THERMALLY CURED COIL-COATINGS

UTILIZING NOVEL RESINS AND

FATTY ACID METHYL ESTERS
AS REACTIVE DILUENTS

Katarina Johansson

AKADEMISK AVHANDLING

Som med tillstånd av Kungliga Tekniska Högskolan i Stockholm framlägges till

offentlig granskning för avläggande av teknisk doktorsexamen fredagen den
19 september 2008, kl. 10.00 i sal F3 ”Flodis”, Lindstedtsvägen 26, KTH,
Stockholm. Avhandlingen försvaras på engelska.
Copyright © 2008 Katarina Johansson
All rights reserved

Paper I © 2006 Elsevier B.V.

Paper II © 2007 Elsevier B.V.
Paper III © 2007 American Chemical Society Publications
Paper IV © 2008 Elsevier B.V.
Figure 1 © National Centre of Prehistory

TRITA-CHE Report 2008:56

ISSN 1654-1081
ISBN 978-91-7415-097-1
ABSTRACT
Solvent-borne thermally cured coil-coating resins contain large amounts
of volatile organic solvents in order to obtain suitable flow for film application.
This work describes how the expensive and environmental hazardous volatile
organic solvent content of a solvent-borne thermally cured polyester/melamine
coil-coating system can be reduced by introduction of fatty acid methyl esters
(FAMEs) as reactive diluents and modification of the polyester binder resin. The
evaluated reactive diluents, two rape seed methyl esters (RMEs), two linseed oil
methyl esters (LMEs), and a tall oil methyl ester (TOME) have been evaluated both
in a fully formulated clear-coat system and via model studies.
Viscosity measurements of wet paint mixtures showed that formulations
with hyperbranched polyester binder hold lower viscosity than conventional
polyester binder resins and that FAME works as a diluent. Fully formulated clear-
coats were cured under simulated industrial coil-coating cure conditions and in a
convection oven at lower temperatures respectively. FAME increases the mobility
of the system enhancing the film formation process. Free-standing clear-coat films
were analyzed with Raman, carbon-14 dating, extraction, dynamic mechanical
analysis, and visual observation. Incorporation of FAME could not be confirmed
by Raman analysis. However, carbon-14 dating indicated the presence of FAME
that could not be extracted from the films. The mechanical properties of the films
were also affected by the addition of FAMEs, oven temperature, choice of co-
solvent, and flash-off period. Conventional film characterization tests on substrate
supported coatings indicated that binder resin structure and cure conditions affect
the final film properties.
Model studies were performed to clarify how FAME can chemically react
through transesterification with the hydroxyl-groups of the polyester. The
transesterification reaction between different FAMEs and primary alcohols with
and without tertiary hydrogen was monitored with 1H-NMR and real time IR.
Evaporation and side reactions, e.g. alkene reactions, are competing factors to the
transesterification reaction. The study showed that fatty acid structure, reaction
time, and temperature affect the transesterification conversion, degree of side
reactions, and choice of catalyst.
A pigmented fully formulated polyester/melamine formulation with a
reactive diluent was cured at full scale in an industrial coil-coating production
facility. Evaluation of the final film properties showed that the coating fulfills the
specification of conventional polyester/melamine coil-coating systems.
SAMMANFATTNING
Lösningsmedelsburna och termiskt härdande bandlackeringsfärger
innehåller stora kvantiteter dyra och miljöfarliga organiska lösningsmedel för att
erhålla lämpliga reologiska egenskaper för applicering. Detta arbete beskriver hur
lösningsmedelsinnehållet kan reduceras genom att använda fettsyrametylesterar
(FAME) som reaktiv spädare och modifiera polyester-bindemedlet i ett
lösningsmedelsburet och termiskt härdande polyester/melamin
bandlackeringssystem. De använda reaktiva spädarna, två rapsmetylestrar (RME),
två linoljemetylestrar (LME) och en talloljemetylester (TOME), har utvärderats
både i ett komplett formulerat klarlackssystem och genom modellstudier.
Reologiska mätningar på ohärdade färgformuleringar visar på att
hyperförgrenade polyestrar har lägre viskositet än konventionella polyestrar samt
att FAME späder färgsystemet. Färgfilmer av kompletta klarlacksformuleringar
har härdats under simulerade industriella härdningsförhållanden och vid lägre
temperatur i en konvektionsugn. Tillsatts av FAME förbättrar filmbildningen och
ökar antalet felfria filmer, eftersom mobiliteten i färgen ökar under torkförloppet.
Fristående färgfilmer har analyserats med Raman-spektroskopi, kol-14-datering,
extraktion, dynamisk mekanisk analys och genom visuella observationer.
Inkorporering av FAME genom tvärbindning kunde inte verifieras med Raman-
analys. Kol-14-datering och extraktionsförsök visar dock på att FAME
inkorporerats i färgfilmen. Även färgfilmernas mekaniska egenskaper påverkas av
tillsatts av FAME, härdningsbetingelser, lösningsmedelsval och flash-off före
härdning. Konventionella filmkarakteriseringstester av färgfilmer på substrat
visar på att bindemedelsstruktur och härdningsförhållanden påverkar
filmegenskaperna.
För att förstå hur omförestring mellan FAME och hydroxyl-grupper i en
polyester kan ske har modellstudier genomförts. Reaktion mellan FAME och
primära alkoholer, med och utan tertiära väten, har studerats med 1H-NMR och
realtids-IR. Avdunstning och sidoreaktioner, t.ex. alken-reaktioner, konkurrerar
med omförestringsreaktionen. Studien har visat att fettsyrastruktur, reaktionstid
och temperatur påverkar omförestringens omsättning, mängd sidoreaktioner samt
val av katalysator.
En pigmenterad polyester/melamin färgformulering med reaktiv spädare
har härdats i en fullskalig industriell bandlackeringslina. Utvärdering av
filmegenskaperna bekräftar att bandlackeringsfärgen uppfyller kraven för
industriellt använda polyester/melamin bandlackeringsfärger.
LIST OF PAPERS
The thesis is a summary of the following papers:

I. “A Model Study on Fatty Acid Methyl Esters as Reactive Diluents in

Thermally Cured Coil Coating Systems”, K. Johansson, M. Johansson,
Progress in Organic Coatings, 55 (2006) 382-387*

II. “The Effect of Fatty Acid Methyl Esters on the Curing Performance and
Final Properties of Thermally Cured Solvent-Borne Coil Coatings”,
K. Johansson, M. Johansson, Progress in Organic Coatings, 59 (2007) 146-151

III. “Fatty Acid Methyl Esters as Reactive Diluents in Coil-Coatings”,

K. Johansson, M. Johansson, Polymer Preprints, 48(2) (2007) 857-858

IV. “Fatty Acid Methyl Ester as Reactive Diluent in Thermally Cured Solvent-
Borne Coil-Coatings - The Effect of Fatty Acid Pattern on the Curing
Performance and Final Properties”, K. Johansson, M. Johansson, Progress
in Organic Coatings, 63 (2008) 155-159
V. “Hyperbranched Aliphatic Polyesters as Thermoset Resins for Thermally
Cured Coil-Coatings”, K. Johansson, T. Bergman, M. Johansson,
manuscript

The thesis also contains unpublished results.

* For corrections to paper I see page 59.

My contribution to the appended papers:

I. All the experimental work and most of the preparation of the manuscript.

II. All the experimental work and most of the preparation of the manuscript.

III. All the experimental work and most of the preparation of the manuscript.

IV. All the experimental work and most of the preparation of the manuscript.

V. A majority of the experimental work and most of the preparation of the

manuscript.
TABLE OF CONTENTS
1 PURPOSE OF THE STUDY ....................................................................................1
2 INTRODUCTION .....................................................................................................2
2.1 COATINGS.............................................................................................................2
2.1.1 Components .................................................................................................2
2.1.2 Reactive diluents..........................................................................................4
2.1.3 Film properties.............................................................................................4
2.1.4 Coil-coatings................................................................................................5
2.2 VEGETABLE OILS ..................................................................................................7
2.2.1 Chemistry of oils..........................................................................................8
2.2.1.1 Auto-oxidation..................................................................................................... 8
2.2.1.2 Esterification...................................................................................................... 11
2.2.1.3 Alkene reactions ................................................................................................ 11
2.2.1.4 Oil modification................................................................................................. 11
2.2.2 Vegetable oils in coatings ..........................................................................12
2.2.3 Fatty acid methyl esters .............................................................................13
2.3 DENDRITIC POLYMERS ........................................................................................13
3 EXPERIMENTAL...................................................................................................15
3.1 MATERIALS ........................................................................................................15
3.2 SAMPLE PREPARATION .......................................................................................16
3.2.1 Fully formulated clear-coat systems ..........................................................16
3.2.1.1 Conventional polyester resin.............................................................................. 16
3.2.1.2 Hyperbranched polyester resin .......................................................................... 18
3.2.1.3 Curing procedures.............................................................................................. 18
3.2.2 Model study ...............................................................................................19
3.2.2.1 Model reactions in thin films ............................................................................. 19
3.2.2.2 Model reactions in bulk ..................................................................................... 20
3.2.3 Full scale industrial evaluation ..................................................................20
3.3 TECHNIQUES .......................................................................................................21
3.3.1 Thermal gravimetric analysis.....................................................................21
3.3.2 Rheological measurements ........................................................................21
3.3.3 Spectroscopical analysis ............................................................................21
1
3.3.4 H-NMR.....................................................................................................22
3.3.5 Carbon-14 dating .......................................................................................22
 3.3.6 Extraction...................................................................................................22
3.3.7 Dynamic mechanical analysis....................................................................22
3.3.8 Differential scanning calorimetry ..............................................................22
3.3.9 Micro hardness...........................................................................................22
3.3.10 Conventional film characterization............................................................23
4 RESULTS AND DISCUSSION ..............................................................................24
4.1 MISCIBILITY .......................................................................................................24
4.2 RHEOLOGY .........................................................................................................25
4.3 FILM FORMATION ...............................................................................................27
4.4 INCORPORATION OF FAME ................................................................................27
4.4.1 Fully formulated clear-coats ......................................................................27
4.4.2 Model study ...............................................................................................29
4.4.2.1 Effect of FAME structure .................................................................................. 31
4.4.2.2 Effect of catalyst ................................................................................................ 32
4.4.2.3 Effect of alcohol structure.................................................................................. 33
4.5 FILM PROPERTIES ................................................................................................34
4.5.1 Free-standing clear-coats ...........................................................................34
4.5.1.1 Effect of FAME structure .................................................................................. 36
4.5.1.2 Effect of oven temperature................................................................................. 37
4.5.1.3 Effect of co-solvent............................................................................................ 37
4.5.1.4 Effect of flash-off .............................................................................................. 40
4.5.2 Substrate supported clear-coats..................................................................41
4.5.2.1 Effect of resin structure...................................................................................... 41
4.5.2.2 Effect of substrate .............................................................................................. 42
4.5.2.3 Effect of oven temperature................................................................................. 43
4.6 FULL SCALE INDUSTRIAL EVALUATION ...............................................................45
5 CONCLUSIONS ......................................................................................................46
6 FUTURE WORK.....................................................................................................48
7 ACKNOWLEDGEMENTS ....................................................................................49
8 REFERENCES ........................................................................................................51
CORRECTIONS .............................................................................................................59
 Purpose of the Study

1 PURPOSE OF THE STUDY

Increased environmental awareness and governmental legislations during the past
decades have resulted in an increased industrial interest to reduce the use of
volatile organic compounds (VOC) in organic coatings. The use and subsequent
emission of VOC is expensive and environmentally hazardous as it e.g.
contributes to global warming. Solvent-borne thermally cured coil-coating resins
contain large amounts of solvents in order to obtain suitable viscosity, flow
behavior and wetting of the substrate for film application. During cure in large
convection ovens, the solvent is evaporated and thereafter incinerated for energy
recovery. The purpose of this work was to reduce the volatile solvent content in a
solvent-borne thermally cured polyester/melamine coil-coating system. It is here
proposed that a reduction in solvent content can be achieved by both introduction
of reactive diluents and modification of the polyester binder resin. The main
issues are how new resins and specific reactive diluents function and chemically
react under the very special cure conditions associated with coil-coatings i.e. high
cure speed, high temperature, and high material through-put.

Fatty acid methyl esters (FAMEs) have been evaluated as reactive diluents as they
generally have low viscosity, good solubility properties, and possess a functional
group, the ester acyl moiety, which can react. Furthermore, if the reactive diluent
could be derived from a renewable resource it would have twofold positive
environmental impact since it would not only reduce the VOC content but also
introduce a renewable material into the final film. Moreover, FAMEs are cheaper
than many of the conventional solvents used today in coil-coating formulations.

Reduction of the solvent content by exchanging conventional polyesters with

hyperbranched polyester as the binder resin has also been studied.
Hyperbranched polymers are known to exhibit lower viscosities than
conventional polymers at equivalent molecular weights because their globular
structure hampers chain entanglements.

1
Introduction

2 INTRODUCTION
2.1 COATINGS

As early as 20-30 000 BC, mankind knew how to produce paint by mixing colored
minerals, e.g. ferric oxide, and titanium dioxide, with blood, milk or sap. These
old kind of paintings are still around for us to see, like for example rock-paintings
such as the ones in the cave of Lascaux in France, Figure 1. The coating industry
has developed vastly since then and a lot of research has been conducted and
endless systems are sill to be evaluated. Coatings of today are not only used for
decorative and aesthetic purposes but also for several other applications such as
corrosion protection, impact and abrasion resistance, and hygienic properties.

