........................................................................................116
6.2.3
Effect of added catalysts .....................................................................................117
6.3
CONCLUSIONS ..........................................................................................................119
6.4
EXPERIMENTAL ........................................................................................................120
6.4.1
Chemicals ..........................................................................................................120
6.4.2
Apparatus and methods.......................................................................................120
6.4.3
Analyses ............................................................................................................120
REFERENCES ...........................................................................................................................121
Contents
SUMMARY...............................................................................................................................123
SAMENVATTING....................................................................................................................127
ACKNOWLEDGEMENTS.......................................................................................................131
CURRICULUM VITAE............................................................................................................133
Chapter 1
General introduction
1
Chapter 1
1.1 Energy in the future
The main source of energy for life on earth is the sun. The sun applies nuclear
fusion and the amount of energy it delivers is enormous. The earth and the atmosphere
immediately reflect about 55% of this energy back into space. The remaining 45% is
absorbed and mainly converted to heat, which in turn is mostly radiated back into space.
From all the solar energy that arrives on earth only about 0.1% is converted by plants.
Nevertheless this leeds to a yearly biomass yield of 120×109 t.
In the 20th century the world population has increased from 1.65 to over 6 billion
and it continues to increase to level off at 9 billion people according to UN expectations.
This large increase in population has resulted in a large increase of energy consumption.
The world energy consumption is derived from various sources, but currently
fossil fuels (oil, gas and coal) are the major sources of energy. They constitute more than
85% of the total amount of energy needed. Using current energy systems, CO2 emissions
will continue to rise and will be 52% higher in 2030 than nowadays [1]. Most of that
increase comes from countries like China that remain big users of coal. Particularly power
stations, cars and trucks will give off most of the increased energy-related CO2 emissions.
The concerns about global warming call for a rapid decrease of the emission of CO2 but this
is extremely difficult to realize using existing technology. Furthermore, the quest for
alternative sources is stimulated by the political situation in the world, and new
environmental regulations.
The key to solving the problem seems to be technology development in the areas
of advanced nuclear reactor design and renewable technology that is clean, environmentally
friendly and free of CO2 emission. Several renewable energy sources can be considered,
including hydroelectricity, wind energy, solar energy, and biomass. The latter one seems
very attractive, at least for the immediate future.
Use of biomass, as a source of energy, has some advantages. Biomass is a CO2
neutral feedstock that may decrease global warming problems. By efficiently exploring
biomass as a source of energy the availability of fossil fuel as the main source of energy
may be extended. Furthermore, using biomass as a source of energy may make the world
less dependent on oil-exporting countries, and in this way the political tension may be
2
General Introduction
reduced. Finally, the economies of agricultural regions benefit from new jobs that are
created by this new technology.
Unfortunately, most of the recent studies dealing with the substitution of gasoline
and diesel oil by biofuels did not consider the fact that an increase in the amount of the
feedstock requires, besides the use of waste land, the conversion of forest and grassland to
new cropland. This process has a very negative carbon-balance since not only less carbon is
biologically bound but also the burning of large areas of e.g. rain-forest releases a lot of
CO2. When these facts are included in the calculations, corn-based ethanol doubles
greenhouse emission over 30 years and increases greenhouse gasses for more than 160
years. This negative balance can only be overcome by not creating new cropland by
substitution but by using waste land and by using already existing waste streams instead
[2].
Figure 1.1 Energy sources, carriers and conversion [3]. FC = fuel cell, ICE = internal
combustion engine, DME = dimethyl ether.
3
Chapter 1
Considering all available energy sources there is a necessity for energy carriers
that make possible simple energy conversion for power and automotive applications. There
is a complex network between energy sources, carriers and conversion (see Fig.1.1) [3].
Catalysts play an increasing role in creating flexibility in this network by providing new
routes [4- 6].
1.2 Traditional sources of energy
1.2.1 Coal
Coal is currently a major traditional source of energy. Coal’s share in production
of energy is projected to increase from 26 % in 2004 to 28 % of the total energy required
in 2030 [7], in all developed countries of the world. It may be expected that coal will
continue to be a major source of electricity generation, and also will be the major source of
hydrogen. In addition, coal has the potential to become an important source of liquid fuels
because it can be converted into a variety of fuels. Some advantages of liquid fuels from
coal, so-called CTL (coal-to-liquid) fuels, are:
1.
Coal is worldwide easily available. Many countries have access to domestic coal
reserves and using fuels based on coal may decrease the need of oil imports and thus
improve energy security.
2.
Coal liquids can be used for transport, cooking, stationary power generation, and in the
chemical industry.
There are two different methods for converting coal into liquid fuels, direct and indirect
liquefaction [8]. Direct liquefaction works by dissolving the coal in a H-donating solvent at
high temperature and H2-pressure. This process is highly efficient, but the liquid products
require further refining to achieve high-grade fuel characteristics. Indirect liquefaction
gasifies the coal to form syngas (a mixture of hydrogen and carbon monoxide). Starting
with syngas one can enter Fischer Tropsch synthesis of hydrocarbons or one can make
methanol, ethanol, etc.
4
General Introduction
1.2.2 Oil
Currently, global oil reserves are very difficult to measure. Still new oil deposits
are being found at different locations. No one can accurately estimate how much oil exists.
The values given in the Table 1.1 for oil reserves indicate that oil will be exhausted within
25 years [9].
Table 1.1 Two estimates of remaining oil reserves in the world [9].
Region
OECD North America
Canada
USA
Mexico
OECD Europe
Norway
UK
OECD Pacific
Australia
Russian Federation
Azerbaijan
Kazakhstan
China
South Asia
East Asia
Indonesia
Latin America
Brazil
Venezuela
Middle East
Kuwait
Iran
Iraq
Saudi Arabia
UAE
Africa
Algeria
Angola
Libya
Nigeria
World
Remaining reserves [Gb*]
EWG **
IHS ***
84
17
41
26
25.5
11
8
2.5
2.4
105
9.2
33
27
5.5
16.5
6.8
52.5
13.2
21.9
362
35
43.5
41
181
39
125
14
19
33
42
854
67.6
15.3
31.9
20.4
23.5
11.6
7.8
5.1
4.8
128
14
39
25.5
5.9
24.1
8.6
129
24
89
678.5
51
134
99
286
57
104.9
13.5
14.5
27
36
1.255
* Giga barrels, ** Environmental Watch Group statistics, *** Industry database
5
Chapter 1
The world use of petroleum is expected to grow from 83 million barrels oil equivalent per
day in 2004 to 97 million barrels per day in 2015 [7].
1.2.3 Natural gas
The world natural gas consumption is expected to increase on average 1.9 % per
year from a world total of 2.8×1012 m3 in 2004 to 3.6×1012 m3 in 2015 and 4.6×1012 m3 in
2030 [7]. Among the major fuels, in the high energy consuming countries, natural gas is
expected to provide the greatest increase in energy consumption, due to its use in industrial
sectors and in electric power generation.
The gas reserves are very large [10]. Natural gas may be also used as a feedstock
for several industrial processes. There are many chemical streams e.g. hydrogen, ammonia,
methanol that can be obtained during the gas utilisation. There are plenty opportunities for
the gas to be used as fuel for power generation plants.
The current technological development indicates that the future production of
energy will strongly depend on gas as a feedstock.
1.3 Alternative sources of energy
1.3.1 Hydroelectricity
Currently, hydroelectricity supplies about 19% of the world's electricity [11].
Small hydroelectricity plants (order 1-1.5 MW) are popular in China, which has over 50%
of the world's small hydroelectricity [11].
Hydroelectricity has a lot of advantages, of which the major one is the elimination
of the cost of fuel. Hydroelectric plants that have been built about 100 years ago are still in
service for energy production; they have a substantially longer lifetime than the generators,
which burn fuel. The cost of energy production is low thanks to automated operating,
whereas little personnel is required during normal operation. Another advantage is that the
hydroelectric plants have relatively low construction costs. Since no fossil fuel is
consumed, emission of CO, CO2, SO2, NOx, from burning such fuels is eliminated:
6
General Introduction
hydroelectricity generates no waste. Unfortunately, hydroelectricity generation is strongly
dependent on the rainfall.
1.3.2 Wind energy
Energy production using wind instead of mechanical energy appeared in the
beginning of 20th century. The first oil price shock increased significantly the interest in
wind as a source of energy. The early 1970s brought the development of the first turbine for
electricity generation [12].
Figure 1.2 Total installed wind power in EU countries [13]. Wind power is presently the
fastest growing renewable energy technology. The yearly worldwide installed power grew
from 0.24 GW in 1991 to 19.92 GW in 2007. In the end of 2007 globally 94.2 GW of wind
energy was installed (Europe 57.1 GW, US/Canada 18.9 GW and Asia 14.3 GW) [13].
7
Chapter 1
For EU countries Fig. 1.2 gives the following data: the totally installed power
(peak), the in the year 2007 installed power and when a third number is given: the
decommissioned power in 2007. Denmark, Germany and Spain are forerunners in wind
energy. In 2007, wind turbines accounted for 6.6 % of the electricity in Germany (see Fig.
1.3) [14]. In Spain this figure was even 10 %.
This fast development prompted the Dutch government to create regulations
regarding the price per kWh that the local distribution company must pay. These
regulations reduce the price of electricity and therefore provide a long-term secure income
to investors.
Wind energy has many advantages and therefore, has potential to play an
important role in the future energy supply in many areas of the world. Wind energy is one
of the lowest-priced renewable energy technologies available today. Wind turbines do not
produce atmospheric emissions or other harmful products that cause acid rain or
greenhouse gasses. Wind turbines use only a small part of the land, therefore this is another
economic benefit regarding to the high price of the ground in the Netherlands. One major
disadvantage is that there is not always enough wind available.
During last 10 years wind turbine technology has reached a very high level. The
growing international market will lead to further improvements, such as large wind
turbines. Nowadays wind turbines up to 3 MW are constructed. In the R&D stage are
turbines up to 6 MW. This will cause further cost reductions and then wind energy will be
able to compete with conventional fossil fuel power generation technology. Moreover offshore wind farms come to the fore.
8
General Introduction
Figure 1.3 Contribution of various energy sources to the net electricity production
of Germany in 2007 (in %) [14].
1.3.3 Solar energy
Through the process of photosynthesis, the energy of sunlight can be used to
create the biomass on our planet (yearly 120×109 tons). Also the energy of sunlight can be
used for electricity production using solar cells [16] and for heating of water using solar
collectors.
With a solar cell efficiency of 10% and 65% efficiency for the electrolysis of
water, the overall efficiency of hydrogen production would be 6.5%. Electrolysis relies on
platinum or other catalysts for gas evolution, which are in limited supply.
9
Chapter 1
Figure 1.4 Hydrogen cycle of power generation [15].
Notwithstanding the still relatively high costs of photovoltaic solar cells, solar
energy is developing fast. See Table 1.2 for the worldwide yearly installed photovoltaic
(PV) power in 2007; 3733 MW was installed of which 1541 MW in the EU. Leading
countries are Germany, Japan, China and Spain. Vehicles powered by sun energy are
suitable for short distances.
In a future scenario (Fig. 1.4) electricity provided by photovoltaics (or
hydropower) is used to split water into H2 and O2. The H2 then enters the hydrogen cycle.
10
General Introduction
Table 1.2 PV power per inhabitant (wp/inhabitant) for each EU country in 2007 [13].
Countries
Luxembourg
Germany
Spain
Austria
Netherlands
Cyprus
Italy
Portugal
Finland
Greece
France
Sweden
Belgium
Denmark
Czech Rep.
Slovenia
UK
Malta
Ireland
Hungary
Bulgaria
Poland
Romania
Slovakia
Lithuania
Latvia
Estonia
EU
Wp/Inhabitant
51.20
46.50
11.74
3.49
3.34
2.25
1.71
1.68
0.95
0.82
0.77
0.68
0.59
0.57
0.39
0.32
0.29
0.24
0.10
0.03
0.02
0.02
0.01
0.01
0.01
0.00
0.00
8.49
1.3.4 Nuclear energy
The expected increase in electricity generation from nuclear power is projected
from 2.619×109 kWh in 2004 to 3.619×109 kWh in 2030 [3]. Nuclear energy produces
about 17 % of the energy used in the world today. That is an increase less than the 6,000
plants that were projected in 1997 [17].
Nuclear energy will not displace coal, oil, and gas. Nuclear energy plants operate
at low efficiency, and generate nuclear waste. Storing nuclear waste is a big problem
because of its extremely long decay time [18]. After the Chernobyl accident in many
11
Chapter 1
countries a lot of strict regulations have been introduced, which make nuclear energy power
plants not cost-effective. Other countries (e.g. France and Belgium) are much more
accepting nuclear power [19].
Nuclear fusion, as applied by the sun, would be a final solution. It will take a long
time, however, before this can be executed on earth in a controlled way.
1.3.5 Biomass
In general terms, “biomass” is an organic resource that can be used to produce
energy using different processes. During photosynthesis, plants combine carbon dioxide
from the air and water from the ground to generate carbohydrates, which form the building
blocks of biomass. In this way, the solar energy is stored in the chemical bonds of the
structural components of biomass. Biomass has a variety of sources including:
1. Agricultural waste: such as cereal straw, fruit trees trimmings, leaves, etc. For some
properties see Table 1.3 [20].
2. Agricultural crops: such as sugarcane, sugar beet, corn and sweet sorghum.
3. Forestry waste: including under-utilised wood, logging residues, imperfect commercial
trees or non-commercial trees.
4. Industrial waste: residues are of organic nature e.g. from the beverage and food industries
or black liquor, a waste stream from the paper industry, food industry.
5. Organic domestic waste.
12
General Introduction
Table 1.3 Properties of various biomasses [20].
Biomass
HHV *
(MJ/kg)
Moisture
(wt%)
Chlorine
Ash
Sulfur
(wt% dry) (wt% dry) (wt % dry)
Charcoal
25-32
1-10
0.5-6
Wood
Coconut shell
10-20
18-19
10-60
8-10
0.25-1.7
1-4
0.01
0.01
Straw
Ground nut
shells
Coffee husks
14-16
17
10
2-3
4-5
10
0.07
0.49
16
10
0.6
Cotton
residues
(stalks)
Cocoa husks
16
10-20
0.1
13-16
7-9
7-14
0.05
1.48
0.07
0.49
Palm oil
residues
(shells)
Rice husk
15
15
13-14
9-15
15-20
Soya straw
15-16
8-9
5-6
Cotton residue
(gin trash)
14
9
12
Maize (stalk)
13-15
10-20
2 (3-7)
Palm oil
residues
(fibres)
11
40
Sawdust
11
35
2
40-60
63
1-4
5
Bagasse
8-10
Palm oil
5
residues (fruit
stems)
Bark
* high heating value
13
Chapter 1
Figure 1.5 Simple images of lignocellulosic biomass composition and their fractions [22].
The chemical composition of biomass varies among species and sources, an
average composition based on dry wood is: 40-50 wt % cellulose, 25 wt % hemicelullose
and 25 wt % lignin (see Figure 1.5) [21].
The carbohydrates portion is formed by the polysaccarides cellulose and
hemicellulose. Cellulose is a linear homopolymer of over 10,000 1-4 linked β-D
glucopyranose units (see Fig 1.6).
14
General Introduction
OH
OH
OH
O
HO
O
O
HO
O
O
O
OH
O
HO
OH
OH
Figure 1.6 Chemical structure of cellulose.
Hemicelullose is not a single specific substance but a copolymer of carbohydrate units
(pentoses: arabinose, xyloses, hexoses: galactose, mannose) with an amorphous structure
and side chains. The degree of polymerisation is 70-280. Hemicellulose is linked with
lignin in a lignin-carbohydrate complex.
The lignin fraction consists of polymerized aromatic units (e.g. coniferyl alcohol)
linked together in large sheet like structures. A lignin impression and common lignin
linkages based on various phenylpropane units are given in Fig. 1.7.
HO
O
MeO
OH
HO
OMe
OH
O
O
OMe
O
OMe
OMe
OH
Lignin
(Mw ~ 20*10 3 )
Figure 1.7 Lignin impression [23].
15
Chapter 1
Lignin is a difficult material to upgrade. Nevertheless some commercial ligninderived products are known: lignosulfonate (by sulfonation of lignin) and vanillin (by
oxidation, Borregaard, Norway). Besides cellulose, hemicelullose and lignin, biomass also
contains extractives and inorganic components. Extractives include a lipophilic fraction
containing terpenoids, resin acids and sterols [24]. Compounds like: tannins, flavonoids,
lignans and stilbene can be found in a hydrophilic fraction. The inorganic components
include elements such as Ca, K, Mg, Mn, Na, Al, Fe, Zn, that are concentrated in the ash
after combustion of wood.
Dry wood contains about 50 weight % carbon, 6 weight % hydrogen and 44
weight % oxygen [25]. Other lignocellulosic biomasses vary in a limited area around the
dry wood composition (Fig 1.8).
Figure 1.8 Elementary composition (in mole %) of biomass and conversion routes [26].
16
General Introduction
An excellent review on the technological and economical aspects of lignocellulose
conversions has been published by Lange [27]. The three major routes are discussed and
compared:
-
pyrolysis to bio-crude
-
gasification to syngas and its subsequent conversion to alkanes or methanol
-
hydrolysis to sugars and subsequent conversion via chemical or fermentation
routes
1.4 Biomass as a solid fuel
Primary energy production from solid biomass (wood, wood waste and other solid
vegetal and animal materials) in the EU amounted to 62.4 million tons oil equivalent
(Mtoe) in 2006. This was 3.1 Mtoe more than in 2005. The primary energy coming from
the combustion of renewable origin solid urban waste can also be added to this figure. In
2006, this was a production of 5.3 Mtoe.
In the important paper-producing countries (Sweden, Finland), black liquor
constituted an important component of the biomass fuel. Thus in Finland black liquor
accounts for 50% of the primary energy furnished by direct biomass burning. An elemental
analysis of black liquor showed the presence of 36.4 wt % C, 3.5 % H, 34.3% O and of
course quite some inorganics.
It may be noted that biomass fueling of power plants may involve exclusive
biomass feeding (e.g. the plant at Cuijk, the Netherlands) or co-feeding of biomass in coalbased power plants.
For a per capita comparision of the EU countries in this biomass use see Fig. 1.9.
[28]
17
0.13 5
0.0 12
UK
Cyprus
Total EU
0 .031
0.0 13
0 .034
Luxembourg
Italy
0.034
Netherlands
Slovakia
0.04 2
0.0 84
0.0 76
Greece
0.04 2
0 .099
Spain
Ireland
0.105
Hungary
Belgium
0.11 3
0.107
Poland
0.15 3
France
Germany
0.1 53
Czech Rep.
0.224
0 .212
Slovenia
Llithuania
0.25 8
0.235
Portugal
Denmark
0.4 05
Austria
Estonia
Latvia
Sweden
Finland
0 .525
0 .866
0.988
1.413
Chapter 1
Figure 1.9 Primary energy production from solid biomass in 2006 in ton equivalents petrol
per inhabitant [28].
1.5 Conversion of biomass into fuels
Biomass can be converted to products, which can be used for the production of
effective energy carriers such as liquid fuels. This can be achieved using various processes
under different conditions.
Depending on the conditions applied (pressure, temperature and residence time in
the reactor) during biomass conversion, different product distributions can be obtained (see
Fig 1.10). Considering current industrial production and future trends the most important
processes nowadays regarding to biomass conversion are: gasification, fermentation and
hydrothermal upgrading (HTU).
18
General Introduction
Figure 1.10 Strategies for production of fuels from lignocellulosic biomass [26].
1.5.1 Gasification
Gasification is a process in which biomass reacts with oxygen (present in air) at a
high temperature (up to 1300 oC) to produce syngas or producer gas that contains different
proportions of CO, H2, CO2, CH4 and N2 [29, 30]. During gasification of biomass many
chemical reactions take place. Some examples are shown in Table 1.4.
Syngas that is produced in industry from natural gas and coal [31] and producer
gas (that has lower concentrations of CO, H2, CO2, CH4 and higher content of N2 than
syngas) are very important for industrial syntheses. There are many industrial routes for
using syngas for chemicals production and utilizations such a production of H2, methanol
and methanol-derived fuels and higher alcohols. The most well known chemical reaction,
using syngas to produce mostly straight chain alkanes up to waxes is the Fischer-Tropsch
synthesis (FT) [32] (see Scheme 1.1). Early FT catalysts were iron-based but present-day
catalysts are cobalt-based. Slurry as well as fixed bed parallel pipe processes are applied.
19
Chapter 1
Table 1.4 Fundamental reactions and enthalpy of selected cellulose gasification reactions
[30].