Figure 1. 17,000 years old rock-painting from the cave of Lascaux, France. Picture published with
permission from the National Centre of Prehistory, France.1

2.1.1 Components

Paints are generally composed of an organic binder, volatile components,

additives and in most cases pigments.2-4 The final coating, after application and
evaporation of volatile compounds, consist of solidified binder and pigments. The
solidification process can progress in several different ways, e.g. coalescence of

2
 Introduction

thermoplastic latex, air-drying of alkyds and chemical curing of two-pack systems

such as phenolic/epoxy.

Additives2, 3 are included in small quantities to modify different coating

properties. Catalysts and driers are added to accelerate cure rates. The appearance
of the coating can be improved with e.g. wetting agents, flow modifiers,
dispersants, coalescent and film forming aids. Further, the film properties can be
modified with inter alia plasticizers and stabilizers. These additives are often solid
organic compounds in solution.

Pigments2, 3, 5, 6 are generally insoluble compounds that are added to provide color
and opacity. However, they can also supply other functions such as exterior
durability, corrosion resistance, fire retardation, and mechanical strength.2, 3, 5, 6
Extender pigments and fillers, e.g. talc and clay, do not give color, but are used in
high quantities as they are cheap, and to add mechanical strength and to adjust
the gloss. Titanium dioxide, carbon black, and iron oxide are commonly used
colored pigments. Coatings with little or no pigments are transparent and called
clear-coats.

Volatile compounds2, 3 are used in paints to obtain desired flow and leveling for
application and film formation. Evaporation rate of the volatile material during
application and film formation is dependent on temperature, vapor pressure,
surface/volume ratio, rate of air flow over the surface, and diffusion through the
film.2, 3 Choice of volatile compound does not only influence the cure rate but can
also affect film properties and aesthetic appearance.7 Too rapid solvent release
entails poor wetting with subsequent adhesion problems and uneven films with
surface defects like orange peel, and frothing. Uneven coatings with e.g. sag and
curtain defects are obtained if the evaporation rate instead is too slow. Moreover,
the solvency property of the volatile component must remain adequate
throughout the entire film formation process to yield defect free films. Particularly
important is that the final solvent released, i.e. the tail solvent, is a powerful true
solvent, as mixtures of solvents often are utilized.

The composition of the paint components is adjusted to obtain desirable coating

properties to withstand use without damage. Coating application can be
performed with several different methods depending on the flow properties of the
paint and substrate to be coated. Industrial application techniques are e.g. by
brush, large rollers, and air assisted spray.

3
Introduction

2.1.2 Reactive diluents

A reactive diluent4, 8, 9 is a compound that acts as a solvent in the liquid paint,

lowering the viscosity, and further chemically reacts into the final film during
cure. By introducing a reactive diluent, the solvent content and thereby the VOC
content, of a paint can be reduced. To be able to act as a good diluent the reactive
diluent must be compatible with the matrix and have low viscosity. The reactive
diluent must also have a functional group that can react with the other coating
components during cure. The functional groups must be of the right amount and
reactivity to prevent in-can storage problems, retardation of the drying, and to
avoid remains of non-reacted reactive diluent working as plasticizer in the final
film. For high temperature applications, e.g. coil-coatings, the volatility must be
low enough to prevent evaporation.

2.1.3 Film properties

Mechanical properties of coatings are not only related to the paint components but
also to e.g. thickness, adhesion, heterogeneity, pigment volume concentration,
modulus, and glass transition temperature (Tg).3 There are numerous different
methods3 to evaluate film properties such as adhesion, abrasion, hardness,
formability, exterior durability, deformation and chemical resistance. Hardness
can be determined with pendulum tests, indentation, scratch resistance, and pencil
hardness. Adhesion is commonly rated with cross hatch adhesion test, but also
with direct pull test or wear of a penknife. Formability and flexibility at slow
deformation rate are studied e.g. with Erichsen cupping, T-bend, and Mandrel
bend test and at high deformation rate with impact tests. Chemical resistance is
generally assessed with double rub with methyl ethyl ketone (MEK) until film
rupture. Exterior and corrosion durability are both evaluated with outdoor
exposure and accelerated tests. Under accelerated tests coatings are subjected to
e.g. UV radiation, water, and salt spray as changes in mechanical properties, gloss
and color retention are measured. However, these methods commonly test
combination of properties of the coating and are empirical, hence they are more
suitable for quality control. International standards, such as ISO10, CEN11, and
ASTM12, regulate the test methods.

Film properties are affected by Tg as it regulates the molecular mobility in the

coatings.3 During film formation, with subsequent increase in Tg, the free volume
is decreased i.e. the mobility is restricted. A lower baking temperature, than the Tg
of the solvent free polymer, retards solvent evaporation for thermoplastic systems.

4
 Introduction

For thermosetting resins the cross-linking rate will be affected by limited polymer
chain motions at a too low cure temperature. Hence, it is important to perform the
film formation at a significantly higher temperature than the Tg of the resin to
obtain solvent-free and fully cross-linked coatings. Tg can be determined with e.g.
differential scanning calorimetry (DSC)3 and dynamic mechanical analysis
(DMA)3, 13. Storage modulus (E'), loss modulus (E"), and tan delta (E"/E'=tan δ) is
obtained from DMA measurements. Tg is defined as the peak of either the loss
modulus or the tan δ plot. Evaluation with DSC and DMA renders different Tg’s,
where the peak value of the loss modulus plot usually corresponds best to the
DSC value. The width of the Tg transition, i.e. the E' slope and tan δ width,
correlates to the heterogeneity of the coating and is e.g. a measure of cross-link
density. The greater sensitivity of DMA measurements makes it a better option
than DSC for cross-linked systems.

2.1.4 Coil-coatings

In the coil-coating processes14-16, i.e. pre-coating of sheet metal, an organic coating

is continuously applied to sheet metal in a closed system. The field of application
of pre-coated steel and aluminum is vast and range from markets such as
construction, automotive, domestic appliance, heating and ventilation, office
furniture, and lighting. Coil-coating is attractive for many end-users since it
permits omitting of paint shops improving efficiency and reducing production
costs. Both thermoplastic and thermosetting coating systems are used. Industrially
used resins are polyvinyl chloride, polyvinyl difluoride, and polyester cross-
linked by melamine or isocyanates. Polyester resins are dominating worldwide.

Top-coat

Primer

Pre-treatment

Steel/aluminum substrate

Figure 2. Schematic description of the different coil-coating layers.

In the coil-coating process, large coils of metal are unwound, pretreated, primed,
top-coated, and recoiled at a constant speed of 60-150 m/min, Figure 2 and Figure
3. The metal coils are 0.6-1.8 m wide and between 600 to 1800 m long with a
weight of 5-6 tonnes for aluminum and up to 20 tones for steel. The substrate is
pretreated and primed to inhibit corrosion and improve adhesion. The top-coat

5
Introduction

improves mechanical and exterior durability. Both primer and top-coat paints are
applied via two or three large rollers on one or both sides. The coating is cured in a
convection oven at approximately 350°C for 20-40 s to reach a peak metal
temperature (PMT) of about 200-250°C. The short dwell time put special demands
on the film formation and curing chemistry to obtain a final product without
defects like frothing, sagging and leveling. The final coated sheet metal is cut,
bent, punched, and drawn into various shapes by the customer. This harsh
forming process puts high demands on the film properties, such as high flexibility
and good adhesion.

A. Bare metal is uncoiled F. Primer curing oven

B. Coil splicing G. Top-coat application
C. Accumulator stack H. Top-coat curing oven
D. Metal degreasing, cleaning, rinsing, and I. Optional lamination or embossing
chemical pretreatment J. Accumulator stack
E. Primer application K. Recoiling of finished pre-coated sheet metal

Figure 3. Schematic depiction of the coil-coating process.

To be able to apply the coating the paint formulations often consist of over 50 wt%
solvent. The solvent is evaporated during the cure process, collected and
incinerated to recycle the energy. This high solvent content is costly and involves
handling of large amounts of environmental hazardous VOC. Reduction of the
solvent content is highly desirable due to increased environmental awareness and
government legislations.17, 18 To achieve this reduction methods widely used in the
post-painting industry have been proposed for coil-coatings, e.g. water-borne3, 19,
powder coatings3, 20-22, and UV curable systems3, 21, 22. However, these methods
have not been applied in large scale in the coil-coating industry due to clean up
difficulties upon color change and cost issues concerning modification of the
production facilities.23 Reactive diluents is a VOC reduction alternative that
utilizes the same production procedures, that has been suggested for several other
coating applications.9, 18, 24-29

6
 Introduction

2.2 VEGETABLE OILS

Vegetable oils2, 3, 30 are triglycerides, i.e. triesters of glycerol and three fatty acids,
Figure 4. The fatty acid structure of a triglyceride can vary in chain length and
amount of double bonds, i.e. vegetable oil contains a mixture of fatty acids. The
fatty acid pattern is dependant on type of crop and growth condition, e.g. greater
level of unsaturation in oil grown in colder climate.31, 32 To prevent oxidation
reactions in the plant, the oil often contains antioxidants such as tocopherol.

O
O
O
O

Figure 4. Triglyceride

Traditionally, vegetable oils have been used in foodstuff and air-drying coating
systems. As vegetable oils are renewable materials, extensive research has been
performed to find new fields of application, e.g. epoxidized oil as plasticizers and
stabilizers for vinyl plastics33-35, biobased materials36-39, biodiesel40, reactive
diluents24, 26, 28, 29, and printing inks33-35.

Vegetable oils can be divided into three groups; non-drying, semi-drying, and
drying oils, depending on their fatty acid pattern. Non-drying oils contains
saturated fatty acids that can not be cross-linked by air oxidation. Olive oil and
castor oil are examples of non-drying oils. Semi-drying oils, e.g. soybean and tall
oil, contain one or two unsaturations that slowly can cross-link through oxidation.
Semi-drying oil films will never be completely tack-free. Drying-oils are highly

7
Introduction

unsaturated oils that air oxidizes to form a tack-free film with time. A drying-oil is
traditionally defined as an oil with an average number of diallylic groups per
molecule greater than 2.2.3 Linseed and tung oil are commonly used drying-oils.
Oils can further be divided into non-conjugated and conjugated depending on
whether the unsaturations are separated by a methylene group or not.

Fatty acids can also contain more unusual functionalities such as conjugated
alkenes33-35, alkyne41, 42, epoxy33-35, 43, and hydroxyl groups33-35. These unusual
structured oils are not grown in large quantities today, but modern crop
development and gene modification techniques38, 44 may in the future make them
an to large extent used environmental friendly alternative as raw materials in
monomer and polymer synthesis45-49.

2.2.1 Chemistry of oils

Vegetable oils can react through many different reaction routes such as auto-
oxidation, esterification, and alkene reaction.

2.2.1.1 Auto-oxidation

Auto-oxidation3, 30, 50-53 is a complex free radical process, Figure 5. This process is
generally described by an induction period, followed by uptake of oxygen, and a
final polymerization stage involving cross-linking and formation of volatile
oxidation products.

During the induction step antioxidants, e.g. tocopherols, are consumed. The
oxidation is proposed to be initiated by abstraction of an allylic hydrogen, forming
a carbon-centered radical. However, how the oxidation reaction is initiated is still
under debate, but it is suggested that singlet oxygen and hydroperoxides causes
the initiation.54, 55 Auto-oxidation can also be initiated by UV-light and γ-radiation.
Hydrogens located between double bonds, i.e. bis-allylic hydrogens, are more
easily abstracted leading to faster drying of fatty acids with greater unsaturation
content. The formed carbon-centered radical rearranges into a conjugated system.
Conjugated fatty acids show a more rapid drying than non-conjugated as this
rearrangement is not necessary. The carbon radical is either terminated with
another radical or, more probably due to high oxygen concentration, reacted with
oxygen to form a peroxy radical. The peroxy radical abstracts hydrogen producing
a hydroperoxide. The unstable hydroperoxide will eventually decompose into an
alkoxy radical. The reactive alkoxy radical can undergo several reactions such as

8
 Introduction

addition to unsaturations or other radicals, hydrogen abstraction, and scission of

the carbon chain. Drying of vegetable oils is conveyed by addition reactions.
Chain scission reactions, on the other hand, degrades the material.

H H

Allylic hydrogen
abstraction

O2

O
O
Peroxy radical
termination
Hydrogen
abstraction
HO
O

Intermolecular Intramolecular Hydrogen

cross-linking chain scission abstraction
reactions reactions

Figure 5. Auto-oxidation mechanism.

9
Introduction

The mechanism of these chain scission reactions is not clear. However, β-scission53,
56-62and Russell termination62, 63 are the dominating theories, Figure 6. The β-
scission mechanism, propose that alkoxy radicals cause the scission producing
both volatile and non-volatile low molecular weight byproducts such as alkanes,
aldehydes, and carboxylic acids.53, 56-62 β-Scission competes with the hydrogen
abstraction mechanism depending on temperature and nature of hydrogen donor
species.53, 62 The Russell termination theory states that peroxy radicals are
terminated by reacting with each other forming ketones and alcohols.62, 63

β-Scission Russell termination

O O
O O
O

O O
O O
O*

O O*
O2
H HO

Figure 6. β-scission and Russell termination

The rather slow oxidative drying rate of fatty acids is often accelerated by addition
of driers and siccatives, i.e. catalysts in form of transition metal salt or soap.2, 3, 30, 64-
66 The catalytic effect of driers is active throughout the service life of the coating,

thus accelerating degradation of the cross-linked material.67 The knowledge of

driers is very old and reaches as far back as Egyptian civilizations, when various
compounds of lead, iron, and manganese metals were used with drying oils.

The oxidative drying process can be followed with e.g. gas chromatography
(GC)68, 69, size exclusion chromatography (SEC)67, 68, 70, nuclear magnetic resonance
(NMR)65, 68, chemiluminescence29, 71-74, infrared (IR)50, 65, 67, 75, real-time IR (RTIR)29, 72-
74, and Raman50, 65 spectroscopy. Rearrangements from non-conjugated to

conjugated unsaturations can be followed with spectroscopic analysis methods.76

Excited carbonyls are generated during the oxidation process, suggestible from
the β-scission or Russell termination.77, 78 These excited carbonyls will eventually
relax and emit photons, i.e. chemiluminescence, that can be measured with a
sensitive photomultiplier.77, 78

10
 Introduction

2.2.1.2 Esterification

The triglyceride of vegetable oils can be reacted with esterification reactions to

form different fatty acid derivatives e.g. free fatty acids, FAMEs, and alkyds.2, 3, 30,
79, 80 Free fatty acids are formed through hydrolysis, i.e. saponification, of the

triglyceride under basic conditions. Presence of alcohol yields transesterification to

the corresponding ester. These esterification reactions can also be enzyme
catalyzed.81, 82 The fatty acid pattern of vegetable oils is often determined by
transesterification to methyl ester followed by gas chromatography analysis.
Transesterification of vegetable oil with polyols and polyesters under acidic
conditions is a common way to synthesize resins, e.g. alkyds. Moreover, resins can
be produced through direct esterification of fatty acids, polyols and polybasic
acids. Depending on the fatty acid structure the resins can be cross-linked with
different curing mechanisms, e.g. oxidative drying, free radical polymerization,
and cationic polymerization.2, 3, 30

2.2.1.3 Alkene reactions

Unsaturations in vegetable oils can not only be polymerized through auto-

oxidation but also via other alkene reactions, such as thiol-ene83, 84 and Diels-Alder
reactions85-89. The thiol-ene reaction is a step-wise free radical process that
proceeds at very high rates.90 This reaction is very versatile and almost any type of
alkenes can be used making it applicable for unsaturated fatty acid derivatives.
Diels-Alder reactions91 take place between 1,3-dienes and alkenes (dienophiles) or
alkynes. The alkenes in fatty acids reacts via Diels-Alder inter-molecularly30, 86 or
e.g. with styrene87, acrylics88, and quinones89.

2.2.1.4 Oil modification

The drying properties of a drying oil can be modified through pre-polymerization,

e.g. heat-bodied, blown or boiled oil.2, 3, 30 The most common modification process
is thermal polymerization at 225°C to 320°C under inert atmosphere to form heat-
bodied i.e. stand oils. Blown oils are produced by heating oil to about 130°C while
passing through air. Heating oil to 250°C in presence of driers produces boiled
oils. All these pre-polymerization methods improve the flow and leveling
properties as the viscosity of the oil is increased. However, the wetting ability
tends to be impaired.

Maleinization2, 3, 30 can increase the drying rate of slow oxidative drying oils, e.g.
soya and safflower oil, and enhance the color appearance of e.g. linseed oil. In

11
Introduction

maleinization a small amount of maleic anhydride is added to the unsaturations of

fatty acids. Water-borne resins can be obtained by controlling the maleic
anhydride concentration.

Another way to modify oils is epoxidation2, 3 of the double bonds with peracetic or
performic acid. Epoxidized oils are used e.g. as plasticizers and in the preparation
of alkyds and polyethers.

2.2.2 Vegetable oils in coatings

Vegetable oils are one of the oldest types of binders and have been used in
different coating applications for centuries. Good wetting properties and low
viscosity, without need for additional solvents, make oils feasible in crude air-
drying coatings, e.g. linseed oil, and as components in resins, e.g. in alkyds,
uralkyds, and epoxy esters.2, 3, 30, 92

Crude oil drying coatings are soft and initially impact and chemically sensitive.2, 3,
30 The coating is degraded and yellows with time due to the radical processes

present. Moreover, the cure speed is generally slow, even for conjugated drying
oils and with driers. These shortcomings can be diminished by use of modified
oils such as bodied, blown, or maleinized oil.

The benefits of vegetable oils can be utilized by using them as raw material in
vehicles. Tall oil fatty acid, linseed, soya, and safflower oil are commonly used for
resin manufacturing.2, 3, 30 Air-drying alkyds generally have good wetting,
adhesion, gloss, flow, leveling, and compatibility properties.2, 3 The resin’s oil
length affects hardness, gloss, and drying properties.2, 3 Alkyds with high oil
content ratio, i.e. long oil alkyds, are slow drying and form soft and flexible
coatings with somewhat low gloss. Short oil alkyd resins, with low oil ratio, on the
contrary, dry faster and render hard coatings with high gloss. Introduction of
diisocyanate yields fast drying uralkyds, with increased hardness, abrasion and
chemical resistance, and somewhat lower gloss.2, 3 Another binder system, epoxy
ester2, 3 is produced from esterification of epoxy resin and vegetable oil. Epoxy
esters possess film properties with greater adhesion, flexibility, hardness and
chemical resistance than conventional alkyds. However, the exterior durability of
epoxy esters is poor.

12
 Introduction

2.2.3 Fatty acid methyl esters

FAMEs are vegetable oil derivatives that have gained utilization. Especially as bio-
diesel40 due to their low viscosity, good storage stability and solubility
properties.93 These properties and the presence of the ester acyl moiety also make
FAME feasible as reactive diluents in coating applications. A reactive diluent
derived from a renewable resource would entail twofold positive environmental
impact since not only the VOC content is reduced but also a renewable compound
is introduced into the final material

FAME is commonly produced through alkaline transesterification with methanol2,

3, 30, 40, but also through acid and enzymatic catalyzed reactions40, 81. Some common

FAMEs are presented in Figure 7.

O O

O O
Methyl stearate Methyl oleate
O O

O O
Methyl linoleate Methyl linolenate
O

Methyl pinolenate

Figure 7. Common fatty acid methyl esters.

2.3 DENDRITIC POLYMERS

Dendritic polymers94 are a class of highly branched macromolecules with a large

number of end-groups. This class of polymer can further be divided into three
subgroups; dendrimers, dendrigrafts and hyperbranched polymers. The
subgroups differ in relative structural control of the branching units. Dendrimers
possess the greatest architectural control and have a polydispersity index (PDI)
between 1.00 and 1.05.94 Dendrigrafts have a arborescent structure with a PDI of
1.1-1.5.94 Hyperbranched polymers have the highest polydispersity of the three
subgroups with a PDI between 2 and 10.94

Dendritic polymers consist of three components; the core, the interior, and end-
groups, Figure 8. The interior is built up of ABx (x ≥ 2) monomer layers, i.e.

13
Introduction

generations. This globular architecture hampers chain entanglements giving

unique rheological properties, inter alia low viscosity in relation to molecular
weight.95, 96 Numerous monomers can be used for dendritic polymer synthesis
resulting in a wide range of properties that can be further tailored with end-group
modification. These properties and the lower production cost of hyperbranched
polymers97, compared to dendrimers, make them industrially applicable as
thermoset coatings98, 99 e.g. in flame retardant100, 101, radiation cured24, 102-108, and
high solid systems109-111.

Figure 8. Basic components of a hyperbranched polymer.

14
 Experimental

3 EXPERIMENTAL
This experimental section will only describe the experimental details briefly. More
detailed descriptions are given in the appended papers (Papers I-V).

3.1 MATERIALS

Two rape seed methyl esters (RME), RME 1 and RME 2, were supplied by Norups
Gård AB, Knislinge, Sweden and Statoil Frescati, Stockholm, Sweden respectively,
two Linutin versions, LME 1 and LME 2, were supplied by Lantmännen,
Stockholm, Sweden, and tall oil methyl ester (TOME), TOME, was supplied by
SunPine AB, Piteå, Sweden. Fatty acid composition and peroxide value of the
evaluated FAMEs are presented in Table 1.

Table 1. FAME composition in wt% and peroxide value of RME 1, RME 2, LME 1, LME 2 and TOME.
Carbon chain length: number of unsaturations.

Fatty acid RME 1 RME 2 LME 1 LME 2 TOME

Methyl palmitate C16:0 4.3 6.2 6.1 6.2 4.6
Methyl oleate C18:1 58.3 66.5 13.4 13.6 22.1
Methyl linoleate C18:2 20.9 15.4 67.2 72.0 47.2
Methyl linolenate C18:3 10.5 4.5 7.7 2.1 1.7
Methyl pinolenate C18:3 --- --- --- --- 10.0
Other FAMEs 5.0 5.4 5.6 6.1 14.4
Peroxide value 27 951 247 198 27

Methyl oleate (MeO), methyl linoleate (MeL), 2-octyl-1-dodecanol, and p-toluene

sulfonic acid (PTSA) were purchased from Sigma–Aldrich. 1-octadecanol (mp 60–
61°C) was purchased from Merck. A conventional polyester resin (CP1),
hexamethoxymethyl melamine (HMMM), dodecylbenzene sulfonic acid (DDBSA),
dibutyltindilaurate (DBTDL), methyl ethyl ketone (MEK), butyldiglycol acetate
(BDGA), pigments, and fillers were obtained from Akzo Nobel Nippon Paint AB,

15
Experimental

Gamleby, Sweden. The hydroxy-functional CP1 with partial aromatic structure

was supplied dissolved in a 3:1 (w/w) mixture of light aromatic solvent naphtha
(CAS 64742-95-6) and butyl glycol (CAS 111-76-2) to a dry content of 70 wt%. Two
hyperbranched aliphatic polyesters, Boltorn H2003 (HP2) and Boltorn H2004
(HP3), based on 2,2-bis(methylol)propionic acid (bis-MPA) as AB2-monomer, were
obtained from Perstorp Specialty Chemicals AB, Perstorp, Sweden. HP2 and HP3
are based on a second generation hydroxy-functional hyperbranched polyester
that have been end-group modified to 25% and 62.5% respectively with eight
carbon long alkyl acids. Experimental data for CP1, HP2, and HP3 as given by the
suppliers are presented in Table 2. Three types of substrates were used; non-stick
coated aluminum sheets, neat hot dipped galvanized (HDG) steel sheets, and
organic primer coated HDG steel sheets. All chemicals were used as received.

Table 2. Experimental data for CP1, HP2, and HP3 as given by the suppliers.

CP1 HP2 HP3

Molecular weight, Mw [g/mol] 4000 2500 3200
Dry weight content [wt%] 70 100 100
Acid value [mg KOH/g resin] 8 7 7
Hydroxyl value [mg KOH/g resin] 121 290 115
Number of hydroxyl end-groups/mol 9 12 6

3.2 SAMPLE PREPARATION

Three different test series have been evaluated; three component fully formulated
clear-coats, two component model systems, and a full scale industrial trial.

3.2.1 Fully formulated clear-coat systems

3.2.1.1 Conventional polyester resin

Formulations with a CP1:HMMM hydroxyl:ether functional group ratio of 1.6:1

with different amounts of FAME were mixed according to Table 3. Samples 5-8
and 9-12 were diluted with BDGA and MEK respectively, to a viscosity of 60±5 s
determined by efflux cup viscosity measurements. Samples 22 and 23 were diluted
with light aromatic solvent naphtha to a dry weight of 60 wt%.

16
 Experimental

Table 3. Formulation details for the three component full system study with conventional resin.

Sample FAME FAME content1 Solvent2 Oven Dry content3

name [wt%] [nr] [wt%]
1 --- --- --- 1,2 73
2 RME 1 5 --- 1 75
3 RME 1 10 --- 1,2 76
4 RME 1 20 --- 1 78
5 --- --- A 1 56
6 RME 1 5 A 1 58
7 RME 1 10 A 1 59
8 RME 1 20 A 1 68
9 --- --- B 1 65
10 RME 1 5 B 1 67
11 RME 1 10 B 1 70
12 RME 1 20 B 1 74
13 LME 1 5 --- 1 75
14 LME 1 10 --- 1,2 76
15 LME 1 20 --- 1 78
16 LME 2 5 --- 1 75
17 LME 2 10 --- 1,2 76
18 LME 2 20 --- 1 78
19 TOME 5 --- 1 75
20 TOME 10 --- 1,2 76
21 TOME 20 --- 1 78
22 --- --- C 1,2 60
23 RME 1 10 C 1,2 60

1 Weight percent added FAME calculated on wet polyester resin and HMMM mixtures prior to

addition of additional solvent.

2 All mixtures contain the solvent present in CP1. The solvents are: A BDGA, B MEK, and C light

aromatic solvent naphtha.

3 The polyester, HMMM, and FAME are considered as dry content and the solvent present in the

polyester resin and additional solvent as wet content.