Classification
Stoichiometry
Enthalpy (kJ/g-mol)
ref temp 300 K
pyrolysis
C6H10O5 5CO + 5H2 + C
C6H10O5 5CO + CH4 + 3H2
C6H10O5 3CO + CO2 + 2CH4 + H2
180
300
-142
partial oxydation
C6H10O5 + 1/2 O2 6CO + 5H2
C6H10O5 + O2 5CO + CO2 + 5H2
C6H10O5 + 2O2 3CO + 3CO2 + 5H2
71
-213
-778
steam gasification
C6H10O5 + H2O 6CO + 6H2
C6H10O5 + 3H2O 4CO + 2CO2 + 8H2
C6H10O5 + 7H2O 6CO2 + 12H2
310
230
64
CO + H2O CO2 + H2
CO + 3H2 CH4 + H2O
-41
-206
water-gas shift
methanation
n CO + (2n + 1) H2 H(CH2)nH + n H2O
Scheme 1.1 Fisher-Tropsch reaction.
Gasification of biomass has three stages that are strongly dependent on the
temperature. In the first stage of the gasification, when the temperature is in the range of
350-500 oC, pyrolysis reactions take place and gas products with low molecular weight,
bio-oils and charcoal are formed. Solid biomass forms oxygenated vapors, primary
oxygenated liquids that contain: cellulose and hemicellulose-derived molecules, ligninderived methoxy-phenols [33], and gaseous H2O and CO2.
The second stage of the gasification has two temperature regimes. The first
temperature regime is above 500 oC and the second one in the range of 700 - 850 oC. The
products formed contain gaseous olefins, aromatics, CO, CO2, H2, H2O and condensed oils
such as phenols.
The third stage of the gasification takes place at temperatures between 850 – 1000
o
C and results in products as CO, CO2, H2, H2O and polynuclear aromatic compounds.
20
General Introduction
Products like benzene, naphthalene, and pyrene, condense to form a liquid phase. Also soot,
coke and ash are formed during the last two stages of gasification. The ash is removed from
the bottom of the reactor and contains products like CaO, K2O, P2O5, MgO, SiO2, adsorbed
SO3, and Na2O [34].
Biomass composition, the presence of oxygen or steam, gasification agent and the
conditions of the gasification process, have a significant influence on the composition of
the gas that is formed [35, 36]. Tar that contains higher molecular weight hydrocarbons,
causes a lot of problems in gasification installations because it condenses in the pipes and
then blocks the stream. The costs of tar removal and its utilization are very high and this is
a bottleneck for the development of commercial gasification technology [37]. The amount
of tar can be decreased by adding solid catalysts (Pd, Pt, Ru and Ni) inside the gasification
reactor or by mixing alkali metal catalysts with the biomass feedstock [38, 39, and 40].
Gasification of biomass can be also performed in supercritical and near-supercritical water
[41].
Currently several small reactors operate on laboratory scale e.g. at the University
of Hawaii, the Pacific Northwest National Laboratory, the Hiroshima University, and in
The Netherlands (University of Twente, TNO-MEP and BTG).
1.5.2 Pyrolysis processes
Various biomass pyrolysis processes have been designed. One approach, called
flash pyrolysis, applies high temperatures and short residence times to prevent the
condensation of the volatile products. A process of this type is commercialized by the
Biomass Technology Group (BTG) in cooperation with the University of Twente.
21
Chapter 1
Figure 1.11 BTG’s biomass pyrolysis process.
This process (see Fig. 1.11) is based on a conical reactor that is fed with biomass and hot
sand and operates as a cyclone to separate the volatile components from the hot sands and
char as soon as they form. The hot sand is fed to the reactor to provide the required reaction
heat. The pyrolysis oil is condensed out of the volatile stream and remaining gaseous
species are fired to provide the required process energy. The sand and char exit the conical
reactor through the bottom and are sent to a combustion reactor, where the char is burnt to
heat up the sand to the required temperature, before recycling to the pyrolysis reactor. The
elegant design by van Swaaij et al. [41] allows yields in pyrolysis oil of up to 70 wt.%.
Three US companies, UOP, the National Renewable Energy Laboratory (NREL)
and the Pacific Northwest National laboratory joined forces in another pyrolysis process.
Here several pyrolysis units would supply a central refinery in which the pyrolysis oil is
subjected to UOPs hydroprocessing technology [42].
Some typical properties of pyrolysis oil, also denoted as bio-oil are contained in
Table 1.5.
22
General Introduction
Table 1.5 Typical properties of bio-oil [42].
Property
Typical values
Moisture content
20-30
%
pH
Density
2-3
1200
kg/m3
C
56
wt%
H
7
wt%
O
37
wt%
N
Ash
Viscosity (40oC, 20% H2O)
0.1
0-0.2
40-100
wt%
wt%
cP
Particulates
<0.3-1
wt%
Heating value
16-18
MJ/kg
Recently, Brem et al. have presented a novel process for flash pyrolysis of wood
(saw dust) and biomass residues (jatropha cake, coconut residue, sunflower residue etc.),
the PyRos process [43], that produces solid-free bio-oil. A pilot plant with a throughput of
30 kg/hr has been built (see Fig. 1.12).
23
Chapter 1
Figure 1.12 PyRos pilot plant [43].
The technology is low-cost and can be used for small and medium scale. The flash
pyrolysis technology focused mainly on quality and stability of the produced oil.
The temperature of this process is 450-550o C. The biomass is transported to the reactor
using a flow of inert material (sand), which allows a good control of the residence time. In
one day about 100 liters of oil can be produced. Depending on the biomass feed, oil yields
of 45-70 % can be obtained. The cost of pyrolysis oil production with this technology is
much lower than that of e.g. fossil crude oil.
It is interesting to mention here the process that is under joint development by
CHOREN (Germany), Shell and Volkswagen (Germany) Fig. 1.12. As outlined by Lange
[27], this process combines the gasification technology developed by CHOREN with
Shell’s Fisher-Tropsch technology. It consists (see Fig. 1.13) of a pyrolysis reactor, which
converts the biomass into volatiles and char.
24
General Introduction
Figure 1.13 CHOREN’s process.
The reactive volatiles are sent to an O2-fed burner for initial gasification while the
less reactive char is injected down-stream in the high-temperature part of the burner flame
for effective gasification. The obtained synthesis gas is cooled, cleaned up, partly shifted
and enters the Fischer-Tropsch reactor for conversion into alkanes. Shell’s Fischer-Tropsch
technology is presently applied in Malaysia at 0.5×106 t a
6
gas into diesel fuel. A unit of 6×10 t a
–1
–1
for the conversion of natural
is under construction in Qatar.
1.5.3 Fermentation and ethanol production
Fermentation of carbohydrates (biological conversion) towards ethanol is currently
worldwide the dominant technology for liquid fuels production from biomass resources.
The largest producers are the US and Brazil, that produced 20 x 109 and 24 x 109 L of
ethanol in 2007, respectively [44].
In the EU the biodiesel production is so far much larger than the bioethanol
production. Expressed in ton oil equivalents (TOE) biodiesel production in 2006 was
4.073904 TOE whereas bioethanol production was 871.673 TOE [45].
Ethanol can be used directly as a fuel or can be blended with gasoline e.g. in a
ratio of 10% ethanol and 90% gasoline (sold mostly in US). In Brazil hydrated ethanol is
used to power vehicles [46]. For conversion of sugars to ethanol by fermentation, yeasts
25
Chapter 1
(e.g. Saccharomyces cerevisiae bacteria or fungi) can be used. The reaction is shown in
Scheme 1.2.
C6H12O6 2C2H5OH + 2CO2
Scheme 1.2 Sugar conversion by fermentation to ethanol.
During this reaction almost half of the mass of sugar is released as CO2. Yeast
ferments sugars like: glucose, mannose, fructose, and galactose. From this reaction beside
ethanol with high yield, small amounts of other products like: glycerol, lactic acid, acetic
acid, and succinic acid, are formed. For large-scale ethanol production, different feedstocks
are used as shown in Fig. 1.14. Here the left side of the Figure shows the present-day
separation and processing of corn and sugar cane into fermentable sugars and by-products.
The right side of Fig. 1.14 pertains to more total crop processing towards ethanol.
Ethanol production requires initial separation and processing of the biomass into
fermentable sugars. This conversion depends on the feedstock. Sugar cane is converted into
water-soluble sugar or cane juice (30 wt % of sugar cane). By-product is bagasse that is
nowadays mainly used as fuel. Microbial contamination in the feedstock is usually removed
by adding ammonium sulfate and / or other salts. The fermentation is performed at
temperatures between 30-38 oC, using a residence time of 28-48 h, at pH 4-5 [44, 46].
Usually, the yield for ethanol production from sugar cane is 160-190 L/metric ton [37].
26
General Introduction
Figure 1.14 Block diagram for ethanol production from corn, sugar cane and cellulosic
biomass [44].
Another feedstock for ethanol production is corn that contains 70 wt % starch, 1011 wt % crude protein, 4.5-6.0 wt % oil, 6 wt % hemicellulose, 2-3 wt % cellulose, 1 wt %
lignin, and 1 wt % ash [44]. Here first enzymatic or acid hydrolosis of starch to glucose is
carried out. The obtained ethanol solution contains 12-14 wt % pure ethanol and the latter is
distilled to the azeotropic level (95 % ethanol). Currently much research is ongoing to
produce ethanol from lignocellulosic biomass in an economical way [46-49]. A design
model of a process for the conversion of corn stover (lignocellulose) to ethanol based on
dilute acid prehydrolysis and enzymatic cellulose hydrolysis already exists [47]. The first
step in this process is to hydrolyze most of the hemicellulose towards C5 and C6 sugars.
27
Chapter 1
This can be done by treating lignocellulose in sulfuric acid for 2 min at 190 oC and a
pressure of 12 atm. In the next step the reaction temperature is decreased to 100 oC and the
sulphuric acid is neutralized by increasing the pH to about 10. The solid fraction is sent to a
saccharification unit. During saccharification cellulose is hydrolyzed to glucose and
cellobiose. This process occurs at 65 oC for 1.5 days and enzymes catalyze the hydrolysis
reactions.
The above process steps are present in a lignocellulosics-to-ethanol demonstration
plant recently started up by the Verenium Corp. (Fig.1.15). The target is 90 gallons (340 L)
of ethanol per ton of dry biomass [48a]. Recently Shell and Iogan announced on extended
commercial alliance to accelerate the development and deployment of cellulosic ethanol
[48b].
Figure 1.15 Lignocellulosics-to-ethanol demonstration plant.
28
General Introduction
1.5.4 Biogas
An interesting bio-gasification technique that comes more and more to the fore is
(small scale) anaerobic microbial gasification of organic waste (e.g. manure) to give biogas.
Biogas consists of 55-75% methane and 25-45% CO2. Biomethane, which is biogas
purified to above 95% methane can be used as transportation fuel or to generate electricity
or can be pumped in national gas grids. In Germany alone, 3500 biogas installations were in
operation in 2007. Here biogas accounted already for 8.9% of the renewable energy in 2007
[49].
1.5.5 Hydrothermal upgrading (HTU)
The HTU process is the hydrothermal conversion of biomass into an organic crude
oil (biocrude). During this process, the oxygen content of the organic material is claimed to
be reduced from about 40% to between 10 and 15%. The removed oxygen ends up mainly
in CO2 and water [21, 50]. HTU was invented long time ago, and by the end of the 1970’s
to early 1980’s many laboratories, mainly universities, were working on it. H.P. Ruyter was
the driving force for the HTU process development to be taken up by the Shell laboratory at
Amsterdam in 1980.
In the HTU process, the reaction temperature ranges from 300 to 360oC. The
pressure is 17-27 MPa and the residence time 5-20 min. The obtained biocrude is not
miscible with water and the heating value is 30-35 MJ/kg. Experimental HTU set-ups will
be described and discussed in chapter 2 of this thesis.
1.6 Water properties under HTU conditions
The HTU process is usually performed at in liquid water at subcritical conditions
(Tc 374.1 oC and Pc 22.1 MPa).
High temperature and high pressure have a significant influence on the properties
of water and the water phase diagram. Changes in viscosity, heat capacity, conductivity, the
structure of H bonds, diffusion coefficients and density significantly influence the transport
29
Chapter 1
characteristics of aqueous solutions. The characteristics of supercritical water are
intermediate between those of the liquid and gas (see Table 1.6).
Table 1.6 Characteristics of water under different conditions [51].
Figure 1.16 Phase diagram of water at elevated temperatures and pressures. T/ K: 553 (1),
573 (2), 593 (3), 623 (4), 633 (5), 647 (6), 653 (7) and 673 (8); G is gas and L is liquid [51].
30
General Introduction
Because at the near-critical region, water has much higher compressibility than
normal liquid water, the density of water may be tuned over a very wide range by small
variations in temperature and pressure. This results in the density of water being halved as
the temperature increases by only 6 K at a pressure of 22 MPa [52]. Generally speaking,
performing chemical reactions in aqueous solutions, parameters such as reagent
concentrations, density, temperature and pressure should be well controlled, because minor
variations of these parameters can significantly influence the distribution of reaction
products [53, 54].
1.7 Model compounds for HTU reaction
The HTU process creates great opportunities for the production of fuel and
chemicals. Its efficiency can be improved in many ways. One of them is to exploit insights
into the mechanism of the chemical reactions during the biomass decomposition.
To study the reaction mechanism of biomass decomposition during HTU process,
a good selection of model compounds is very important. Model compounds that are
representative for biomass should be similar in chemical structure and in their physical
properties e.g. solubility in water. It is known that during the initial stage of the HTU
process, biomass (polysaccharides) hydrolyzes very rapidly to monosaccharides [55].
Therefore, good model compounds to study decomposition of the (hemi-)cellulose fraction
of
biomass
are
C6
monosaccharides like D-glucose,
D-fructose,
D-mannose.
Decomposition of C5 sugars, e.g. D-xylose and D-arabinose that are present in
hemicellulose or simple compounds like alcohols, aldehydes or organic acids might also
generate a lot of useful information regarding the reaction pathways. It would be of value to
study lignin model compounds too.
Results of experiments using a variety of model compounds are presented in this
thesis aiming at substantially enhanced insights into the HTU reaction networks.
31
Chapter 1
1.8 Aim of the thesis
The main goal of this research project is the elucidation of the chemical
mechanism and the determination of the main reaction pathways that play a role during
hydrothermal upgrading of biomass (HTU). The details of the chemistry of the HTU
process might help to optimise the reaction conditions with regard to minimising the
oxygen content of HTU products and to increase the yield of biocrude.
Chapter 2 describes experimental set-ups for the hydrothermal upgrading process
both with continuous and with batch reactors. Set-ups for short as well as for long reaction
times are discussed.
Chapter 3 deals with a study of hydrothermal reactions on some monosaccharides
as model compounds. During the hydrothermal upgrading of biomass, hydrolysis to glucose
is an important step. To elucidate some of the reaction pathways that follow this initial
hydrolysis, the hydrothermal treatment (340 oC, 27.5 MPa, 25–204 s) of dilute (50 mM)
solutions of D-glucose and some other monosaccharides were studied.
The next logical step is performing the hydrothermal treatment on concentrated (500 mM)
solutions of D-glucose to check whether the concentration of the substrate has an influence
on the reaction pathways. Chapter 4 describes the results of this study. The product
distributions are compared with those obtained during the experiments under more diluted
conditions as described in chapter 3.
Since glycolaldehyde is one of the main products obtained from hydrothermal
reaction in both diluted and concentrated glucose solutions, it was of interest to study the
behavior of glycolaldehyde under HTU reactions. The results are described in Chapter 5.
Chapter 6 describes a study on hydrothermal reactions of glycolaldehyde and
glucose in the presence of formic acid as a potential hydride donor at various pH values.
The reactions were also performed with acetic acid, since this compound cannot act as
hydride donor. Furthermore, some metal ions and supported metals were screened as
catalysts.
32
General Introduction
References
1.
Timilsina, G.R.; Atmospheric Stabilization of CO2 Emissions: Near-term
Reductions and Intensity–based Targets, Policy Research Working Paper 4352,
The World Bank Development Reasearch Group Sustainable Rural and Urban
Development Team, September 2007.
2.
Searchinger, T., Heimlich, R., Houghton, R.A., Dong, F., Elobeid, A., Fabiosa, J.,
Takgoz, S., Hayes, D., Yu, T.; Science, 2008, 319, 1238.
3.
Rostrup-Nielsen, J.R.; Fuels and Energy for the Future Cat. Rev. 2004, 46, 247.
4.
Rostrup-Nielsen, J.R. In NATO ASI Sustainable Strategies for the Upgrading of
Natural Gas: Fundamentals, Challenges, and Opportunities; Kluwer: Vilamoura,
July 6–18, 2003.
5.
Mills, G.A., Rostrup-Nielsen, J.R.; Catal. Today 1994, 22, 335.
6.
Rostrup-Nielsen, J.R., Aasberg-Petersen, K., Schoubye, P.S.; Stud. Surf. Sci.
1997, 107, 473.
7.
International Energy Outlook 2007. Report: DOE/EIA-0484 May 2007.
8.
Zhang, Z., Men, Z.; Lianyou Jishu Yu Gongcheng 2003, 33(7), 58.
9.
Report to the Energy Watch Group, October 2007, EWG-Series No 3/2007,
http://www.energywatchgroup.org/fileadmin/global/pdf/EWG_Oilreport_102007.pdf
10. The BP Statistical Review of World Energy 2007, www.bp.com.
11. Renewables
Global
Status
Report
2006
Update,REN21,2006,05-16.
http://www.ren21.net/globalstatusreport/download/re_gsr_2006_Update.pdf.
12. Ackermann, T., Soeder L.; Renew. Sust. Energ. Rev. 4, 2000, 315.
13. Le Journal des Energies Renouvelables – June 2008.
14. Sugar Industry / Zuckerindustrie 133, 2008, 2.
15. Zuttel, A., Borgschulte, A., and Schlapbach, L.; Hydrogen as a Future Energy
Carrier, 2008, 4.
16. Takahashi, Y., Ono, Y.; PCT Int. Appl. 2005.
17. The World Offshore Renewable Energy Report 2004-2008
33
Chapter 1
18. Viola, V. E.; Handbook of Nuclear Chemistry 2003, 1, 137. Kluwer Academic
Publishers.
19. Sobajima, M.; JAERI-Review 1999, 99-011, i-vi, 1.
20. Quaak, P., Koef, H., and Stassen, H.; Energy from Biomass, 2002.
21. Petrus, L., Noordermeer, M.A.; Green Chem., 2006, 8, 861.
22. Hsu, T. A., Ladisch, M. R., Tsao, G. T.; Chem. Technol. 1980, 10, 315.
23. Hammerschlag, R.; Environ. Sci. Technol. 2006, 40, 1744.
24. Sjostrom E.; Wood Chemistry. Fundamentals and Applications. New York,
academic press, 1993.
25. Huber, G. W.; Dumesic, J. A.; Catal. Today 2006, 111, 119
26. Van der Toorn, B.; Shell Selected Papers, 1998 / Luijkx G.C.A. 1994,
Hydrothermal conversion of carbohydrates and related compounds. Ph.D.
Thesis, Delft University of Technology.
27. Centi, G. and van Santen, R.; Book of Catalysis for Renewables, Wiley-VCH,
Weinheim, 2007, 21.
28. Le Journal des Energies Renouvelables, 2007, 182, 57.
29. Beenackers, A. A. C. M., Swaaij, W. P. M. v. In Thermochemical Processing of
Biomass; Bridgwater, A. V., Ed.; Butterworth: London, U.K., 1984.
30. Klass, D. L.; Biomass for Renewable Energy, Fuels and Chemicals; Academic
Press: San Diego, 1998.
31. Rostrup-Nielsen, J. R.; Catal. Today 2002, 71, 243.
32. Paisley, M.A.; Kirk-Othmer Encyclopedia of Chemical Technology, 5th edition,
2004, p. 683.
33. Milne, T. A., Evans, R. J.; Abatzoglou, N. Biomass Gasifier Tars: Their Nature,
Formation and Conversion; Report No. NREL/TP-570-25357; National
Renewable Energy Laboratory: Golden, CO, 1998.
34. Klass, D. L. Biomass for Renewable Energy, Fuels and Chemicals; Academic
Press: San Diego, 1998.
35. Bauen, A.; Encyclopedia of Energy; Cleveland, C. J., Ed.; Elsevier: Amderdam,
2004, 1.
36. Devi, L., Ptasinski, K. J. Janssen, F. J. J. G.; Biomass Bioenergy 2003, 24, 125.
34
General Introduction
37. Dayton, D.; A Review of the Literature on Catalytic Biomass Tar Destruction,
Report No. NREL/TP-510-32815; National Renewable Energy Laboratory:
Golden, CO, 2002.
38. Rapagna, S., and N., Kiennemann, A., Foscolo, P. U.; Biomass Bioenergy 2000,
19, 187.