17
Experimental

3.2.1.2 Hyperbranched polyester resin

Formulations with a polyester:HMMM hydroxyl:ether functional group ratio of

1.6:1 with different amounts of RME 1 were prepared according to Table 4.

Table 4. Formulation details for the three component full system study with hyperbranched resin.

Sample Polyester1 RME 1 content2 Oven Dry content3

name [wt%] [nr] [wt%]
24 HP2 --- 1,2 100
25 CP1 HP2 --- 1,2 81
26 HP3 --- 1,2 100
27 CP1 HP3 --- 1,2 85
28 HP2 HP3 --- 1,2 100
29 HP2 10 1,2 100
30 CP1 HP2 10 1,2 83
31 HP3 10 1,2 100
32 CP1 HP3 10 1,2 87
33 HP2 HP3 10 1,2 100

1 The number of hydroxyl groups is equally divided by the polyester in the case of two polyesters.
2 Weight percent added FAME calculated on wet polyester resin and HMMM mixtures.
3 The polyester, HMMM, and FAME are considered as dry content and the solvent present in the

polyester resin and additional solvent as wet content.

3.2.1.3 Curing procedures

Thermal cure were performed both in an automated laboratory oven built to

simulate coil-coating cure from Tryckluftsteknik i Västervik AB, Västervik,
Sweden (oven 1), and in a conventional convection T6 function line oven from
Heraeus instruments, Hanau, Germany (oven 2). Peak metal temperature (PMT)
for films cured in oven 1 was determined with temperature-tape-indicators
mounted on the back of the test panels. 3 wt% DDBSA was added to all
formulations before film application and cure.

A non-stick coated substrate was used to obtain free-standing films with a dry
film thickness of 20±5µm of formulations 1-21, Table 3. Formulations 1-12 were
cured in oven 1 directly after application at 300, 400, and 500°C respectively to a
PMT of 232-241°C. One set of samples 1-4 were also left to flash-off at ambient

18
 Experimental

conditions for 15 minutes before cure at 400°C. Films of samples 1-4 and 13-21
were cured in oven 1 at 255°C to reach a PMT of 232-241°C. Different oven
temperatures and air flows in oven 1 entail different dwell time to obtain equal
PMT, i.e. 45s at 255°C, 48s at 300°C, 24s at 400°C and 15s at 500°C. Free-standing
films of samples 1, 3, 14, 17, and 20 were cured in oven 2 at 130°C for 10 min.

Samples 1, 3, and 24-33 (Table 3 and Table 4), diluted to a dry weight content of
73 wt% with light aromatic solvent naphtha, and samples 22 and 23 (Table 3) were
applied with a dry film thickness of 23±3µm on untreated and organic primer
coated HDG substrates respectively. Films on untreated HDG substrates were
cured in oven 1 at 360°C for 43s and in oven 2 at 130°C for 10 min. Films on
primed HDG substrate were cured in oven 1 at 300°C for 39s. Different
temperature and cure time was employed for cure in oven 1 to obtain a PMT of
232-241°C, due to differences in heat reflectivity of the two substrates.

3.2.2 Model study

Composition of the studied two component formulations are presented in Table 5.

The formulations were prepared with a FAME/alcohol molar ratio of 1:1.

3.2.2.1 Model reactions in thin films

Before reaction in oven, 3 wt% DDBSA was added to the model study mixtures.
The samples were reacted in both oven 1 and oven 2 (see section 3.2.1.3 Curing
procedures) respectively whereas the sample names will be followed by 1 for
reaction in oven 1, and by 2 for reaction in oven 2. For example sample B2 is
reacted in oven 2 and consists of a RME 1/1-octadecanol mixture with 3 wt%
DDBSA. Untreated HDG steel sheets were used as substrates. The substrates were
pre-heated to 68°C for samples A–K containing 1-octadecanol (mp 60–61°C) in
order to avoid crystallization. Films were applied with a wet film thickness of
60 µm. Samples C, and G-Q were reacted in oven 1 at 330°C for 55 and 50 s,
respectively, for the ambient and pre-heated substrate to a PMT of approximately
230°C. The samples were then left to cool down at ambient conditions. 1H-NMR
samples were taken from the reacted films. Samples B, D-F were reacted in oven 2
for up to 60 min at 130°C and samples C, G-Q were reacted in oven 2 for up to
60 min at 110, 130, 150 and 170°C respectively. Approximately ten 1H-NMR
samples were taken at regular interval during the reaction period.

19
Experimental

Table 5. Composition of the two component model system samples.

Sample name FAME Alcohol Catalyst

A RME 2 1-octadecanol PTSA
B RME 1 1-octadecanol DDBSA
C RME 2 1-octadecanol DDBSA
D LME 1 1-octadecanol DDBSA
E LME 2 1-octadecanol DDBSA
F TOME 1-octadecanol DDBSA
G MeO 1-octadecanol DDBSA
H MeL 1-octadecanol DDBSA
I RME 2 1-octadecanol DBTDL
J MeO 1-octadecanol DBTDL
K MeL 1-octadecanol DBTDL
L RME 2 2-octyl-1-dodecanol DDBSA
M MeO 2-octyl-1-dodecanol DDBSA
N MeL 2-octyl-1-dodecanol DDBSA
O RME 2 2-octyl-1-dodecanol DBTDL
P MeO 2-octyl-1-dodecanol DBTDL
Q MeL 2-octyl-1-dodecanol DBTDL

3.2.2.2 Model reactions in bulk

Reaction in bulk was conducted in inert atmosphere. Mixtures of 0.5 g of

formulations C, and G–Q were prepared in a round bottom flask placed under
argon (Ar) flow. Samples C, and G-K, containing 1-octadecanol (mp 60–61°C),
were left at 68°C for 30 min and samples L-Q, containing 2-octyl-1-dodecanol,
were left at room temperature for 30 min. The temperature was then raised to
130°C and 3 wt% DDBSA was added. The system was left to react for 15 min
under Ar flow before being analyzed with 1H-NMR.

3.2.3 Full scale industrial evaluation

A pigmented fully formulated coil-coating formulation with 10 wt% RME (PCF)
has been evaluated in an industrial full scale trial. The PCF, containing CP1,
HMMM, pigments, fillers, and solvents, was prepared in the same manner as
conventional coil-coating paint by Akzo Nobel Nippon Paint AB, Gamleby,

20
 Experimental

Sweden. Conventional organic primer coated HDG steel sheets were industrially
coated with PCF and cured in the pre-coated sheet metal production facility at
SSAB Tunnplåt AB, Borlänge, Sweden.

3.3 TECHNIQUES

3.3.1 Thermal gravimetric analysis

Volatility properties of FAMEs were analyzed with thermal gravimetric analysis

(TGA) performed on a Mettler Toledo SDTA/TGA851 equipped with a water
cooler. Samples were heated from 30°C to 300°C with a heating rate of 20°C/min in
nitrogen atmosphere.

3.3.2 Rheological measurements

Rheological measurements were performed with a Gardco/DIN 4mm viscosity

cup and on a TA-instruments ARES Rheometer. The Gardco/DIN 4mm viscosity
cup was conformed according to the flow characteristics specified by Deutsche
Normen DIN 53211. The viscosity value is given by the time for a Gardco/DIN
4mm viscosity cup brimful with formulation to efflux at 23°C, until the first break
of the stream, in seconds. Rheological measurements on the ARES Rheometer
equipped with 25 mm parallel plates were performed in the linear region at 30°C
with a frequency ramp from 0.1 to 10Hz.

3.3.3 Spectroscopical analysis

IR and real time infrared (RTIR) spectra were obtained on a Perkin Elmer
Spectrum 2000 instrument equipped either with a single reflection ATR accessory
(Golden Gate) or a heat controlled ATR from Specac Ltd. The IR measurements
were performed in reflection mode.

For the RTIR measurements, the heat controller was set to desired temperature
(110, 130, 150, or 170°C) and left to equilibrate for 10 min. Two drops (~75µl) of the
sample was then placed on the ATR crystal and the measurement was
immediately started. Spectra were recorded every 35 s over a period of 60 min
using TimeBase® software from Perkin Elmer. The reaction was followed by
monitoring the shift of the carbonyl peak from 1744 to 1740 cm-1.

Raman analysis was performed on a Perkin Elmer Spectrum 2000 FT-Raman.

21
Experimental

3.3.4 1 H-NMR

H-NMR were performed on both a 400MHz Bruker Aspect NMR and a Bruker
1

Avance 400MHz using CDCl3 as solvent and internal standard.

3.3.5 Carbon-14 dating

Carbon-14 dating on free-standing films was performed by the Ångström

Laboratory, Uppsala, Sweden.

3.3.6 Extraction

Free standing films were extracted with hexane for three days. Gas
chromatography (GC) analysis on the rotaevaporated extraction phase was
performed by the SW laboratory, Svalöf, Sweden.

3.3.7 Dynamic mechanical analysis

Dynamic mechanical analysis (DMA) was performed on TA-instrument Q800

equipped with a film fixture for tensile testing. DMA measurements were
performed on rectangular free standing film samples (20x6 mm) with a heating
rate of 2°C/min. The tests were performed in controlled strain mode with a
frequency of 1 Hz.

3.3.8 Differential scanning calorimetry

Differential scanning calorimetry (DSC) analysis were performed using a Mettler
Toledo DSC820 equipped with a Mettler Toledo Sample Robot TSO801RO
calibrated using standard procedures. The samples were heated from -40°C to
180°C, cooled down to -40°C and then reheated to 180°C at a rate of 10°C/min.
Midpoint Tg was determined during the second heating.

3.3.9 Micro hardness

Micro hardness was determined with a Fischerscope H100 equipped with a

Vickers indenter according to ISO 6441. The measurements were performed at a
10% indentation depth with 60 s recover periods at constant load and after
removal of the load.

22
 Experimental

3.3.10 Conventional film characterization

Adhesion, gloss, impact resistance, Erichsen flexibility, T-bend, chemical

resistance, corrosion resistance, pencil hardness, and scratch resistance of the
coating were tested according to ISO, CEN, or ASTM standards.

23
Results and Discussion

4 RESULTS AND DISCUSSION

The aim of the present study was to decrease the use of volatile solvents in
solvent-borne thermally cured coil-coating resins. This was done by modification
of the resin and introduction of reactive diluents. The use of reactive diluent in
coating applications will affect the performance for the entire production chain
from paint manufacture to the final film properties. The effect upon introduction
of FAMEs as reactive diluents and modification of the polyester resin on the wet
paint characteristics, film formation, cross-linking chemistry, and final film
properties in a thermally cured solvent-borne polyester/melamine coil-coating
system will here be addressed.

4.1 MISCIBILITY

Miscibility of the whole paint system, e.g. resin, catalyst, and diluents, is
important to obtain leveled and defect free coatings. Miscibility generally depends
on compatibility, polarity, molecular weight and intermolecular interactions
between the components.3 The used reactive diluents, i.e. FAMEs, are in average
eighteen carbon long molecules, un-polar, and miscible with both un-polar and
amphifilic structures. The initially used catalyst, PTSA, had poor miscibility with
FAME and 1-octadecanol (sample A) and phase separated. Therefore, PTSA was
exchanged for DDBSA and DBTDL that are miscible with fatty compounds due to
their more un-polar structures. The conventional polyester, CP1, was designed to
be compatible with both HMMM and FAME.

The evaluated hydroxy-functional aliphatic hyperbranched polyesters based on

bis-MPA, HP2 and HP3, were modified to different extent with alkyl acid by the
supplier to increase compatibility with FAME. Modification of the polar
hyperbranched polyesters decreases polarity and reduces intra- and
intermolecular secondary bonds associated with a large quantity of hydroxyl end-
groups. Compatibility with FAME was improved with extent of modification of
the hyperbranched polyester, i.e. HP3 is more compatible with FAME than HP2
although HP2 and FAME, samples 29 and 30, formed transparent formulations

24
 Results and Discussion

and cured coatings. However, HP2 has poor compatibility with CP1 and HP3 as
indicated by opaque wet mixtures of formulations 25, 28, 30, and 33 although clear
single phase films were attained after cure.

4.2 RHEOLOGY

In order to be able to apply coil-coating paint industrially with large rollers the
wet paint must have suitable rheological properties. Viscosity of paints can be
adjusted by type and amount of conventional solvent. One of the properties of
reactive diluents is that they work as diluents lowering the viscosity of mixtures.
Therefore, solvents can be exchanged for reactive diluents in paint formulations.
FAMEs are suitable as reactive diluents as they have low viscosity and generally
good solubility properties.93 Viscosity measurements of samples 1-4 showed that
FAME has diluting properties in a fully formulated conventional
polyester/melamine clear-coat system, Figure 9. However, FAMEs were not as
powerful solvents as conventional solvents, established with comparison of
viscosity measurements of samples 1-12. Of the two conventional solvents
evaluated, MEK is a more powerful solvent than BDGA. The results show that
introduction of 10 wt% reactive diluent reduces the use of conventional solvent by
35 wt% in order to retain equal viscosity.

3,5

3,0
Viscosity [Pa s]

2,5

2,0

1,5

1,0

0 5 10 15 20
Reactive diluent [wt%]

Figure 9. Viscosity versus reactive diluent content.