39. Baker, E. G., Mudge, L. K., Brown, M. D.; Ind. Eng. Chem. Res. 1987, 26, 1335.
40. Sutton, D., Kelleher, B., Ross, J. R. H.; Fuel Process. Technol. 2001, 73, 155.
41. Matsumura, Y., Minowa, T., Potic, B., Kersten, S. R. A., Prins, W., Swaaij, W. P.
M. v., Beld, B. v. d., Elliott, D. C., Neuenschwander, G. G., Kruse, A., Antal, M.
J.; Biomass Bioenergy 2005, 29, 269.
42. Higman, C., v. d. Burgt, M.; Gasification 2nd Edition, Elsevier 2008, 82.
43. NPT procestechnologie, 2008, 15 februari.
44. Wyman, C.; In Encyclopedia of Energy; Cleveland, C. J., Ed.; Elsevier: London,
2004, 2.
45. Weber, J.; Journal des Energies Renouvelables, 185, 2008, 49.
46. Wooley, R., Ruth, M., Glassner, D., Sheehan, J.; Biotechnol. Prog. 1999, 15, 794.
47. Aden, A., Ruth, M., Ibsen, K., Jechura, J., Neeves, K., Sheehan, J., Wallace, B.,
Montague, L., Slayton, A., Lukas, J.; Lignocellulosic Biomass to Ethanol Process
Design and Economics Utilizing Co-Current Dilute Acid Prehydrolysis and
Enzymatic Hydrolysis for Corn StoVer; Report No. NREL/TP-510-32438;
National Renewable Energy Laboratory: Golden, CO, 2002.
48. a. Chem. Ing. May 2008.
b. Shell, press release 15 July, 2008.
49. Speers, P.; Chem Ind. 2008, p. 23.
50. Goudriaan, F., Peferoen, D. G. R.; Chem. Eng. Sci., 1990, 45, 2729.
51. Galkin, A.A., Lunin, L.L.; Russ. Chem Rev. 2005, 74, 21.
52. Schmidt, Properties of Water and Steam in SI-Units, Berlin 1969. Springer
Verlag, Berlin, Germany.
53. Van Eldik, R., Hubbard, C.D.; Chemistry under Extreme or Non-Classical
Conditions, Wiley & Spektrum. Akademischer Verlag, New York, 1997.
54. Eckert, C.A., Chandler, K.J.; Supercrit. Fluids, 1998, 13, 187.
35
Chapter 1
55. Kabyemela, B.M., Takigawa, M., Adschiri, T., Malaluan, R.M., and Arai, K.;
Ind. Eng. Chem. Res. 1998, 37, 357.
36
Chapter 2
Experimental set-ups for studying reaction pathways
during hydrothermal reactions
37
Chapter 2
2.1 Introduction
Hydrothermal liquefaction is a process for the conversion of biomass into organic
product oil. It has been widely studied in the early 1980s [1]. The early process
development was mainly based on experiments in autoclaves and in a continuous bench
scale unit. The set ups used in that research have been reviewed previously [1]. A recent
conceptual design of a plant for hydrothermal upgrading of biomass is shown in Fig 2.1 [2].
Figure 2.1 The conceptual HTU process scheme and the product applications [2].
Since the results of the initial laboratory scale experiments were promising, a
continuous pilot plant with an intake capacity of 20 kg/h (dry basis) has been designed and
built during 1999 in Apeldoorn [1]. The schematic layout of this pilot plant is shown in Fig.
2.2.
38
Experimental set-ups for studying reaction pathways during hydrothermal reactions
Figure 2.2 The schematic layout of the continuous pilot plant in Apeldoorn.
The pilot plant process operates in a temperature range from 300 to 350 oC and at
pressures from 10 to 18 MPa. Biomass is transported via a high-pressure pump to reactor 1,
where the biomass mixture is brought to the desired reaction temperature. At this
temperature, the reaction mixture is injected in reactor 2 and stays there between 5 and 20
min (residence time). Finally the reaction mixture is depressurized and ran down in
collector 1. The reaction mixture is initially cooled (cooler 1) and gas liquid separation
takes a place. The gas phase (mainly CO2) goes to a condenser and the liquid phase is
cooled further (cooler 2) and finally collected (collector 1 and 2). With this pilot plant
biomass was converted to a heavy organic liquid with a lower heating value of 30-35 MJ/kg
and it has been used to optimize the process conditions [2].
The aim of the research described in this thesis is to provide some insight into the
reaction pathways that play a role in the HTU process. Knowledge of the reaction paths will
allow a more rational fine-tuning of the HTU process.
39
Chapter 2
The reaction systems involved in the HTU process are very complex and,
therefore, it is important to have a detailed picture of the initial reaction steps, which means
that equipment is needed that allows very short reaction times under the reaction conditions
of the HTU process. For the design of this equipment the following requirements should be
considered:
-
The reactor should allow a very fast heating-up to the desired reaction temperature
(about 340 oC) and also a very fast cooling down to room temperature. The heatup and cool-down times should be short (2 min) compared to the reaction time (15
min). The reactor should have possibilities to monitor the temperature.
-
The reactor should operate at high pressure (about 280 bar) and this pressure
preferably should be controllable.
-
Good mixing in the reactor should be possible. For continuous types of reactors
the Reynolds number preferably should be such that the flow is turbulent.
-
The design of continuous reactors should be such that the risk of blocking by tar
and char formed during the reaction is minimal.
-
Leaching of metals from the reactors should be minimal, since these may influence
the reaction paths.
In this Chapter first some typical examples of equipment for the study of
hydrothermal reactions reported in the literature will be reviewed and then the set-ups used
in the research described in this thesis will be described in detail.
2.2 Reported systems in the literature
Currently there are a few descriptions of experimental set-ups available in the
literature. Kabyemela et al. [3] have reported a continuous flow system to study initial
products of the HTU process at extremely short residence times (between 0.02 – 2.0 s) with
a very good temperature controlling (see Fig. 2.3). The heart of the reactor was made of
stainless steel (ss 316) and consisted of several pieces of metal pipe connected by screw-
40
Experimental set-ups for studying reaction pathways during hydrothermal reactions
connectors. The length of the central part of the reactor determined the residence time. The
reactor was placed in a metal salt bath, which allowed rather fast heating up.
The temperature was measured at several places inside of the reactor. A cooler
located at the end of the reactor stopped the reaction due to decreasing the temperature of
the reaction mixture. The pressure inside the reactor was controlled by a back pressure
regulator. Water was added to a concentrated solution of the feedstock via a pre-heater. The
feedstock solution was entered by another pump without pre-heating. The temperature
profile of the reactor has not been reported.
Figure 2.3 Schematic diagram of experimental set-up [3]
(extremely short residence times 0.02 – 2.0 s).
41
Chapter 2
The internal diameter of the reactor (1.18 mm) is small. The Reynolds number
(assuming flow rate as 20 ml/min) is about 360 and, therefore, the stream inside the reactor
is laminar. Because of the small internal diameter of the reactor, this system is suitable only
for experiments with low concentrations of the feedstock. Tar and char, which usually are
formed during reactions at high concentration, will block the reactor.
A continuous reactor set-up to study HTU reactions on some biomass model
compounds with a low concentration (50 mM) and with somewhat longer residence times
has been reported by Luijkx [4], see Fig 2.4.
Figure 2.4 Continuous tubular reactor set-up for hydrothermal conversion [4]. 1.Water,
2.Reactant
solution,
3.Waters 6000A HPLC pump, 4.Electrical heating tape,
5.Coolingwater jacket, 6.Hastelloy C-276 reactor tube (i.d., 1.43 mm; length, 3.18 m),
7.Perkin Elmer F11 GC oven, 8.Tradinco dome-loaded back pressure regulator, 9.Barnet
deadweight tester, PI, pressure indicator, TI, temperature indicator.
42
Experimental set-ups for studying reaction pathways during hydrothermal reactions
It consisted of a Hastelloy C-276 tube (id. 1.43 mm: length, 3.18 m) fed by a
Waters 6000A HPLC pump (0.1 to 9.9 ml/min, maximum pressure 40 MPa). The tube
reactor was heated in a Perkin-Elmer F11 GC oven with an additional heating via a tightly
fitted electrical heating tape at the first 10 cm of the reactor. The temperature was measured
between the heating tape and the tube reactor as well as in the oven itself. The reactor is
equipped with a cooling-water jacket at both ends of the reactor to assure fast cooling down
at the outlet, and to prevent heating up before the entrance of the reactor tube. The
residence time was depending on the flow of the HPLC pump and the density of the
feedstock under the applied conditions. A Tradinco dome-loaded pressure regulator with a
teflon membrane controls the pressure in this reactor. The surface temperature of the
reactor tube was measured at several positions near the exit and the entrance of the reactor
using an Everest Interscience 2400 IR temperature meter. From the dimensions of the
reactor it can be estimated that the Reynolds number in this reactor (assuming flow rate as 5
ml/min) is 74 and, therefore, the stream inside the reactor is laminar. Due to the small
diameter of the reactor tube blocking by tar and char will be a major problem when high
concentrations of the feedstock are used.
Jakab et al. used a very interesting reactor system for the thermal decomposition of
wood and cellulose in the presence of solvent vapours [5] (see Fig 2.5). The solid sample
was placed inside the reactor and heated up to 400 oC. The fluid medium (methanol, water,
or methanol +5% water) was continuously pumped through the reactor. The outlet of the
reactor was connected via a solenoid valve to a gas chromatograph that allows analyzing
immediately the formed gas products.
Kruse et al. studied the reaction of wet biomass using an experimental set-up
shown in Fig. 2.6 [6]. The experiments were performed in a stirred batch reactor made from
a nickel based alloy Nimonic 90. Cartridge type heaters heated up the reactor and the
temperature was controlled by a thermocouple in the reactor wall.
43
Chapter 2
Figure 2.5 Experimental set-up for thermal decomposition of wood and cellulose in the
presence of solvent vapours [5].
Another thermocouple was placed in the interior of the reactor to measure the temperature
of the reaction mixture. In this set-up water was heated to the desired temperature
separately and a horizontally opposed double screw press pressed the biomass into the
reactor. This was done to avoid chemical reaction by heating the biomass for a too long
period of time. Valve (V12) allows expanding the reaction mixture to the normal pressure
and cooler (K1) to cool the reaction mixture to room temperature. The reaction mixture was
separated into gaseous and liquid phases in separator (B5). A wet gas meter measured the
total amount of gases. A significant advantage of this set-up is that the content of the
reactor can be mixed and that the total amount of gases can be measured. Since the
diameter of the reactor is rather large, solid products formed during the reaction have no
risk of blocking it.
44
Experimental set-ups for studying reaction pathways during hydrothermal reactions
Figure 2.6 Experimental set-up for study reaction on wet biomass [6].
Knezevic et al. used a quartz capillary of 2 mm ID, 4 mm OD and ca. 90 mm in length (V ≈
0.3 ml) for batch micro reactions [7]. The capillaries were filled with a known amount of
feedstock, the air in the capillaries was replaced by nitrogen, after which the capillaries
were sealed in a hydrogen flame. The capillaries were heated to the reaction temperature in
an electrical oven (see Fig 2.7). The oven had several loopholes to allow observation of the
quartz capillary and to have the possibility of taking photographs during the reaction. The
capillaries were attached via a metal ring to a motor that allows shaking capillaries during
the reaction, to facilitate phase separation and to prevent the occurrence of elongated gas
bubbles in the capillaries.
45
Chapter 2
Figure 2.7 Schematic picture of the capillary set-up described by Knezevic et al. [7].
The temperature in the oven was controlled by a thermocouple connected to the
temperature controller. After reaction, the capillaries were rapidly cooled in a water bath.
This is an excellent set-up to perform batch micro reactions. The small amount of chemicals
that are used for the reaction and the low cost of the capillaries make the experiments very
cheap. Quartz itself is resistant to corrosion, has almost no catalytic activity and can survive
high temperature and high pressure. Rapid heating up and cooling down is no problem.
The possibility of taking photos and visual inspection of the reaction progress may
provide useful information. Unfortunately, the pressure during the reaction in this set-up
cannot be controlled. Due to the small volume of the reactor, the analysis of the reaction
products could be problematic.
Potic et al. used quartz capillaries (150 mm in length) heated in a fluid bed with nitrogen
supply [8]. Using a sand fluid bed, very fast heating up of the reaction mixture was
achieved. The temperature in the system was controlled by several thermocouples. After
reaction, the capillary was cooled to room temperature and then placed in a metal chamber.
46
Experimental set-ups for studying reaction pathways during hydrothermal reactions
Then this chamber was flushed with helium and closed at both sides with valves.
Subsequently the capillary was crushed with a hammer mechanism. A gas sample was
taken from the chamber and introduced into a gas chromatograph for the analysis. This setup is particularly suitable for studying gaseous products formed during HTU reactions.
Figure 2.8 Quartz capillaries heated in the fluid bed and gas analysis system described by
Potic et al. [8]. 1. Capillary, 2. Nitrogen supply for the fluidized bed, 3. Oven, 4. Fluidized
bed, 5. Thermocouples, 6. Capillary holder, 7. Sampling chamber, 8. Pressure indicator,
9.Gas chromatograph, 10. Temperature controller for the oven.
47
Chapter 2
2.3 The experimental set-ups applied in the research described in this
thesis
2.3.1 Small autoclave
For batch reactions on concentrated feedstocks a small autoclave of stainless steel
(ss 316) was made (see Fig. 2.9). The length of the reactor was 320 mm and internal
diameter 4 mm. The reactor was heated in a fluid sand bed. A thermocouple was located
inside the reactor and another one was placed in the fluid sand bed. The desired temperature
was reached in about 2 min. and cooling down took about 10 s.
The autoclave was filled with 2 ml of 0.5 M substrate, using a syringe with a long
needle. After closing the top cover, the reactor was placed for the desired period of time
(residence time) in the fluid bed. After that, the reactor was cooled down by ice water to 1
o
C. Then, the autoclave was taken out of the cooling bath and stabilized at room
temperature for a period of 20 min. After that, the valve was connected to a gas burette,
which was used to collect any gasses formed during the reaction. All the other products
were taken out of the reactor manually.
Figure 2.9 Small batch autoclave for HTU reaction. A – single autoclave, B – autoclave
placed in fluid bed.
48
Experimental set-ups for studying reaction pathways during hydrothermal reactions
The advantage of this set-up is that the temperature can be measured inside the
reactor, the fluid bed allows a very short heating up time and gasses, which are formed
during the reaction, can be released via a valve. Due to the simple construction of the
reactor, solid products could be easily removed. A significant disadvantage of this set-up is
the impossibility of controlling the pressure during the reaction.
2.3.2 Batch micro reactors
For batch experiments on micro scale, a set-up similar to that described by
Knezevic et al. (see above) [7] was applied. Quartz capillaries of 2 mm ID, 4 mm ED and
90 mm in length (V ≈ 0.3 ml) were used as batch micro reactors. The capillary was charged
with a known amount of feedstock and sealed in a hydrogen flame.
To perform the reaction, the capillary was placed in a fluid bed of sand for 15 min
(residence time) and after reaction, it was cooled in ice water. Due to the application of a
fluid bed for heating, the heat-up time was extremely short, whereas the cool-down time
was also extremely short due to the cooling of the capillaries in ice water. Other advantages
of the system are that the capillaries have a small size and low price. Therefore, the reaction
can be performed on many samples at the same time. Since the reactors are made of quartz,
leaching of metals is no problem. The disadvantage of the system is that the volume of the
reaction mixture is very small (can cause difficulties during analysis) and that the pressure
cannot be controlled.
2.3.3 Continuous reactor set-up
Hydrothermal reactions with highly diluted solutions of the model compounds (50
mM) were performed using a continuous flow type reactor (see Fig. 2.10) [9]. This reactor
(made of stainless steel 316, length 350 mm, internal diameter 4 mm) allowed the short
residence times required to study the initial products.
49
Chapter 2
Figure 2.10 Continuous flow type reactor set-up for HTU reaction.
A solid cylinder of aluminium with an outer diameter of 25 mm surrounded the
reactor. The reactor and the aluminium housing were heated in an oven. The temperature
inside the reactor was measured by a chromel-alumel thermocouple. The inlet and outlet of
the reactor were cooled by cooling water jackets, which decreased the temperature to less
than 30 ºC. The feedstock had usually a concentration of 50 mmol/L of the component
under study and was fed to the reactor using an HPLC pump at a rate of 1.0 – 9.9 mL/min
depending on the residence time required. The pressure in the reactor was controlled by a
Tradinco dome-loaded backpressure regulator with a teflon membrane to which a hydraulic
reference oil pressure was applied with a Barnet dead-weight tester (maximum pressure 70
MPa). The pressure regulator is designed to allow a continuous flow that is depressurized to
atmospheric pressure. It can be used for both liquids and liquid/gas mixtures. No pressure
fluctuations could be observed as long as the membrane was clean.
50
Experimental set-ups for studying reaction pathways during hydrothermal reactions
Reactor shut-down was performed by flushing with water during cooling-down, and if tar
had been deposited in the reactor, it was dissolved by pumping acetone through the reactor
after cooling down to room temperature.
The internal diameter of the reactor is 4 mm. The Reynolds number (assuming flow rate 5
ml/min) is about 27 and, therefore, the stream inside the reactor is laminar.
This set-up is proper to study HTU reactions using low concentrations (50 mM) of model
compounds representative for biomass. Due to an internal diameter (4 mm) of the reactor,
there is some risk of reactor blocking by solid products formed during the reaction,
particularly for the longer residence times (> 150 s). Unfortunately, performing HTU
reaction with high concentration of the feedstock, even with the shortest residence time (25
s), results in blocking the reactor. Some of the solid particles can be transported with the
flow to the membrane, therefore it was necessary to place a filter between the reactor outlet
and the membrane. Frequent cleaning of the membrane was necessary for a reliable
pressure control of the reactor.
2.3.4 Batch reactor set-up
The hydrothermal reactions of the highly concentrated solutions of the model
compounds were performed using a batch reactor set-up (see Fig. 2.11).
This reactor was constructed from stainless steal (metal ss 316). The reactor was filled by
means of an HPLC pump. The temperature was measured inside of the reactor by a
thermocouple. The reactor was heated by three heating elements on the outside of the
reactor. Another thermocouple was placed between the heating elements and the reactor
wall. The temperature profile is shown in Fig. 2.12.
51
Chapter 2
Figure 2.11 Batch reactor set-up for HTU reaction
A back pressure regulator controlled the pressure during the reaction. After
heating to the desired temperature, the reactor was kept at this temperature for a period 0-15
min. The preheating time in this system was about 20 min. To avoid overheating by
radiation, the reactor was isolated by water-cooling jackets on the both sides of the reactor.
A stream of a cold air cooled the reactor. A backpressure regulator controlled the pressure
in the system. In the described batch experimental set-up, the HTU reaction could be
studied with no limit on the residence time. A significant disadvantage in this set-up is that
heat up and cool down time were very long, (see Fig. 2.12).
52
Experimental set-ups for studying reaction pathways during hydrothermal reactions
400
Temperature (C)
350
300
250
200
150
100
50
0
0
20
40
60
80
Time (min)
Figure 2.12 Thermal profile for a batch reactor.
In this Chapter several experimental set-ups that are suitable to study HTU
reactions of different feedstocks have been described. In Table 2.1, a comparison of the
most important properties is given.
Table 2.1 Comparison of the properties of experimental set-ups described in this chapter.
Set-ups
Residence time
Mixing
Blocking
Figure 2.3
Extremely short
(0,02 – 2 s)
Short
(30 s - 51 min)
Good,
(laminar flow)
Good,
(laminar flow)
Very easy
Figure 2.5
Very long
(no limit)
Medium
Easy
Figure 2.6
Very long
(no limit)
Very long
(no limit)
Very good
No
No
No
Figure 2.8
Very long
(no limit)
No
No
Figure 2.9
Very long
(no limit)
Long
(25 – 204 s)
No
No
Good,
(laminar flow)
Easy
No
No
Figure 2.4
Figure 2.7
Figure 2.10
Figure 2.11
Very long
(no limit)
Very easy
53
Chapter 2
References
1.
Hydrothermal Conversion of Wet Biomass, A Review, Report GAVE-9919,
Utrecht 2000, 6.6, 77.
2.
Goudriaan, F., van de Beld, B., Boerefijn, F. R., Bos, G. M., Naber, J. E., van der
Wal, S., Zeevalkink, J. A.; Thermal efficiency of the HTU process for biomass
liquefaction. Progress in Thermochemical Biomass Conversion, [Conference],
5th, Tyrol, Austria, Sept. 17, 2000 (2001), Meeting Date 2000, 2, 1312.
3.
Kabyemela, B. M., Takigawa, M., Adschiri, T., Malaluan, R. M., Arai, K; Ind.
Eng. Chem. Res. 1997, 36, 1552.
4.
Luijkx G.C.A. 1994, Hydrothermal conversion of carbohydrates and related
compounds. Ph.D. Thesis, Delft University of Technology.
5.
Jakab, E., Liu, K., Meuzelaar, H. L. C.; Ind. Eng. Chem. Res. 1997, 35, 2087.
6.