25
Results and Discussion

The conventional solvent content can be further decreased by modifying the

polyester resin. Hyperbranched polymers have unique rheological properties and
exhibit lower viscosity than conventional branched polymers at equal molecular
weight.95, 96 Viscosity versus dry weight content for CP1, HP2, and HP3 resin
formulations with and without 10 wt% RME 1 is presented in Figure 10. The
conventional solvent content can be reduced by 50 wt% by replacing the
conventional polyester CP1 with hyperbranched HP3 in order to obtain a suitable
application viscosity of 200 mPa·s. The reduction in viscosity is minor with HP2
due to the high degree of hydroxyl end-groups that present strong intermolecular
secondary hydrogen bonds. These hydrogen bonds and the amphifilic character of
the resin further entail minor reduction in viscosity upon addition of reactive
diluents (sample 29), although FAMEs are hydrogen bond acceptors. The viscosity
was though decreased upon introduction of reactive diluent in the HP3 resin,
sample 31. It is suggested that a paint formulation without any conventional
organic solvent can be obtained by further modification of the polyester resin and
use of greater amounts of reactive diluents.

100 1 3
24 29
26 31

10
Viscosity [Pa s]

0,1

50 60 70 80 90 100
Dry weight [wt%]

Figure 10. Viscosity versus dry weight content for CP1 (samples 1 and 3), HP2 (samples 24 and 29), and
HP3 (samples 26 and 31) resin formulations with and without 10 wt% RME 1.

26
 Results and Discussion

4.3 FILM FORMATION

Curing is affected both by evaporation of solvent and the chemical cross-linking

process. Solvents and other volatiles, including added reactive diluents, must
evaporate at a controlled rate to avoid film defects like frothing, sagging and
leveling defects.2, 3 No direct studies of the film formation could be performed as
the cure rate in coil-coatings is very high. Hence, only dry films were studied.

FAME enhanced the film formation properties of the system as more leveled films
with less frothing were obtained with higher FAME content. It is suggested that
the reactive diluent acts as a plasticizer and a so called tail solvent2, 3, increasing
the mobility of the system during the final stages of the film formation, providing
good flow and leveling of the film. This increased mobility of the system also
permits a more controlled evaporation of the conventional solvents. The low
polarity of FAME furthermore suggests a reduction in surface tension. The
wetting of the non-stick coated aluminum sheet was somewhat poor, resulting in
some contraction around the edges of the films left to flash-off for 15 min.
Otherwise, the film formation was not noticeably effected by the flash-off period.

4.4 INCORPORATION OF FAME

A reactive diluent should not only affect the flow properties of a paint but also be
incorporated in the coating network during cure. The methyl ester group of FAME
is suggested to react with the hydroxyl groups of the polyester, in the
polyester/melamine system, chemically bonding FAME into the final film.
However, FAME can also be considered as volatile as they generally starts to
evaporate at 200°C and coil-coatings are cured to a PMT of 200-250°C. The volatile
property of FAME is not only negative as it allows evaporation of non-reacted
FAME that otherwise would remain un-reacted in the dry film, i.e. FAME either
reacts into the film or evaporates.

4.4.1 Fully formulated clear-coats

Fully formulated clear-coat films are chemically cross-linked and can therefore not
be dissolved, limiting possible analyze methods. Raman analysis, carbon-14
dating, and extraction were performed to confirm incorporation of FAME in cured
films.

27
Results and Discussion

Double bonds present in unsaturated fatty acids show a strong signal at

approximately 1650 cm-1 in Raman spectra. This signal is the only one in FAME
that is not overlapped by peaks from either the polyester or HMMM. However,
the peak at 1650 cm-1 could only be detected for the free-standing films cured at
lower temperatures, i.e. in oven 2. This indicates that all FAMEs evaporates or the
unsaturations reacts, e.g. oxidizes or thermally polymerizes, under simulated
industrial cure conditions, i.e. in oven 1.

Carbon-14 dating can be used to determine presence of FAMEs as they are derived
from crops grown during the last few years while the other components in the
coating (polyester, HMMM, and solvents) are derived from crude oil, an old
carbon resource, according to the suppliers. Analysis of samples 1-4 showed
remaining RME 1 in the films as the estimated age decreased with increased
reactive diluent content, Table 6. A clear trend is seen with a decreased average
age with increased FAME content in the formulation. However, the method does
not confirm incorporation of FAME via transesterification but just the presence.

Table 6. Carbon-14 dating data.

Sample name 14C age BP1

1 >40 000
2 37 555 ± 1 890
3 34 560 ± 1 300
4 25 850 ± 440

1 BP=before present, defined as AD 1950.

Extraction tests showed that no significant amount of FAME could be extracted

from the films.

The Raman, carbon-14 dating, and extraction test results strongly suggest that
FAMEs react into the film via the acyl group and that non-reacted FAMEs
evaporate. The results furthermore indicate that the unsaturations react during
cure although it is not clear if this leads to further cross-linking or chain scission. It
is well known that unsaturated fatty acids can contribute to both reaction routes3,
30, 50-53 and further tests is needed to clarify this. The model studies presented below

in section 4.4.2 Model study and results in section 4.5 Film properties furthermore
support incorporation of FAMEs into the coating.

28
 Results and Discussion

4.4.2 Model study

The reactivity between the methyl ester of FAME and hydroxyl groups in the
polyester was simulated with model studies between RME 1, RME 2, LME 1,
LME 2, TOME, MeO, MeL, and primary alcohols. Kinetics of the
transesterification reaction was followed with RTIR and 1H-NMR.

RTIR measurements performed at low temperatures (130, 150, and 170°C) show
that the carbonyl peak (1744 cm-1) shifts to lower wavelengths (1740 cm-1) as the
methyl esters transesterifies with alcohol, Figure 11. Time to complete shift of the
carbonyl peak is decreased with increased temperature, i.e. the reaction proceeds
faster at higher temperature, Figure 12. The reaction time to full conversion
(~5 min at 170°C), for these low temperatures, is very long compared to the
industrial process (~37 s). However, extrapolating of the curves to higher
temperatures indicate that full conversion can be obtained within desired time
frame for coil-coatings.

0.11
0.10
0.09
0.08
0.07
0.06
Abs

Increased time

0.05
0.04
0.03
0.02
0.01
0.00
1760 1750 1740 1730 1720
Wavenumber [cm -1]

Figure 11. RTIR shift of the carbonyl peak at approximately 1740 cm-1 during 60 min for mixture H at
130°C.

29
Results and Discussion

1,0

Normalized absorbans 0,9

0,8 130°C
150°C
170°C
0,7

0,6

0 5 10 15 20 25
Time [min]

Figure 12. Normalized absorbans for fully shifted carbonyl peak versus time for mixture G at 130°C,
150°C, and 170°C measured with RTIR.

1H-NMR also confirms transesterification as a new peak appears at 4 ppm

(-CH2CH2OCO-) from the new ester and the peaks from the reactants at 3.6 ppm
(FAME CH3OCO-), 3.6 ppm (octadecanol –CH2CH2OH), and 3.5 ppm (2-octyl-1-
dodecanol –CHCH2OH) decreases, Figure 13. 1H-NMR spectra also show that the
alkenes in FAME thermal and/or oxidative reacts as the integrals from the double
bond peaks at 5.3 ppm (-CH=CH-) and 2.7 ppm (-CH=CHCH2CH=CH-) decrease
when the reaction was performed in air. It is difficult to exactly estimate the
amount of alkene reactions as oxidation processes are very complex and
numerous reactions occur simultaneously.65, 70 Decrease of the signal at 5.3 ppm
relative to the protons on the carbon beside the acyl group at 2.3 ppm
(-OCOCH2CH2-) was used as a relative measure of the alkene reactions, although
this only partly describes the overall thermal and oxidative alkene processes.

30
 Results and Discussion

O
O
1.
2.
OH

3. O
O

1.
3.
2.

5.0 4.0 3.0 2.0 1.0

ppm

Figure 13. 1H-NMR spectra for sample G2 reacted at 130°C for 30 min.

4.4.2.1 Effect of FAME structure

Degree of oxidation is affected by fatty acid pattern, presence of oxygen,

temperature, and time. Fatty acid pattern vary depending on crop source and
cultivation origin. A high content of unsaturations render high degree of auto-
oxidation with subsequent polymerization or chain scission. These oxidation
processes are competing factors to the transesterification reaction with alcohol as
lower conversions were obtained for FAMEs with high number of double bonds,
Figure 14. The cause of the competition could be that an oxidative cleavage of a
FAME molecule would render volatile fragments with subsequent change in
stoichiometry of the system. Nevertheless, TOME showed a high
transesterification conversion despite it contains 11 wt% methyl pinolenate with
three double bonds. Methyl pinolenate (methyl cis,cis,cis-5,9,12-
octadecatrienoate), is however less prone to oxidize than methyl linolenate
(methyl cis,cis,cis-9,12,15-octadecatrienoate), due to the placement of the double
bonds in the carbon chain.112 Further, FAMEs must contain at least one
unsaturation in order to be usable as reactive diluents as they otherwise will
crystallize at ambient conditions.

31
Results and Discussion

100
B2 C2
D2 E2
F2
80
Conversion [%]

60

40

20

0
0 10 20 30 40 50 60
Time [min]

Figure 14. Conversion versus reaction time for samples B2, C2, D2, E2, and F2 reacted at 130°C.

4.4.2.2 Effect of catalyst

Catalyst efficiency alters with changed cure condition. DBTDL is a more efficient
catalyst than DDBSA under industrial cure conditions, Table 7. However, opposite
efficiency, with less differences between the catalysts, was attained for cure at
lower temperatures (110, 130, 150, and 170°C), Figure 15. These discrepancies, due
to cure conditions, may be due to different temperature dependence113 for the two
catalysts or the fact that the cure time was much shorter with industrial cure
conditions somewhat preventing competing alkene reactions.

Table 7. Conversion for samples C1, G1-K1.

Sample Name Conversion [%]

C1 42
G1 60
H1 54
I1 94
J1 87
K1 97

32
 Results and Discussion

100 C2 I2
G2 J2
H2 K2
80
Conversion [%]

60

40

20

0 10 20 30 40 50 60
Time [min]

Figure 15. Conversion versus time for sample C2, G2-K2 reacted at 130°C.

4.4.2.3 Effect of alcohol structure

A minor model study, regarding choice of polyester monomer, shows that care
must be taken to not introduce easily oxidized monomers when using methyl
linoleate esters as reactive diluents. Two primary alcohols, 2-octyl-1-dodecanol
with tertiary hydrogen and 1-octadecanol without tertiary hydrogen, Figure 16,
were evaluated. The degree of alkene reactions, and subsequent low
transesterification conversion, was significantly greater for 2-octyl-1-dodecanol
compared to 1-octadecanol when reacted with MeL, Figure 17. This difference
indicates that not only MeL but also the alcohol is susceptible to oxidation due to
presence of tertiary hydrogen which can be abstracted in an auto-oxidation
process. No extra care in choice of polyester monomer has to be taken when using
FAMEs with low methyl linolenate content as no greater oxidation was noticed
with RME 2 and MeO as reactive diluents.

HO

OH

1-Oktadekanol 2-Octyl-1-dodecanol

Figure 16. 1-Octadecanol and 2-octyl-1-dodecanol.

33
Results and Discussion

C2 L2
100
G2 M2
H2 N2
80

Conversion [%]
60

40

20

0
0 10 20 30 40 50 60
Time [min]

Figure 17. Conversion versus reaction time for samples C2, G2, H2, L2, M2, and N2 reacted at 130°C.

4.5 FILM PROPERTIES

Final pre-coated metal sheets are cut, bent, punched, and drawn into desired
shape, putting high demands on the film properties. These properties should not
be negatively affected by modification of the binder resin or introduction of a
reactive diluent. Free standing films of fully formulated clear-coats have been
analyzed with DMA to reveal how the incorporation of FAME affects the
mechanical properties. Further, coated metal test panels have been characterized
with conventional film characterization test methods.

4.5.1 Free-standing clear-coats

Introduction of FAME decreases the Tg as its long un-polar chain (average 18

carbons) plasticizes, changes the polarity and alters the cross-link density of
coatings, Figure 18. However, a slight increase in Tg was noticed for sample 2
containing 5 wt% RME 1. This increase may be due to changes in cross-linking
chemistry and polarity, antiplasticization, thermal and/or oxidative cross-linking.
Antiplasticization13, 114, 115, stiffening of the material instead of plasticization, is
often observed when small amounts of plasticizer are added to polymers. The
consumption of unsaturations, as seen from the Raman spectra, indicates that
oxidative and thermal cross-linking occurs to some extent. Furthermore, the
network is more heterogeneous upon introduction of FAMEs, as indicated by the
decreased magnitude and broadening, i.e. width of the tan δ curve, of the glass
transition.116

34
 Results and Discussion

A.
10000

Storage Modulus (MPa)

1000

100

10

1
30 40 50 60 70 80 90 100
Temperature (°C)
B.
1000

100
Loss Modulus (MPa)

10

0.1
30 40 50 60 70 80 90 100
Temperature (°C)
C.
1.2

1.0

0.8
Tan Delta

0.6

0.4

0.2

0.0
30 40 50 60 70 80 90 100
Temperature (°C)

Figure 18. A. Storage modulus, B. Loss modulus, C. Tan δ versus temperature for samples 1 ( ), 2
( ), 3 ( ), and 4 ( ) cured in oven 1 at 400°C.

35
Results and Discussion

4.5.1.1 Effect of FAME structure

The model study regarding transesterification conversion (section 4.4.2 Model

study) showed that the FAME structure affects the degree of incorporation
indicating that the mechanical properties also are affected by the FAME structure.
This was confirmed with DMA measurements on coatings containing FAMEs with
different fatty acid pattern, Figure 19. The most significant affect on the
mechanical properties, implying the highest amount of incorporation, was
obtained upon introduction of LME 1, i.e. samples 13-15 showed greatest decrease
in Tg. The lowest shift in Tg was attained with TOME as reactive diluent.
However, these results are not consistent with the model study results where the
highest degree of transesterification conversion was obtained with TOME and the
lowest with LME 1 as FAME. These contradicting results can be connected to the
FAME structure. Incorporation of FAME via transesterification would render a
dangling chain end that will plasticize the coating, i.e. decrease the Tg. Alkene
cross-linking on the other hand would restrict the mobility of the material, i.e.
increase the Tg. The lower shift in Tg than expected for TOME is presumably due
to its content of methyl pinolenate with an isolated double bond that can act as an
additional cross-linking site. The present study cannot discriminate between the
possible reaction routes. However, it clearly shows the complexity of the curing
process and that the fatty acid pattern of the reactive diluents has to be taken into
consideration.

1.5

1.0
Tan Delta

0.5

0.0
30 40 50 60 70 80 90 100
Temperature (°C)

Figure 19. Tan δ versus temperature for samples 1 ( ), 3 ( ), 14 ( ), 17 ( ), and 20

( ) cured in oven 1 at 255°C.

36
 Results and Discussion

4.5.1.2 Effect of oven temperature

An increased oven temperature with subsequent line speed increase in the coil-
coating production can render cost reductions. However, larger costs are
associated with higher energy consumption upon raised oven temperatures. The
mechanical properties of fully formulated clear-coat systems without any reactive
diluent are not affected by change of oven temperature during cure in oven 1,
Figure 20, diagram A. There are some minor variations in Tg with changed oven
temperature for the films with high amounts (20 wt%) of reactive diluent, Figure
20, diagram B. However, these variations are not logical with highest Tg at 300°C
and lowest Tg at 400°C. This inconsistence may be due to different time-
temperature dependences for degree of evaporation and alkene polymerization,
which would increase the Tg, and incorporation of FAME, which would decrease
the Tg. Further reduction of the oven temperature to 130°C (cure in oven 2)
renders a greater drop in Tg, Figure 21. This major decrease confirms the time-
temperature dependence rendering less alkene reactions and lower degree of
evaporation at an inferior cure temperature, i.e. more FAMEs are left to soften the
coating.

4.5.1.3 Effect of co-solvent

Co-solvents are often added to paint systems in order to attain a suitable

application viscosity. Addition of co-solvents with different vapor pressure and
polarity affects the film formation process that e.g. is shown by shift in Tg, Figure
22, diagram A. The greatest shift in Tg is obtained with BDGA as co-solvent and
the least Tg shift for the sample without any additional solvent for the evaluated
films without any reactive diluent. The homogeneity of the initial system is not
affected by choice of solvent.

37
Results and Discussion

A.
1.2

1.0

Tan Delta 0.8

0.6

0.4

0.2

0.0
30 40 50 60 70 80 90 100
Temperature (°C)
B.
1.2

1.0

0.8
Tan Delta

0.6

0.4

0.2

0.0
30 40 50 60 70 80 90 100
Temperature (°C)
Figure 20. Tan δ versus temperature for films cured in oven 1 at 300°C ( ), 400°C ( ), and
500°C ( ) respectively. A. Sample 1, B. Sample 4.

1.5

1.0
Tan Delta

0.5

0.0
0 10 20 30 40 50 60 70 80 90 100
Temperature (°C)
Figure 21. Tan δ versus temperature for samples 1 and 17 cured in oven 1 (1 ( ) and 17 ( ))
and in oven 2 (1 ( ) and 17 ( )).

38
 Results and Discussion

A.
1.2

1.0

0.8
Tan Delta

0.6

0.4

0.2

0.0
30 40 50 60 70 80 90 100
Temperature (°C)
B.
1.2

1.0

0.8
Tan Delta

0.6

0.4

0.2

0.0
30 40 50 60 70 80 90 100
Temperature (°C)

Figure 22. A. Tan δ versus temperature for samples 1 ( ), 5 ( ), and 9 ( ) cured in oven 1
at 400°C. B. Tan δ versus temperature for samples 4 ( ), 8 ( ), and 12 ( ) cured in oven 1
at 400°C.

Co-solvent will not only affect the film formation process but can also cause co-
evaporation of FAME. Various variations depending on solvent were observed
upon addition of FAME, Figure 22, diagram B. The glass transitions for samples
with reactive diluents were shifted compared to the initial transitions. The shift
was greatest and the homogeneity significantly decreased for the sample without
any additional solvent, indicating that a higher degree of plasticizing FAME is
present in the film. This confirms co-evaporation of FAME in the diluted samples.

39
Results and Discussion

4.5.1.4 Effect of flash-off

As choice of co-solvent has shown to affect the mechanical properties of the cured
coatings it is highly likely that a flash-off time also will do so. DMA measurements
show that the film forming process is merely slightly affected by a flash-off period
of 15 minutes, cf. Figure 18 diagram C and Figure 23. Comparison shows that the
films left to flash-off have a slightly decreased Tg and are more homogenous.

1.2

1.0

0.8
Tan Delta

0.6

0.4

0.2

0.0
30 40 50 60 70 80 90 100
Temperature (°C)

Figure 23. Tan δ versus temperature for samples 1 ( ), 2 ( ), 3 ( ), and 4 ( ) with a

flash-off period of 15 minutes prior cure in oven 1 at 400°C.

40
 Results and Discussion

4.5.2 Substrate supported clear-coats

The rough forming process associated with pre-coated sheet metal products
requires hard coatings with retained good adhesion, flexibility, and impact
resistance to prevent inter alia flaking and film ruptures. Adequate properties are
e.g.; impact value ≥8 J, pencil hardness ≥HB, scratch resistance ≥35 N, chemical
resistance ≥100 MEK double rub.117 Coil-coatings used in outdoor applications
must further have excellent corrosion protection and exterior durability to redeem
warranty times of up to 20 years.117 Usually pre-coated sheet metal substrates are
pretreated and primed to improve film properties such as corrosion protection
and adhesion. Polyester/melamine clear-coats with and without reactive diluents
on neat and organic primer coated HDG steel substrates respectively were
characterized with conventional film characterization techniques and DSC. All
evaluated formulations, samples 1, 3, and 24-33, were prepared with a dry weight
content of 73 wt%. However, the viscosity of sample 1 was too high to obtain
leveled and defect free films when cured under industrial cure conditions on neat
HDG steel substrate. Therefore, samples 22 and 23 with a dry weight content of
60 wt% were also prepared. Test results of adhesion, pencil hardness, flexibility,
micro hardness, chemical and scratch resistance for films cured on neat HDG steel
substrates under industrial cure conditions are presented in Table 8. The
conventional polyester resin, sample 22, has good adhesion, chemical and scratch
resistance but is harder, less flexible and scratch resistant than desired to
withstand the forming process of pre-coated sheet metal. Introduction of reactive
diluents, samples 3 and 23, acting as plasticizers, improves pencil hardness and
flexibility at low deformation rates, i.e. Erichsen flexibility, whereas impact
resistance is substantially impaired.

4.5.2.1 Effect of resin structure

Film properties are affected by e.g. resin structure and polarity, heterogeneity, and
film thickness.3 The evaluated conventional polyester, CP1, contains aromatic
structures whereas the hyperbranched polymers, HP2 and HP3, have aliphatic
structures. Comparison of the film properties of the different formulations in
Table 8 shows that the properties were impaired for the aliphatic hyperbranched
resins. HP2 containing sample 24 has poor impact resistance and Erichsen
flexibility. This poor formability is due to the aliphatic structure and high
hydroxy-functionality, average of 12 hydroxyls per molecule, causing a brittle and
densely cross-linked network. On the contrary, sample 26, containing HP3 with an
average of 6 hydroxyl groups per molecule, renders loosely cross-linked and tacky

41
Results and Discussion

films with poor scratch and chemical resistance. The hydroxy-functionality of

sample 28, with HP2 and HP3 polyester resin, is equal to sample 22, with only
CP1 as polyester resin, and should form equally cross-linked networks. However,
the different network architecture with dangling chain-ends and aliphatic
structure of the hyperbranched polyesters render softer films with poor scratch
resistance. It is furthermore proposed that the dendritic back-bone structure
affects the overall network properties although it is not clear at present how the
network looks like. The dangling chain-ends in the hyperbranched polyester
resins further decreases the plasticizing effect upon introduction of reactive
diluent, samples 29-33, i.e. the film properties is not significantly affected upon
addition of FAME.

Table 8. Film properties for samples 3, and 22-33 on neat HDG substrate cured in oven 1.

Sample Adhesion Impact Erichsen Chemical Pencil Scratch Micro

name resist. flexibility resist. hardness resist. hardness
[J] [mm] [N] [N/mm2]
24 1 0 <1 >200 F 6 90,4
25 0 10 4 >200 HB 6 72,4
26 0 >18 5 18 <2B 1 0,8
27 0 16 5 >200 B 4 1,3
28 0 4 3-4 >200 <2B 0 5,2
22 0 8 3-4 >200 HB 8 113,6
29 0 0 <1 >200 HB 7 94,0
30 0 6 5-6 >200 F 10 89,6
31 0 16 5 76 <2B 1 1,0
32 0 >18 6 >200 <2B 4 1,2
33 0 6 4 >200 <2B 1 5,4
3 0 2 7 >200 H 12 126,8
23 0 0 7 >200 2H 7 111,9

4.5.2.2 Effect of substrate

Film properties for the films cured on organic primer coated HDG steel substrates
are presented in Table 9. Comparison of Table 8 and Table 9 shows that impact
resistance and Erichsen flexibility are inferior for films cured on primed than on
neat HDG steel substrate. The scratch resistance is improved for primed substrate
supported films as both the clear-coat and the primer are evaluated in the scratch

42
 Results and Discussion

resistance test. It should be noted that the formability were limited to the organic
primer performance, i.e. film rupture occurred at the steel/primer interface. It is
proposed that substitution to a primer system with improved formability
performance would render good formability properties of the evaluated top-coat
on organic primer coated steel substrates.

Table 9. Film properties for samples 1, 3, and 24-33 on primed HDG substrate cured in oven 1.

Sample Adhesion Impact Erichsen Chemical Pencil Scratch Hardness

name resist. flexibility resist. hardness resist. [N/mm2]
[J] [mm] [N]
24 1-2 0 <1 >200 2H 20 84,3
25 0 0 2 >200 H-2H 18 108,3
26 0 2 1 16 2B 16 1,5
27 0 4 2 49 HB 17 2,7
28 0 0 1 89 HB-F 17 8,7
1 0 0 3 >200 3H 22 131,9
29 0 0 <1 >200 2-3H 19 87,6
30 0 0 2 >200 2H-3H 21 107,5
31 0 0 2 19 2B 20 2,3
32 0 2 4 26 F 17 3,1
33 0 0 1 69 B-HB 19 8,4
3 0 0 2-3 >200 2H 17 137,9

4.5.2.3 Effect of oven temperature

As established in section 4.5.1.2 Effect of oven temperature, the mechanical

properties of free standing clear-coats is affected by oven temperature upon
introduction of reactive diluent. The film properties, established with conventional
film characterization techniques, are also affected by oven temperature, cp. Table 8
and Table 10. Decreased oven temperature entails uncured coatings with poor
chemical resistance. Number of hydroxyl groups is only enough to obtain
sufficient cross-linked networks to be chemical resistant for the HP2 containing
formulations 24, 25, and 29. Further, less FAMEs evaporate rendering softer films
upon introduction of reactive diluent. Moreover, DSC measurements indicate a
greater presence of either incorporated or non-reacted FAME in the films cured at
a lower temperature in oven 2, Table 11.

43
Results and Discussion

Table 10. Film properties for samples 1, 3, and 24-33 on neat HDG substrate cured in oven 2.

Sample Adhesion Erichsen Chemical Pencil Scratch Hardness

name flexibility [mm] resistance hardness resist. [N] [N/mm2]
24 5 <1 >200 HB 8 95,2
25 3 5 >200 H-2H 16 115,4
26 3 5 15 <2B 0 1,7
27 1 8 23 HB 9 4,0
28 3 3 74 2B 6 15,1
1 1 9 88 2H-3H 13 130,7
29 4 <1 >200 2B 5 51,5
30 1 6 60 B 13 10,2
31 3 5 21 <2B 1 1,1
32 2 7 16 <2B 3 1,3
33 1 5 28 <2B 2 4,9
3 0 9 41 HB 14 4,7

Table 11. Tg determined with DSC for samples 1, 3, 22, and 23.