Kruse, A., Gawlik, A.; Ind. Eng. Chem. Res. 2003, 42, 267.
7.
Knezevic, D., Rep, M., Kersten, S.R.A., Prins, W., van Swaaij, W.P.M.; Science in
Thermal and Chemical Biomass Conversion [Conference], Vancouver Island, BC,
Canada, 30.08-2.09. 2004, 2, 1082.
8.
Potic, B., Kersten, S.R.A., Prins, W., van Swaaij, W.P.M.; Ind. Eng. Chem. Res.
2004, 43, 4580.
9.
Srokol, Z., Bouche, A-G., van Estrik, A., Strik, R.C.J., Maschmeyer, T., Peters
J.A.; Carbohydr. Res. 2004, 339, 1717.
54
Chapter 3
Hydrothermal upgrading of biomass to biofuel;
studies on some monosaccharide model compounds
55
Chapter 3
3.1 Introduction
The anticipated depletion of fossil fuels and the concerns about global warming
ask for initiatives to search for renewable energy sources. Particularly interesting in this
respect are forms of biomass including wood, grass, agricultural waste and domestic waste.
Several technologies have been developed to convert biomass into a biofuel with a higher
heating value, such as gasification, fast pyrolysis [1], and hydrothermal upgrading (HTU)
[2, 3]. In the latter process, the biomass is treated during 5-20 min with water under
subcritical conditions (300-350°C, 10-18 MPa) to give a heavy organic liquid ("biocrude")
with a heating value of 30-35 MJ/kg. During this process, the oxygen content of the organic
material is reduced from about 40% to between 10 and 15%. The removed oxygen ends up
in CO2, H2O, and CO. The main components of biomass resources typically are lignin,
cellulose, hemicellulose, and minerals. Sasaki et al have studied the decomposition of
cellulose in water under subcritical and supercritical conditions [4]. After 1.6 s at 320 ºC
and 25 MPa, 47% conversion was obtained yielding hydrolysis products (cellobiose,
glucose etc., 44%) and decomposition products of glucose (erythrose, 1,6-anhydroglucose,
5-hydroxymethylfurfural, 3%) [4]. Furthermore, it has been shown that cellobiose
decomposes via hydrolysis to glucose and via pyrolysis to glycosylerythrose and
glycosylgycolaldehyde, which are further hydrolysed into glucose, erythrose, and
glycolaldehyde [5].
Therefore, glucose may be a good model compound for studying the reaction paths
of the HTU reaction of cellulose in biomass. Various groups have studied reactions of
glucose in water under HTU conditions [4-10]. Usually, these reactions were performed at
short reaction times (0.02-10 s). Remarkably, all studies indicate that the amount of CO or
CO2 that is formed is negligible, under the conditions applied, whereas the reaction
products identified usually can be ascribed to fragmentations and dehydratations. Recently,
a study of hydrothermal treatment under HTU conditions (310-410 °C, 30-50 MPa, 15 min)
of baby food as a model for phytomass (lignin free) was reported [11]. Large amounts of
CO2 were evolved under these conditions. A gap exists between the studies on glucose at
56
Hydrothermal upgrading of biomass to biofuel; studies on some monosaccharide model compounds
very short residence times and the latter study, which is closer to the conditions of the
actual HTU process.
Therefore, we report here the results of a systematic study on the hydrothermal
reactions of diluted solutions (50 mM) of glucose, fructose, mannose and galactose for
longer reaction times (25-200 s). To obtain more insight into the reaction paths, the C3-C5
sugars glyceraldehyde, erythrose, and arabinose, glycolaldehyde and the identified initial
reaction products of these compounds were treated under similar conditions. For
comparison, sorbitol was included as substrate. The results are discussed in relation to the
HTU process of biomass.
3.2 Results and discussion
The hydrothermal reactions of the model compounds were performed with 50 mM
solutions in water using a continuous flow-type reactor (see Chapter 2, Fig. 2.10). This
allowed short residence times (25-204 s) and rapid heating and cooling of the reaction
reaction mixture.
3.2.1
D-Glucose
(1), D-mannose (2), D-fructose (3), D-galactose (4), and
D-sorbitol
Hydrothermal treatment at 340 ºC and 27.5 MPa of 50 mM aqueous solutions of
these compounds resulted in brown solutions with pH values between 2.5 and 2.8 and small
amounts of tar. The amount of tar increased with the residence time (τ). No gases were
observed during these reactions. The liquid samples were complex mixtures of low
molecular weight compounds, most of which could be identified by HPLC and 1H NMR.
The 14 identified compounds are compiled in Table 3.1. These compounds have also been
detected in studies using shorter residence times, reported in the literature [4-9]. The major
initial compounds are glycolaldehyde (10) and 5-hydroxymethylfurfural (HMF, 7). These
compounds are clearly intermediates, as their concentrations initially increase until they
reach a maximum at τ = 30-100 s, after which they decrease (see Fig. 3.1).
57
Chapter 3
Table 3.1 Compounds that have been identified after HTU reactions of monosaccharides.
Name
Formula
5-hydroxymethylfurfural (7)
2-furaldehyde (9)
glycolaldehyde (10)
dihydroxyacetone (12)
glyceraldehyde (13)
1,2,4-benzenetriol (18)
pyruvaldehyde (20)
lactic acid (21)
acrylic acid (22)
acetaldehyde (23)
formic acid (24)
acetic acid (26)
glycolic acid (28)
acetone (30)
C6H6O3
C5H4O2
C2H4O2
C3H6O3
C3H6O3
C6H6O3
C3H4O2
C3H6O3
C3H4O2
C2H4O
CH2O2
C2H4O2
C2H4O3
C3H6O
Total amount of identified carbon
Total organic carbon
a
Concentration
(mmol/L)a
8.0
4.6
15.3
0.5
<0.1
2.5
0.6
1.6
<0.1
1.1
4.3
1.2
3.1
0.5
141.2 mmol C/L
234 mmol C/L
In reaction mixture of glucose (50 mM, 340 °C, 27.5 MPa, 120 s).
The concentration of all other compounds initially steadily increased to an almost
constant value at a τ-value of about 100 s. Qualitatively, no differences existed in the
products formed from the sugars investigated. However, the amounts of the various
compounds and their rates of formation depend strongly on the nature of the starting sugar.
The reaction mixtures obtained from these sugars contain compounds that are
characteristic for acid degradation of sugars next to others that are characteristic for basic
degradation.
For example, HMF (7) is a product that typically is formed upon acid
degradation of sugars [12], whereas glycolaldehyde (10) has been observed as an important
product in alkaline degradation [13-17]. The occurrence of both acidic and base catalysed
reactions can be ascribed to the increase of the value of the ion product of water, Kw,, near
the critical point [18-19]. Under those conditions Kw is about three orders of magnitude
higher than under ambient conditions.
58
Hydrothermal upgrading of biomass to biofuel; studies on some monosaccharide model compounds
Figure 3.1 Formation of 5-hydroxymethylfurfural (▲) and glycolaldehyde (●) during
the hydrothermal reactions of some monosaccharides (50 mM, 340° C, 27.5 MPa).
The similarity of the hydrothermal reaction products of D-glucose (1), D-mannose
(2), D-fructose (3), and D-galactose (4) may be explained by the base-catalysed Lobry de
Bruyn-Alberda van Ekenstein rearrangement, which results in interconversion of glucose
(1), mannose (2), and fructose (3) via the 1,2-enediol anion (5, see Scheme 3.1) [20]. It
should be noted that mannose and fructose have been observed in the reaction mixtures
obtained upon hydrothermal treatment of glucose using very short residence times [9].
Possibly, successive rearrangements via other enediol anions (e.g. 6) result in
interconversion of all C6-monosaccharides.
59
Chapter 3
OH
OH
O
HO
HO
D-mannose
2
OH
HO
HO
O
OH OH
D-glucose
1
HO
HC
OH
C
OH
OH
HO H2C
OH
O
OH
CH
HC
OH
HC
OH
CH 2OH
HO
D-fructose
3
CH 2OH
5
H 2C
OH
C
OH
OH
OH
O
HO C
HC
OH
HC
OH
CH2OH
HO
OH
OH
D-galactose
4
6
Scheme 3.1 Rearrangements of monosaccharides via enediol during the hydrothermal
reaction. For convenience, each sugar is presented by only one of its anomeric forms.
Although the reaction products of the four monosaccharides 1-4 are qualitatively
similar, a remarkable difference exists between the product distribution from fructose (3)
and those of the other sugars. For fructose (3), HMF (7) is the major reaction product and
glycolaldehyde (10) is much less abundant than with the other sugars. Of the four sugars
studied, fructose (3) has the energetically most favourable furanose form. Under ambient
conditions, 28% of this anomeric form occurs, whereas glucose (1), mannose (2), and
galactose (4) occur almost exclusively in pyranose forms. Furthermore, the isomerisation
rate of glucose to fructose is much faster than the reverse reaction [8]. This supports the
formation of HMF (7) via successive acid catalysed dehydrations of fructofuranose (3) as
proposed by Antal et al. (see Scheme 3.2) [12].
60
Hydrothermal upgrading of biomass to biofuel; studies on some monosaccharide model compounds
O
OH
HOH 2C
HO
OH
OH
OH
O
OH
CH 2O H
HO
3
CH 2O H
- 3 H2 O
O
HO H2 C
CHO
7
Scheme 3.2 Formation of furan derivatives during the hydrothermal reactions of
monosaccharides. Here the reaction of D-fructose is shown as an example.
Glycolaldehyde (10) can be formed from glucose in the open form via a retro-aldol
reaction (see below, Scheme 3.4). Formation of large amounts of this compound is
remarkable, since Kabyemela et al. have observed glyceraldehyde (13) as an important
reaction product after extremely short reaction times (τ = 0.02-2 s). Under the conditions
that we applied, glyceraldehyde was not observed (13); only minor amounts of its
isomerization product dihydroxyacetone (12, ≤ 1 mM) were detected. Both the formation of
glyceraldehyde (13) and its consecutive reactions are very fast [8-21], and, therefore,
possibly only the secondary reaction products are observed at the relative long residence
times that we applied (see also below).
During the hydrothermal reactions of the monosaccharides only minor amounts of
gases were formed; the amount was negligible for D-glucose (1). It should be noted that the
solubility of CO2 in water is about 33 mmol/L. However, a 13C NMR spectrum of a sample
taken at τ = 400 s showed that the amount of bicarbonate in the sample is negligible. The
molecular formulae of most of the identified compounds can be given as Cn(H2O)m (see
Table 3.1). Only some minor compounds had a different H/O ratio (acetaldehyde (23),
formic acid (24), glycolic acid (28), and acetone (30)). This supports the suggestion that
decarboxylation and decarbonylation are relatively unimportant pathways under the
conditions applied.
61
Chapter 3
The exchange of the OH and COOH protons and the water protons is rapid on the
1
H NMR time scale. Consequently only an averaged signal can be observed. For the
reaction of glucose, the total integral of the other (non-exchangeable) protons indicated that
the amount of this type of protons decreased from 350 to about 267 milliatom/L during the
first 120 seconds of the reaction (see Fig. 3.2). Similar behaviour was observed for the
other sugars. This indicates that about 24 mol % dehydration occurs under these conditions,
which corresponds with a decrease of the amount of oxygen with the same percentage.
Figure 3.2 The total amount of non-exchangeable protons of the organic substrate during
the hydrothermal reaction of glucose.
The total organic carbon number (TOC) of the samples at the longest residence
times showed that in the liquid samples between 60 and 96% of the total amount of carbon
was recovered. The losses can be ascribed to tar deposited on the wall of the reactor. Upon
increase of the concentration of the substrate the amount of tar increased, which frequently
resulted in blocks of the reactor.
Surprisingly,
D-sorbitol
remained unreacted when it was subjected to similar
hydrothermal conditions (50 mM solution, 240 s at 340 ºC and 27.5 MPa). This indicates
that the hemi-acetal function of the sugars is essential for their conversion. Most likely, the
enediol (and/or its anion) is an important intermediate for the basic catalysed pathways,
whereas cyclic structures are important in the acid catalysed reactions.
62
Hydrothermal upgrading of biomass to biofuel; studies on some monosaccharide model compounds
3.2.2
D-Arabinose
(8)
When the C5 monosaccharide D-Arabinose (8) was subjected to the hydrothermal
treatment, mainly glycolaldehyde (10) and 2-furaldehyde (9) were obtained. Smaller
amounts of glycolic acid (28), lactic acid (21), acetic acid (26), and acetaldehyde (23) were
identified in the reaction mixture (see Fig. 3.3). Apparently, similar reaction paths occur as
with the C6 monosaccharides. Once again, initially dehydration and retro-aldol
condensation seem to be important.
Figure 3.3 Distribution of the most important species during the hydrothermal treatment of
D-arabinose
acetone,
(50 mmol/l, 340° C, 27.5 MPa),
glycolic acid,
formic acid,
1,2,4-benzenetriol,
2-furaldehyde,
acetic acid,
glycolaldehyde,
lactic acid.
3.2.3 Reaction pathways via 5-hydroxymethylfurfural (7)
To untangle the reaction pathways of the reaction of
D-glucose
(1), we also
performed hydrothermal reactions on compounds that were identified as the initial reaction
products of this sugar. HMF (7) showed some decomposition; at τ = 400 s, the
concentration of HMF (7) in the reaction mixture was decreased from 55 to 28 mM. The 1H
NMR spectrum showed the presence of several minor reaction products of which 1,2,4benzenetriol (18) was by far the most abundant (1.3 mM at τ = 400 s). This compound was
63
Chapter 3
identified before by Luijkx et al [7]. It has been suggested that this compound is formed
from HMF (7) by hydrolysis of the furan ring followed by a rearrangement to a hexatriene
(17), electrolytic rearrangement and dehydration (see Scheme 3.3). 1,2,4-Benzenetriol (18)
is stable under HTU conditions, so it is one of the end products of this reaction. However,
aqueous samples of this compound are not stable in air. They slowly convert into a dimer
via an oxidative coupling.
HO
O
O
HO
HO
HO
OH
O
O
H2O
HO
OH
7
HO
HO
OH
HO
OH
-H 2O
HO
OH
HO
18
OH
HO
OH
17
Scheme 3.3 Pathway proposed for the rearrangement of HMF (7) to 1,2,4-benzenetriol (18).
No 2-furaldehyde (9) could be detected in the reaction mixtures from the
hydrothermal reaction of HMF. This implies that the 2-furaldehyde (9) that has been
observed in the reaction mixture from the C6 monosaccharides (see Table 3.1) has been
formed via another pathway, most likely through C5 ketoses as already proposed previously
by Kallury et al [8]. The latter compounds may be formed via several pathways. For
example, ketose 14 (see Scheme 3.4), which can fragment via a retro- aldol reaction into
formaldehyde (15) and a C5-monosaccharide (16).
Alternatively, the initially formed 1,2-enediol may be converted into the
corresponding 1,2-diketone by an α-elimination. Subsequent α-dicarbonyl cleavage results
in formic acid (24) and the C5-monosaccharide. The C5-monosaccharides can then
dehydrate to 2-furaldehyde (9) as has been demonstrated in the hydrothermal treatment of
D-arabinose
(8, see above). Most likely, first a Lobry de Bruyn-Alberda van Ekenstein
rearrangement to ribulose takes place. Then, following the conclusions of the work of Antal
et al. [12], three consecutive dehydration steps may lead to 2-furaldehyde (9).
64
Hydrothermal upgrading of biomass to biofuel; studies on some monosaccharide model compounds
H
O
C
H
HC
HO
OH
(ii)
CH
HC
OH
HC
OH
H
O
C
O
C
+
CH2OH
10
HC
OH
HC
OH
H
(ii)
O
C
2
CH 2OH
10
CH 2OH
11
CH 2OH
1
(i)
CH 2OH
C
HO
H
O
(ii )
CH
C
HC
OH
HC
OH
O
C
CH2OH
O
+
HC
CH 2OH
OH
CH 2OH
13
CH2OH
12
(i)
CHO
3
CHOH
(i)
CH 2OH
HC
C
HC
HC
CH 2OH (i )
C O
(ii )
H2C
OH
O
+
15
HC
OH
HC
OH
O
OH
CH 2OH
14
(ii)
H
O
C
CH2OH
+
CH 2OH
10
HC
C
OH
HC
OH
CH 2OH
(ii)
CH 2OH
16
OH
HC
H
CH2OH
O
C
+
CH 2OH
10
C
O
CH2OH
13
OH
O
CH2OH
Scheme 3.4 Fragmentation of monosaccharides via (i) Lobry de Bruyn-Alberda van
Ekenstein rearrangements and (ii) retro-aldol condensations. For convenience the
saccharides are represented in their linear forms.
Quantitatively, the compounds identified with NMR and HPLC cannot account for
the decrease in the concentration of HMF (7) observed during the reaction. We assume that
tar deposited on the wall of the reactor tube is responsible for this.
65
Chapter 3
3.2.4 Reaction pathways via glycolaldehyde (10) and D-erythrose (11)
An important pathway of the hydrothermal reaction of D-glucose and the other C6
sugars starts with a retro-aldol condensation to form glycolaldehyde (10, see Scheme 3.4).
The second product should be a C4 sugar, such as D-erythrose (11), but these sugars could
not be identified by 1H NMR because of the presence of many overlapping signals in the
region of the spectrum where the resonances of these compounds should be expected. Also
in the HPLC chromatogram no C4 sugars could be identified. To get some insight in
possible reactions starting from this type of sugars, we performed an analogous reaction
with D-erythrose (11). The main reaction product was glycolaldehyde, which can be formed
via a retro-aldol reaction.
It should be noted glycolaldehyde (10) cannot be formed directly from a retroaldol condensation from D-fructose (3). This compound should first isomerise to glucose to
enable this reaction. This is reflected in the low initial rate of formation of glycolaldehyde
from
D-fructose.
Glycolaldehyde (10), when subjected to the same reaction conditions,
converted with small yield into a complex mixture of compounds with NMR resonances at
chemical shifts (3-5 ppm) that are characteristic for polyhydroxy compounds (see Fig. 3.4).
Most likely these products are the result of condensation reactions.
Figure 3.4 Expanded 1H NMR spectrum of the reaction product of the hydrothermal
treatment of glycolaldehyde (13), showing the presence of a complex mixture of
polyhydroxy compounds.
66
Hydrothermal upgrading of biomass to biofuel; studies on some monosaccharide model compounds
3.2.5 Reaction pathways via glyceraldehyde (13)
Another initial product of the hydrothermal reactions of monosaccharides is, once
again as the result of a retro-aldol condensation reaction, glyceraldehyde (13, see Scheme
3.4). This reaction can only occur in 2-ketoses, such as fructose (3). Therefore, starting
from D-glucose (1) a Lobry de Bruyn-Alberda van Ekenstein rearrangement should precede
this route. Only minor amounts of glyceraldehyde (13) were detected, but up to 1 mM of
dihydroxyacetone (12) was observed at τ = 400 s during the hydrothermal treatment of
glucose. This compound can be formed from glyceraldehyde (13) via another Lobry de
Bruyn-Alberda van Ekenstein rearrangement.
When glyceraldehyde (13) was applied as the feedstock for the hydrothermal
reaction, a rapid conversion into a mixture of compounds was observed that contained most
minor compounds observed in the reaction of
D-glucose
(1): dihydroxyacetone (12),
pyruvaldehyde (20), lactic acid (21), acetaldehyde (23), formic acid (24), acrylic acid (22),
acetic acid (26), glycolic acid (28), and acetone (30). After 200 s, lactic acid (21) was the
major component (25 mM, see Fig. 3.5). The concentrations of pyruvaldehyde (20) and
acetone (30) have maxima after about 50 s, suggesting that these compounds are
intermediates. Pyruvaldehyde (20) can be formed from glyceraldehyde (13) by
dehydratation to compound 19 (see Scheme 3.5), followed by a keto-enol rearrangement.
Pyruvaldehyde (20) in turn can be converted into lactic acid (21) by means of a benzilic
rearrangement. An α-dicarbonyl cleavage can explain the formation of formic acid (24) and
acetaldehyde (23) and of acetic acid (26). The latter compound should be accompanied by
formaldehyde (25), which due to its high reactivity probably has escaped observation. It
may be concluded that the pathway glyceraldehyde (13) → pyruvaldehyde (20) →
acetaldehyde (23) + formic acid (24), formaldehyde (25) + acetic acid (26) is an important
route to compounds that have a decreased O-content with respect to glucose.
67
Chapter 3
Figure 3.5 Distribution of the major reaction products during the hydrothermal treatment of
glycolaldehyde,
a 50 mM solution of glyceraldehyde (13).
acetone,
glycolic acid,
lactic acid, o pyruvaldehyde.
H
O
H
C
HC
-H2O
OH
O
H
C
C
CH 2OH
13
O
C
OH
CH 3
20
OH -
C
OO
HC
OH
2H +
H +, -H 2O
COOCH
CH2
OH
O
HC
O-
CH 3
CH 3
21
C
O
C
CH2
19
H
C
OH
O-
C
O
H
C
OH
O-
C
O
CH3
CH3
OH
H
C
OH
HC
OH
CH3
O H-
O H-
O
C
-H 2O,-CO
CH 3
23
+
HCOO-
H
O
C
C
OH
OCH 3
CH2O
25
+
CH3COO26
24
22
Scheme 3.5 Possible pathways of the reactions during the hydrothermal treatment of
glyceraldehyde (13).
68
Hydrothermal upgrading of biomass to biofuel; studies on some monosaccharide model compounds
Experimentally, starting from lactic acid as feedstock (21), next to acrylic acid
(22), once again acetaldehyde (23) was observed. Both reactions are acid catalysed.
Dehydration results in acrylic acid (22), whereas an acid catalysed decarbonylation may
afford acetaldehyde (see Scheme 3.5).
3.2.6 Effects of added acid and base
Since both acid and base catalysed reactions were observed during the
hydrothermal reactions of D-glucose, we decided to investigate the effect of added acid and
base. In the presence of 6 mM NaOH, the acid catalysed reaction paths via HMF (7) and
furfural (9) appeared to be completely suppressed (see Fig. 3.6 b), neither HMF (7), 2furaldehyde (9) nor 1,2,4-benzenetriol (18) were observed. The amounts of lactic acid (21),
acetic acid (26), and acetaldehyde (23) were considerably higher, indicating that the
reaction paths via glyceraldehyde (10) are relatively important under basic conditions.
Major products were glycolic acid (28) and acetone (30), which were observed in all other
reactions only in minor amounts. Many possible pathways can be envisaged for the
formation of glycolic acid (28) and acetone (30), for example starting from the 2,3-enediol
6. An α-elimination followed by enol-keto rearrangement gives diketone (27), which may
give glycolic acid (28) and compound (29) after an α-dicarbonyl cleavage (see Scheme
3.6). The latter could be converted into acetone (30) and formic acid (24) via dehydratation,
enol-keto rearrangement followed by a base catalysed fragmentation (see Scheme 6).
In the presence of added HCl (6 mM), initially HMF (7) is formed (see Fig. 3.6c).
However, it rapidly decomposes to unidentified products. Most likely tars are being formed.
After the reaction a large amount of deposited tar was observed in the reactor. The base
catalysed reactions are not fully suppressed, glycolaldehyde (10) is still present in
considerable amounts. The most abundant products of the reaction in the presence of HCl
are acetaldehyde (23) and formic acid (24), which may be explained by reaction pathways
starting from initially formed glyceraldehyde (13, see Scheme 3.5).
69
Chapter 3
Figure 3.6 Distribution of the most important products during the hydrothermal reaction of
D
-glucose (50 mM, 340° C, 27.5 MPa), (a) without addition of acid or base, (b) in the
presence of 6 mM NaOH, (c) in the presence of 6 mM HCl
1,2,4-benzenetriol,
furaldehyde,
70
glycolaldehyde,
acetic acid,
acetone,
lactic acid.
5-hydroxymethylfurfural,
glycolic acid,
formic acid,
2-
Hydrothermal upgrading of biomass to biofuel; studies on some monosaccharide model compounds
H
O
C
HC O H
HC O H
C
OH
CH 2OH
CH 2OH
C
C
OH
C
OH OH -
C
OH
HC
OH
CH
HC
OH
HC
OH
CH2OH
HC
CH 2OH
O
OH
C
O
C
CH2
OH
CH2OH
29
+
H , -H2O
O OH
-
OH
HC
CH 2OH
OH
H
H
O
C
CH 2
CH 2
C
C
OH
CH 2
O
CH 3
OH -
C
O
CH 3
HC
OH
CH2OH
29
OH
OC
CH 2
COO 28
+
H
O
C
CH2
CH 2OH
27
H
C
CH 2OH
O
CH 2
6
O
HC
O-
C
CH2OH
H
CH 2OH
C
HC
HO CH
HC O H
CH2OH
H+
HCOOH
25
+
CH3
C
O
CH3
30
Scheme 3.6 Possible pathways during the hydrothermal treatment of glucose.
3.3 Conclusions
During the initial stages of the hydrothermal treatment (340 ºC, 27.5 MPa) of a low
concentration of C6 monosaccharides (50 mM) both acid and base catalysed reactions play
a role. The acid catalysed reactions are mainly dehydrations with HMF as a major
intermediate. Several base catalysed reactions lead to fragmentations via, for example,
retro-aldol condensation and beta elimination. Decarbonylation and decarboxylation
reactions play a minor role at the high dilution of the feedstocks used in this research. Some
highly reactive compounds, including glycolaldehyde (10) and glyceraldehyde (13) are key
intermediates in the pathways that were observed in this investigation. It may be expected
that, at higher concentrations, condensation reactions may occur similar to those observed
for the alkaline degradation of sugars at lower temperature [12-15]. The occurrence of
condensation reactions was already observed during the reaction of glycolaldehyde (10),
which resulted in significant amounts of condensation products. Therefore, at the higher
71
Chapter 3
concentrations that are applied for the HTU reaction of biomass it may be expected that
fragments formed subsequently enter in condensation reactions with other fragments to
afford a complex mixture of compounds. In this dynamic system irreversible reactions
similar to those observed in this study may occur, including the formation of 1,2,4benzenetriol (18). Reactions similar to that of the pathway glucose (1) → glyceraldehyde
(13) → pyruvaldehyde (20) → acetaldehyde (23) + CO / formaldehyde (25) + acetic acid
(26) may ultimately lead to the desired reduction of the O-content of the liquid reaction
products. These pathways seem to be favoured under more acidic conditions, whereas
under more basic conditions pathways similar to those observed for basic degradation of
sugars seem to be dominant. Then, routes via diketones to glycolic acid (28) and acetone
(30) seem to be of special interest with regard to the reduction of the O-content during the
HTU reaction. Furthermore, it may be expected that condensation reactions of the initially
formed products may afford new compounds and maybe also new pathways may become
operative. This might include pathways that result in formation of CO2.
Another important conclusion of this research is that the HTU reaction mixture of
biomass may contain considerable amounts of highly reactive aldehydes, which will have a
bad influence on the stability of the biocrude. Further upgrading, for example by
hydrodeoxygenation may be required to obtain a more stable fuel.
3.4 Experimental
3.4.1 Chemicals
All chemicals used were obtained from Aldrich. A 50 mM solution of
glycolaldehyde in water was heated for 10 min at 60 ºC to hydrolyze any oligomers present
just prior to the hydrothermal reaction. All other substrates were used without further
purification. The water used was demineralized and had a conductivity of 18.2 mΩ·cm.
72
Hydrothermal upgrading of biomass to biofuel; studies on some monosaccharide model compounds
3.4.2 Apparatus and method
The home-built apparatus used for the HTU reactions is described in detail and
schematically depicted in chapter 2, Fig. 2.10.
3.4.3 Analyses
At each residence time, three samples of 10 mL were taken. Any solid particles
formed were centrifuged or filtered off and the clear solutions were analysed by HPLC and
1
H NMR. Assignments of peaks were made by comparison of chromatograms and spectra
with those of authentic samples. The quantitative results of the HPLC and 1H NMR
analyses agreed with each other within the experimental error (ca 5%). Some of the samples
were also analysed for total organic carbon (TOC).
The HPLC analyses were carried out with a Millipore-Waters 590 pump and
Phenomenex Rezex organic acid column at 60 ºC. The eluent was 0.01 M aq trifluoroacetic
acid and the flow rate 0.5 mL/min. A refractive index (RI) (Shodex model SE-51) detector
and an ultraviolet (UV) (Shimadzu spectrometric detector model SPD-6A) at 215 nm were
used. The peak areas were determined using an integrator (Kipp & Zonen). HPLC
quantitation was achieved using THF as an internal standard.
1
H NMR spectra were measured on a Varian Unity INOVA-300 spectrometer at
300 MHz). A weighed amount of tert-butanol in D2O was added to the samples to lock and
as an internal standard. The chemical shifts are reported with respect to the CH3 signal of
tert-butanol, which was set at 1.20 ppm. The water resonance was suppressed using
presaturation with the transmitter. The peak areas were determinded by deconvolution of
the spectrum with lorentzian peaks.
Total Organic Carbon analyses were performed using a Shimadzu T.O.C. 5050A
instrument.
73
Chapter 3
Acknowledgements
Thanks are due to E. Wurtz and R. Sloter for assistance with the construction of
the experimental set up and to Dr F. Goudriaan, Dr L. Petrus, and Dr. G.C.A. Luijkx for
helpful discussions. We are also indebted to K. Djanashvili and J. Knoll for performing the
NMR and T.O.C. analysis.
74
Hydrothermal upgrading of biomass to biofuel; studies on some monosaccharide model compounds
References
1.
Bridgwater, A.V.; Chem. Eng. J., 2003, 91, 87.
2.
Goudriaan, F., and Peferoen, D.G.R.; Chem. Eng. Sci. 1990, 45, 2729.
3.
Goudriaan, F., van de Beld, B., Boerefijn, F.R., Bos, G.M., Naber, J.E., van de
Wal, S., and Zeevalkink, J.A., in A.V. Bridgwater (Ed), Proceedings Progress in
Thermochemical Biomass Conversion, Blackwell Science Ltd, Oxford UK, 2001.
4.
Sasaki, M., Fang, Z., Fukushima, Y., Adschiri, T., and Arai, K.; Ind. Eng. Chem.
Res. 2000, 39, 2883.
5.
Kabyemela, B.M., Takigawa, M., Adschiri, T., Malaluan, R.M., and Arai, K.; Ind.
Eng. Chem. Res. 1998, 37, 357.
6.
Bobleter, O., and Bonn, G.; Carbohydr. Res., 1983, 124, 185.
7.
Luijkx, G.C.A, van Rantwijk, F., van Bekkum, H., and Antal, M.J.; Carbohydr.
Res. 1995, 272, 191.
8.
Kabyemela, B.M., Adschiri, T., Malaluan, R.M., and Arai, K.; Ind. Eng. Chem.
Res. 1997, 36, 1552.
9.
Kabyemela, B.M., Adschiri, T., Malaluan, R.M., and Arai, K.; Ind. Eng. Chem.
Res. 1999, 38, 2888.
10. Kallury, R.K.M.R., Ambidge, C., Tidwell, T.T., Boocock, D.G.B., Agblevor, F.A.,
Steward, D.J.; Carbohydr. Res. 1986, 158, 253.
11. Kruse, A., and Gawlik, A.; Ind. Eng. Chem. Res. 2003, 42, 267.
12. Antal, M.J.Jr., Mok, W.S.L., and Richards, G.N.; Carbohydr. Res. 1990, 199, 91.
13. De Bruijn, J.M., Kieboom, A.P.G., van Bekkum, H., and van der Poel, P.W.;
Sugar Technol. Rev. 1986, 13, 21.
14. De Bruijn, J.M., Kieboom, A.P.G., and van Bekkum, H.; Starch 1987, 39, 23.
15. De Bruijn, J.M., Touwslager, F., Kieboom, A.P.G., and van Bekkum, H.; Starch
1987, 39, 49.
16. De Bruijn, J.M., Kieboom, A.P.G., van Bekkum, H., van der Poel, P.W., de
Visser, N.H.M., de Schutter, M.A.M.; Int. Sugar J. 1987, 89, 208.
17. Ellis, A.V., and Wilson, M.A.; J. Org. Chem. 2002, 67, 8469.
75
Chapter 3
18. Akiya, N.A., and Savage, P.E.; Chem. Rev. 2002, 102, 2725.
19. Marshall, W.L., and Franck, E.U.; J. Phys. Chem. Ref. Data 1981, 10, 295.
20. Lobry de Bruyn, C.A., and Alberda van Ekenstein, W.; Recl. Trav. Chim. PaysBas, 1895, 14, 201.
21. Kabyemela, B.M., Adschiri, T., Malaluan, R., and Arai, K.; Ind. Eng. Chem. Res.
1997, 36, 2025.
22. Luijkx, G.C.A., van Rantwijk, F., and van Bekkum, H.; Carbohydr. Res. 1993,
242, 131.
76
Chapter 4
Hydrothermal reactions of glucose; the influence of
concentration on reaction pathways
77
Chapter 4
4.1 Introduction
In view of the concerns about global warming and the expected shortage of fossil
fuels in about 60 years from now, it is necessary to develop alternatives for the production
of energy from cheap and renewable energy sources. One of the possibilities is the use of
biomass particularly in the form of wood, grass, agriculture and domestic waste. Promising
technologies to produce energy from biomass include gasification [1], fast pyrolysis [2] and
hydrothermal upgrading (HTU) [3]. The latter is a hydrothermal conversion of biomass into
an organic crude oil (biocrude) with a heating value approaching that of fossil fuels [4].
During this process the oxygen content of the organic material is reduced from about 40%
to 10-15%. The removed oxygen ends up in CO2 and water.
The main components of biomass resources are lignin, cellulose, hemicelullose
and minerals. It has been reported that during the initial stage of the HTU process,
cellobiose and other polysaccharides hydrolyze to glucose. Therefore, glucose is a good
model compound to elucidate the reaction pathways of the HTU process.
Several groups have studied the decomposition of glucose in water under HTU
conditions [5-11] using short reaction times and low concentrations. Previously we have
studied the reaction pathways of a 50 mM aqueous solution of glucose at 340 oC and 280
bar [12]. Water soluble products were obtained under these conditions. These products
could be explained by both acid and base catalysed reactions as a result of the relatively
high value of Kw under subcritical conditions. The main products are formed via
dehydrations,
rearrangements
and
fragmentations.
Glycolaldehyde
and
5-
hydroxymethylfurfural (HMF) are major initial products obtained from diluted glucose
solution (50 mM) under HTU conditions. Surprisingly, no carbon dioxide and tar were
detected in the reaction products under these conditions.
However, Kruse et al. have reported that baby food (lignin free) when treated
under HTU conditions (310-410 oC, 30 –50 MPa, 15 min) gives large amounts of CO2 [13].
Minowa et al. have demonstrated that cellulose, under hydrothermal conditions, is initially
converted in water-soluble low molecular weight compounds, which subsequently are
forming gases, tar and char [14]. Therefore, condensation reactions of these initial reaction
products may play an important role in the formation of biofuel (tar). This suggests that the
78
Hydrothermal reactions of glucose; the influence of concentration on reaction pathways
concentration of the biomass and possibly also the reaction time may be important
parameters in relation to the reaction pathways leading to biofuel.
In this Chapter we report the results of a study of the effect of the concentration on the
pathways in the hydrothermal reactions of the model compound glucose.
4.2 Results and discussion
Previously we have studied hydrothermal reactions of glucose using diluted
solutions (50 mM) in a continuous flow reactor at 340o C and 280 bar and reaction times of
25-200 s [12]. Under those conditions the amounts of tar and char formed were negligible.
However, an attempt to perform similar reactions at higher concentrations in this setup
failed due to blocking of the reactor by the considerable amounts of tar and char formed
under those conditions. Therefore, we performed the reactions at higher concentrations in a
small batch reactor (6 ml, reactor 1) without pressure control and in a larger reactor (150
ml) with pressure control (reactor 2, for a detailed description of these set-ups, see Chapter
2). The disadvantage of the latter set-up was that the heat-up and cool-down times of this
reactor were large (both about 20 min) and therefore, only relatively long reaction times
could be studied. With the small reactor (reactor 1) heat-up and cool-down times were only
about 1.5 min and 10 s, respectively. Hydrothermal reactions with these setups with diluted
glucose solutions (50 mM) gave similar results as obtained previously with the continuous
flow reactor [12]. At higher concentrations (250-500 mM) completely different product
compositions were obtained: now in addition to water-soluble products, substantial amounts
of tar, char, and gases were produced. Gas production started at about 240 °C. Here, we
define tar as tetrahydrofuran (THF) soluble and char as THF insoluble products.
Unfortunately, due to experimental problems, we have not been able to offer
accurate mass balances. However, we feel that the trends reported below give important
information about the reaction paths during hydrothermal reactions of glucose.
79
Chapter 4
TAR
150
mmol C
100
50
0
0
5
10
Residence time (min)
CHAR
150
mmol C
100
50
0
0
5
10
Residence time (min)
GAS
20
mmol C
15
10
5
0
0
5
10
Residence time (min)
Figure 4.1 Amount of tar, char, and gases obtained during the hydrothermal reactions of
150 mL of an aqueous solution of glucose (• 50 mM, □ 250 mM, ♦ 500 mM, corresponding
to an intake of 45, 225, 450 mmol C), reaction temperature 340 °C, pressure 17.5 MPa. The
reactions were performed in reactor 2. The residence time is defined as the reaction time
after reaching the desired temperature. The curves are guides to the eye.
80
Hydrothermal reactions of glucose; the influence of concentration on reaction pathways
The amounts of tar, char, and gases obtained in reactions with 250 and 500 mM
glucose were considerable (see Figure 4.1). The amount of these phases appeared to be
strongly dependent on the concentration; upon increase of the glucose concentration the
amounts increased. Figure 4.1 shows that, for high initial glucose concentrations the
amounts of char and gases increase at the expense of those of tar upon increasing the
reaction time.
4.2.1 Water phase
The pH of the water layers varied between 2.3 and 2.8; upon increase of the initial
concentration of the glucose, the pH value of the reaction product decreased. The
composition of the water phase was analyzed by 1H NMR spectroscopy. The majority of
the reaction products could be identified by comparison with NMR spectra of the
corresponding authentic samples. Only low molecular weight products were observed, most
of which are the same as those identified previously in the reaction products obtained after
very short reaction times (< 4 min) with a continuous reactor [12]. However, the
distribution of the reactions products differs significantly. In Figure 4.2, the distribution of
the most important products in the water layer during 15 min hydrothermal reaction of 500
mM glucose solution is compared with that of the previously studied reaction of a 50 mM
solution during 2 min. [12]. It can be seen that the concentrations in the water layer of the
reaction with the 500 mM glucose solution are an order of magnitude lower. Since, only
low molecular weight compounds were detected, it may be concluded that the initial
products formed from glucose are reacting to higher molecular weight water-insoluble
compounds (tar and char) and gases.
81
Chapter 4
30
Concentration (mmol/l)
Concentration (mmol/l)
A
20
10
B
2.5
2.0
1.5
1.0
0.5
0.0
0
0
50
100
0
5
Residence time (s)
10
15
Residence time (min)
Figure 4.2 Distribution of the most important products in the water phase during the
hydrothermal reaction of D-glucose, (A) 50 mM solution in continuous reactor with 25-120
s residence time at 340 °C and 27.5 MPa, (B) 150 ml of a 500 mM solution in batch reactor
at 340 °C and 17.5 MPa,
benzenetriol,
●
glycolaldehyde,
▲
5-hydroxymethylfurfural,
■
1,2,4-
acetic acid, ▼ acetone, ∆ dihydroxyacetone, + formic acid, ∇ levulinic acid,
□
2-furfural.
A striking difference between the two reactions is that glycolaldehyde, which is
the major initial product of a 50 mM solution is decreasing rapidly in concentration during
the reaction of the 500 mM solution in reactor 1 (heat-up time 1.5 min; cool-down time 10
s). In the products of a similar reaction in set-up 2 (heat-up and cool-down time both about
20 min), glycolaldehyde was even completely absent. A possible explanation could be that
glycolaldehyde is rapidly converted into other products during hydrothermal reactions at
higher concentrations. In the water layer, no products were found that could be formed
from glycolaldehyde by means of the common fragmentation or dehydration reactions.
Therefore, it is most likely that the highly reactive initially formed glycolaldehyde is
involved in bimolecular reactions leading to high-molecular weight compounds that end up
in the tar or the char fraction.
82
Hydrothermal reactions of glucose; the influence of concentration on reaction pathways
OH
HOH 2C
OH
O
O
HO
HO
- 3 H 2O
OH
OH
HO H2C
CHO
O
CH 2OH
OH
HO
H+, H2O
OH
OH
OH
CH 3
O
OH
H 2C
HO H2C
O
CHO
O
CHO
CHO
OH
CH 3
CHO
O
H 2O
CH 3
O
- CO
CHO
O
CH 3
O
CH
O
O
+ H2O
CH 3
COOH
CH 3
O
CH(OH)2
O
Scheme 4.1 Possible pathway leading to the formation of levulinic acid during the
hydrothermal treatment of 500 mM D-glucose in water.
On the other hand, the major product of the reaction with 500 mM glucose in
reactor 2 is levulinic acid, which was not observed in the reaction with the 50 mM solution,
even not after prolonged reaction times. Formation of levulinic acid from glucose requires
splitting off of water and either CO or formic acid. Glucose easily equilibrates with fructose
under
the
HTU
conditions.