Sample name Oven Tg [°C]

1 2 21
3 2 -1
22 1 25
23 1 24

44
 Results and Discussion

4.6 FULL SCALE INDUSTRIAL EVALUATION

The best way to evaluate a new coil-coating system is a full scale test trial with a
pigmented fully formulated resin. Such a formulation with a 10 wt% RME reactive
diluent content, sample PCF, was prepared, industrially cured, and analyzed
regarding adhesion, flexibility (T-bend, Erichsen flexibility, and impact), pencil
hardness, accelerated corrosion, scratch and chemical resistance. The test results
together with the properties of a polyester/melamine reference are presented in
Table 12. The film properties of PCF are equivalent to the reference, except for the
T-bend value. However, a flexibility value of 1 T is still in the range of what is
accepted for conventional polyester/melamine coil-coatings. An accelerated
corrosion test, i.e. a Cleveland test, has been performed by exposing test panels to
continuous condensed water at a 100% relative humidity. After being exposed for
1000 h gloss retention and adhesion were determined on both a flat and pre-
strained area of the test panel. The Cleveland corrosion properties of PCF also
fulfill the demands for coil-coatings.

Table 12. Film properties for a pigmented fully formulated industrially cured coil-coating resin with
10 wt% reactive diluent (PCF) and a conventional polyester/melamine reference.

Sample name PCF Reference

T-bend [T] 1 0
Erichsen flexibility [mm] 9 9
Impact [J] >16 >16
Pencil hardness F HB
Scratch resistance [N] 45 27
Chemical resistance >100 >100
Adhesion 0 0
Gloss 39 44
Cleveland 1000 h
Adhesion 0 0
Adhesion (Impact 11J) 0-0 0-0
Gloss 40 44

45
Conclusions

5 CONCLUSIONS
The volatile organic solvent content in a solvent-borne thermally cured
polyester/melamine coil-coating paint system has been decreased by introduction
of reactive diluents and modification of the polyester binder resin. The evaluated
FAMEs work as diluents, and are incorporated into the film during cure without
negatively affecting the mechanical properties of the final coat.

The reactive diluents reduce the viscosity of a fully formulated

polyester/melamine clear-coat system; it is thereby possible to exchange
conventional solvent for FAMEs. However, FAMEs are not as powerful diluents
as organic solvents and large amounts are required to obtain a suitable application
viscosity without any addition of traditional solvents. A conventional solvent free
formulation may still be attained by introduction of hyperbranched polyester
binders that possess low intrinsic viscosity.

Introduction of FAME in the paint system enhances the film formation process
entailing leveled and defect free films. This enhancement is achieved as FAME
acts as a plasticizer and tail solvent increasing the mobility during the final stages
of the film formation.

Incorporation of FAME in a fully formulated clear-coat network is difficult to

confirm. However, carbon-14 dating has shown that the final films contain FAME
that can not be extracted. The appearance and mechanical properties of the
coatings are also significantly affected by addition of FAMEs.

It has been confirmed with model studies that FAMEs can transesterify with the
hydroxyl groups of the polyester under industrial coil-coating cure conditions.
The transesterification reaction is affected by FAME structure, polyester polyol-
monomer structure, catalyst, and heating rate. Both evaporation and alkene
reactions of FAME are competing factors with the transesterification reaction
making it important to have oxidation-stable reactants. The degree of alkene
reactions increases with higher number of unsaturations in the fatty acid structure.

46
 Conclusions

Moreover, the placement of the double bonds is of importance as bis-allylic

hydrogen is more prone to participate in oxidative reactions.

The mechanical properties of free standing films are affected by addition of

FAMEs, oven temperature, co-solvent, and flash-off period. Incorporation of
FAME and extent of evaporation affect the glass transition. Tg is decreased by
incorporation of FAME and increased by evaporation, thermal and oxidative
polymerization. Further, FAME pattern affects the mechanical properties as
greater number of double bonds leads to a greater extent of alkene reactions
resulting in either polymerization or chain scission with subsequent evaporation
of the cleaved products. It is proposed that the isolated alkene of a triene FAME,
e.g. methyl pinolenate, acts as an additional cross-linking site instead of
contributing to the auto-oxidation. Introduction of FAMEs also increase the
heterogeneity of the coating.

Conventional film characterization tests on substrate supported coatings have

shown that the final film properties are affected by binder resin structure, and
oven temperature. The aliphatic structure of the hyperbranched polyesters entails
softer and less flexible films than eligible to withstand the rough forming process
associated with pre-coated sheet metal production. On the other hand, the
aromatic structure of the conventional polyester renders harder and less flexible
coatings than desired. However, introduction of reactive diluent in the
conventional polyester binder formulation results in coatings with good
formability properties. The hyperbranched polyester binder formulations are not
affected upon addition of reactive diluent due to their content of dangling chain
ends from the alkyl acid modification. A too low oven temperature has shown to
lead to uncured films with a greater content of reacted and non-reacted FAME due
to less evaporation.

Industrial evaluation of a fully formulated pigmented coil-coating coating

containing reactive diluents has shown that parts of the conventional volatile
solvent content can be exchanges for reactive diluent. The industrially cured
coating fulfills the film property demands specified for coil-coatings.

47
Future Work

6 FUTURE WORK
The present study shows that FAME can be used as reactive diluent in thermally
cured coil-coating systems. The evaluated system has not been optimized and
further studies on resin structure, and catalyst effect could improve the
compatibility, incorporation of reactive diluent, and final film properties. The
oxidation process of FAMEs under coil-coating cure conditions has not been fully
established with the performed studies. Extended studies could clarify the route of
the present oxidation and thermal alkene reactions.

Coil-coating products have long warranty-times, up to 20 years, and are seldom

used in non exposed conditions making durability tests, e.g. hydrolysis,
photolysis, and biological degradation durability, necessary. The industrial
collaborations partners have already commenced durability tests that will
proceed.

The use of a vegetable oil derived reactive diluent is environmental compliant as a

reduction of traditionally used solvents is possible and a renewable material is
introduces into the final film. The amount of incorporated reactive diluent can be
increased if the reactivity and functionality of the resin or the reactive diluent are
increased. FAMEs with other functionalities such as epoxy or hydroxyl groups can
for example be utilized. Reactive diluents with two functionalities could also be
used for dual cure systems where a coating can be cured a second time after the
forming process. The amount of renewable material in the coating could also be
increased by introduction of renewable monomers in the binder resin.

The effect of cure conditions could be further evaluated in order to introduce the
concept of reactive diluent in thermally cured coatings in other applications such
as original equipment manufacture (OEM).

48
 Acknowledgements

7 ACKNOWLEDGEMENTS
First of all I would like to thank my supervisor Prof. Mats Johansson for accepting
me as his PhD student, for his wide knowledge and for always having time for me
and my questions. You have been a great supervisor!

Prof. Anders Hult, Prof. Eva Malmström, Assist. Prof. Anna Carlmark Malkoch,
and Assist. Prof. Michael Malkoch are thanked for making Ytgruppen such a nice
research group to work in.

I would like to acknowledge the present and former heads of the department, and
all senior scientists at FPT for their hard work and creating an inspiring and
creative atmosphere. The administrative personnel, especially Inger Lord and
Margareta Andersson are thanked for taking care of all important paperwork.

EcoBuild, VINNOVA (The Swedish Agency for Innovation Systems), SSAB

Tunnplåt AB, Lantmännen, and Akzo Nobel Nippon Paint AB are gratefully
acknowledged for financial support. Knut och Alice Wallenbergs stiftelse, Knut
och Amalia Styffes stiftelse, and Stiftelsen Peter Klasons stipendiefond are
acknowledged for travel grants.

Per-Erik Sundell, Tina Bergman, Martin Svensson, Tomas Deltin, Irina Tran, and
Glenn Svensson are thanked for introducing me to the industrial perspective of
science and always having great interest in my work. Zrinka Kristic at Akzo Nobel
Nippon Paint AB is thanked for help with viscosity measurements and film
formation.

All PhD students at the department are thanked for all help and making this such
a great place to work at.

To all present and former members of Ytgruppen: You have made my time as a
PhD student really enjoyable. I will never forget all the fun and memorable
moments we have had together. Thanks to Robert for being a truly great friend, I

49
Acknowledgements

will miss not having you around, Camilla for being my constant travel companion
on our trips around the world, Linda for being almost better than me on
presenting my project, Hanna for being so competitive when I am not, Emma for
not completely leaving the group, Josefina for proving that you can live in a u-
land, Pontus for knowing how to behave late at night in corridors with really thin
walls, Pelle for being a truly great roomie and introducing me to all sorts of music,
Danne for all nice chats, Daniel for sharing my love and hate for the DMA, Petra
for daring to scream with me on the rollercoaster, Magnus for being a proud
Ducknose-winner, Andreas for always having an explanation, Peter for
personifying a real PhD, Robert V for his dancing skills, Yvonne for being the new
column expert, Susanne and Lina for being so chic, Johan and Ci for being my
idols in the vegetable oil-field, Alireza for continuing the work in the vegetable
oil-field, Mauro for always being on a sugar-rush, Sara for bearing to live in two
cities, Stacy for embracing the Swedish culture and language, Maribel for
introducing traditional Spanish food, George for being proud to be Greek, Neil for
teaching us weird English expressions, and Jarmo for knowing how to wear a
beer-hat. I would also like to thank all the diploma workers and project workers,
especially Andreas Schuler, who have visited the group over the years.

My friends outside the department are thanked; La Familia for being great friends
and believing in me, Majsan for knowing what it’s like to be a PhD student,
Eleonor, Christian, Anne, Jonas, and Engblom for all the parties.

Finally I would like to thank my family; The Ohlsson-Bengtsson-Andersson family

for always making me feel welcome in your homes; mamma, pappa, Jossan, and
Jocke for all your support and love.
Puss puss, Bojen!

50
 References

8 REFERENCES
[1] La Grotte de Lascaux, http://www.culture.gouv.fr/culture/arcnat/
lascaux/en, cited 2008-06-30.

[2] C. H. Hare, Protective coating; Fundamentals of chemistry and

composition, SSPC: The Society for Protective Coatings, Pittsburgh, PA,
U.S. (1998).

[3] Z. W. Wicks Jr., F. N. Jones, S. P. Pappas, D. A. Wicks, Organic coatings:

Science and technology, John Wiley & Sons, Inc., Hoboken, NJ, U.S.
(2007).

[4] A. R. Marrion, The chemistry and physics of coatings, The Royal Society
of Chemistry, Cambridge, UK (1994).

[5] D. Y. Perera, Prog. Org. Coat. 50 (2004) 247.

[6] A. Toussaint, Prog. Org. Coat. 2 (1974) 237.

[7] M. Dušková-Smrčková, K. Dušek, P. Vlasák, Macromol. Symp. 198 (2003)

259.

[8] J. Samuelsson, Ph.D. Thesis, Department of Polymer Technology, Royal

Institute of Technology, Stockholm, Sweden (2004).

[9] K. H. Zabel, R. P. Klaasen, W. J. Muizebelt, B. P. Gracey, C. Hallett, C. D.

Brooks, Prog. Org. Coat. 35 (1999) 255.

[10] ISO - International Organization for Standardization,

http://www.iso.org, cited 2008-06-30.

[11] CEN - European Committee for Standardization, http://www.cen.eu,

cited 2008-06-30.

51
References

[12] ASTM International - Standards Worldwide, http://www.astm.org, cited

2008-06-30.

[13] K. P. Menard, Dynamic Mechanical Analysis : A Practical Introduction,

CRC Press LLC, Boca Raton, Florida, USA (1999).

[14] European Coil Coating Association, http://www.eccacoil.com, cited

2008-06-17.

[15] F. Graziano, Metal Finishing 98 (2000) 175.

[16] P.-Å. Remming, L. Schön, L. Ståhlberg, G. Hügard, P. O. Alm, M.

Östlund, Boken om Prelaqproduktion, Pruntus/Sjuhärads, Borås,
Sweden (1994).

[17] The European Commission - Environment, http://ec.europa.eu/

environment, cited 2008-06-25.

[18] J. Lindeboom, Prog. Org. Coat. 34 (1998) 147.

[19] G. Toulemonde, G. Clausen, S. Vigneron, 2nd International Symposium

on Characterization and Control of Odours and VOC in the Process
Industries, Louvain-la-Neuve, Belgium, 61 (1993) 239.

[20] T. A. Misev, R. van der Linde, Prog. Org. Coat. 34 (1998) 160.

[21] P.-E. Sundell, C. Miller, C. Hasselgren, S. Jönsson, PRA RADnews 25

(1998) 3.

[22] P.-E. Sundell, A. Hult, C. Miller, C. Hasselgren, K. Lindgren, S. Jönsson,

8th Fusion UV Seminar, Japan, Paper 5. (2000).

[23] R. Drufke, Coil coating part III - The coatings and their markets, May
2005, http://www.specialchem4coatings.com, cited 2008-06-17.

[24] J. Samuelsson, P.-E. Sundell, M. Johansson, Prog. Org. Coat. 50 (2004)

193.

52
 References

[25] J. van Haveren, E. A. Oostveen, F. Miccichè, B. A. J. Noordover, C. E.

Koning, R. A. T. M. van Benthem, A. E. Frissen, J. G. J. Weijnrn, J. Coat.
Technol. Res. 4 (2007) 177.

[26] E. D. Casebolt, B. E. Mote, D. L. Trumbo, Prog. Org. Coat. 44 (2002) 147.