Dehydratation
of
fructose
is
known
to
give
hydroxymethylfurfural [12], which can form levulinic acid via elimination of CO [15]. A
possible reaction mechanism is depicted in Scheme 4.1. The reaction has shown to be acid
catalysed [12]. As a matter of fact, levulinic acid could also be detected in a hydrothermal
reaction of 50 mM glucose in the presence of 6 mM HCl. Apparently, the slightly lower pH
of the reaction mixture from 500 mM glucose as compared to that of 50 mM glucose is
83
Chapter 4
sufficient to catalyse the formation of levulinic acid. It should be noted that the initial
concentrations of formic acid (see Figure 4.2A) are rather high, which may explain the
decrease in pH. Probably formic acid is initially formed via isomersation of glucose to
fructose followed by a retro-aldol condensation to dihydroxyacetone and glyceraldehyde
[6], [12]. The latter can form pyruvaldehyde by dehydration. An α-dicarbonyl cleavage
then gives formic acid and acetaldehyde. Possibly, formic acid leaves the water layer after
dehydration to carbon monoxide.
Luijkx et al. have shown that levulinic acid reacts only very slowly under
hydrothermal conditions. In a 1 M solution only 10% conversion was observed after 15 min
[16]. These authors observed small amounts of 2-butanone and succinic acid in the water
layer and in ether extracts small amounts of aromatic and aliphatic products could be
identified, such as toluene, xylenes, dihydro-dimethylindenes, dimethyltetralins, 3-methyl2-cyclopentanone, and 2-butanone.
Hydroxymethylfurfural (HMF) was the second initial main product observed in
the hydrothermal treatment of 50 mM glucose. This product was also present in the water
layers of the presently studied reactions of 500 mM glucose, but at relatively low
concentrations. Conversion of HMF into levulinic acid is one of the possible pathways that
may account for this. Another route is the previously described conversion into 1,2,4benzenetriol [8].
A sample obtained from Knezevic et al. of a water layer obtained by the
hydrothermal treatment of a woodstick [1] was also analyzed by 1H NMR. In Table 4.1, the
composition of this water layer is compared with the water layer obtained from the
hydrothermal upgrade of D-glucose in the present study. As can be seen, the compositions
of these water layers are very similar, which suggests that mainly products from
hydrothermal reactions of cellulose end up in the water layer.
84
Hydrothermal reactions of glucose; the influence of concentration on reaction pathways
Table 4.1 Comparison of the relative amounts of the various compounds present in the water layer
after hydrothermal treatment of glucose (500 mM, 340 °C, 17.5 MPa) and of a wood stick (400 o C in
a sealed capillary).
Glucose a
Wood stick b
1
1
1,2,4 – trihydroxybenzene
2.22
0.96
Furfural
0.57
0.59
Acetic acid
2.56
3.80
Formic acid
1.81
1.24
Acetone
0.64
0.33
Compounds
HMF
a
present work, b sample obtained from HTU wood stick treatment [1].
4.2.2 Tar phase
Quantitative recovering of the tar and char fractions from the small reactor (set-up
1) was impossible. Therefore, only tar and char fractions obtained from reactor 2 were
analyzed. Gel permeation chromatography (GPC, see Figure 4.3) showed that the molecular
weight distribution of the tar was almost independent of the residence time and of the
temperature. The majority of the products have molecular weights between 500 and more
than 12000.
Figure 4.3 Molecular weight distributions as determined by GPC analysis of tar fractions
obtained from hydrothermal treatment of a 500 mM solution of D-glucose in reactor 2 at
250 and 340 oC.
85
Chapter 4
Figure 4.4 (A) 1H NMR spectrum and (B)
13
C NMR spectrum of a solution in THF-d8 of
the tar fraction obtained from hydrothermal treatment of a 500 mM solution of D-glucose in
reactor 2 at 340 °C and 17.5 MPa. The asterisks indicate resonances due to (THF-d8).
The quantitative 1H and 13C NMR spectra of these fractions confirmed that their
composition is independent of reaction time and temperature. Figure 4.4 shows typical
NMR spectra. The spectra consisted mainly of very broad unresolved resonances, which
suggests, that the tar consists predominantly of complex high molecular weight compounds.
The 13C NMR spectra also showed a number of narrow resonances, which can be ascribed
to low molecular weight compounds. However, the total intensity of those resonances is
negligible in comparison to that of the broad resonances. Those results are in agreement
with the results of the GPC analysis. The low integral of the region between 52 and 85 ppm
in the
13
C NMR spectra indicates that polyol and sugar type compounds are almost
completely converted into less oxygen containing compounds. The low intensity of
aldehyde peaks (9-11 ppm) in the 1H NMR spectrum shows that the aldehydes that were
initially detected in the hydrothermal reaction of diluted D-glucose samples are completely
converted in other products under the conditions applied in the present study. The integrals
of the various spectral regions together with assignments are compiled in Table 4.2.
86
Hydrothermal reactions of glucose; the influence of concentration on reaction pathways
Table 4.2 Integrals of regions in 1H and 13C NMR spectra of a solution of the tar fraction of
the hydrothermal treatment of a 500 mM solution of D-glucose in reactor 2 at 340 °C and
17.5 MPa; reaction time 15 min.
Chemical shift (ppm)
13
Integral
(% of total)
Assignment
C
0-52
42
aliphatic Cs
52-85
2
sugar and polyol Cs
90-168
50
olefinic and aromatic Cs
168-185
2
ester carbonyl and carboxyl Cs
4
aldehyde and ketone carbonyl Cs
0-5.2
73
aliphatic Hs
5.2-9
26
olefinic and aromatic Hs
9-11
1
aldehyde Hs
> 185
1
H
The tar from the reaction with a concentrated glucose (500 mM water solution)
was distilled under vacuum. Only a small fraction (1.0 wt %), with a boiling point of 40 oC
/ 0.1 mm Hg could be obtained. A GC analysis showed that this fraction consisted of
numerous compounds. This fraction was subjected to GC-MS analysis. Some of the
compounds were identified by comparison of their mass spectra with those in a library of
mass spectra data (see Figure 4.5). The nature of most compounds indicates that they are
formed from phenol derivatives and from HMF. Figure 4.5 gives also the structures of
some compounds obtained from a distillation of biocrude obtained from a HTU reaction of
biomass (temp 300 to 3500 C the pressure 100-180 bar and the residence time 5 to 20 min)
in the HTU pilot plant. The resemblance of these compositions once again demonstrates
that reaction products originating from cellulose play an important role in the HTU of
biomass. It should be noted that it cannot be ruled out that the distillation products of the tar
fraction may be due to cracking. In our opinion, however, their structures are reflecting the
structures of compounds present in the tar.
87
Chapter 4
HO
OH
O
O
C
H
CH3
O
O
OH
O
C O
O
OH
HO
O
O
O
C
H
O
O
CH3
O
Figure 4.5 A Molecular structures of structures identified in the distillate of the tar fraction
of the hydrothermal treatment of D-glucose.
88
Hydrothermal reactions of glucose; the influence of concentration on reaction pathways
O
O
H
N
O
H
N
O
H
H
N
N
O
H
H
H
N
N
N
O
H
OH
O
N
O
N
HO
N
N
NH2
O
O
HO
N
O
H
COOH
O
HO
N
N
N
N
H
OH
O
N
O
N
O
H
O
H
COOH
O
O
S
O
O
O
OH
OH
O
O
OH
Figure 4.5B Molecular structures of structures identified in the distillate of the tar fraction
of the hydrothermal treatment of onion pulp.
89
Chapter 4
4.2.3 Gas phase
The formation of gases parallels that of the tar. The main gases formed are CO,
CO2, H2, and propene. The amount formed was increasing with the residence time. The
amounts of the gases obtained from a reaction of a 500 mM glucose solution as a function
of the reaction time is given in Figure 4.6. The gases obtained at residence time RT=10 min
were analyzed by GC. It contained CO, CO2, ethene and propene in a molar ratio: 1 : 0.9 :
0.001 : 0.009.
500
o
340 C
Amount (ml)
400
o
250 C
300
200
100
0
0
2
4
6
8
10
12
14
16
Residence time (min)
Figure 4.6 Amounts of the gases (measured at atmospheric pressure and room temperature)
produced during the hydrothermal reaction of 500 mM D-glucose in reactor 2 at 250 oC,
340 °C and 17.5 MPa as a function of the reaction time.
The amount of H2 was about equal to that of CO. Therefore, the formation of CO2
and hydrogen possibly can be ascribed to the water gas shift reaction. Apparently, large
amounts of CO are being formed during the hydrothermal treatment of glucose, indicating
that decarbonylation is an important reaction mechanism. Carboxylic acids can
decarbonylate under highly acidic conditions under the formation of a carbenium ion that
subsequently can deprotonate to an alkene or hydrate to an alcohol (see Scheme 4.2) [17].
90
Hydrothermal reactions of glucose; the influence of concentration on reaction pathways
Most likely propene has been formed also in this way from butanoic acid. The presence of
propene in the reaction medium, the occurrence of decarbonylations, and the acidity of the
reaction medium (pH 2-3) make it very likely that carbenium ion reactions play a role in the
later stages of the HTU process and, consequently, polymerisation reactions via carbenium
ion mechanisms may play a role in the formation of the high molecular weight compounds
observed in the tar.
OH
O
R
C
+
R
H
C
OH
OH
-
H2O
R C O
R
+
CO
Polymers
Alcohols
Alkenes
Scheme 4.2 Formation of CO through decarbonylation of carboxylic acids.
4.3 Conclusions
The results of this investigation show that the concentration of the starting
compound is important for the reduction of the oxygen content of biomass by the HTU
process. Only at higher concentrations (> 250 mM), the model compound glucose forms
91
Chapter 4
gases and tar with a higher C/O ratio. This confirms our previous suggestion that
bimolecular reactions are important pathways [12].
The picture that emerges from the studies described in Chapters 3 and 4 resembles
in many aspects the pathways that have been proposed for the alkaline degradation of
sugars. Thus, initially, D-glucose degrades to a complex dynamic mixture of interconverting
aldehydes as a result of base catalysed isomersations (Lobry de Bruyn - Alberda van
Ekenstein rearragements), retro-aldol reactions, β-eliminations, and α-dicarbonyl
cleavages, and aldol condensations [18]. Irreversible base catalysed formation of carboxylic
acids via benzilic rearrangements and α-dicarbonyl cleavages leads to a decrease of the
amount of aldehydes. After 15 min reaction of a 500 mM D-glucose solution at 340 ºC and
17.5 MPa, aldehydes could no longer be detected in either the water or the tar layer. Next to
the base catalysed reactions, acid catalysed reactions including dehydratation/hydratation
and alkylation occur. Initially, dehydration to e.g. HMF is an important pathway. The
importance of the acid catalysed reactions increases during the course of the reaction
because the acidity of the reaction mixture is increasing as a result of the formation of acids
like formic acid and acetic acid.
Carboxylic acids decarbonylate to olefins, which subsequently may polymerize
under the acidic reaction conditions. In the final tar fraction the amount of carbonyl
compounds is very low (only 4% of the C-atoms is in a carbonyl function).
The major pathways in the hydrothermal reaction of D-glucose are depicted in a
simplified and schematic way in Scheme 4.3.
92
Hydrothermal reactions of glucose; the influence of concentration on reaction pathways
Glucose
Condensations
Fragments
Biocrude
- H2 O
Char
“Irreversible products”
(CO, CO2, phenols)
Scheme 4.3 Major reaction pathways during the hydrothermal treatment of D-glucose.
The elemental analyses of the char and tar fraction of the reaction of a 500 mM solution
show that char contains 72.88 wt % C, 3.34 wt % H, and 23.78 wt % O. Tar obtained from
residence time 5 min contains 63.67 wt % C, 6.34 wt % H, 28.44 wt % O, and tar obtained
from residence time 20 min contains 64.84 wt % C, 5.64 wt % H, 27.24 wt % O. In Figure
4.7 the composition is compared with that of tar and of the starting compound D-glucose.
Since glucose, H2O and char are almost on a straight line, it can be concluded that most of
the char has been formed by dehydration reactions. The location of the points for tar and
gas on different sides of this line suggest that these are formed by a disproportionation
process.
93
Chapter 4
H
CnH2n
Tar
H2O
Glucose
Char
C
CO Gas CO
2
O
Figure 4.7 Elementary composition (in mole %) of char.
4.4 Experimental
4.4.1 Chemicals
All chemicals used were obtained from Aldrich and were used without further
purification. The water used was demineralised and had a conductivity of 18.2 mΩ·cm.
Distillate of biocrude obtained from the HTU reaction of biomass was obtained through
Biofuel B.V.
4.4.2 Apparatus and methods
The set-ups used for performing the HTU reactions have been described in detail
in Chapter 2. Reactor 1 is described in Fig. 2.9 and reactor 2 in Fig. 2.11.
94
Hydrothermal reactions of glucose; the influence of concentration on reaction pathways
4.4.3 Analyses
HPLC analyses were carried out with a Millipore-Waters 590 pump and
Phenomenex Rezex organic acid column at 60 ºC. The eluent was 0.01 M aq trifluoroacetic
acid and the flow rate 0.5 mL/min. A refractive index (RI) (Shodex model SE-51) detector
and an ultraviolet (UV) (Shimadzu spectrometric detector model SPD-6A) at 215 nm were
used. The peak areas were determined using an integrator (Kipp & Zonen). HPLC
quantitation was achieved using THF as an internal standard.
Gel Permeation Chromatography (GPC) analyses were performed with two
columns in series packed with µSTYRAGEL with pore size of 100 and 500 Å, respectively.
The eluent was THF (tetrahydrofuran). A refractive index (RI) (Shodex model SE-51)
detector was used.
1
H NMR spectra were measured on a Varian Unity INOVA-300 spectrometer at
300 MHz). A weighed amount of tert-butanol in D2O was added to the samples to lock and
as an internal standard. The chemical shifts are reported with respect to the CH3 signal of
tert-butanol, which was set at 1.20 ppm. The water resonance was suppressed using
presaturation with the transmitter. The peak areas were determined by deconvolution of the
spectrum with lorentzian peaks.
Mass spectral analyses were done on LC Quattro ESP+ spectrometer.
Total Organic Carbon analyses were performed using a Shimadzu T.O.C. 5050A
instrument.
Acknowledgements
Thanks are due to E. Wurtz, and R. Sloter for their assistance during the
construction of the experimental setup and to Dr. F. Goudriaan and Dr. L. Petrus for helpful
discussions. We are grateful to J. Knoll for performing T.O.C. analyses, to A. van Estrik
and K. Djanashvili for performing NMR analyses, and to D. Knezevic (University of
Twente) for providing us with the reaction mixture of an HTU reaction of a woodstick in a
quartz capillary.
95
Chapter 4
References
1.
Matsumura, Y., Minowa, T., Potic, B., Kersten, S.R.A., Prins, W., van Swaaij,
W.P.M., van de Beld, B., Elliott, C.D., Neuenschwander, G.G., Kruse, A., Antal
Jr., M.J.; Biomass and Bioenergy 2005, 29, 269.
2.
Bridgwater, A. V., Czernik, S., Piskorz, J. An overview of fast pyrolysis. Progress
in Thermochemical Biomass Conversion, [Conference], 5th, Tyrol, Austria, Sept.
17-22, 2000.
3.
Goudriaan, F., Peferoen D.G.R.; Chem. Eng. Sci. 1990, 45, 2729.
4.
Goudriaan, F., Naber, J. E., Zeevalkink, J. A.; NPT Procestechnologie 2004, 11,
31.
5.
Sasaki, M., Fang, Z., Fukushima, Y., Adschiri, T., Arai, K.; Ind. Eng. Chem. Res.
2000, 39, 2883.
6.
Kabyemela, B. M., Takigawa, M., Adschiri, T., Malaluan, R. M., Arai, K.; Ind.
Eng. Chem. Res. 1999, 38, 2888.
7.
Bobleter, O., Bonn, G.; Carbohydr. Res. 1983, 124, 185.
8.
Luijkx, G. C. A., van Rantwijk, F., van Bekkum, H., Antal, M. J.; Carbohydr. Res.
1995, 272, 191.
9.
Kabyemela, B. M., Adschiri, T., Malaluan, R. M., Arai, K.; Ind. Eng. Chem. Res.
1997, 36, 1552.
10. Kabyemela, B. M., Adschiri, T., Malaluan, R. M., Arai, K.; Ind. Eng. Chem. Res.
1999, 38, 2888.
11. Kallury, R. K. M. R., Ambidge, C., Tidwell, T. T., Boocock, D. G. B., Agblevor,
F. A., Stewart, D. J.; Carbohydr. Res. 1986, 158, 253.
12. Srokol, Z., Bouche, A-G., van Estrik, A., Strik, R.C.J., Maschmeyer, T., Peters
J.A.; Carbohydr. Res. 2004, 339, 1717.
13. Kruse, A., Gawlik, A.; Ind. Eng. Chem. 2003, 42, 267.
14. Minowa, T., Fang, Z., Ogi, T., Varhegyi, G.; J.Chem.Eng.Jpn. 1998, 31, 131.
15. Horvat, J., Klaic, B., Metelko, B., Sunijc, V.; Tetrahedron Lett. 1985, Vol. 26,
2111.
96
Hydrothermal reactions of glucose; the influence of concentration on reaction pathways
16. Luijkx G.C.A. (1994) Hydrothermal conversion of carbohydrates and related
compounds. Ph.D. Thesis, Delft University of Technology.
17. Deno, N.C., Pittman, C. U. Jr., Wisotsky, M. J.; J. Am. Chem. Soc. 1964, Vol. 86,
4370.
18. De Bruijn J.M. (1986) Monosaccharidies in alkaline medium: isomerisation,
degradation, oligomerization. Ph.D. Thesis, Delft University of Technology.
97
Chapter 5
Hydrothermal reactions of glycolaldehyde
99
Chapter 5
5.1 Introduction
Hydrothermal treatment of biomass is a very promising technology to upgrade
biomass to a fuel with a heating value approaching that of fossil fuels [1]. During this
hydrothermal upgrading (HTU) the oxygen content of the organic material is reduced from
about 40% to 10-15%. The main components of biomass resources are lignin, cellulose,
hemicelullose and minerals. The reduction of the oxygen content is achieved mainly by
pathways by which the oxygen atoms are expelled as CO, CO2 and water.
It has been reported that during the initial stage of the HTU process, cellulose and
other polysaccharides hydrolyze very rapidly to monosaccharides [2]. Consequently,
D-
glucose is a good model compound to elucidate the reaction pathways of cellulose during
the HTU process [3].
In our previous research, we have shown that during short (< 4 min) hydrothermal
treatment of diluted glucose solutions (50 mM) almost exclusively water soluble products
were obtained [3]. Under these conditions almost no gases and tar or char were obtained.
The products observed could be explained by both acid and base catalysed reactions as a
result of the relatively high value of Kw under subcritical conditions. The main products are
formed via dehydrations, rearrangements and fragmentations.
However, at higher concentrations (500 mM) and using reaction times up to 15
min, a substantial amount of gas, a tar fraction and char were obtained. The phenomena
observed and the composition of the reaction products was explained by an initial
degradation of D-glucose into a complex dynamic mixture of interconverting aldehydes [4].
Initially the two main products are glycolaldehyde and hydroxymethylfurfural (HMF).
About 15 % of the carbon atoms of glucose end up in each of these compounds during the
initial stages of the reaction under diluted conditions. At higher concentrations,
recombinations of the aldehyde fragments occur. Irreversible base catalysed formation of
carboxylic acids via benzilic rearrangements and α-dicarbonyl cleavages leads to a
decrease of the amount of aldehydes. After 15 min reaction of a 500 mM D-glucose solution
at 340 ºC and 17.5 MPa, no aldehydes could be detected anymore in either the water or the
tar layer. Next to the base catalysed reactions, acid catalysed reactions including
dehydratation/hydratation and alkylation occur.
100
Hydrothermal reactions of glycolaldehyde
Here, we report on a more detailed study of one of the key compounds of this
reaction, glycolaldehyde.
5.2 Results and discussion
The hydrothermal treatments of glycolaldehyde were performed with the
continuous flow reactor described previously [3]. The effect of concentration of
glycolaldehyde in the feedstock was investigated for the range 50-500 mM, at a
temperature of 340o C and 27.5 MPa of pressure. The reaction time was 200 s in all cases.
Under those conditions the amounts of gases, tar and char were too low to allow analysis.
B
60
40
20
0
0
100
200
300
400
500
Concentration of the products (mM)
Concentration of the products (mM)
A
8
6
4
2
0
0
Initial concentartion of glycolaldehyde (mM)
100
200
300
400
500
Initial concentration of glycolaldehyde (mM)
Figure 5.1 Concentration of the products as a function of the initial concentration of
glycolaldehyde (t=340o C, p=27.5 MPa, 200 s reaction time), A - products with high
concentration;
glycolaldehyde (1),
glycolic acid,
formic acid (7),
hydroxyacetone
(5), 1 ‫׀‬-hydroxy-2-butanone (9), B – products with low concentration; lactic acid, acetic
acid,
acetone,
dihydroxyacetone (6), ◊ acetic aldehyde, ○ ethylene glycol.
101
Chapter 5
The water layer was very complex; numerous small peaks were observed in the 1H
NMR spectrum. The line widths of all resonances were very small, suggesting that the
molecular weights of the products in the water layer were low. The major products could be
identified by comparison with spectra of authentic samples or, in the case of 1-hydroxy-2butanone by analogy with the spectrum of hydroxyacetone. Figure 5.1 gives the
concentration of these products as a function of the initial concentration of glycolaldehyde.
The shapes of the curves in Fig 5.1 suggest that the rates of the reactions are increasing
with increasing concentration. Exceptions are the curve for hydroxyacetone and
acetaldehyde, which suggest that the compounds are intermediates involved in bimolecular
reactions. Furthermore, the presence of reaction products with more than two C atoms
(lactic acid, acetone, hydroxyacetone, dihydroxyacetone, 1-hydroxy-2-butanone) proves
that oligomerization reactions play an important role. Hydroxyacetone most likely is
formed via one of the C4 sugars formed by aldol condensation of glycolaldehyde. This
sugar may subsequently be converted into hydroxyacetone by a mechanism proposed by
Shafizadeh and Lai for the formation of this compound during the thermal degradation of
1,6-anhydro-β-glucopyranose [5]. Thus the C4-sugar dehydrates to hydroxyketoaldehyde
(4) and then looses CO to give hydroxyacetone (5), (see Scheme 5.1).
O
2
OH
A ld o l
OH
c o n d e n s a t io n
H
O
OH
1
2
OH
- H
O
O
OH
k e to -e n o l
OH
2
O
OH
H
3
4
- CO
OH
O
CH
3
5
Scheme 5.1 Possible reaction mechanisms formation of hydroxyacetone from
glycolaldehyde via C4 sugars.
102
Hydrothermal reactions of glycolaldehyde
Dihydroxyacetone could be formed via the aldol reaction of glycolaldehyde (1) to
C4 sugars, followed by isomerisation (via Lobry de Bruyn-Alberda van Ekenstein
rearrangements) and retro aldol reaction, (see Scheme 5.2).
O
2
OH
OH O
Aldol
OH
H
HO
HO
OH
H
condensation
1
Isomerisation
2
O
OH
Retro
aldol
OH
OH
+
O
H2C
O
7
6
Scheme 5.2 Formation of dihydroxyacetone from glycolaldehyde.
The formation of 1-hydroxy-2-butanone (9) is surprising, since it cannot be
explained by only condensation and fragmentation reactions. Somewhere in the reaction
paths a reduction should occur. A possible pathway is the Cannizzaro reaction, which is
consistent with the presence of glycolic acid. A possible pathway to 9 could be aldol
condensation of hydroxyacetone (5) with formaldehyde (8) followed by reduction and
dehydration (see Scheme 5.3).
103
Chapter 5
O
O
Aldol
CH 2OH
H2C
+
O
OH
condensation
OH
8
5
Reduction
OH
- H2 O
OH
O
OH
OH
9
Scheme 5.3 Formation of 1-hydroxy-2-butanone.
Dimerisation or oligomerisation of glycolaldehyde will result in a complex
mixture of diastereoisomers and probably these compounds are present in relatively low
amounts. This is supported by the presence of numerous small peaks in the 1H NMR
spectra, particularly in the shift range of sugars [3].
It is known that hydroxyacetone and 1-hydroxy-2-butanone are precursors of
cyclic diketones [6], the compounds that have typical caramel like odours. All reaction
products of hydrothermal reactions of glycolaldehyde had these odours.
These compounds were not observed in the reaction products of the hydrothermal
treatment of D-glucose, studied previously [2]. However if the latter reaction was performed
in the presence of some acid, ketones were among the reaction products. Probably, the
additional acids favours the dehydration required for the formation of these ketones.
The presence of other major products, including lactic acid, acetone, and
dihydroxyacetone could be explained by several mechanisms. Previously, we have given
several suggestions for pathways by which these compounds could be formed from
monosaccharides [3]. The presence of ethylene glycol is remarkable. Formation of this
product from glycolaldehyde requires a reduction for example, a Cannizzaro reduction.
104
Hydrothermal reactions of glycolaldehyde
5.2.1 Effect of added acid and base
Both acid and base might catalyse reactions during the hydrothermal reaction of
glycolaldehyde. Therefore, we decided to investigate the effect of added acid and base. In
the presence of 1 mM HCl, the products distribution is not much different from the
products distribution in the presence of 1 mM NaOH.
The pH after action in the case of addition of HCl was 2.5 but in the case of addition of
base, the pH after reaction was 3.85. This might indicate that much more base is needed to
achieve a non-acidic medium during the reaction. Formation of large amount of organic
acids in the reaction products might keep the pH low. Increasing the amount of NaOH in
the feedstock lead to an increased formation of organic acids as lactic acid, formic acid,
acetic acid, while the amount of glycolic acid formed decreased, see Table 5.1.
Table 5.1 Concentration of the products of hydrothermal reaction (340 °C) of
glycolaldehyde after adding NaOH.
Base addition to 50 mM glycolaldehyde
Compound (mM)
Solution
1 mM HCl
1 mM NaOH
10 mM NaOH
50 mM NaOH
Glycolaldehyde
19.26
16.18
1.26
2.42
Lactic acid
0.76
0.69
2.16
5.45
Glycolic acid
1.98
4.85
2.53
0.86
Acetic acid
0.33
0.55
1.02
3.81
Formic acid
1.69
0.85
0.77
5.67
Acetone
0.11
0.14
0.18
0.49
Hydroxyacetone
4.21
4.45
3.37
-
Dihydroxyacetone
0.60
0.54
2.04
-
Acetic aldehyde
1.21
1.41
-
-
Ethylene glycol
0.09
0.23
0.13
1.49
1-hydroxy-2-butanone
1.48
1.94
0.50
0.62
105
Chapter 5
5.2.2 Autoclave experiments
Some autoclave experiments were performed to study HTU reactions of
glycoladehyde with long residence time (30 min) and for comparison with the results
obtained from the continuous flow reactor. The concentrations of the products obtained are
shown in Table 5.2.
Table 5.2 Concentration of the products from autoclave experiments (see Fig. 2.9A, 340
°C, residence time 30 min)
Compound
Concentration (mM)
Glycolaldehyde
49.55
Lactic acid
-
Acetic acid
5.87
Formic acid
-
Acetone
0.55
Hydroxyacetone
4.86
Dihydroxyacetone
1.66
Acetic aldehyde
3.53
Ethylene glycol
0.75
1-Hydroxy-2-butanone
1.67
2-Furaldehyde
0.83
Pyruvaldehyde
0.41
Surprisingly, after 30 min, the reaction mixture, besides compounds that have been
observed before, contains pyruvaldehyde and furfural (see Table 5.2). The presence of
pyruvaldehyde suggests glyceraldehydes as an intermediate in the reaction. Most likely it
has been formed by a retro aldol reaction of sugars with 4 or more C-atoms. The presence
of furfural indicates C5 sugars as intermediates. Previously, we have shown that this
compound can be formed from arabinose [3].
106
Hydrothermal reactions of glycolaldehyde
CH2OH
C
O
CHOH
(CHOH) n
CH2OH
Figure 5.2 General structure of sugar that might give rise to formation of glyceraldehydes
by retro-aldol reaction (n ≥ 0).
5.3 Conclusions
In the reaction mixture obtained from hydrothermal treatment of glycolaldehyde in
a continuous flow reactor no aromatic products have been observed. The products obtained
show that dehydrations and retro-aldol reactions are important reactions. In addition, many
products were obtained that only can be explained with intermediates that require
condensations reactions. The presence of products like ethylene glycol, hydroxyacetone,
and hydroxyl-2-butanone demonstrated that reductions, most likely Cannizzaro reactions,
also play a role in the HTU reactions. Reductions will be discussed in more detail in the
next chapter.
5.4 Experimental
5.4.1 Chemicals
All chemicals used were obtained from Aldrich and were used without further
purification. The water used was demineralised and had a conductivity of 18.2 mΩ·cm.
107
Chapter 5
5.4.2 Apparatus and methods
The HTU reactions of glycolaldehyde were performed in two different
experimental set-ups (continuous flow reactor and a small autoclave).
The home-built continuous flow reactor is described in detail and schematically depicted in
chapter 2, Fig. 2.10, and a small autoclave in chapter 2, Fig. 2.9.
5.4.3 Analysis
HPLC analyses were carried out with a Millipore-Waters 590 pump and
Phenomenex Rezex organic acid column at 60 ºC. The eluent was 0.01 M aq trifluoroacetic
acid and the flow rate 0.5 mL/min. A refractive index (RI) (Shodex model SE-51) detector
and an ultraviolet (UV) (Shimadzu spectrometric detector model SPD-6A) at 215 nm were
used. The peak areas were determined using an integrator (Kipp & Zonen). HPLC
quantitation was achieved using THF as an internal standard.
1
H NMR spectra were measured on a Varian Unity INOVA-300 spectrometer at
300 MHz). A weighed amount of tert-butanol in D2O was added to the samples to lock and
as an internal standard. The water resonance was suppressed using presaturation with the
transmitter. The peak areas were determinded by deconvolution of the spectrum with
lorentzian peaks.
Acknowledgements
Thanks to Nicole de Groot for performing most of the experimental assistance.
Anton van Estrik and Kristina Djanashvili for performing NMR analyses, and Dr. Frans
Goudriaan and Dr. Leo Petrus for helpful discussions.
108
Hydrothermal reactions of glycolaldehyde
References
1.
Goudriaan, F., van de Beld, B., Boerefijn, F.R., Bos, G.M., Naber, J.E., van de Wal, S.,
and Zeevalkink, J.A., in A.V. Bridgwater (Ed), Proceedings Progress in
Thermochemical Biomass Conversion, Blackwell Science Ltd, Oxford UK, 2001.
2.
Kabyemela, B.M., Takigawa, M., Adschiri, T., Malaluan, R.M., and Arai, K.; Ind. Eng.
Chem. Res. 1998, 37, 357.
3.
Srokol, Z., Bouche, A-G., van Estrik, A., Strik, R.C.J., Maschmeyer, T., Peters J.A.;
Carbohydr. Res. 2004, 339, 1717.
4.
This thesis, Chapter 4.
5.
Shafizadeh, F., Lai, Y.Z. J.; Org. Chem.1972, 37, 2.
6.
Shaw, P.E., Tatum, J.H., and Berry, R.E., Agric, J.; Food Chem. 1968, 16, 979.
109
Chapter 6
Hydrogen transfers during hydrothermal reactions of
glycolaldehyde and glucose
111
Chapter 6
6.1 Introduction
The average chemical composition of biomass based on dry weight for wood is:
40-50 % cellulose, 20-30 % hemicellulose and 25-30 % lignin [1]. To covert biomass into a
useful biofuel it is necessary to reduce the oxygen content of the starting material. One of
the possibilities to achieve this is by means of hydrothermal upgrading (HTU), a treatment
of wet biomass at temperatures of 300 to 350 ºC and 10 – 18 MPa during 20 min [2]. Such
a treatment leads to the desired reduction of the O-content through loss of CO and CO2. It
has been reported that during the initial stage of the HTU process, polysaccharides
hydrolyze very rapidly to monosaccharides [3]. Retro-aldol reactions of the
monosaccharides to glycolaldehyde have been shown to be important secondary pathways
[3,4]. Therefore, both
D-glucose
and glycolaldehyde are good model compounds to
elucidate reaction pathways starting with cellulose during the HTU process [2].
Previously, we have shown that during short (< 4 min) hydrothermal treatment of
diluted glucose solutions (50 mM) almost exclusively water-soluble products were obtained
[4]. The products observed could be explained by both acid and base catalysed reactions
occurring as a result of the relatively high value of Kw under subcritical conditions. The
main products are formed via dehydrations, rearrangements and fragmentations. Initially,
the two main products are glycolaldehyde and hydroxymethylfurfural (HMF). About 15 %
of the carbon atoms of glucose end up in each of these compounds during the initial stages
of the reaction under diluted conditions.
At higher concentrations, recombinations of the aldehyde fragments occur [5]. In a
detailed study of hydrothermal treatment of glycolaldehyde in a continuous flow reactor, no
aromatic and olefinic products have been observed [6]. Many compounds that were
obtained can only be explained with intermediates that require condensation reactions. The
presence of ethylene glycol, hydroxyacetone and hydroxy-2-butanone showed that in
addition reductions should take place. A possible pathway could be a Cannizzaro reaction
or a crossed Cannizzaro reaction of aldehydes formed during the initial stages of the HTU
process [7-9]. During the HTU process, a substantial amount of formic acid is formed. It is
known that this compound is an efficient hydride donor in Meerwein-Ponndorf-Verley
reductions [10]. Formic acid oxidizes to CO2 under these conditions and as a result the
112
Hydrogen transfers during hydrothermal reactions of glycolaldehyde and glucose
equilibrium in the reduction shifts to the alcohol. Therefore, we started a study on
hydrothermal reactions of glycolaldehyde and glucose with formic acid at various pH
values. For comparison, the reactions were also performed with acetic acid, since this
compound cannot act as hydride donor. Furthermore, some metal ions and supported metals
were screened as catalysts for redox reactions. In this Chapter the results of these studies
are reported.
6.2 Results and discussion
The hydrothermal treatments of glycolaldehyde and glucose in the presence of
formic acid or acetic acid were performed in sealed quartz capillaries (see chapter 2, p. 3.2).
The concentration of glycolaldehyde or glucose in the starting mixture was 500 mM and the
reactions were performed at 340o C during 15 min. The experiments were performed at pH
values between 1.2 to 5.4 by using mixtures of sodium formate and formic acid or sodium
acetate and acetic acid.
6.2.1 Reactions of glycolaldehyde
Under the conditions applied, the amounts of tar and char formed during the
reactions were very small and, since the previous research (see Chapter 4) showed that the
composition of these fractions is not very informative regarding the reaction pathways, only
the aqueous phase was analyzed. The composition of the water layer was very complex;
numerous small peaks were observed in the 1H NMR spectrum. The line widths of all
resonances were very small, suggesting that the molecular weights of the products in the
water layer were low. The major products in the mixture were identified by comparison
with 1H NMR spectra of authentic samples. Figure 6.1 gives the concentration of these
products obtained during the HTU reactions of glycolaldehyde as a function of pH.
113
A
B
50
Concentration of the products (mmol/l)
Concentration of the products (mmol/l)
Chapter 6
40
30
20
10
0
1
2
3
4
5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
6
1
2
3
pH
C
5
6
D
80
8
Concentration of the products (mmol/l)
Concentration of the products (mmol/l)
4
pH
70
60
50
40
30
20
10
0
2
3
4
5
6
7
8
6
4
2
0
1
2
3
4
pH
5
6
7
8
pH
Figure 6.1 Concentration of the products obtained from a hydrothermal reaction of a 500
mM solution of glycolaldehyde performed in a sealed capillary (t=340o C), A – major
products for the reaction in the presence of HCOOH/HCOONa, B – minor products for the
reaction in the presence of HCOOH/HCOONa, C – major products for the reaction in the
presence of CH3COOH/CH3COONa, D – minor products for the reaction in the presence of
CH3COOH/CH3COONa;
hydroxyacetone,
acetic acid,
glycolaldehyde,
acetone,
glycolic acid,
formic acid,
dihydroxyacetone, ◊ acetic aldehyde, I acrylic
acid, ○ ethylene glycol.
Figure 6.1 shows that the conversion rate of glycolaldehyde increases with the pH.
At higher pHs condensation reactions appear to play an important role as witnessed by the
formation of hydroxyacetone and a complex mixture of small amounts of unidentified
compounds with 1H chemical shifts between 0.7 and 4 ppm.
Surprisingly, a large amount of formic acid has been formed in the reactions with
acetic acid, particularly at higher pHs. Possibly, the formic acid is formed after
114
Hydrogen transfers during hydrothermal reactions of glycolaldehyde and glucose
condensations and fragmentation reactions of glycolaldehyde leading via pyruvaldehyde to
formaldehyde and formic acid [5, 11].
A comparison of the results of the HTU reactions of glycolaldehyde with formate
and acetate confirms that HCOOH is involved in the reduction of glycolaldehyde to glycol;
at pH values of about 3, considerable amounts of glycol were formed in the reaction with
formic acid, whereas almost no glycol was present in the reaction mixtures with acetic acid.
It should be noted that the amount of ethylene glycol formed in the absence of
HCOOH/HCOONa was also negligible (see Chapter 4). Figure 6.1A shows that at higher
pHs (pH > 3), the amount of glycol formed is decreasing with the pH. Most likely,
condensations are competing with the reduction of glycolaldehyde and higher molecular
weight alcohols are formed under these conditions.
At pH > 3, the amount of glycolic acid obtained in both the reaction with formic
acid and acetic acid is significant. Since almost no glycol is observed in the reaction in the
presence of acetic acid, glycolic acid is most likely not formed through a Cannizzaro
reaction of glycolaldehyde. Possibly a cross-Cannizzaro reaction between glycolaldehyde
and an aldehyde formed by condensation/fragmentation reactions can explain these
observations. The polyols that should also be formed in these reactions could explain the
unassigned peaks between 0.7 and 4 ppm in the 1H spectrum of the reaction mixture (see
Scheme 6.1).
R C
O
O
+
H
H
R C
O
RCOOH
+
HOCH2CH2OH
C CH2OH
RCH2OH
+
HOCH2COOH
C CH2OH
CO2
+
HOCH2CH2OH
O
+
H
H
HO C
C CH2OH
O
O
+
H
H
Scheme 6.1 Possible pathways leading to ethylene glycol and glycolic acid.
115
Chapter 6
At pH 2-4, a considerable amount of acrylic acid was formed in all reactions. This
compound can be formed by acid catalyzed dehydration of lactic acid. This once again
demonstrates that condensation reactions of glycolaldehyde are very important pathways,
even at low pH.
6.2.2 Reactions of glucose
Figure 6.2 gives the concentration of products obtained after HTU reaction of
glucose, as a function of the pH of the starting reaction mixture.
B
A
90
80
70
60
50
40
30
20
10
0
1
2
3
4
5
6
7
8
24
Concentration of the products (mmol/l)
Concentration of the products (mmol/l)
100
21
18
15
12
9
6
3
0
1
2
3
4
pH
C
6
7
8
D
100
25
80
60
40
20
0
1
2
3
4
5
6
7
8
Concentration of the products (mmol/l)
Concentration of the products (mmol/l)
5
pH
20
15
10
5
0
1
2
3
pH
4
5
6
7
8
pH
Figure 6.2 Concentration of the products obtained from a hydrothermal reaction of a 500
mM solution of glucose performed in a sealed capillary (t=340o C), A – major products for
reaction in the presence of HCOOH/HCOONa, B – minor products for the reaction in the
presence of HCOOH/HCOONa, C – major products for the reaction in the presence of
CH3COOH/CH3COONa, D – minor products for the reaction in the presence of
CH3COOH/CH3COONa;
116
glycolaldehyde,
glycolic acid,
formic acid,
Hydrogen transfers during hydrothermal reactions of glycolaldehyde and glucose
hydroxyacetone,
▲
acetic acid,
acetone, ◊ acetic aldehyde, □ 2-furfural, ○ ethylene glycol,
5 hydroxymethylfurfural, ∇ levulinic acid.
During the reaction also some tar, char and gases have been formed. However, the
amount was so small that an accurate analysis was not possible.
The aqueous layer contained the same products as obtained from glycolaldehyde,
but now in addition to those products, 5-hydroxymethylfurfural, 2-furaldehyde, and
levulinic acid were obtained. These three products occur particularly when the starting
mixture had a low pH (pH < 5), which is in agreement with the acid catalyzed reactions
involved in the formation of these compounds from glucose. The products obtained from
reactions starting from neutral or basic mixtures are very similar to those obtained from
glycolaldehyde under similar conditions. Most likely, under basic conditions, glucose
rapidly fragments to glycolaldehyde via retro-aldol reactions.
6.2.3 Effect of added catalysts
The effects of metal catalysts (5 wt %) were screened for reactions with
glycolaldehyde and with glucose. No formic acid or acetic acid was added to the reaction
mixture in this case. The reaction products obtained were qualitatively similar to those
obtained in the absence of metals. We focused our attention to products that could be
involved in hydrogen transfer reactions of glycolaldehyde, glycolic acid ethylene glycol.
The results are shown in Table 6.1, and 6.2.
117
Chapter 6
Table 6.1 Concentration of some products obtained from glycolaldehyde with metal
catalysts (500 mM, reaction temperature: 340o C, reaction time 15 min).