[27] W. J. Muizebelt, J. C. Hubert, M. W. F. Nielen, R. P. Klaasen, Prog. Org.

Coat. 40 (2000) 121.

[28] P. Muturi, D. Wang, S. Dirlikov, Prog. Org. Coat. 25 (1994) 85.

[29] C. Stenberg, M. Svensson, E. Wallström, M. Johansson, Surf. Coat. Int. B:

Coat. Trans. 88 (2005) 1.

[30] D. H. Solomon, The chemistry of organic film formers, Robert E. Krieger

Publishing Company, Malabar, FL, U.S. (1982).

[31] Z. Yaniv, D. Schafferman, M. Zur, Ind. Crops Prod. 3 (1995) 247.

[32] A. Overeem, G. J. H. Buisman, J. T. P. Derksen, F. P. Cuperus, L.

Molhoek, W. Grisnich, C. Goemans, Ind. Crops Prod. 10 (1999) 157.

[33] G. H. Hutchinson, Surf. Coat. Int. B: Coat. Trans. 85 (2002) 1.

[34] J. T. P. Derksen, F. P. Cuperus, P. Kolster, Prog. Org. Coat. 27 (1996) 45.

[35] J. T. P. Derksen, F. P. Cuperus, P. Kolster, Ind. Crops Prod. 3 (1995) 225.

[36] F.-L. Jin, S.-J. Park, Polym. Int. 57 (2008) 577.

[37] G. Lligadas, J. C. Ronda, M. Galià, V. Cádiz, Biomacromolecules 7 (2006)

2420.

[38] U. Biermann, W. Friedt, S. Lang, W. Lühs, G. Machmüller, J. O. Metzger,

M. Rüsch gen. Klaas, H. J. Schäfer, M. P. Schneider, Angew. Chem. Int. Ed
39 (2000) 2206.

[39] C. K. Hong, R. P. Wool, J. Appl. Polym. Sci. 95 (2005) 1524.

[40] P. T. Vasudevan, M. Briggs, J. Ind. Microbiol. Biotechnol. 35 (2008) 421.

53
References

[41] K. I. Mikolajczak, C. R. Smith, Jr., M. O. Bagby, I. A. Wolff, Journal of

Organic Chemistry 29 (1964) 318.

[42] V. Spitzer, S. A. D. Bordignon, E. P. Schenkel, F. Marx, J. Am. Oil Chem.

Soc. 71 (1994) 1343.

[43] I. Ncube, J. S. Read, P. Adlercreutz, B. Mattiasson, Photochemistry 47

(1998) 723.

[44] M. Lee, M. Lenman, A. Banas, M. Bafor, S. Singh, M. Schweizer, R.

Nilsson, C. Liljenberg, A. Dahlqvist, P. O. Gummeson, S. Sjodahl, A.
Green, S. Stymne, Science 280 (1998) 915.

[45] F. O. Ayorinde, F. T. Powers, L. D. Streete, R. L. Shepard, D. N. Tabi, J.

Am. Oil Chem. Soc. 66 (1989) 690.

[46] F. O. Ayorinde, G. Osman, R. L. Shepard, F. T. Powers, J. Am. Oil Chem.

Soc. 65 (1988) 1774.

[47] F. O. Ayorinde, E. Y. Nana, P. D. Nicely, A. S. Woods, E. O. Price, C. P.

Nwaonicha, J. Am. Oil Chem. Soc. 74 (1997) 531.

[48] K. A. A. Bryant, C. P. Nwaonicha, M. Hassan, M. A. Anderson, F. O.

Ayorinde, J. Am. Oil Chem. Soc. 70 (1993) 457.

[49] R. Subramanian, L. D. Schmidt, Angew. Chem. Int. Ed 44 (2005) 302.

[50] J. Mallégol, J.-L. Gardette, J. Lemaire, J. Am. Oil Chem. Soc. 76 (1999) 967.

[51] D. T. Coxon, K. R. Price, H. W.-S. Chan, Chem. Phys. Lipids 28 (1981) 365.

[52] E. N. Frankel, Prog. Lipid Res. 19 (1980) 1.

[53] N. A. Porter, S. E. Caldwell, K. A. Mills, Lipids 30 (1995) 277.

[54] J. C. Hubert, R. A. M. Vanderbosch, W. J. Muizebelt, R. P. Klaasen, K. H.

Zabel, Prog. Org. Coat. 31 (1997) 331.

[55] W. J. Muizebelt, J. C. Hubert, R. P. Klaasen, R. A. M. Vanderbosch, JCT

CoatingsTech 69 (1997) 59.

54
 References

[56] N. A. Porter, B. A. Weber, H. Weenen, J. A. Khan, J. Am. Chem. Soc. 102

(1980) 5597.

[57] N. A. Porter, L. S. Lehman, B. A. Weber, K. J. Smith, J. Am. Chem. Soc.

103 (1981) 6447.

[58] N. A. Porter, D. G. Wujek, J. Am. Chem. Soc. 106 (1984) 2626.

[59] E. N. Frankel, Prog. Lipid Res. 22 (1983) 1.

[60] E. N. Frankel, E. Selke, W. E. Neff, K. Miyashita, Lipids 27 (1992) 442.

[61] P. Fjällström, B. Andersson, C. Nilsson, K. Andersson, Ind. Crops Prod.

16 (2002) 173.

[62] H. W. Gardner, Free Radical Biology & Medicine 7 (1989) 65.

[63] G. A. Russell, J. Am. Chem. Soc. 79 (1957) 3871.

[64] J. Mallégol, J. Lemaire, J.-L. Gardette, Prog. Org. Coat. 39 (2000) 107.

[65] Z. O. Oyman, Ph. D. Thesis, Materials and Interface Chemistry,

Eindhoven University of Technology, Eindhoven, The Netherlands
(2005).

[66] R. van Gorkum, E. Bouwman, Coordination Chemistry Rewievs 249 (2005)

1709.

[67] M. Lazzari, O. Chiantore, Polymer Degradation and Stability 65 (1999) 303.

[68] W. J. Muizebelt, J. C. Hubert, R. A. M. Venderbosch, Prog. Org. Coat. 24

(1994) 263.

[69] A. Tekin, E. G. Hammond, J. Am. Oil Chem. Soc. 81 (2004) 481.

[70] W. J. Muizebelt, M. W. F. Nielen, J. Mass Spectrom. 31 (1996) 545.

[71] B. W. Matthäus, J. Am. Oil Chem. Soc. 73 (1996) 1039.

[72] C. Stenberg, M. Svensson, M. Johansson, Ind. Crops Prod. 21 (2005) 263.

55
References

[73] J. Samuelsson, M. Johansson, J. Am. Oil Chem. Soc. 78 (2001) 1191.

[74] J. Samuelsson, M. Johansson, J. Am. Oil Chem. Soc. 80 (2003) 491.

[75] J. Mallégol, J.-L. Gardette, J. Lemaire, J. Am. Oil Chem. Soc. 77 (2000) 257.

[76] G. Socrates, Infrared and Raman Characteristic Group Frequencies, John

Wiley & Sons Ltd., Chichester, UK (2001).

[77] R. L. Clough, N. C. Billingham, K. T. Gillen, Polymer Durability,

American Chemical Society, Washington, U.S. (1996).

[78] I. Blakey, G. A. George, Macromolecules 34 (2001) 1873.

[79] U. Schuchardt, R. Sercheli, R. M. Vargas, J. Braz. Chem. Soc. 9 (1998) 199.

[80] H. E. Hoydonckx, D. E. De Vos, S. A. Chavan, P. A. Jacobs, Topics in

Catalysis 27 (2004) 83.

[81] R. Lortie, Biotechnology Advances 15 (1997) 1.

[82] M. D. Bhabhe, V. D. Athawale, Prog. Org. Coat. 30 (1997) 207.

[83] J. Samuelsson, M. Jonsson, T. Brinck, M. Johansson, J. Polym. Sci. Part A:

Polym. Chem. 42 (2004) 6346.

[84] U. Bexell, M. Olsson, M. Johansson, J. Samuelsson, P.-E. Sundell, Surf.

Coat. Tech. 166 (2003) 141.

[85] J. Chen, M. D. Souceka, W. J. Simonsickb, R. W. Celikay, Polymer 43

(2002) 5379.

[86] C. Wang, S. Erhan, J. Am. Oil Chem. Soc. 76 (1999) 1211.

[87] R. S. Balakrishna, B. G. K. Murthy, J. S. Aggarwal, J. Am. Oil Chem. Soc.

48 (1971) 689.

[88] G. S. R. Sastry, B. G. K. Murthy, J. S. Aggarwal, J. Am. Oil Chem. Soc. 49

(1972) 192.

56
 References

[89] A. Behr, M. Flene, F. Naendrup, K. Schürmann, Eur. J. Lipid Sci. Technol.

102 (2000) 342.

[90] C. Walling, W. Helmreich, J. Am. Chem. Soc. 81 (1959) 1144.

[91] J. M. Jones, Organic Chemistry, W. W. Norton & Company, New York,

NY, U.S. (1997).

[92] H. Hintze-Brüning, Ind. Crops Prod. 1 (1993) 89.

[93] J. Samuelsson, M. Johansson, J. Am. Oil Chem. Soc. 78 (2001) 1191.

[94] J. M. J. Fréchet, D. A. Tomalia, Dendrimers and other dendritic

polymers, John Wiley & Sons, Ltd., Chichester, UK (2001).

[95] H. Magnusson, E. Malmström, A. Hult, M. Johansson, Polymer 43 (2002)

301.

[96] C. J. Hawker, E. Malmström, C. W. Frank, J. P. Kampf, J. Am. Chem. Soc.

119 (1997) 9903.

[97] E. Malmström, A. Hult, Polymer Reviews 37 (1997) 555.

[98] P. N. Mehta, Surf. Coat. Int. B: Coat. Trans. 89 (2006) 333.

[99] D. K. Chattopadhyay, K. V. S. N. Raju, Prog. Polym. Sci. 32 (2007) 352.

[100] Z. Huang, W. Shi, European Polymer Journal 43 (2007) 1302.

[101] S. S. Mahapatra, N. Karak, Prog. Org. Coat. 60 (2007) 328.

[102] P. Löwenhielm, D. Nyström, M. Johansson, A. Hult, Prog. Org. Coat. 54

(2005) 269.

[103] M. Johansson, T. Glauser, A. Jansson, A. Hult, E. Malmström, H.

Claesson, Prog. Org. Coat. 48 (2003) 194.

[104] M. Johansson, T. Glauser, G. Rospo, A. Hult, J. Appl. Polym. Sci. 75 (2000)

612.

57
References

[105] H. Claesson, E. Malmström, M. Johansson, A. Hult, Polymer 43 (2002)

3511.

[106] M. Johansson, E. Malmström, A. Hult, J. Polym. Sci. Part A: Polym. Chem

31 (1993) 619.

[107] X. Cheng, Z. Huang, J. Liu, W. Shi, Prog. Org. Coat. 59 (2007) 284.

[108] Z. Chen, D. C. Webster, Polym. Int. 56 (2007) 754.

[109] K. Mańczyk, P. Szewczyk, Prog. Org. Coat. 44 (2002) 99.

[110] C. Karakaya, G. Gündüz, L. Aras, İ. A. Mecidoğlu, Prog. Org. Coat. 59

(2007) 265.

[111] R. A. T. M. van Benthem, Prog. Org. Coat. 40 (2000) 203.

[112] T. Takagi, K. Miyashita, J. Am. Oil Chem. Soc. 64 (1987) 407.

[113] P. Sykes, A Guidebook to Mechanism in Organic Chemistry, Longman

Groups Ltd., New York, U.S. (1981).

[114] W. J. Jackson Jr., J. R. Caldwell, J. Appl. Polym. Sci. 11 (1967) 211.

[115] W. J. Jackson Jr., J. R. Caldwell, J. Appl. Polym. Sci. 11 (1967) 227.

[116] J. Boye, J. J. Martinez, C. Lacabanne, P. Perret, B. Chabert, J. F. Gerard,

Polymer 33 (1992) 323.

[117] SSAB Tunnplåt AB - Prelaq, http://www.prelaq.com, cited 2008-07-15.

58
 Corrections

CORRECTIONS
Corrections to paper I.

1. The figure text for Figure 4 should read “Fig. 4. Conversion vs. reaction time
for mixtures B2, F2, and J2 at 110°C.” instead of “Fig. 4. Conversion vs.
reaction time for mixtures B2, D2, and F2 at 110°C.”

2. The legend in Figure 4 should read “■ B2, ● F2, ▲ J2” instead of “■ B2, ● D2,
▲ F2”.

3. The figure text for Figure 5 should read “Fig. 5. Degree of oxidation vs.
reaction time for mixtures B2, F2, and J2 at 110°C.” instead of “Fig. 4.
Conversion vs. reaction time for mixtures B2, D2, and F2 at 110°C.”

4. The legend in Figure 5 should read “■ B2, ● F2, ▲ J2” instead of “■ B2, ● D2,
▲ F2”.

5. Figure 7 should be replace by

100

90

80

70 B2
Conversion [%]

D2
60
F2
50 H2
J2
40
L2
30

20

10

0
0 10 20 30 40 50 60
Time [min]

Fig. 7. Conversion vs. time for B2, D2, F2, H2, J2, and L2 at 130°C.

59