500 mM glycolaldehyde
Compound
(mM)
Glycolic acid
Ethylene
glycol
Catalyst (5 % mass)
No catalyst
Ni 2+
Cu 2+
Al 3+
Pd/C
Ru/C
Pt/C
1.59
11.95
13.52
2.65
2.22
1.79
1.90
0.42
0.69
0.94
0.92
0.32
0.59
0.38
Table 6.2 Concentration of the products obtained from glucose with metal catalysts (500
mM, temp. 340o C, residence time 15 min).
500 mM glucose
Compound
(mM)
Glycolic acid
Ethylene
glycol
Catalyst (5% mass)
No catalyst
Ni 2+
Cu 2+
Al 3+
Pd/C
0.47
16.39
19.22
1.53
0.09
100.36
0.33
0.33
Ru/C
Pt/C
15.84
1.92
2.00
1.12
0
0.36
The concentration of ethylene glycol and, particularly that of glycolic acid, after
HTU reactions in the presence of Ni 2+ and Cu 2+ is significantly higher than with the other
metals/metal ions. Most likely, the catalytic effect of these metal ions can be ascribed to
activation of metal hydride transfer reactions in mixed ligand complexes of the metal ions
and starting compounds (see scheme 6.2). A similar effect has been reported for Cannizzaro
reactions of glyoxylic acid [12].
118
Hydrogen transfers during hydrothermal reactions of glycolaldehyde and glucose
H
O
H
O
HO
O
HO
O
+n
M
+n
M
O
O
H
OH
HO
Scheme 6.2 Mixed ligand complexes of the metal ions and starting compounds.
Surprisingly, also the amount of glycolic acid obtained after reaction of the glucose solution
with Pd/C was relatively large. In this case also a relatively large amount of acetic acid (42
mM) was formed. It is known that Pd/C is a good catalyst for dehydrogenation of glucose
to gluconic acid [13]. Possibly glycolic acid and acetic acid are formed via gluconic acid as
intermediate.
6.3 Conclusions
In the reaction mixture obtained from hydrothermal reaction of glycolaldehyde
and glucose in the presence of both formic and acetic products have been observed, which
should be due to hydrogen transfer reactions. The amount of ethylene glycol observed in
the reaction with HCOOH/HCOONa is higher than in the presence of CH3COOH/
CH3COONa, which suggests that formate acts as a hydride donor. A maximum conversion
to ethylene glycol was observed for reaction with an initial pH of 3-4. At higher pH values
the amount of ethylene glycol formed was significantly lower and condensation products of
glycolaldehyde increased. Under more basic conditions, a relatively large amount of
hydroxyacetone is formed. Catalytic effects on the hydrogen transfer reactions occur with
Ni 2+ and Cu 2+. With these metal ions a large amount of glycolic acid was produced in the
reaction of glycolaldehyde, whereas glucose with Ni
2+
gave a large amount of ethylene
glycol.
119
Chapter 6
6.4 Experimental
6.4.1 Chemicals
All chemicals used were obtained from Aldrich and were used without further
purification. The water used was demineralised and had a conductivity of 18.2 mΩ·cm.
6.4.2 Apparatus and methods
The HTU reactions of glycolaldehyde and glucose were performed in a batch
micro reactor. For a detailed description see Chapter 2, section 2.3.2. The reactor was
heated to 340 oC in a fluid sand bed.
After cooling the capillaries, they were kept at room temperature for about 30 min
and then the top was broken off. The content of each capillary was transferred to an NMR
sample tube by means of a syringe with a long needle. They were rinsed with D2O, which
was also added to the NMR sample tube.
6.4.3 Analyses
1
H NMR spectra were measured on a Varian Unity INOVA-300 spectrometer at
300 MHz). A weighed amount of tert-butanol in D2O was added to the samples to lock and
as an internal standard. The chemical shifts are reported with respect to the CH3 signal of
tert-butanol, which was set at 1.20 ppm. The water resonance was suppressed using
presaturation with the transmitter. The peak areas were determinded by deconvolution of
the spectrum with Lorentzian peaks.
120
Hydrogen transfers during hydrothermal reactions of glycolaldehyde and glucose
References
1.
Pettersen, R.C.; in The Chemistry of Solid Wood, Advances in Chemistry Series
207, R.M. Rowell (Ed.), American Chemical Society, Washington (1984) p.57
2.
Goudriaan, F., van de Beld, B., Boerefijn, F.R., Bos, G.M., Naber, J.E., van de
Wal, S., and Zeevalkink, J.A.; in A.V. Bridgwater (Ed), Proceedings Progress in
Thermochemical Biomass Conversion, Blackwell Science Ltd, Oxford UK, 2001.
3.
Kabyemela, B.M., Takigawa, M., Adschiri, T., Malaluan, R.M., and Arai, K.; Ind.
Eng. Chem. Res. 1998, 37, 357.
4.
Srokol, Z., Bouche, A-G., van Estrik, A., Strik, R.C.J., Maschmeyer, T., Peters
J.A.; Carbohydr. Res. 2004, Vol. 339, 1717.
5.
This thesis, Chapter 4.
6.
This thesis, Chapter 5.
7.
Cannizzaro, S.; Liebigs Ann. 1853, 88, 129.
8.
Geissman, T.A.; Org.React. 1944, 2, 94.
9.
Sominsky, L., Rozental, E., Gottlieb, H., Gedenken, A., Hoz, S.; J. Org.Chem.
2004, 69, 1492.
10. Klomp, D., Peters, J.A., Hanefeld, U.; Transfer hydrogenation including the
Meerwein-Ponndorf-Verley
reduction,
Handbook
of
Homogeneous
Hydrogenation, Eds. J.G. de Vries, C.J. Elsevier, Wiley, Weinheim, 2007, 3, 20,
585.
11. Kabyemela, B. M., Takigawa, M., Adschiri, T., Malaluan, R. M., Arai, K.; Ind.
Eng. Chem. Res. 1999, 38, 2888.
12. Hoefnagel, A. J., Peters, J. A., van Bekkum, H.; Recl. Trav. Chim. Pays-Bas.
1988, 107, 242.
13. De Wit, G., de Vlieger, J.J., Kock-van Dalen, A.C., Heus, R., Laroy R., van
Hengstum, A.J., Kieboom, A.P.G., van Bekkum, H.; Carbohydr. Res. 1981, 91,
125.
121
Summary
Chapter 1 is a general introduction. The chapter describes aspects of energy in the
future and renewable and non-renewable sources of energy. Gasification, fermentation and
hydrothermal upgrading as most popular biomass conversion technologies are discussed.
Chapter 2 deals with the equipment required for the study of reaction mechanisms
of hydrothermal reactions. Since it is important to obtain information on the initial reaction
steps, the set-ups should enable short reaction times under the extreme conditions of the
hydrothermal upgrading process (340 oC, 27.5 MPa). First, some typical examples of
equipment for the study of hydrothermal reactions reported in the literature are reviewed
and then the set-ups used in the research described in this thesis are described in detail.
Chapter 3 describes a study of hydrothermal reactions on some monosaccharides
as model compounds. During the hydrothermal upgrading of biomass, hydrolysis to glucose
is an important step. To elucidate some of the reaction pathways that follow this initial
hydrolysis, the hydrothermal treatment (340 oC, 27.5 MPa, 25–204 s) of dilute (50mM)
solutions of D-glucose and some other monosaccharides were studied. As a result of the
increase of Kw under subcritical conditions, both acid and base catalysed reactions occur.
The acid catalysed reactions are mainly dehydrations leading initially to 5hydroxymethylfurfural. Important base catalysed reactions result in glycolaldehyde and
glyceraldehyde. Further fragmentations and dehydrations lead to a variety of low molecular
weight compounds such as formic acid, acetic acid, lactic acid, acrylic acid, 2-furaldehyde
and 1,2,4-benzenetriol. Important pathways leading to a decrease of the O-content of the
liquid reaction products start from the intermediate glyceraldehyde, which forms
pyruvaldehyde, which in its turn is converted into formic acid and acetaldehyde. The latter
compound can also be formed via isomerisation of glyceraldehyde into lactic acid followed
by decarbonylation.
Chapter 4 gives the results of study on the influence of the concentration of
glucose on its reaction pathways in the HTU reaction. Only at concentrations higher than
250 mM, glucose forms gases, tar and char in addition to a water phase with predominantly
low molecular weight compounds. This indicates that bimolecular reactions play an
important role in the pathways of the HTU reaction of this model compound. The products
123
Summary
that have been detected in the water phase suggest that initially glucose degrades to
aldehydes through base catalysed isomersations, retro-aldol reactions, β-eliminations, αdicarbonyl cleavages, and aldol condensations. Formation of carboxylic acids via benzilic
rearrangements and α-dicarbonyl cleavages leads to a decrease of the amount of aldehydes.
After a relatively long residence time (15 min) in a reaction of a 500 mM
D-glucose
solution at 340 ºC and 17.5 MPa, aldehydes could no longer be detected. Also acid
catalysed reactions including dehydratation/hydratation and alkylation occur. Formation of
acids during the course of the reaction leads to increasing the acidity of the reaction
mixture, which may favor the acid catalyzed decarbonylation follwed by polymerization of
the resulting olefins. In the final tar layer, the concentration of aldehydes is very low. By
means of GC-MS analysis after destillation of this fraction a number of products could be
identified, which were very similar to those obtained from a HTU reaction of biomass in a
pilot plant, which demonstrates that glucose is a good model compound to study reaction
pathways of the HTU reaction of biomass.
In Chapter 5 a detailed study of one of the main initial compounds obtained during
HTU reactions on some monosaccharidies as a model compounds, glycolaldehyde, is
presented. The reactions were performed with a continuous flow reactor. In contrast to the
previously studied reactions with monosaccharides, no aromatic products have been
observed in the reaction mixture obtained from glycolaldehyde. The products obtained
show that aldol condensations and retro-aldol reactions are important. Surprisingly, some
products were obtained, which can only be explained with reaction paths involving
reductions of intermediate aldehydes, including ethylene glycol, hydroxyacetone, and 1hydroxy-2-butanone. Most likely, Cannizzaro reactions of intermediate aldehydes can
account for these products.
Finally, chapter 6 deals with hydrogen transfer reactions during the HTU reactions
of glycolaldehyde and glucose in the presence of formic and acetic buffers. A comparison
of results of the HTU reactions of glycolaldehyde with formate and acetate shows that the
hydride from HCOOH is involved in the reduction of glycolaldehyde to glycol. Most likely,
in addition (cross)-Cannizzaro reactions, involving aldehydes formed by condensation of
glycolaldehyde are occurring. Under more basic conditions (pH starting mixture > 6), base
catalyzed reactions dominate over the hydride transfer reactions. Hydride reactions can be
124
Summary
catalyzed by applying Ni2+ or Cu2+) as catalysts. In the presence of these metal ions, a large
amount of glycolic acid and ethylene glycol was formed.
125
Samenvatting
Hoofdstuk 1 is een algemene inleiding. Het hoofdstuk beschrijft aspecten van
energie in de toekomst en hernieuwbare en niet-hernieuwbare bronnen van energie. De
populairste technologieën voor de omzetting van biomassa ´´gasification´´, fermentatie en
´´hydrothermal upgrading´´ worden besproken.
Hoofdstuk 2 handelt over de apparatuur, die nodig is voor het bestuderen van
hydrothermische reacties. Omdat het belangrijk is om inzicht te krijgen in de eerste
reactiestappen van dit proces, is een vereiste voor apparatuur dat korte reactietijden onder
extreme condities (340 °C, 27.5 MPa) mogelijk zijn. Eerst wordt een overzicht gegeven van
karakteristieke voorbeelden van de in de literatuur voor dit doel gerapporteerde apparatuur
en daarna wordt de apparatuur die gebruikt is bij het in dit proefschrift beschreven
onderzoek in detail beschreven.
Hoofdstuk 3 beschrijft een studie van hydrotherme reacties aan een aantal
monosacchariden als modelstof. Hydrolyse naar glucose is een belangrijke stap. Ten einde
enkele reactiepaden, die volgen op deze hydrolyse, te ontrafelen werden de
reactieproducten van hydrotherme behandeling (340 °C, 27.5 MPa) van verdunde
oplossingen van D-glucose en enkele andere monosacchariden bestudeerd. Omdat Kw onder
sub-kritische condities relatief hoog is, treden er zowel zuur- als base-gekatalyseerde
reacties op. De zuur-gekatalyseerde reacties zijn voornamelijk dehydrataties, die in eerste
instantie 5-hydroxymethylfurfural geven. Belangrijke reactiepaden via base-gekatalyseerde
reacties resulteren in glycolaldhyde en glyceraldehyde. Verdere fragmentaties en
dehydrataties leiden tot een variëteit aan laag-moleculaire verbindingen zoals mierenzuur,
azijnzuur, melkzuur, acrylzuur, furfural, en 1,2,4-benzeentriol. Glycolaldehyde is een
belangrijk intermediair in reactiepaden, die resulteren in afname van het O-gehalte van de
vloeibare producten. Het vormt pyruvaldehyde, dat op zijn beurt wordt omgezet in
mierenzuur en acetaldehyde. De laatste verbinding kan ook gevormd worden via
isomerisatie van glyceraldehyde naar melkzuur gevolgd door decarbonylering.
Hoofdstuk 4 geeft de resultaten van een studie naar de invloed van de glucose
concentratie op zijn reactiepaden tijdens de HTU reactie van deze modelverbinding. De in
de waterlaag waargenomen producten suggereren dat glucose in eerste instantie degradeert
127
Samenvatting
naar aldehyden via base-gekatalyseerde isomerisaties, retro-aldol reacties, β-eliminaties, αdicarbonyl splisingen, en aldol condensaties. Vorming van carbonzuren via benzil
omleggingen en α-dicarbonyl splitsingen leidt to afname van de hoeveelheid aldehyden. Na
relatief lange verblijftijden in de reactor (15 min) bij 340 ºC en 17.5 MPa konden geen
aldehydeproducten meer waargenomen worden in het reactiemengsel van een 500 mM Dglucose oplossing. Daarnaast treden er ook zuur-gekatalyseerde reacties op, zoals
dehydratatie/hydratatie en alkylering. De vorming van zuren leidt to verhoging van de
zuurgraad van het reactiemengsel, waardoor decarbonylering van de carbonzuren bevorderd
wordt. De ontstane olefinen kunnen vervolgens polymeriseren. In de uiteindelijk gevormde
teerfase is de concentratie van de aldehyden erg laag. Door middel van GC-MS analyse van
destillatiefracties van deze fractie konden een aantal componenten geïdentificeerd worden.
Analoge verbindingen werden aangetoond in een teerfractie verkregen uit een HTU reactie
van biomassa in een proeffabriek. Dit bevestigt dat glucose een goede modelstof is voor de
bestuderen van de reactiepaden vanuit biomassa in de HTU reactie.
In hoofdstuk 5 wordt een gedetailleerde studie naar het gedrag van
glycolaldehyde, een initieel intermediair in het het HTU proces, gepresenteerd. De reacties
werden uitgevoerd met een continue flow-reactor. In tegenstelling tot de in hoofdstuk 3
beschreven reacties met monosacchariden, werden geen aromatische producten verkregen.
De verkregen producten tonen aan dat condensaties, aldol en retro-aldolreacties een
belangrijke rol spelen. Daarnaast zijn er een aantal verbindingen aangetoond, zoals glycol,
hydroxyaceton en 1-hydroxy-2-butanon, die alleen ontstaan kunnen zijn door middel van
reductie van intermediaire aldehyden. Zeer waarshijnlijk zijn Cannizzaro reacties
verantwoordelijk voor het ontstaan van deze producten.
Hoofdstuk 6, tenslotte handelt over hydride transfers gedurende de HTU reacties
van glycolaldehyde en glucose in aanwezigheid van formiaat en acetaat buffers.
Vergelijking van de reactieproducten van de reacties in aanwezigheid van formiaat met die
in aanwezigheid van acetaat leidt tot de conclusie dat formiaat fungeert als hydride donor,
Daarnaast spelen (cross)-Cannizzaro reacties met aldehyden, gevormd via condensatie van
glycolaldehyde, een rol. Onder meer basische omstandigheden (pH uitgangsmengsel > 6)
domineren
128
base-gekatalyseerde
condensaties.
Hydride
transfer
reacties
blijken
Samenvatting
gekatalyseerd te kunnen worden door Ni2+ en Cu2+. In aanwezigheid van deze metaalionen
worden aanzienlijke hoeveelheden glycolzuur en glycol gevormd.
129
130
Acknowledgements
First of all I would like to thank to my direct supervisor Dr. Joop A. Peters. Joop,
you were like a father for me during all the years we were working together. Thanks for
your patience and wisdom. Besides chemistry and NMR I have learned a lot about life from
you.
Thanks are due to Prof. dr. Thomas Maschmeyer for the opportunity to become a
member of his team and for believing in me. Thomas I will never forget your jokes!
Prof. dr. Roger Sheldon for accepting to be my promotor. Prof. dr. Herman van Bekkum for
big help with chapter 1.
My Ph.D. would not be possible without the excellent work that was done by my first
teacher of chemistry. So big thanks due to mgr Małgorzata Krywult. Your lectures were full
of passion and scientific curiosity. I will never forget the first chemical experiments that I
have seen during your lessons. Those experiments have charmed me so much, that I knew
that I will become a chemist.
I am grateful to my teachers at my high-school (Chemical College in Oświęcim),
mgr Zofia Kłys, mgr inż. Urszula Janik for the opportunity they gave me to develop my
skills and interest in chemistry, mgr Ewa Oplustil and mgr Jarosław Kolka, for teaching me
physics and mathematics.
Thanks are due to my Professors at the Cracow University of Technology: Dr.
Barbara Żmudzińska-Żurek, Dr. inż. Mieczysław Chmura, Dr. inż. Andrzej Wyczesany, Dr.
inż. Zdzisław Borowiec, Dr. inż. Krystyna Porzycka-Semczuk for knowledge, great
atmosphere and friendship during my study there.
Thanks to Dr. Frans Goudriaan and Dr. Gerard Luijkx for helpful discussion
during all the project. I am particularly grateful to the late Dr. Leo Petrus, with whom I had
many fruitful and motivating discussions. I always will remember his enthusiasm. He
deceased too young.
Kristina, thank you very much for all your help, your NMR analyses and for
sharing the office with me. I am grateful that you could stand me and for listening to all my
stories. Anton van Estrik for your NMR analysis, friendship and for teasing me.
Fortunately, you went for your retirement before I finished this thesis.
131
Acknowledgements
Thanks to my students: Anne-Gaëlle Bouche, Nicole de Groot, Rob Strik, Jeroen
van Luijtelaer for performing some experiments.
Luuk van Langen and Michiel van Vliet for great humor, willingness to help,
optimism and friendship. Adrie and Johan, for LC-MS analyses. Fred van Rantwijk for help
with HPLC. Ernst Wurtz and Rien Sloter for assistance with the construction of the
experimental set-ups. Ricardo for discussions about computer technologies, Bart for help
with web design and friendship.
Ton Hubregste for nice discussions and for joining me at dinner in the Aula.
Big thanks to all nice people I met in the lab: Leen, Koos, Mieke, Mieke J., Frank, Lars,
Gerd-Jan, Arné, Scoob, Annemieke, Chrétien, Carlos P., Carlos G., Erwin, Luca G., Xavier,
Leon, Rafaella, Sandrine, Martijn, César, Andrea, Menno, Michiel H., Dirk, Li, Jan-Kees,
Petra, Jan K., Eva, Vojta, Silvia P. (for dancing together at the parties), Dean, Moira,
Bruno, Pedro, Hilda, Rute, Paulo P., Stefan (thanks for stories about England and playing
chess with me), Jacek, Daniel S. (thanks for help with ch.1 and friendship).
Thanks to my Polish colleagues: Ola, Arek, Andrzej, for friendship, jokes and
laughing together. I hope that my advices helped you.
Thanks to sponsors: NWO, Shell Global Solutions.
Finally, thanks due to my parents for believing in me and for all help.
It was great time,
THANK YOU ALL !
132
Curriculum Vitae
Zbigniew Wojciech Srokol was born in BielskoBiała, Poland, on August 12th, 1973.
In primary school he became interested in
chemistry and therefore in October 1988 he joined
the Chemical College in Oświęcim.
His passion for chemistry has brought him as a
student to the Cracow University of Technology in
Poland where he studied at the Faculty of
Chemical
Engineering and Technology.
He
focused and specialised on Crude Oil Distillation
Processes and Design. He graduated in June 2000 after having defended his Master’s
Thesis and obtained his Master of Science title.
From March 2001 to April 2006 he worked as a Ph.D. student at the Faculty of
Applied Science in Laboratory of Organic Chemistry and Catalysis at the Delft University
of Technology. His scientific work was focused on determination of the reaction pathways
during hydrothermal upgrading biomass to biofuel, under the supervision of Dr.ir. Joop A.
Peters. The research described in this thesis was supported with a grant from the
Netherlands Organization for Scientific Research (NWO) with financial contributions of
Shell Global Solutions, The Dutch Ministry of Economic Affairs (Senter) and Novem.
133
134